Nonmelanoma Skin Cancer

The calcium channel blocker amlodipine besylate was effective in reducing the number of muscle cramps caused by vismodegib, a basal cell carcinoma drug, according to a research letter from Dr. Mina Ally and her associates.

Patients who took amlodipine had the number of muscle cramps halved after 2 weeks of treatment, and this level was maintained for the 8-week medication regimen. No significant change in cramp severity, duration, or frequency of nighttime awakenings was seen. Side effects only appeared in too patients, with one reporting mild intermittent dizziness and another reporting grade 1 peripheral edema.

The control group saw a nonsignificant increase in cramp frequency, compared with the significant decrease in the amlodipine group. No change was seen in cramp severity, duration, or number of nighttime awakenings in the control group.

“Amlodipine may be effective in vismodegib-induced muscle cramps because it blocks voltage-gated calcium channels and inhibits the transport of extracellular calcium into muscle that is required for contraction,” the investigators noted.

Find the full research letter in JAMA Dermatology (doi:10.1001/jamadermatol.2015.1937).

lfranki@frontlinemedcom.com

The calcium channel blocker amlodipine besylate was effective in reducing the number of muscle cramps caused by vismodegib, a basal cell carcinoma drug, according to a research letter from Dr. Mina Ally and her associates.

Patients who took amlodipine had the number of muscle cramps halved after 2 weeks of treatment, and this level was maintained for the 8-week medication regimen. No significant change in cramp severity, duration, or frequency of nighttime awakenings was seen. Side effects only appeared in too patients, with one reporting mild intermittent dizziness and another reporting grade 1 peripheral edema.

The control group saw a nonsignificant increase in cramp frequency, compared with the significant decrease in the amlodipine group. No change was seen in cramp severity, duration, or number of nighttime awakenings in the control group.

“Amlodipine may be effective in vismodegib-induced muscle cramps because it blocks voltage-gated calcium channels and inhibits the transport of extracellular calcium into muscle that is required for contraction,” the investigators noted.

Find the full research letter in JAMA Dermatology (doi:10.1001/jamadermatol.2015.1937).

lfranki@frontlinemedcom.com

The calcium channel blocker amlodipine besylate was effective in reducing the number of muscle cramps caused by vismodegib, a basal cell carcinoma drug, according to a research letter from Dr. Mina Ally and her associates.

Patients who took amlodipine had the number of muscle cramps halved after 2 weeks of treatment, and this level was maintained for the 8-week medication regimen. No significant change in cramp severity, duration, or frequency of nighttime awakenings was seen. Side effects only appeared in too patients, with one reporting mild intermittent dizziness and another reporting grade 1 peripheral edema.

The control group saw a nonsignificant increase in cramp frequency, compared with the significant decrease in the amlodipine group. No change was seen in cramp severity, duration, or number of nighttime awakenings in the control group.

“Amlodipine may be effective in vismodegib-induced muscle cramps because it blocks voltage-gated calcium channels and inhibits the transport of extracellular calcium into muscle that is required for contraction,” the investigators noted.

Find the full research letter in JAMA Dermatology (doi:10.1001/jamadermatol.2015.1937).

lfranki@frontlinemedcom.com

What does your patient need to know at the first visit?

Patients need a realistic approach to photoprotection based on their genetics, including Fitzpatrick skin type and family history of melanoma and nonmelanoma skin cancer; skin examination for photodamage and photoaging as well as number and type of pigmented lesions; and lifestyle history, which should include location of residence as well as occupation and recreational pursuits. This discussion should, as usual, include questions about general health, systemic and skin disease, and medication usage, with particular focus on photoaggravated diseases such as lupus and melasma as well as ongoing use of topical agents and systemic photosensitizers. These inquiries should lead to a frank discussion of the patient’s risk for developing photodamage and skin cancer and other specific conditions that alter the advice you would give.

What are your go-to treatments? Is your recommendation anecdotal or evidence based? What are the side effects?

I always recommend that my patients use a product that they like, which may sound simplistic. But if the patient doesn’t like the feel and look of the sunscreen, he/she won’t use it. Patients routinely should use a sunscreen with a sun protection factor (SPF) of 30 or higher that also carries a “broad spectrum” label. At the beach or during sweaty sports, patients should use one with a water-resistant SPF.

I prefer spray sunscreens for application on the back if the patient is alone without someone to help apply sunscreen to hard-to-reach areas and for male scalps. But you never know how much spray to use, so use a lot!

If patients are at the beach, playing sports, or watching sports outside, then they should reapply sunscreen every 2 hours. If patients work indoors and use a facial sunscreen in the morning, that’s sufficient.

Although there is no evidence that sunscreens are harmful for children older than 6 months of age and pregnant women, if patients in these special populations have concerns, I recommend using agents with inorganic compounds (physical blockers) such as titanium dioxide and zinc oxide only. Children are best protected with clothing and hats.

The evidence supports this approach. Patients really don’t need SPF 30 protection, but no one uses the amount of product that will result in the SPF listed on the bottle. So if patients use an SPF 30 or greater, they will get at least an SPF 15, which is sufficient everywhere but at the equator. Using SPF 30 the way we all apply it will give SPF 15–level protection.

There is evidence that sunscreens prevent squamous cell carcinoma, actinic keratosis, and photoaging. Early evidence, less strong but positive, also suggests protection against basal cell carcinoma and melanoma.

The biggest side effect is not using the sunscreen. Others include irritation and allergy. Irritation is common, but finding a product to use without irritation should be easy. Allergy is rarer, and when it occurs, it is usually due to the preservative or fragrance, not the active ingredients. If allergy does occur, patch testing by a dermatologist is necessary to determine the allergen.

Although it is still controversial, wearing sunscreens religiously can lead to vitamin D insufficiency or deficiency, which is particularly true for individuals with skin of color—Fitzpatrick skin types IV, V, and VI—and those cancer patients who adhere to rigorous photoprotection. These patients should be encouraged to take supplemental vitamin D₃ and I suggest 2000 IU; this recommendation is my opinion and is not evidence based.

As to the literature in the laypress about hormonal changes from benzophenone, cancer from retinoids, and nanoparticle toxicity: There is no evidence to support those claims.

How do you keep patients compliant with treatment?

Keep telling them, and then tell them again. 

What do you do if they refuse treatment?

Tell them to see someone else. 

What resources do you recommend to patients for more information?

Consult the American Academy of Dermatology Web site (www.aad.org) and the Skin Cancer Foundation (www.skincancer.org).

Editorial Note

Practical Pearls From the Cutis^® Board is a new feature that will appear in print and online (www.cutis.com). Each month a member of the Cutis Editorial Board will provide pearls relating to the practice needs of dermatologists. Future topics will include:

Looking for pearls on a specific topic? The Editorial Board welcomes your feedback on potential topics. Send an e-mail to the Editorial Office (cutis@frontlinemedcom.com) with your suggestions.

What does your patient need to know at the first visit?

Patients need a realistic approach to photoprotection based on their genetics, including Fitzpatrick skin type and family history of melanoma and nonmelanoma skin cancer; skin examination for photodamage and photoaging as well as number and type of pigmented lesions; and lifestyle history, which should include location of residence as well as occupation and recreational pursuits. This discussion should, as usual, include questions about general health, systemic and skin disease, and medication usage, with particular focus on photoaggravated diseases such as lupus and melasma as well as ongoing use of topical agents and systemic photosensitizers. These inquiries should lead to a frank discussion of the patient’s risk for developing photodamage and skin cancer and other specific conditions that alter the advice you would give.

What are your go-to treatments? Is your recommendation anecdotal or evidence based? What are the side effects?

I always recommend that my patients use a product that they like, which may sound simplistic. But if the patient doesn’t like the feel and look of the sunscreen, he/she won’t use it. Patients routinely should use a sunscreen with a sun protection factor (SPF) of 30 or higher that also carries a “broad spectrum” label. At the beach or during sweaty sports, patients should use one with a water-resistant SPF.

I prefer spray sunscreens for application on the back if the patient is alone without someone to help apply sunscreen to hard-to-reach areas and for male scalps. But you never know how much spray to use, so use a lot!

If patients are at the beach, playing sports, or watching sports outside, then they should reapply sunscreen every 2 hours. If patients work indoors and use a facial sunscreen in the morning, that’s sufficient.

Although there is no evidence that sunscreens are harmful for children older than 6 months of age and pregnant women, if patients in these special populations have concerns, I recommend using agents with inorganic compounds (physical blockers) such as titanium dioxide and zinc oxide only. Children are best protected with clothing and hats.

The evidence supports this approach. Patients really don’t need SPF 30 protection, but no one uses the amount of product that will result in the SPF listed on the bottle. So if patients use an SPF 30 or greater, they will get at least an SPF 15, which is sufficient everywhere but at the equator. Using SPF 30 the way we all apply it will give SPF 15–level protection.

There is evidence that sunscreens prevent squamous cell carcinoma, actinic keratosis, and photoaging. Early evidence, less strong but positive, also suggests protection against basal cell carcinoma and melanoma.

The biggest side effect is not using the sunscreen. Others include irritation and allergy. Irritation is common, but finding a product to use without irritation should be easy. Allergy is rarer, and when it occurs, it is usually due to the preservative or fragrance, not the active ingredients. If allergy does occur, patch testing by a dermatologist is necessary to determine the allergen.

Although it is still controversial, wearing sunscreens religiously can lead to vitamin D insufficiency or deficiency, which is particularly true for individuals with skin of color—Fitzpatrick skin types IV, V, and VI—and those cancer patients who adhere to rigorous photoprotection. These patients should be encouraged to take supplemental vitamin D₃ and I suggest 2000 IU; this recommendation is my opinion and is not evidence based.

As to the literature in the laypress about hormonal changes from benzophenone, cancer from retinoids, and nanoparticle toxicity: There is no evidence to support those claims.

How do you keep patients compliant with treatment?

Keep telling them, and then tell them again. 

What do you do if they refuse treatment?

Tell them to see someone else. 

What resources do you recommend to patients for more information?

Consult the American Academy of Dermatology Web site (www.aad.org) and the Skin Cancer Foundation (www.skincancer.org).

Editorial Note

Practical Pearls From the Cutis^® Board is a new feature that will appear in print and online (www.cutis.com). Each month a member of the Cutis Editorial Board will provide pearls relating to the practice needs of dermatologists. Future topics will include:

Looking for pearls on a specific topic? The Editorial Board welcomes your feedback on potential topics. Send an e-mail to the Editorial Office (cutis@frontlinemedcom.com) with your suggestions.

What does your patient need to know at the first visit?

Patients need a realistic approach to photoprotection based on their genetics, including Fitzpatrick skin type and family history of melanoma and nonmelanoma skin cancer; skin examination for photodamage and photoaging as well as number and type of pigmented lesions; and lifestyle history, which should include location of residence as well as occupation and recreational pursuits. This discussion should, as usual, include questions about general health, systemic and skin disease, and medication usage, with particular focus on photoaggravated diseases such as lupus and melasma as well as ongoing use of topical agents and systemic photosensitizers. These inquiries should lead to a frank discussion of the patient’s risk for developing photodamage and skin cancer and other specific conditions that alter the advice you would give.

What are your go-to treatments? Is your recommendation anecdotal or evidence based? What are the side effects?

I always recommend that my patients use a product that they like, which may sound simplistic. But if the patient doesn’t like the feel and look of the sunscreen, he/she won’t use it. Patients routinely should use a sunscreen with a sun protection factor (SPF) of 30 or higher that also carries a “broad spectrum” label. At the beach or during sweaty sports, patients should use one with a water-resistant SPF.

I prefer spray sunscreens for application on the back if the patient is alone without someone to help apply sunscreen to hard-to-reach areas and for male scalps. But you never know how much spray to use, so use a lot!

If patients are at the beach, playing sports, or watching sports outside, then they should reapply sunscreen every 2 hours. If patients work indoors and use a facial sunscreen in the morning, that’s sufficient.

Although there is no evidence that sunscreens are harmful for children older than 6 months of age and pregnant women, if patients in these special populations have concerns, I recommend using agents with inorganic compounds (physical blockers) such as titanium dioxide and zinc oxide only. Children are best protected with clothing and hats.

The evidence supports this approach. Patients really don’t need SPF 30 protection, but no one uses the amount of product that will result in the SPF listed on the bottle. So if patients use an SPF 30 or greater, they will get at least an SPF 15, which is sufficient everywhere but at the equator. Using SPF 30 the way we all apply it will give SPF 15–level protection.

There is evidence that sunscreens prevent squamous cell carcinoma, actinic keratosis, and photoaging. Early evidence, less strong but positive, also suggests protection against basal cell carcinoma and melanoma.

The biggest side effect is not using the sunscreen. Others include irritation and allergy. Irritation is common, but finding a product to use without irritation should be easy. Allergy is rarer, and when it occurs, it is usually due to the preservative or fragrance, not the active ingredients. If allergy does occur, patch testing by a dermatologist is necessary to determine the allergen.

Although it is still controversial, wearing sunscreens religiously can lead to vitamin D insufficiency or deficiency, which is particularly true for individuals with skin of color—Fitzpatrick skin types IV, V, and VI—and those cancer patients who adhere to rigorous photoprotection. These patients should be encouraged to take supplemental vitamin D₃ and I suggest 2000 IU; this recommendation is my opinion and is not evidence based.

As to the literature in the laypress about hormonal changes from benzophenone, cancer from retinoids, and nanoparticle toxicity: There is no evidence to support those claims.

How do you keep patients compliant with treatment?

Keep telling them, and then tell them again. 

What do you do if they refuse treatment?

Tell them to see someone else. 

What resources do you recommend to patients for more information?

Consult the American Academy of Dermatology Web site (www.aad.org) and the Skin Cancer Foundation (www.skincancer.org).

Editorial Note

Practical Pearls From the Cutis^® Board is a new feature that will appear in print and online (www.cutis.com). Each month a member of the Cutis Editorial Board will provide pearls relating to the practice needs of dermatologists. Future topics will include:

Looking for pearls on a specific topic? The Editorial Board welcomes your feedback on potential topics. Send an e-mail to the Editorial Office (cutis@frontlinemedcom.com) with your suggestions.

Skin cancer patients should beware of products available online that fraudulently claim to prevent and cure cancer, including melanoma and nonmelanoma skin cancers, according to the US Food and Drug Administration (FDA). These products often are marketed as natural treatments or dietary supplements. They have not gained FDA approval and therefore are not proven to be safe or effective. Rather, they can cause more harm to patients and delay the effects of conventional cancer treatments.

Firms that illegally market fraudulent cancer treatments often use exaggerated unsubstantiated claims to promote their products. The FDA has provided consumer health information with several phrases that consumers should recognize as warning signs for fraudulent cancer treatments:

• “Scientific breakthrough”

• “Miraculous cure”

• “Ancient remedy”

• “Treats all forms of cancer”

• “Skin cancers disappear”

• “Shrinks malignant tumors”

• “Nontoxic”

• “Doesn’t make you sick”

• “Avoid painful surgery, radiotherapy, chemotherapy, or other conventional treatments”

• “Treat nonmelanoma skin cancers easily and safely”

• “Target cancer cells while leaving healthy cells alone”

Undocumented case histories or personal testimonials from patients or physicians claiming amazing results; suggestions that a product can treat serious or incurable diseases; and promises of no-risk, money-back guarantees also are signs of health fraud.

The FDA has cited black salves as one of the fake cancer remedies that have proven to be harmful. In a June 2015 Cutis article “Black Salve and Bloodroot Extract in Dermatologic Conditions,” Hou and Brewer reported an increased popularity of self-treatment with black salves in curing skin cancers and healing other skin conditions due to extensive advertising of its effectiveness. According to the FDA, black salves are sold with false promises that they will cure melanoma and nonmelanoma skin cancers by “drawing out” the disease from beneath the skin. However, Hou and Brewer warned that some black salves contain escharotics such as zinc chloride and bloodroot, which could cause damage to healthy tissue.

“Despite the information and testimonials that are widely available on the Internet, black salve use has not been validated by rigorous studies,” the authors reported. “[It] is not regulated by the US Food and Drug Administration, resulting in poor quality control and inconsistent user instructions.”

Dermatologists should be aware that skin cancer patients may be attracted to alternative treatments such as black salves. Health care professionals should educate patients about fraudulent cancer treatments versus investigational treatments.

For a complete list of fake cancer cures consumers should avoid, consult the FDA. 

Skin cancer patients should beware of products available online that fraudulently claim to prevent and cure cancer, including melanoma and nonmelanoma skin cancers, according to the US Food and Drug Administration (FDA). These products often are marketed as natural treatments or dietary supplements. They have not gained FDA approval and therefore are not proven to be safe or effective. Rather, they can cause more harm to patients and delay the effects of conventional cancer treatments.

Firms that illegally market fraudulent cancer treatments often use exaggerated unsubstantiated claims to promote their products. The FDA has provided consumer health information with several phrases that consumers should recognize as warning signs for fraudulent cancer treatments:

• “Scientific breakthrough”

• “Miraculous cure”

• “Ancient remedy”

• “Treats all forms of cancer”

• “Skin cancers disappear”

• “Shrinks malignant tumors”

• “Nontoxic”

• “Doesn’t make you sick”

• “Avoid painful surgery, radiotherapy, chemotherapy, or other conventional treatments”

• “Treat nonmelanoma skin cancers easily and safely”

• “Target cancer cells while leaving healthy cells alone”

Undocumented case histories or personal testimonials from patients or physicians claiming amazing results; suggestions that a product can treat serious or incurable diseases; and promises of no-risk, money-back guarantees also are signs of health fraud.

The FDA has cited black salves as one of the fake cancer remedies that have proven to be harmful. In a June 2015 Cutis article “Black Salve and Bloodroot Extract in Dermatologic Conditions,” Hou and Brewer reported an increased popularity of self-treatment with black salves in curing skin cancers and healing other skin conditions due to extensive advertising of its effectiveness. According to the FDA, black salves are sold with false promises that they will cure melanoma and nonmelanoma skin cancers by “drawing out” the disease from beneath the skin. However, Hou and Brewer warned that some black salves contain escharotics such as zinc chloride and bloodroot, which could cause damage to healthy tissue.

“Despite the information and testimonials that are widely available on the Internet, black salve use has not been validated by rigorous studies,” the authors reported. “[It] is not regulated by the US Food and Drug Administration, resulting in poor quality control and inconsistent user instructions.”

Dermatologists should be aware that skin cancer patients may be attracted to alternative treatments such as black salves. Health care professionals should educate patients about fraudulent cancer treatments versus investigational treatments.

For a complete list of fake cancer cures consumers should avoid, consult the FDA. 

Skin cancer patients should beware of products available online that fraudulently claim to prevent and cure cancer, including melanoma and nonmelanoma skin cancers, according to the US Food and Drug Administration (FDA). These products often are marketed as natural treatments or dietary supplements. They have not gained FDA approval and therefore are not proven to be safe or effective. Rather, they can cause more harm to patients and delay the effects of conventional cancer treatments.

Firms that illegally market fraudulent cancer treatments often use exaggerated unsubstantiated claims to promote their products. The FDA has provided consumer health information with several phrases that consumers should recognize as warning signs for fraudulent cancer treatments:

• “Scientific breakthrough”

• “Miraculous cure”

• “Ancient remedy”

• “Treats all forms of cancer”

• “Skin cancers disappear”

• “Shrinks malignant tumors”

• “Nontoxic”

• “Doesn’t make you sick”

• “Avoid painful surgery, radiotherapy, chemotherapy, or other conventional treatments”

• “Treat nonmelanoma skin cancers easily and safely”

• “Target cancer cells while leaving healthy cells alone”

Undocumented case histories or personal testimonials from patients or physicians claiming amazing results; suggestions that a product can treat serious or incurable diseases; and promises of no-risk, money-back guarantees also are signs of health fraud.

The FDA has cited black salves as one of the fake cancer remedies that have proven to be harmful. In a June 2015 Cutis article “Black Salve and Bloodroot Extract in Dermatologic Conditions,” Hou and Brewer reported an increased popularity of self-treatment with black salves in curing skin cancers and healing other skin conditions due to extensive advertising of its effectiveness. According to the FDA, black salves are sold with false promises that they will cure melanoma and nonmelanoma skin cancers by “drawing out” the disease from beneath the skin. However, Hou and Brewer warned that some black salves contain escharotics such as zinc chloride and bloodroot, which could cause damage to healthy tissue.

“Despite the information and testimonials that are widely available on the Internet, black salve use has not been validated by rigorous studies,” the authors reported. “[It] is not regulated by the US Food and Drug Administration, resulting in poor quality control and inconsistent user instructions.”

Dermatologists should be aware that skin cancer patients may be attracted to alternative treatments such as black salves. Health care professionals should educate patients about fraudulent cancer treatments versus investigational treatments.

For a complete list of fake cancer cures consumers should avoid, consult the FDA. 

Approximately 2 million basal cell carcinomas occur in the United States each year, a number that has increased only modestly during the past 15 years, according to a report published online June 3 in JAMA Dermatology.

The epidemiology of BCCs has been difficult to pin down because these tumors are excluded from cancer registries and national cancer surveillance programs, and because when they are tracked they are usually lumped together with squamous cell carcinomas. The most recent National Cancer Institute–funded survey of BCC was done more than 30 years ago.

©Kelly Nelson/National Cancer Institute

So the current incidence of BCC “is not well characterized.” Nevertheless. some researchers have posited that the incidence has risen 80%-200% during the past 10-20 years, said Dr. Maryam M. Asgari of Kaiser Permanente Northern California, Oakland, and the department of dermatology at the University of California, San Francisco.

To estimate incidence more accurately, Dr. Asgari and her associates analyzed data from every electronic pathology report in a registry covering the 3.2 million participants in the HMO from 1998 through 2012. They identified 221,624 cases of BCC in patients aged 2-105 years. The annual incidence increased from 513 to 600 cases per 100,000 population during the study period – a 17% rise that was deemed “not remarkable.”

“In extrapolating our data to the United States, we estimate that approximately 2 million individuals develop at least one BCC in the U.S. in a given year,” Dr. Asgari and her associates wrote (JAMA Dermatol. 2015 June 3 [doi:10.1001/jamadermatol.2015.1188]).

This total is higher than that cited on the National Institutes of Health website, which estimates the annual incidence of all nonmelanoma skin cancers, not just BCC, at 2 million per year.

Males were at higher risk than were females for developing BCC (incidence rate ratio, 1.65), and risk increased with increasing patient age. As expected, whites were at 8- to 70-fold greater risk than were Hispanics, Asians, or blacks. Contrary to one previous report, the incidence of BCC did not increase among Hispanics during the study period.

The National Cancer Institute and the National Institutes of Health supported the study. Dr. Asgari reported receiving research grants from Kaiser Permanente, Pfizer, and Valeant Pharmaceuticals, but not for this project.

Approximately 2 million basal cell carcinomas occur in the United States each year, a number that has increased only modestly during the past 15 years, according to a report published online June 3 in JAMA Dermatology.

The epidemiology of BCCs has been difficult to pin down because these tumors are excluded from cancer registries and national cancer surveillance programs, and because when they are tracked they are usually lumped together with squamous cell carcinomas. The most recent National Cancer Institute–funded survey of BCC was done more than 30 years ago.

©Kelly Nelson/National Cancer Institute

So the current incidence of BCC “is not well characterized.” Nevertheless. some researchers have posited that the incidence has risen 80%-200% during the past 10-20 years, said Dr. Maryam M. Asgari of Kaiser Permanente Northern California, Oakland, and the department of dermatology at the University of California, San Francisco.

To estimate incidence more accurately, Dr. Asgari and her associates analyzed data from every electronic pathology report in a registry covering the 3.2 million participants in the HMO from 1998 through 2012. They identified 221,624 cases of BCC in patients aged 2-105 years. The annual incidence increased from 513 to 600 cases per 100,000 population during the study period – a 17% rise that was deemed “not remarkable.”

“In extrapolating our data to the United States, we estimate that approximately 2 million individuals develop at least one BCC in the U.S. in a given year,” Dr. Asgari and her associates wrote (JAMA Dermatol. 2015 June 3 [doi:10.1001/jamadermatol.2015.1188]).

This total is higher than that cited on the National Institutes of Health website, which estimates the annual incidence of all nonmelanoma skin cancers, not just BCC, at 2 million per year.

Males were at higher risk than were females for developing BCC (incidence rate ratio, 1.65), and risk increased with increasing patient age. As expected, whites were at 8- to 70-fold greater risk than were Hispanics, Asians, or blacks. Contrary to one previous report, the incidence of BCC did not increase among Hispanics during the study period.

The National Cancer Institute and the National Institutes of Health supported the study. Dr. Asgari reported receiving research grants from Kaiser Permanente, Pfizer, and Valeant Pharmaceuticals, but not for this project.

Approximately 2 million basal cell carcinomas occur in the United States each year, a number that has increased only modestly during the past 15 years, according to a report published online June 3 in JAMA Dermatology.

The epidemiology of BCCs has been difficult to pin down because these tumors are excluded from cancer registries and national cancer surveillance programs, and because when they are tracked they are usually lumped together with squamous cell carcinomas. The most recent National Cancer Institute–funded survey of BCC was done more than 30 years ago.

©Kelly Nelson/National Cancer Institute

So the current incidence of BCC “is not well characterized.” Nevertheless. some researchers have posited that the incidence has risen 80%-200% during the past 10-20 years, said Dr. Maryam M. Asgari of Kaiser Permanente Northern California, Oakland, and the department of dermatology at the University of California, San Francisco.

To estimate incidence more accurately, Dr. Asgari and her associates analyzed data from every electronic pathology report in a registry covering the 3.2 million participants in the HMO from 1998 through 2012. They identified 221,624 cases of BCC in patients aged 2-105 years. The annual incidence increased from 513 to 600 cases per 100,000 population during the study period – a 17% rise that was deemed “not remarkable.”

“In extrapolating our data to the United States, we estimate that approximately 2 million individuals develop at least one BCC in the U.S. in a given year,” Dr. Asgari and her associates wrote (JAMA Dermatol. 2015 June 3 [doi:10.1001/jamadermatol.2015.1188]).

This total is higher than that cited on the National Institutes of Health website, which estimates the annual incidence of all nonmelanoma skin cancers, not just BCC, at 2 million per year.

Males were at higher risk than were females for developing BCC (incidence rate ratio, 1.65), and risk increased with increasing patient age. As expected, whites were at 8- to 70-fold greater risk than were Hispanics, Asians, or blacks. Contrary to one previous report, the incidence of BCC did not increase among Hispanics during the study period.

The National Cancer Institute and the National Institutes of Health supported the study. Dr. Asgari reported receiving research grants from Kaiser Permanente, Pfizer, and Valeant Pharmaceuticals, but not for this project.

When follicular mucinosis (FM) is defined as an epithelial reaction pattern characterized by intrafollicular and perifollicular mucin accumulation, it cannot be considered a distinct disease entity, as this pattern is ubiquitously present in various inflammatory and neoplastic skin conditions.^1,2 The distinction between idiopathic FM and lymphoma-associated follicular mucinosis (LAFM) was made several years ago by authors who evaluated the differences in the clinical presentation of these entities, including patient age at onset, number of lesions, pattern of distribution, and most importantly clinical progression.¹ In this article, we discuss the importance of close clinical follow-up in patients with FM or patch-stage mycosis fungoides (MF) in whom histopathologic evaluation is ambiguous or nondiagnostic. We also highlight the value of ancillary testing, including T-cell receptor gene rearrangement, flow cytometry, and immunohistochemistry, as a component in the diagnostic process rather than the sole diagnostic moiety. A review of the pertinent literature also is performed.

History of FM and MF
Pinkus³ first described an entity he termed alopecia mucinosa in 1957. Pinkus noted 3 distinct patterns: an idiopathic form of alopecia mucinosa, lymphoblastoma with associated FM, and alopecia mucinosa that later transformed into lymphoblastoma.⁴ In 1983, however, Pinkus⁴ described uncertainty if alopecia mucinosa represented the first stage of MF or if patients with alopecia mucinosa were simply at an increased risk for developing lymphoma. He believed there were too many cases of lymphoma following a diagnosis of alopecia mucinosa for the relation to be coincidental, yet he noted that many of the cases resolved either spontaneously or following treatment with x-rays or topical steroids. He concluded his report with a sentiment that is echoed in many current studies regarding this entity: “Many questions surrounding this entity are as unanswerable today as they were 25 years ago.”⁴

Jablonska et al⁵ were the first to coin the term mucinosis follicularis, now known as FM, to replace alopecia mucinosa because they felt the description was more accurate, as lesions also arise on non–hair-bearing skin. Although there is general agreement that there is a form of MF that has associated FM, this is where the agreement ends with regard to the diagnosis of MF versus FM. Böer et al⁶ discussed the historic evolution of these terms, mostly to highlight the origins of the confusion. The investigators proposed that FM should only be used as a descriptive term and that all cases of alopecia mucinosa represent MF. They also concluded that many benign dermatoses associated with a risk for evolution to MF (eg, small and large plaque psoriasis [LPP]) should simply be diagnosed as MF.⁶ Subsequently, the proposal that idiopathic FM and LAFM are not 2 distinct entities but rather a clinicopathologic continuum and that idiopathic FM is simply a variant of MF along this spectrum has gained some approval.^6,7 However, this belief is not shared among all authorities in the field, and attempts to define diagnostic criteria that distinguish between a benign clinical course and a course that is more progressive and fatal continue. Currently, it is agreed upon that when distinguishing between these 2 clinical courses, primary (idiopathic) follicular mucinosis refers to a benign course with no overt sign of malignancy, and lymphoma-associated follicular mucinosis refers to a diagnostic malignant condition. Lymphoma-associated follicular mucinosis refers to FM associated with cutaneous T-cell lymphoma, the most common form of which is FM. Many authors^8-15 have sought ancillary methodologies in addition to clinical parameters to assist in the evaluation between both disease courses. Methodologies have included assessment of T-cell receptor gene rearrangements, flow cytometry, and immunohistochemical staining, mostly as an effort to establish monoclonality as a defining characteristic of LAFM; however, monoclonality in cutaneous T-cell infiltrates should be interpreted with caution and should not be considered as a confirmation of malignancy due to recent findings of monoclonality in benign inflammatory dermatoses such as lichen planus. The Table outlines several of the most common benign inflammatory dermatoses that demonstrate monoclonality, but this list should not be considered exhaustive, as there are many others in which monoclonality is sometimes seen.^8-15 The lack of definitive criteria to distinguish between the 2 groups has led to confusion and consternation regarding the diagnosis of idiopathic FM versus LAFM and has led many in the field to consider the 2 conditions to be one and the same.

Diagnosis of FM and MF: Clinicopathologic Features
The World Health Organization (WHO) defined MF as an epidermotropic primary cutaneous T-cell lymphoma (CTCL) characterized by infiltrates of small- to medium-sized T lymphocytes with cerebriform nuclei. Further, the WHO stated that the term mycosis fungoides should be exclusively reserved for classical cases typified by the evolution of cutaneous patches, plaques, and tumors, or for variants that show a similar clinical course.¹⁶ Mycosis fungoides is divided into 3 stages—patch, plaque, and tumor—which are solely clinical descriptors.¹⁷ The WHO also described a clinical staging system with pathologic emphasis placed only on lymph node involvement and identification of Sézary cells.¹⁶ It lists folliculotropic MF as a variant, with only some cases presenting with mucinous degeneration of hair follicles. A lack of consensus among pathologists regarding criteria for diagnosis in patch-stage MF remains, but diagnosis of plaque-stage disease is not regularly debated due to its more reliably present, well-developed histologic features (eg, haloed lymphocytes, epidermotropism of lymphocytes, lymphocytes with convoluted nuceli, Pautrier microabscesses).¹⁸ Although there have been specific histologic findings reported to be associated with patch-stage MF, they have only been present in a few cases and are therefore of limited usefulness in practice.^1,19 The categorization of patients with subtle histologic features common to both MF and inflammatory conditions such as parapsoriasis en plaques (the term plaque in this case is a misnomer because the word plaque means patch in French) continues to be elusive. A lack of agreement regarding LPP persists in the current literature in the same manner as FM. Some researchers have contended for many years that LPP is a type of MF, while others remain unconvinced, mainly due to the lack of evidence that lumping a benign condition (LPP) with an increased risk for malignant transformation and a known malignancy (MF) together is of any benefit to the patient. Assessment of clinicopathologic correlation, immunohistochemistry, clonality, and T-cell gene rearrangement have failed to positively identify patients who are at risk for disease progression, whether the diagnosis is called LPP or early patch-stage MF.²⁰

Mycosis fungoides is more common in males and its incidence increases with age; however, diagnosis should not be ruled out based on age or gender. Typical presentation of early-stage disease includes erythematous patches or plaques, often with light scaling.¹⁹ Lesions routinely are of long-standing duration (months to years), are located in areas that are infrequently exposed to sunlight, and often are 5 cm in diameter or larger with irregular borders.²¹ Associated poikiloderma is relatively specific to MF but rarely is seen in other CTCLs, connective-tissue diseases, and some genodermatoses. Poikiloderma commonly is identified in LPP, which shows the same telangiectasia, mottled pigmentation, and epidermal atrophy as MF-associated poikiloderma, leading some to believe that there is no separation between the 2 conditions. In all stages of MF, lesions frequently are numerous and occur on multiple sites. Plaques and tumors can show spontaneous ulceration. When lesions are folliculotropic, they can cause localized alopecia, follicular-based papules, and fungating pseudotumors in more advanced stages.¹ The clinical presentation of FM substantially overlaps with folliculotropic MF, and although FM lesions often are solitary and are located on the face or scalp, they also can present as multiple lesions located elsewhere on the body. It also has been proposed that folliculotropic MF should not be separated from FM-associated MF (or LAFM).²²

The characteristic histologic picture of LAFM in patch or plaque stage shows mucin deposition within hair follicles, similar to idiopathic FM. On histology, both conditions demonstrate dense lymphoid infiltrates around and within hair follicles as well as in the dermis (Figure). Most cases of LAFM show epidermotropism of lymphocytes between follicles, but this finding is not present in every case and often disappears when the disease advances to the tumor stage.^1,19 Although Pautrier microabscesses (collections of lymphocytes within the superficial epidermis) are considered to be somewhat specific to MF, they are only present in a minority of cases.²⁰ In a study by the International Society for Cutaneous Lymphomas,²¹ the only histopathologic criteria that showed any appreciable sensitivity or specificity in the diagnosis of MF were the presence of lymphoid cells with variable nuclear and cytoplasmic features and/or strikingly irregular nuclear contours with the presence of lymphocytes larger than those usually seen in inflammatory dermatoses. Despite these criteria, the study reported a high misclassification rate. A complicated scoring system for diagnosis of MF in patch- or early plaque-stage disease was proposed by the International Society for Cutaneous Lymphomas,²¹ which integrates clinical, histopathologic, molecular, and immunophenotypic criteria. However, these criteria have been continually debated in the literature and are only discussed in this article in relation to the association between MF and FM. Diagnosis of tumor-stage MF is not addressed in this article, as it is readily identified as lymphoma and is not easily confused with idiopathic FM.

Histopathology of 2 separate cases, both diagnosed as follicular mucinosis pending further follow-up. After close clinical follow-up, one case was diagnosed as mycosis fungoides (A) and the other as follicular mucinosis (B)(both H&E, original magnification ×2). Immunohistochemical stains were noncontributory, and T-cell gene rearrangement was positive only in the case of mycosis fungoides. Histopathology revealed mucin deposition in the hair follicles and a dense lymphoid infiltrate around and within the follicles in both cases. Pautrier microabscesses, cerebriform nuclei, and epidermotropism of lymphocytes between follicles were absent in the case of mycosis fungoides.

Clinical assessment of a patient’s medical history to identify persistent and progressive disease is paramount to the diagnosis of MF. Although MF lesions tend to increase in size and number over time, this presentation is not without exception.²¹ In early patch-stage disease, eliminating some of the patient’s current medications may be sufficient in clearing cutaneous patches that cannot be conclusively identified as either MF or a benign inflammatory lymphoid infiltrate, which further emphasizes the importance of clinical assessment of the patient’s medical history in the diagnosis of MF. The shape of the lesions also is helpful in distinguishing between MF and other skin disorders, such as digitate dermatosis or LPP; unlike the latter, the waxing and waning nature of MF lesions often produces irregularly shaped patches with little coalescence. Again, there are some investigators who believe that these lesions represent varying presentations of MF.⁶

In a study by Cerroni et al,¹ 44 patients with FM were divided into 2 groups: (1) a cohort of 16 patients with no history or clinical evidence of MF or Sézary syndrome (ie, LAFM), and (2) a cohort of 28 patients with clinicopathologic evidence of CTCL. Patients in both groups were followed for a maximum of 20 years. Results indicated that that the presence of perifollicular or intrafollicular mucin, epidermotropism of lymphocytes, monoclonality, and epidemiologic characteristics (eg, age, sex, race) cannot reliably distinguish the 2 disease forms. Furthermore, it was suggested that these conditions are not mutually exclusive entities and are actually variants of CTCL. The observation that the 2 diseases share prognostic overlap adds further credence to the already puzzling conundrum. Nineteen of 28 patients with MF were alive and well at follow-up, and all patients in the idiopathic FM group were alive, with only 9 of 16 patients showing residual disease and none with CTCL.¹

Other clinical factors that may be helpful in the diagnosis of MF are the presentation of lesions in non–sun-exposed areas of the skin and multiple lesions, as unilesional MF is exceedingly uncommon.²¹ No histologic features have been proven to predict which early patch- or plaque-stage MFs will progress to full-blown CTCL versus benign idiopathic FM; thus, great caution should be taken in patients with early-stage disease to ensure they are not prematurely and/or incorrectly classified as CTCL. Such a diagnosis has medical, social, and economical ramifications that should not be overlooked.

If idiopathic FM and LAFM were considered distinct disease processes, the ambiguity in making a definitive diagnosis should give the physician pause, and a diagnosis of LAFM may only be appropriate when there is unequivocal clinicopathologic evidence. Otherwise, a lymphoma diagnosis is somewhat superfluous and potentially harmful. Definitive diagnosis of LAFM also is complicated by reports of other hematologic malignancies presenting with FM-like histopathologic findings, such as chronic myelogenous leukemia, leukemia-associated eosinophilic folliculitis, and acute myeloblastic leukemia.^23,24 Although MF is the most common malignancy associated with FM, it is important to consider other less common malignancies that also may be present.

Diagnosis: Patient Consequences
Accurate diagnosis of idiopathic FM versus LAFM is critical, as the ramifications of a cancer diagnosis can have broad implications. For example, patients who receive cancer diagnoses often experience emotional trauma and social stigma, even when adequate patient education has been provided. The incidence of depression and anxiety also can increase following a cancer diagnosis and can be complicated by medical treatments (eg, systemic steroids, interferon),²⁵ which are known to increase the frequency of these psychological disturbances. Health insurance premiums likely will be higher if a patient is diagnosed with cancer versus a benign inflammatory condition. Hesitation of the pathologist to assign a cancer diagnosis when unequivocal evidence is not present should not be regarded as trickery, malpractice, or deceit of the health care bylaws, as benign language with suggestion of close clinical follow-up in the setting of diagnostic uncertainty will “first, do no harm” and secondly, serve as a vehicle for patient advocacy.

If there is a definitive distinction between idiopathic FM and LAFM, it requires further research before it can be fully understood. Currently, the WHO does not recognize a diagnosis of FM-associated MF (or LAFM) and acknowledges that folliculotropic MF is not always associated with FM.^16,26 Given uncertainty and repercussions associated with a cancer diagnosis, however indolent, it may be morally responsible and medically favorable for physicians to consider FM in the differential diagnosis when applicable rather than making a diagnosis of MF outright. Given the importance of both clinical and histologic factors, it may be beneficial for definitive diagnosis of FM versus MF to lie with the clinician, while the pathologist serves as an adjunct in the diagnostic process. Because the prognosis of idiopathic FM often is marred by possible transformation into MF or other CTCLs, therapeutic decisions should be dictated by close clinical follow-up. Additionally, stage of disease, patient age, treatment compliance, comorbidities, and possible side effects of medications should all be considered when evaluating potential therapeutic regimens.²⁷

Conclusion
Research is underway to more accurately identify patients with FM who are at risk for progression to LAFM versus those with benign remitting FM. Once the required diagnostic criteria are established to accurately classify these patients, with an emphasis on prognosis and suggested treatments, it might be necessary to establish new, less debated terminology so pathologists and clinicians alike can improve patient care. Continued histopathologic scrutiny, use of sophisticated molecular techniques, and knowledge of other currently undiscovered modalities will shed light on this important disease process and aid in proper disease management, which may be advantageous to both patients and physicians.

When follicular mucinosis (FM) is defined as an epithelial reaction pattern characterized by intrafollicular and perifollicular mucin accumulation, it cannot be considered a distinct disease entity, as this pattern is ubiquitously present in various inflammatory and neoplastic skin conditions.^1,2 The distinction between idiopathic FM and lymphoma-associated follicular mucinosis (LAFM) was made several years ago by authors who evaluated the differences in the clinical presentation of these entities, including patient age at onset, number of lesions, pattern of distribution, and most importantly clinical progression.¹ In this article, we discuss the importance of close clinical follow-up in patients with FM or patch-stage mycosis fungoides (MF) in whom histopathologic evaluation is ambiguous or nondiagnostic. We also highlight the value of ancillary testing, including T-cell receptor gene rearrangement, flow cytometry, and immunohistochemistry, as a component in the diagnostic process rather than the sole diagnostic moiety. A review of the pertinent literature also is performed.

History of FM and MF
Pinkus³ first described an entity he termed alopecia mucinosa in 1957. Pinkus noted 3 distinct patterns: an idiopathic form of alopecia mucinosa, lymphoblastoma with associated FM, and alopecia mucinosa that later transformed into lymphoblastoma.⁴ In 1983, however, Pinkus⁴ described uncertainty if alopecia mucinosa represented the first stage of MF or if patients with alopecia mucinosa were simply at an increased risk for developing lymphoma. He believed there were too many cases of lymphoma following a diagnosis of alopecia mucinosa for the relation to be coincidental, yet he noted that many of the cases resolved either spontaneously or following treatment with x-rays or topical steroids. He concluded his report with a sentiment that is echoed in many current studies regarding this entity: “Many questions surrounding this entity are as unanswerable today as they were 25 years ago.”⁴

Jablonska et al⁵ were the first to coin the term mucinosis follicularis, now known as FM, to replace alopecia mucinosa because they felt the description was more accurate, as lesions also arise on non–hair-bearing skin. Although there is general agreement that there is a form of MF that has associated FM, this is where the agreement ends with regard to the diagnosis of MF versus FM. Böer et al⁶ discussed the historic evolution of these terms, mostly to highlight the origins of the confusion. The investigators proposed that FM should only be used as a descriptive term and that all cases of alopecia mucinosa represent MF. They also concluded that many benign dermatoses associated with a risk for evolution to MF (eg, small and large plaque psoriasis [LPP]) should simply be diagnosed as MF.⁶ Subsequently, the proposal that idiopathic FM and LAFM are not 2 distinct entities but rather a clinicopathologic continuum and that idiopathic FM is simply a variant of MF along this spectrum has gained some approval.^6,7 However, this belief is not shared among all authorities in the field, and attempts to define diagnostic criteria that distinguish between a benign clinical course and a course that is more progressive and fatal continue. Currently, it is agreed upon that when distinguishing between these 2 clinical courses, primary (idiopathic) follicular mucinosis refers to a benign course with no overt sign of malignancy, and lymphoma-associated follicular mucinosis refers to a diagnostic malignant condition. Lymphoma-associated follicular mucinosis refers to FM associated with cutaneous T-cell lymphoma, the most common form of which is FM. Many authors^8-15 have sought ancillary methodologies in addition to clinical parameters to assist in the evaluation between both disease courses. Methodologies have included assessment of T-cell receptor gene rearrangements, flow cytometry, and immunohistochemical staining, mostly as an effort to establish monoclonality as a defining characteristic of LAFM; however, monoclonality in cutaneous T-cell infiltrates should be interpreted with caution and should not be considered as a confirmation of malignancy due to recent findings of monoclonality in benign inflammatory dermatoses such as lichen planus. The Table outlines several of the most common benign inflammatory dermatoses that demonstrate monoclonality, but this list should not be considered exhaustive, as there are many others in which monoclonality is sometimes seen.^8-15 The lack of definitive criteria to distinguish between the 2 groups has led to confusion and consternation regarding the diagnosis of idiopathic FM versus LAFM and has led many in the field to consider the 2 conditions to be one and the same.

Diagnosis of FM and MF: Clinicopathologic Features
The World Health Organization (WHO) defined MF as an epidermotropic primary cutaneous T-cell lymphoma (CTCL) characterized by infiltrates of small- to medium-sized T lymphocytes with cerebriform nuclei. Further, the WHO stated that the term mycosis fungoides should be exclusively reserved for classical cases typified by the evolution of cutaneous patches, plaques, and tumors, or for variants that show a similar clinical course.¹⁶ Mycosis fungoides is divided into 3 stages—patch, plaque, and tumor—which are solely clinical descriptors.¹⁷ The WHO also described a clinical staging system with pathologic emphasis placed only on lymph node involvement and identification of Sézary cells.¹⁶ It lists folliculotropic MF as a variant, with only some cases presenting with mucinous degeneration of hair follicles. A lack of consensus among pathologists regarding criteria for diagnosis in patch-stage MF remains, but diagnosis of plaque-stage disease is not regularly debated due to its more reliably present, well-developed histologic features (eg, haloed lymphocytes, epidermotropism of lymphocytes, lymphocytes with convoluted nuceli, Pautrier microabscesses).¹⁸ Although there have been specific histologic findings reported to be associated with patch-stage MF, they have only been present in a few cases and are therefore of limited usefulness in practice.^1,19 The categorization of patients with subtle histologic features common to both MF and inflammatory conditions such as parapsoriasis en plaques (the term plaque in this case is a misnomer because the word plaque means patch in French) continues to be elusive. A lack of agreement regarding LPP persists in the current literature in the same manner as FM. Some researchers have contended for many years that LPP is a type of MF, while others remain unconvinced, mainly due to the lack of evidence that lumping a benign condition (LPP) with an increased risk for malignant transformation and a known malignancy (MF) together is of any benefit to the patient. Assessment of clinicopathologic correlation, immunohistochemistry, clonality, and T-cell gene rearrangement have failed to positively identify patients who are at risk for disease progression, whether the diagnosis is called LPP or early patch-stage MF.²⁰

Mycosis fungoides is more common in males and its incidence increases with age; however, diagnosis should not be ruled out based on age or gender. Typical presentation of early-stage disease includes erythematous patches or plaques, often with light scaling.¹⁹ Lesions routinely are of long-standing duration (months to years), are located in areas that are infrequently exposed to sunlight, and often are 5 cm in diameter or larger with irregular borders.²¹ Associated poikiloderma is relatively specific to MF but rarely is seen in other CTCLs, connective-tissue diseases, and some genodermatoses. Poikiloderma commonly is identified in LPP, which shows the same telangiectasia, mottled pigmentation, and epidermal atrophy as MF-associated poikiloderma, leading some to believe that there is no separation between the 2 conditions. In all stages of MF, lesions frequently are numerous and occur on multiple sites. Plaques and tumors can show spontaneous ulceration. When lesions are folliculotropic, they can cause localized alopecia, follicular-based papules, and fungating pseudotumors in more advanced stages.¹ The clinical presentation of FM substantially overlaps with folliculotropic MF, and although FM lesions often are solitary and are located on the face or scalp, they also can present as multiple lesions located elsewhere on the body. It also has been proposed that folliculotropic MF should not be separated from FM-associated MF (or LAFM).²²

The characteristic histologic picture of LAFM in patch or plaque stage shows mucin deposition within hair follicles, similar to idiopathic FM. On histology, both conditions demonstrate dense lymphoid infiltrates around and within hair follicles as well as in the dermis (Figure). Most cases of LAFM show epidermotropism of lymphocytes between follicles, but this finding is not present in every case and often disappears when the disease advances to the tumor stage.^1,19 Although Pautrier microabscesses (collections of lymphocytes within the superficial epidermis) are considered to be somewhat specific to MF, they are only present in a minority of cases.²⁰ In a study by the International Society for Cutaneous Lymphomas,²¹ the only histopathologic criteria that showed any appreciable sensitivity or specificity in the diagnosis of MF were the presence of lymphoid cells with variable nuclear and cytoplasmic features and/or strikingly irregular nuclear contours with the presence of lymphocytes larger than those usually seen in inflammatory dermatoses. Despite these criteria, the study reported a high misclassification rate. A complicated scoring system for diagnosis of MF in patch- or early plaque-stage disease was proposed by the International Society for Cutaneous Lymphomas,²¹ which integrates clinical, histopathologic, molecular, and immunophenotypic criteria. However, these criteria have been continually debated in the literature and are only discussed in this article in relation to the association between MF and FM. Diagnosis of tumor-stage MF is not addressed in this article, as it is readily identified as lymphoma and is not easily confused with idiopathic FM.

Histopathology of 2 separate cases, both diagnosed as follicular mucinosis pending further follow-up. After close clinical follow-up, one case was diagnosed as mycosis fungoides (A) and the other as follicular mucinosis (B)(both H&E, original magnification ×2). Immunohistochemical stains were noncontributory, and T-cell gene rearrangement was positive only in the case of mycosis fungoides. Histopathology revealed mucin deposition in the hair follicles and a dense lymphoid infiltrate around and within the follicles in both cases. Pautrier microabscesses, cerebriform nuclei, and epidermotropism of lymphocytes between follicles were absent in the case of mycosis fungoides.

Clinical assessment of a patient’s medical history to identify persistent and progressive disease is paramount to the diagnosis of MF. Although MF lesions tend to increase in size and number over time, this presentation is not without exception.²¹ In early patch-stage disease, eliminating some of the patient’s current medications may be sufficient in clearing cutaneous patches that cannot be conclusively identified as either MF or a benign inflammatory lymphoid infiltrate, which further emphasizes the importance of clinical assessment of the patient’s medical history in the diagnosis of MF. The shape of the lesions also is helpful in distinguishing between MF and other skin disorders, such as digitate dermatosis or LPP; unlike the latter, the waxing and waning nature of MF lesions often produces irregularly shaped patches with little coalescence. Again, there are some investigators who believe that these lesions represent varying presentations of MF.⁶

In a study by Cerroni et al,¹ 44 patients with FM were divided into 2 groups: (1) a cohort of 16 patients with no history or clinical evidence of MF or Sézary syndrome (ie, LAFM), and (2) a cohort of 28 patients with clinicopathologic evidence of CTCL. Patients in both groups were followed for a maximum of 20 years. Results indicated that that the presence of perifollicular or intrafollicular mucin, epidermotropism of lymphocytes, monoclonality, and epidemiologic characteristics (eg, age, sex, race) cannot reliably distinguish the 2 disease forms. Furthermore, it was suggested that these conditions are not mutually exclusive entities and are actually variants of CTCL. The observation that the 2 diseases share prognostic overlap adds further credence to the already puzzling conundrum. Nineteen of 28 patients with MF were alive and well at follow-up, and all patients in the idiopathic FM group were alive, with only 9 of 16 patients showing residual disease and none with CTCL.¹

Other clinical factors that may be helpful in the diagnosis of MF are the presentation of lesions in non–sun-exposed areas of the skin and multiple lesions, as unilesional MF is exceedingly uncommon.²¹ No histologic features have been proven to predict which early patch- or plaque-stage MFs will progress to full-blown CTCL versus benign idiopathic FM; thus, great caution should be taken in patients with early-stage disease to ensure they are not prematurely and/or incorrectly classified as CTCL. Such a diagnosis has medical, social, and economical ramifications that should not be overlooked.

If idiopathic FM and LAFM were considered distinct disease processes, the ambiguity in making a definitive diagnosis should give the physician pause, and a diagnosis of LAFM may only be appropriate when there is unequivocal clinicopathologic evidence. Otherwise, a lymphoma diagnosis is somewhat superfluous and potentially harmful. Definitive diagnosis of LAFM also is complicated by reports of other hematologic malignancies presenting with FM-like histopathologic findings, such as chronic myelogenous leukemia, leukemia-associated eosinophilic folliculitis, and acute myeloblastic leukemia.^23,24 Although MF is the most common malignancy associated with FM, it is important to consider other less common malignancies that also may be present.

Diagnosis: Patient Consequences
Accurate diagnosis of idiopathic FM versus LAFM is critical, as the ramifications of a cancer diagnosis can have broad implications. For example, patients who receive cancer diagnoses often experience emotional trauma and social stigma, even when adequate patient education has been provided. The incidence of depression and anxiety also can increase following a cancer diagnosis and can be complicated by medical treatments (eg, systemic steroids, interferon),²⁵ which are known to increase the frequency of these psychological disturbances. Health insurance premiums likely will be higher if a patient is diagnosed with cancer versus a benign inflammatory condition. Hesitation of the pathologist to assign a cancer diagnosis when unequivocal evidence is not present should not be regarded as trickery, malpractice, or deceit of the health care bylaws, as benign language with suggestion of close clinical follow-up in the setting of diagnostic uncertainty will “first, do no harm” and secondly, serve as a vehicle for patient advocacy.

If there is a definitive distinction between idiopathic FM and LAFM, it requires further research before it can be fully understood. Currently, the WHO does not recognize a diagnosis of FM-associated MF (or LAFM) and acknowledges that folliculotropic MF is not always associated with FM.^16,26 Given uncertainty and repercussions associated with a cancer diagnosis, however indolent, it may be morally responsible and medically favorable for physicians to consider FM in the differential diagnosis when applicable rather than making a diagnosis of MF outright. Given the importance of both clinical and histologic factors, it may be beneficial for definitive diagnosis of FM versus MF to lie with the clinician, while the pathologist serves as an adjunct in the diagnostic process. Because the prognosis of idiopathic FM often is marred by possible transformation into MF or other CTCLs, therapeutic decisions should be dictated by close clinical follow-up. Additionally, stage of disease, patient age, treatment compliance, comorbidities, and possible side effects of medications should all be considered when evaluating potential therapeutic regimens.²⁷

Conclusion
Research is underway to more accurately identify patients with FM who are at risk for progression to LAFM versus those with benign remitting FM. Once the required diagnostic criteria are established to accurately classify these patients, with an emphasis on prognosis and suggested treatments, it might be necessary to establish new, less debated terminology so pathologists and clinicians alike can improve patient care. Continued histopathologic scrutiny, use of sophisticated molecular techniques, and knowledge of other currently undiscovered modalities will shed light on this important disease process and aid in proper disease management, which may be advantageous to both patients and physicians.

When follicular mucinosis (FM) is defined as an epithelial reaction pattern characterized by intrafollicular and perifollicular mucin accumulation, it cannot be considered a distinct disease entity, as this pattern is ubiquitously present in various inflammatory and neoplastic skin conditions.^1,2 The distinction between idiopathic FM and lymphoma-associated follicular mucinosis (LAFM) was made several years ago by authors who evaluated the differences in the clinical presentation of these entities, including patient age at onset, number of lesions, pattern of distribution, and most importantly clinical progression.¹ In this article, we discuss the importance of close clinical follow-up in patients with FM or patch-stage mycosis fungoides (MF) in whom histopathologic evaluation is ambiguous or nondiagnostic. We also highlight the value of ancillary testing, including T-cell receptor gene rearrangement, flow cytometry, and immunohistochemistry, as a component in the diagnostic process rather than the sole diagnostic moiety. A review of the pertinent literature also is performed.

History of FM and MF
Pinkus³ first described an entity he termed alopecia mucinosa in 1957. Pinkus noted 3 distinct patterns: an idiopathic form of alopecia mucinosa, lymphoblastoma with associated FM, and alopecia mucinosa that later transformed into lymphoblastoma.⁴ In 1983, however, Pinkus⁴ described uncertainty if alopecia mucinosa represented the first stage of MF or if patients with alopecia mucinosa were simply at an increased risk for developing lymphoma. He believed there were too many cases of lymphoma following a diagnosis of alopecia mucinosa for the relation to be coincidental, yet he noted that many of the cases resolved either spontaneously or following treatment with x-rays or topical steroids. He concluded his report with a sentiment that is echoed in many current studies regarding this entity: “Many questions surrounding this entity are as unanswerable today as they were 25 years ago.”⁴

Jablonska et al⁵ were the first to coin the term mucinosis follicularis, now known as FM, to replace alopecia mucinosa because they felt the description was more accurate, as lesions also arise on non–hair-bearing skin. Although there is general agreement that there is a form of MF that has associated FM, this is where the agreement ends with regard to the diagnosis of MF versus FM. Böer et al⁶ discussed the historic evolution of these terms, mostly to highlight the origins of the confusion. The investigators proposed that FM should only be used as a descriptive term and that all cases of alopecia mucinosa represent MF. They also concluded that many benign dermatoses associated with a risk for evolution to MF (eg, small and large plaque psoriasis [LPP]) should simply be diagnosed as MF.⁶ Subsequently, the proposal that idiopathic FM and LAFM are not 2 distinct entities but rather a clinicopathologic continuum and that idiopathic FM is simply a variant of MF along this spectrum has gained some approval.^6,7 However, this belief is not shared among all authorities in the field, and attempts to define diagnostic criteria that distinguish between a benign clinical course and a course that is more progressive and fatal continue. Currently, it is agreed upon that when distinguishing between these 2 clinical courses, primary (idiopathic) follicular mucinosis refers to a benign course with no overt sign of malignancy, and lymphoma-associated follicular mucinosis refers to a diagnostic malignant condition. Lymphoma-associated follicular mucinosis refers to FM associated with cutaneous T-cell lymphoma, the most common form of which is FM. Many authors^8-15 have sought ancillary methodologies in addition to clinical parameters to assist in the evaluation between both disease courses. Methodologies have included assessment of T-cell receptor gene rearrangements, flow cytometry, and immunohistochemical staining, mostly as an effort to establish monoclonality as a defining characteristic of LAFM; however, monoclonality in cutaneous T-cell infiltrates should be interpreted with caution and should not be considered as a confirmation of malignancy due to recent findings of monoclonality in benign inflammatory dermatoses such as lichen planus. The Table outlines several of the most common benign inflammatory dermatoses that demonstrate monoclonality, but this list should not be considered exhaustive, as there are many others in which monoclonality is sometimes seen.^8-15 The lack of definitive criteria to distinguish between the 2 groups has led to confusion and consternation regarding the diagnosis of idiopathic FM versus LAFM and has led many in the field to consider the 2 conditions to be one and the same.

Diagnosis of FM and MF: Clinicopathologic Features
The World Health Organization (WHO) defined MF as an epidermotropic primary cutaneous T-cell lymphoma (CTCL) characterized by infiltrates of small- to medium-sized T lymphocytes with cerebriform nuclei. Further, the WHO stated that the term mycosis fungoides should be exclusively reserved for classical cases typified by the evolution of cutaneous patches, plaques, and tumors, or for variants that show a similar clinical course.¹⁶ Mycosis fungoides is divided into 3 stages—patch, plaque, and tumor—which are solely clinical descriptors.¹⁷ The WHO also described a clinical staging system with pathologic emphasis placed only on lymph node involvement and identification of Sézary cells.¹⁶ It lists folliculotropic MF as a variant, with only some cases presenting with mucinous degeneration of hair follicles. A lack of consensus among pathologists regarding criteria for diagnosis in patch-stage MF remains, but diagnosis of plaque-stage disease is not regularly debated due to its more reliably present, well-developed histologic features (eg, haloed lymphocytes, epidermotropism of lymphocytes, lymphocytes with convoluted nuceli, Pautrier microabscesses).¹⁸ Although there have been specific histologic findings reported to be associated with patch-stage MF, they have only been present in a few cases and are therefore of limited usefulness in practice.^1,19 The categorization of patients with subtle histologic features common to both MF and inflammatory conditions such as parapsoriasis en plaques (the term plaque in this case is a misnomer because the word plaque means patch in French) continues to be elusive. A lack of agreement regarding LPP persists in the current literature in the same manner as FM. Some researchers have contended for many years that LPP is a type of MF, while others remain unconvinced, mainly due to the lack of evidence that lumping a benign condition (LPP) with an increased risk for malignant transformation and a known malignancy (MF) together is of any benefit to the patient. Assessment of clinicopathologic correlation, immunohistochemistry, clonality, and T-cell gene rearrangement have failed to positively identify patients who are at risk for disease progression, whether the diagnosis is called LPP or early patch-stage MF.²⁰

Mycosis fungoides is more common in males and its incidence increases with age; however, diagnosis should not be ruled out based on age or gender. Typical presentation of early-stage disease includes erythematous patches or plaques, often with light scaling.¹⁹ Lesions routinely are of long-standing duration (months to years), are located in areas that are infrequently exposed to sunlight, and often are 5 cm in diameter or larger with irregular borders.²¹ Associated poikiloderma is relatively specific to MF but rarely is seen in other CTCLs, connective-tissue diseases, and some genodermatoses. Poikiloderma commonly is identified in LPP, which shows the same telangiectasia, mottled pigmentation, and epidermal atrophy as MF-associated poikiloderma, leading some to believe that there is no separation between the 2 conditions. In all stages of MF, lesions frequently are numerous and occur on multiple sites. Plaques and tumors can show spontaneous ulceration. When lesions are folliculotropic, they can cause localized alopecia, follicular-based papules, and fungating pseudotumors in more advanced stages.¹ The clinical presentation of FM substantially overlaps with folliculotropic MF, and although FM lesions often are solitary and are located on the face or scalp, they also can present as multiple lesions located elsewhere on the body. It also has been proposed that folliculotropic MF should not be separated from FM-associated MF (or LAFM).²²

The characteristic histologic picture of LAFM in patch or plaque stage shows mucin deposition within hair follicles, similar to idiopathic FM. On histology, both conditions demonstrate dense lymphoid infiltrates around and within hair follicles as well as in the dermis (Figure). Most cases of LAFM show epidermotropism of lymphocytes between follicles, but this finding is not present in every case and often disappears when the disease advances to the tumor stage.^1,19 Although Pautrier microabscesses (collections of lymphocytes within the superficial epidermis) are considered to be somewhat specific to MF, they are only present in a minority of cases.²⁰ In a study by the International Society for Cutaneous Lymphomas,²¹ the only histopathologic criteria that showed any appreciable sensitivity or specificity in the diagnosis of MF were the presence of lymphoid cells with variable nuclear and cytoplasmic features and/or strikingly irregular nuclear contours with the presence of lymphocytes larger than those usually seen in inflammatory dermatoses. Despite these criteria, the study reported a high misclassification rate. A complicated scoring system for diagnosis of MF in patch- or early plaque-stage disease was proposed by the International Society for Cutaneous Lymphomas,²¹ which integrates clinical, histopathologic, molecular, and immunophenotypic criteria. However, these criteria have been continually debated in the literature and are only discussed in this article in relation to the association between MF and FM. Diagnosis of tumor-stage MF is not addressed in this article, as it is readily identified as lymphoma and is not easily confused with idiopathic FM.

Histopathology of 2 separate cases, both diagnosed as follicular mucinosis pending further follow-up. After close clinical follow-up, one case was diagnosed as mycosis fungoides (A) and the other as follicular mucinosis (B)(both H&E, original magnification ×2). Immunohistochemical stains were noncontributory, and T-cell gene rearrangement was positive only in the case of mycosis fungoides. Histopathology revealed mucin deposition in the hair follicles and a dense lymphoid infiltrate around and within the follicles in both cases. Pautrier microabscesses, cerebriform nuclei, and epidermotropism of lymphocytes between follicles were absent in the case of mycosis fungoides.

Clinical assessment of a patient’s medical history to identify persistent and progressive disease is paramount to the diagnosis of MF. Although MF lesions tend to increase in size and number over time, this presentation is not without exception.²¹ In early patch-stage disease, eliminating some of the patient’s current medications may be sufficient in clearing cutaneous patches that cannot be conclusively identified as either MF or a benign inflammatory lymphoid infiltrate, which further emphasizes the importance of clinical assessment of the patient’s medical history in the diagnosis of MF. The shape of the lesions also is helpful in distinguishing between MF and other skin disorders, such as digitate dermatosis or LPP; unlike the latter, the waxing and waning nature of MF lesions often produces irregularly shaped patches with little coalescence. Again, there are some investigators who believe that these lesions represent varying presentations of MF.⁶

In a study by Cerroni et al,¹ 44 patients with FM were divided into 2 groups: (1) a cohort of 16 patients with no history or clinical evidence of MF or Sézary syndrome (ie, LAFM), and (2) a cohort of 28 patients with clinicopathologic evidence of CTCL. Patients in both groups were followed for a maximum of 20 years. Results indicated that that the presence of perifollicular or intrafollicular mucin, epidermotropism of lymphocytes, monoclonality, and epidemiologic characteristics (eg, age, sex, race) cannot reliably distinguish the 2 disease forms. Furthermore, it was suggested that these conditions are not mutually exclusive entities and are actually variants of CTCL. The observation that the 2 diseases share prognostic overlap adds further credence to the already puzzling conundrum. Nineteen of 28 patients with MF were alive and well at follow-up, and all patients in the idiopathic FM group were alive, with only 9 of 16 patients showing residual disease and none with CTCL.¹

Other clinical factors that may be helpful in the diagnosis of MF are the presentation of lesions in non–sun-exposed areas of the skin and multiple lesions, as unilesional MF is exceedingly uncommon.²¹ No histologic features have been proven to predict which early patch- or plaque-stage MFs will progress to full-blown CTCL versus benign idiopathic FM; thus, great caution should be taken in patients with early-stage disease to ensure they are not prematurely and/or incorrectly classified as CTCL. Such a diagnosis has medical, social, and economical ramifications that should not be overlooked.

If idiopathic FM and LAFM were considered distinct disease processes, the ambiguity in making a definitive diagnosis should give the physician pause, and a diagnosis of LAFM may only be appropriate when there is unequivocal clinicopathologic evidence. Otherwise, a lymphoma diagnosis is somewhat superfluous and potentially harmful. Definitive diagnosis of LAFM also is complicated by reports of other hematologic malignancies presenting with FM-like histopathologic findings, such as chronic myelogenous leukemia, leukemia-associated eosinophilic folliculitis, and acute myeloblastic leukemia.^23,24 Although MF is the most common malignancy associated with FM, it is important to consider other less common malignancies that also may be present.

Diagnosis: Patient Consequences
Accurate diagnosis of idiopathic FM versus LAFM is critical, as the ramifications of a cancer diagnosis can have broad implications. For example, patients who receive cancer diagnoses often experience emotional trauma and social stigma, even when adequate patient education has been provided. The incidence of depression and anxiety also can increase following a cancer diagnosis and can be complicated by medical treatments (eg, systemic steroids, interferon),²⁵ which are known to increase the frequency of these psychological disturbances. Health insurance premiums likely will be higher if a patient is diagnosed with cancer versus a benign inflammatory condition. Hesitation of the pathologist to assign a cancer diagnosis when unequivocal evidence is not present should not be regarded as trickery, malpractice, or deceit of the health care bylaws, as benign language with suggestion of close clinical follow-up in the setting of diagnostic uncertainty will “first, do no harm” and secondly, serve as a vehicle for patient advocacy.

If there is a definitive distinction between idiopathic FM and LAFM, it requires further research before it can be fully understood. Currently, the WHO does not recognize a diagnosis of FM-associated MF (or LAFM) and acknowledges that folliculotropic MF is not always associated with FM.^16,26 Given uncertainty and repercussions associated with a cancer diagnosis, however indolent, it may be morally responsible and medically favorable for physicians to consider FM in the differential diagnosis when applicable rather than making a diagnosis of MF outright. Given the importance of both clinical and histologic factors, it may be beneficial for definitive diagnosis of FM versus MF to lie with the clinician, while the pathologist serves as an adjunct in the diagnostic process. Because the prognosis of idiopathic FM often is marred by possible transformation into MF or other CTCLs, therapeutic decisions should be dictated by close clinical follow-up. Additionally, stage of disease, patient age, treatment compliance, comorbidities, and possible side effects of medications should all be considered when evaluating potential therapeutic regimens.²⁷

Conclusion
Research is underway to more accurately identify patients with FM who are at risk for progression to LAFM versus those with benign remitting FM. Once the required diagnostic criteria are established to accurately classify these patients, with an emphasis on prognosis and suggested treatments, it might be necessary to establish new, less debated terminology so pathologists and clinicians alike can improve patient care. Continued histopathologic scrutiny, use of sophisticated molecular techniques, and knowledge of other currently undiscovered modalities will shed light on this important disease process and aid in proper disease management, which may be advantageous to both patients and physicians.

Primary apocrine adenocarcinoma (AA) is a rare cutaneous malignancy, with most of the available information about this disease consolidated from anecdotal evidence of single case reports and small case series with fewer than 30 patients.^1-11 Although certain histologic and immunohistochemical features have been suggested to be useful in the diagnosis of AA, there is no clear consensus on the required pathologic criteria.^{1,5,6,9,10,12,13} Additionally, the clinical presentation of AA is highly variable, which further adds to the challenge of making an accurate diagnosis.^{1-3,5,9,10,13}

Apocrine adenocarcinoma usually arises in areas of high apocrine gland density such as the axillae or anogenital region.^2,4,6 It also has been reported in areas such as the scalp, ear canal, eyelids, chest, nipples, arms, wrists, and fingers.^4,8,10,14-16 Apocrine adenocarcinoma in unusual locations such as the eyelid and ear canal is thought to arise from modified apocrine glands such as the Moll glands of the eyelid and the ceruminous glands of the ear canal.^9,10 The presence of ectopic apocrine glands may lead to AA in atypical sites such as the wrists and fingers.^5,16 The areola is an apocrine-dense area; therefore, AA may present on the nipples or within supernumerary nipples anywhere along the milk lines.⁴

Apocrine adenocarcinoma clinically presents as an asymptomatic to slightly painful, slowly growing, and erythematous to violaceous nodule or tumor.^4,6,9 However, in a minority of cases the initial presentation consists of a cystic or ulcerated mass with overlying granulation tissue and purulent discharge.^6,9,11 A wide time frame from the onset of symptoms to diagnosis has been reported, ranging from weeks to decades.^4,6-8 The conventional treatment of AA is wide local excision.^2,4,6,9 Although AA often presents with local lymph node metastasis at the time of diagnosis, there is no consensus on the use of sentinel lymph node biopsy (SLNB), nodal dissection, or adjuvant chemoradiation therapy.^1,3,8,9

We report the case of a 49-year-old man with primary AA of the left axilla; the clinical and histologic features of AA as well as the appropriate diagnostic and treatment modalities also are provided.

Case Report

A 49-year-old man with a slowly growing tender mass of the left axilla of 1 year’s duration was referred to our dermatology clinic for evaluation. A review of systems revealed loss of appetite, fatigue, and a 4-month history of unintentional weight loss (15–20 lb). The patient had a history of hepatitis C virus, intravenous drug use, alcohol abuse, and cigarette smoking (1 pack daily) for many years. Additionally, the patient reported a paternal family history of numerous visceral malignancies. Examination of the left axilla revealed a 1.5×5-cm ulcerated tumor that produced serosanguineous discharge and was tender to palpation (Figure 1). Two 1-cm, firm, freely mobile subcutaneous nodules with no overlying skin changes were palpable at the medial border of the ulcerated nodule. There was no additional cervical or axillary lymphadenopathy, and a breast examination was normal.

Figure 1. A 1.5×5-cm ulcerated tumor on the left axilla with firm palpable lymph nodes on the medial border.

The differential diagnosis included primary squamous cell carcinoma or adnexal neoplasm, primary breast carcinoma, lymphoma, scrofuloderma, atypical mycobacterial infection, and cutaneous metastasis from an internal malignancy. Two 4-mm punch biopsies were performed and sent for routine histopathology and bacterial, fungal, and mycobacterial tissue cultures. To exclude a primary visceral malignancy or metastasis, computed tomography of the chest, abdomen, and pelvis; positron emission tomography (PET) from the base of the skull to the thighs; colonoscopy; magnetic resonance imaging of the brain; esophagogastroduodenoscopy; and mammography were conducted. Prominent left axillary lymphadenopathy was noted on computed tomography. Additionally, PET identified extranodal spread in the left axilla, left lateral chest wall, and the left sternocleidomastoid region. Furthermore, a 1-cm hypermetabolic nodule involving the right rectus abdominus muscle was noted on the PET scan. Based on their appearance, the nodules most likely represented metastasis from a primary skin malignancy. The rest of the studies were unremarkable. Serum tumor markers including prostate-specific antigen, cancer antigen 19-9, and carcinoembryonic antigen were within reference range. Immunostaining for estrogen receptor, progesterone receptor, and ERBB2 (formerly HER2/neu) was negative. The only abnormalities noted on serum chemistries were slight elevations in aspartate aminotransferase, alanine aminotransferase, alkaline phosphatase, and the a-fetoprotein tumor marker, which was attributed to chronic hepatitis C infection. Bacterial, fungal, and mycobacterial tissue cultures also were negative. These results ruled out infection and suggested against a primary visceral malignancy with cutaneous metastasis.

Histopathology revealed a moderately differentiated adenocarcinoma adjacent to healthy-appearing apocrine glands (Figure 2A). The normal glands were composed of cuboidal cells with abundant eosinophilic cytoplasm and prominent nuclei. The cells were arranged in a single layer in a glandular formation with prominent decapitation secretion. Adjacent to the normal apocrine glandular tissue was a focus of malignant epithelioid cells that extended to the lateral and inferior margins. The neoplastic cells were cuboidal to angulated in appearance with prominent nuclei and seemed to form ill-defined tubular or glandular structures that partially resembled apocrine glands (Figure 2B). Decapitation secretion is a feature of apocrine differentiation. Examination of additional tissue sections of the tumor did not reveal remarkable decapitation secretion in contrast to the adjacent healthy apocrine glands. Rather, a solid sheet arrangement was primarily noted in several sections (Figure 2B). Neither frequent mitoses nor prominent cellular atypia were seen, and there was no evidence of lymphatic, perineural, or vascular invasion.

Figure 2. A lesional punch biopsy revealed normal apocrine glands adjacent to poorly differentiated glandular and cordlike structures (A)(H&E, original magnification ×10). The neoplastic cells were cuboidal to angulated in appearance with prominent nuclei and abundant cytoplasm demonstrating poorly differentiated apocrine glands (B)(H&E, original magnification ×40).

Immunohistochemically, tumor cells reacted strongly to cytokeratin AE1/AE3 and CAM5.², stains used to identify various cytokeratins present in epithelial tissue. Staining for epithelial membrane antigen and carcinoembryonic antigen revealed focal glandular differentiation, which further supported the epithelial origin of the neoplastic cells. Gross cystic disease fluid protein 15 (GCDFP-15) is a marker of apocrine differentiation and may indicate a carcinoma of apocrine or eccrine origin. In our case, staining for GCDFP-15 was negative in the cutaneous sections but highlighted tumor cells in 6 of 13 ipsilateral lymph nodes from locoregional metastasis. The cellular and structural morphology, immunohistochemistry, and absence of an alternative primary visceral malignancy supported the diagnosis of primary AA.

Initially the patient was not considered to be a candidate for surgery due to the rapid growth of the tumor with metastases, fatigue, weight loss, and pain. Therefore, radiation therapy was started. The patient responded well to treatment with controlled pain and resolution of the palpable mass of the left axilla. Moreover, a follow-up PET scan revealed no residual tumor and persistent, albeit decreased, axillary lymphadenopathy. As the patient’s clinical status had improved, excision of the left axillary tumor with lymph node dissection was performed 10 months after initial presentation.

In this case, the differential diagnosis consisted of various cutaneous neoplasms, primary mammary carcinoma, cutaneous metastasis, and infection. Diagnostic imaging and laboratory testing failed to identify any primary internal malignancies. Similarly, the negative cultures ruled out an infectious process. Furthermore, the axillary mass was noted to be separate from the breast tissue on physical examination and mammography. Histologically, the tumor showed features that were suggestive of an anaplastic process as well as decapitation secretion and glandular formation that clearly resembled apocrine differentiation.

Comment

Apocrine adenocarcinoma arises from apocrine sweat glands and therefore is mostly reported in areas of high apocrine gland density such as the axillae and the anogenital region.^2,4,6 However, AA also has been reported in unusual locations,^1,5,10,14-16 and they may arise from a pre-existing nevus sebaceous or from supernumerary nipples, which can occur anywhere along the milk lines.^4,15 Apocrine adenocarcinoma most commonly arises in individuals aged 40 to 50 years.^3,17 A slight male predominance has been reported but no racial predilection.^1,4-6 Although a few reports have described the development of AAs within pre-existing benign tumors such as apocrine adenomas, apocrine hyperplasias, cylindromas, and nevi sebaceous, they usually are thought to arise de novo.^4-6

Clinical Presentation

Apocrine adenocarcinoma is highly variable in its clinical manifestation.^1,6 Most cases arise as erythematous to violaceous, firm, solitary nodules. Nonetheless, AA also can present as erythematous patches of skin resembling erysipelas and ulcerated nodules with overlying granulation tissue and purulent exudate.^4,6,9,11 Although AA typically is slow growing and indolent, the time frame reported from onset to diagnosis ranges from weeks to decades.^1,6,7 Most cases present asymptomatically; when symptoms do occur, the most common ones are tenderness, purulent discharge, and restricted range of motion from extremely large tumors.^3,9 Incidence of lymph node metastasis is reported at 40% to 50% at the time of presentation.^4,6 Additionally, AA has a high rate of local recurrence, but extranodal metastasis rarely is seen.^2,6 When metastasis does occur, it is via lymphatic and hematogenous spread.^6,9 Metastatic dissemination of AA may occur in the liver, lungs, bone, brain, and parotid glands, as well as the skin via intraepidermal pagetoid spread.^4,6,9,13

Histopathology

The histologic characteristics essential to the diagnosis of primary AA are anaplastic differentiation and apocrine origin.^1,2,9,10,17 Apocrine units include coiled secretory glands that reside in the deep dermis connecting to a straight duct that empties into the isthmus of the hair follicle.^9,13 These secretory glands have a single row of cuboidal secretory cells lining the tubular component and stratified squamous epithelium lining the straight intradermal component that opens onto the hair follicle.⁹ Contractile myoepithelial cells surround the secretory cell layer of the gland.^9,13

The cuboidal secretory cells of the apocrine gland have abundant eosinophilic cytoplasm^1,4,9 and are further characterized by glandular arrangement and decapitation secretion, 2 features that are strongly suggestive of apocrine differentiation.^4-6 In contrast, the tumor cells of AA can be characterized by hyperchromatic nuclei, nuclear pleomorphism, mitotic figures, and a lack of decapitation secretion.^1,2,6 In malignancy, erratic or poorly differentiated ductal structures may be seen,^1,3-6 including papillary, cordlike, solid, or complex glandular patterns that can potentially invade the adjacent tissue without a clearly recognizable myoepithelial layer that contains them.^1,3,4,6 Moreover, AA may progress with lymphatic, vascular, or neural invasion.^1,13

Various stains may be used in immunohistochemical analysis to aid in the diagnosis of AA.^1,5 Cytokeratin AE1/AE3, CAM5.2, epithelial membrane antigen, smooth muscle antigen, periodic acid–Schiff positivity with diastase resistance, and GCDFP-15 are useful in supporting the diagnosis of AA.^2,6,10,17 Cytokeratin AE1/AE3 and CAM5.2 stain various cytokeratins to confirm the epithelial origin of the tissue.² Epithelial membrane antigen is an antigen present on the apical surface of glandular epithelial cells that also has been used to identify epithelial cells in AA.² Additionally, smooth muscle actin may be used to detect the myoepithelial layer of cells surrounding the apocrine glands.¹⁷ The lack of a continuous layer surrounding the secretory cells suggests invasion into the adjacent tissue.^1,9,17 Periodic acid–Schiff staining with diastase resistance can be used to identify the mucin stored in the intracytoplasmic granules of apocrine cells and the lumen.³ Some stains such as GCDFP-15 may highlight cells of multiple origins (eg, apocrine and eccrine).¹⁰ However, there is the possibility that poorly differentiated AAs would fail to be identified as such even with well-established apocrine markers, which may explain the differential GCDFP-15 staining patterns in our patient’s skin and lymph node sections.^1,5 Therefore, there is not a single perfect set of immunohistological criteria to aid in the diagnosis of AA.^6,10,12 Fundamentally, diagnosis requires detection of primary apocrine differentiation with features such as invasion or spread to adjacent tissue to suggest malignancy and rule out an alternate primary malignant process.^1,2,9,10,17

Treatment and Prognosis

Primary treatment of AA consists of wide local excision with adjuvant options that include chemotherapy and radiation.^2,6 Due to the high rate of lymph node metastases at presentation (40%–50%), SLNB is recommended. A positive SLNB should be followed with complete axillary lymphadenectomy^4,6; however, there is a lack of consensus regarding the role of SLNB and lymph node dissection in detecting subclinical lymph node disease, which might improve local recurrence rate and prognosis.⁶ Similarly, research shows variable results with adjunctive treatment such as chemotherapy or radiation therapy.^6,9,13 Adjuvant treatment with chemotherapy or radiation therapy should be considered in cases with large tumor size; perineural, lymphatic, or vascular invasion; or when complete removal of the tumor is not possible due to location or size.^2,6 However, neither the role nor the efficacy of such treatments in AA is well established.^6,9,13

There is little information in the literature regarding the prognosis of AA. Although no specific or well-documented prognostic criteria exist, it is generally believed that patients with well-differentiated AA will have higher cure rates or lower rates of local recurrence and lymph node metastasis than patients with poorly differentiated neoplasms.^3,6,10 A few small case series with long-term follow-up of patients ranging from 2 to 10 years have shown that prognosis may be favorable for AA patients despite local recurrence and regional lymph node metastasis.^1,5

Conclusion

Primary AA is a rare cutaneous neoplasm that most commonly occurs in the axillae and the anogenital region. Apocrine adenocarcinoma presents with highly variable clinical and histopathological findings that make diagnosis a challenge. Clinicians should keep this entity in their differential diagnosis for patients who present with nodules arising in apocrine gland–bearing skin. Ultimately, histopathology is critical to diagnosis, and special stains are often required. To make the diagnosis, a tissue biopsy demonstrating apocrine differentiation and anaplastic features to suggest a malignant process are required. Additionally, a careful workup to rule out other diagnoses should be performed. Testing modalities that detect the presence of useful markers such as apocrine or epithelial origin should be used, and the presence of positive findings should support the diagnosis of AA. However, immunohistochemical findings should be used in the context of the patient’s clinical presentation and other available data. Treatment includes wide local excision, and lymphadenectomy is recommended in the setting of nodal spread. For aggressive tumors or metastases, excision may be followed by radiation therapy and chemotherapy.

Primary apocrine adenocarcinoma (AA) is a rare cutaneous malignancy, with most of the available information about this disease consolidated from anecdotal evidence of single case reports and small case series with fewer than 30 patients.^1-11 Although certain histologic and immunohistochemical features have been suggested to be useful in the diagnosis of AA, there is no clear consensus on the required pathologic criteria.^{1,5,6,9,10,12,13} Additionally, the clinical presentation of AA is highly variable, which further adds to the challenge of making an accurate diagnosis.^{1-3,5,9,10,13}

Apocrine adenocarcinoma usually arises in areas of high apocrine gland density such as the axillae or anogenital region.^2,4,6 It also has been reported in areas such as the scalp, ear canal, eyelids, chest, nipples, arms, wrists, and fingers.^4,8,10,14-16 Apocrine adenocarcinoma in unusual locations such as the eyelid and ear canal is thought to arise from modified apocrine glands such as the Moll glands of the eyelid and the ceruminous glands of the ear canal.^9,10 The presence of ectopic apocrine glands may lead to AA in atypical sites such as the wrists and fingers.^5,16 The areola is an apocrine-dense area; therefore, AA may present on the nipples or within supernumerary nipples anywhere along the milk lines.⁴

Apocrine adenocarcinoma clinically presents as an asymptomatic to slightly painful, slowly growing, and erythematous to violaceous nodule or tumor.^4,6,9 However, in a minority of cases the initial presentation consists of a cystic or ulcerated mass with overlying granulation tissue and purulent discharge.^6,9,11 A wide time frame from the onset of symptoms to diagnosis has been reported, ranging from weeks to decades.^4,6-8 The conventional treatment of AA is wide local excision.^2,4,6,9 Although AA often presents with local lymph node metastasis at the time of diagnosis, there is no consensus on the use of sentinel lymph node biopsy (SLNB), nodal dissection, or adjuvant chemoradiation therapy.^1,3,8,9

We report the case of a 49-year-old man with primary AA of the left axilla; the clinical and histologic features of AA as well as the appropriate diagnostic and treatment modalities also are provided.

Case Report

A 49-year-old man with a slowly growing tender mass of the left axilla of 1 year’s duration was referred to our dermatology clinic for evaluation. A review of systems revealed loss of appetite, fatigue, and a 4-month history of unintentional weight loss (15–20 lb). The patient had a history of hepatitis C virus, intravenous drug use, alcohol abuse, and cigarette smoking (1 pack daily) for many years. Additionally, the patient reported a paternal family history of numerous visceral malignancies. Examination of the left axilla revealed a 1.5×5-cm ulcerated tumor that produced serosanguineous discharge and was tender to palpation (Figure 1). Two 1-cm, firm, freely mobile subcutaneous nodules with no overlying skin changes were palpable at the medial border of the ulcerated nodule. There was no additional cervical or axillary lymphadenopathy, and a breast examination was normal.

Figure 1. A 1.5×5-cm ulcerated tumor on the left axilla with firm palpable lymph nodes on the medial border.

The differential diagnosis included primary squamous cell carcinoma or adnexal neoplasm, primary breast carcinoma, lymphoma, scrofuloderma, atypical mycobacterial infection, and cutaneous metastasis from an internal malignancy. Two 4-mm punch biopsies were performed and sent for routine histopathology and bacterial, fungal, and mycobacterial tissue cultures. To exclude a primary visceral malignancy or metastasis, computed tomography of the chest, abdomen, and pelvis; positron emission tomography (PET) from the base of the skull to the thighs; colonoscopy; magnetic resonance imaging of the brain; esophagogastroduodenoscopy; and mammography were conducted. Prominent left axillary lymphadenopathy was noted on computed tomography. Additionally, PET identified extranodal spread in the left axilla, left lateral chest wall, and the left sternocleidomastoid region. Furthermore, a 1-cm hypermetabolic nodule involving the right rectus abdominus muscle was noted on the PET scan. Based on their appearance, the nodules most likely represented metastasis from a primary skin malignancy. The rest of the studies were unremarkable. Serum tumor markers including prostate-specific antigen, cancer antigen 19-9, and carcinoembryonic antigen were within reference range. Immunostaining for estrogen receptor, progesterone receptor, and ERBB2 (formerly HER2/neu) was negative. The only abnormalities noted on serum chemistries were slight elevations in aspartate aminotransferase, alanine aminotransferase, alkaline phosphatase, and the a-fetoprotein tumor marker, which was attributed to chronic hepatitis C infection. Bacterial, fungal, and mycobacterial tissue cultures also were negative. These results ruled out infection and suggested against a primary visceral malignancy with cutaneous metastasis.

Histopathology revealed a moderately differentiated adenocarcinoma adjacent to healthy-appearing apocrine glands (Figure 2A). The normal glands were composed of cuboidal cells with abundant eosinophilic cytoplasm and prominent nuclei. The cells were arranged in a single layer in a glandular formation with prominent decapitation secretion. Adjacent to the normal apocrine glandular tissue was a focus of malignant epithelioid cells that extended to the lateral and inferior margins. The neoplastic cells were cuboidal to angulated in appearance with prominent nuclei and seemed to form ill-defined tubular or glandular structures that partially resembled apocrine glands (Figure 2B). Decapitation secretion is a feature of apocrine differentiation. Examination of additional tissue sections of the tumor did not reveal remarkable decapitation secretion in contrast to the adjacent healthy apocrine glands. Rather, a solid sheet arrangement was primarily noted in several sections (Figure 2B). Neither frequent mitoses nor prominent cellular atypia were seen, and there was no evidence of lymphatic, perineural, or vascular invasion.

Figure 2. A lesional punch biopsy revealed normal apocrine glands adjacent to poorly differentiated glandular and cordlike structures (A)(H&E, original magnification ×10). The neoplastic cells were cuboidal to angulated in appearance with prominent nuclei and abundant cytoplasm demonstrating poorly differentiated apocrine glands (B)(H&E, original magnification ×40).

Immunohistochemically, tumor cells reacted strongly to cytokeratin AE1/AE3 and CAM5.², stains used to identify various cytokeratins present in epithelial tissue. Staining for epithelial membrane antigen and carcinoembryonic antigen revealed focal glandular differentiation, which further supported the epithelial origin of the neoplastic cells. Gross cystic disease fluid protein 15 (GCDFP-15) is a marker of apocrine differentiation and may indicate a carcinoma of apocrine or eccrine origin. In our case, staining for GCDFP-15 was negative in the cutaneous sections but highlighted tumor cells in 6 of 13 ipsilateral lymph nodes from locoregional metastasis. The cellular and structural morphology, immunohistochemistry, and absence of an alternative primary visceral malignancy supported the diagnosis of primary AA.

Initially the patient was not considered to be a candidate for surgery due to the rapid growth of the tumor with metastases, fatigue, weight loss, and pain. Therefore, radiation therapy was started. The patient responded well to treatment with controlled pain and resolution of the palpable mass of the left axilla. Moreover, a follow-up PET scan revealed no residual tumor and persistent, albeit decreased, axillary lymphadenopathy. As the patient’s clinical status had improved, excision of the left axillary tumor with lymph node dissection was performed 10 months after initial presentation.

In this case, the differential diagnosis consisted of various cutaneous neoplasms, primary mammary carcinoma, cutaneous metastasis, and infection. Diagnostic imaging and laboratory testing failed to identify any primary internal malignancies. Similarly, the negative cultures ruled out an infectious process. Furthermore, the axillary mass was noted to be separate from the breast tissue on physical examination and mammography. Histologically, the tumor showed features that were suggestive of an anaplastic process as well as decapitation secretion and glandular formation that clearly resembled apocrine differentiation.

Comment

Apocrine adenocarcinoma arises from apocrine sweat glands and therefore is mostly reported in areas of high apocrine gland density such as the axillae and the anogenital region.^2,4,6 However, AA also has been reported in unusual locations,^1,5,10,14-16 and they may arise from a pre-existing nevus sebaceous or from supernumerary nipples, which can occur anywhere along the milk lines.^4,15 Apocrine adenocarcinoma most commonly arises in individuals aged 40 to 50 years.^3,17 A slight male predominance has been reported but no racial predilection.^1,4-6 Although a few reports have described the development of AAs within pre-existing benign tumors such as apocrine adenomas, apocrine hyperplasias, cylindromas, and nevi sebaceous, they usually are thought to arise de novo.^4-6

Clinical Presentation

Apocrine adenocarcinoma is highly variable in its clinical manifestation.^1,6 Most cases arise as erythematous to violaceous, firm, solitary nodules. Nonetheless, AA also can present as erythematous patches of skin resembling erysipelas and ulcerated nodules with overlying granulation tissue and purulent exudate.^4,6,9,11 Although AA typically is slow growing and indolent, the time frame reported from onset to diagnosis ranges from weeks to decades.^1,6,7 Most cases present asymptomatically; when symptoms do occur, the most common ones are tenderness, purulent discharge, and restricted range of motion from extremely large tumors.^3,9 Incidence of lymph node metastasis is reported at 40% to 50% at the time of presentation.^4,6 Additionally, AA has a high rate of local recurrence, but extranodal metastasis rarely is seen.^2,6 When metastasis does occur, it is via lymphatic and hematogenous spread.^6,9 Metastatic dissemination of AA may occur in the liver, lungs, bone, brain, and parotid glands, as well as the skin via intraepidermal pagetoid spread.^4,6,9,13

Histopathology

The histologic characteristics essential to the diagnosis of primary AA are anaplastic differentiation and apocrine origin.^1,2,9,10,17 Apocrine units include coiled secretory glands that reside in the deep dermis connecting to a straight duct that empties into the isthmus of the hair follicle.^9,13 These secretory glands have a single row of cuboidal secretory cells lining the tubular component and stratified squamous epithelium lining the straight intradermal component that opens onto the hair follicle.⁹ Contractile myoepithelial cells surround the secretory cell layer of the gland.^9,13

The cuboidal secretory cells of the apocrine gland have abundant eosinophilic cytoplasm^1,4,9 and are further characterized by glandular arrangement and decapitation secretion, 2 features that are strongly suggestive of apocrine differentiation.^4-6 In contrast, the tumor cells of AA can be characterized by hyperchromatic nuclei, nuclear pleomorphism, mitotic figures, and a lack of decapitation secretion.^1,2,6 In malignancy, erratic or poorly differentiated ductal structures may be seen,^1,3-6 including papillary, cordlike, solid, or complex glandular patterns that can potentially invade the adjacent tissue without a clearly recognizable myoepithelial layer that contains them.^1,3,4,6 Moreover, AA may progress with lymphatic, vascular, or neural invasion.^1,13

Various stains may be used in immunohistochemical analysis to aid in the diagnosis of AA.^1,5 Cytokeratin AE1/AE3, CAM5.2, epithelial membrane antigen, smooth muscle antigen, periodic acid–Schiff positivity with diastase resistance, and GCDFP-15 are useful in supporting the diagnosis of AA.^2,6,10,17 Cytokeratin AE1/AE3 and CAM5.2 stain various cytokeratins to confirm the epithelial origin of the tissue.² Epithelial membrane antigen is an antigen present on the apical surface of glandular epithelial cells that also has been used to identify epithelial cells in AA.² Additionally, smooth muscle actin may be used to detect the myoepithelial layer of cells surrounding the apocrine glands.¹⁷ The lack of a continuous layer surrounding the secretory cells suggests invasion into the adjacent tissue.^1,9,17 Periodic acid–Schiff staining with diastase resistance can be used to identify the mucin stored in the intracytoplasmic granules of apocrine cells and the lumen.³ Some stains such as GCDFP-15 may highlight cells of multiple origins (eg, apocrine and eccrine).¹⁰ However, there is the possibility that poorly differentiated AAs would fail to be identified as such even with well-established apocrine markers, which may explain the differential GCDFP-15 staining patterns in our patient’s skin and lymph node sections.^1,5 Therefore, there is not a single perfect set of immunohistological criteria to aid in the diagnosis of AA.^6,10,12 Fundamentally, diagnosis requires detection of primary apocrine differentiation with features such as invasion or spread to adjacent tissue to suggest malignancy and rule out an alternate primary malignant process.^1,2,9,10,17

Treatment and Prognosis

Primary treatment of AA consists of wide local excision with adjuvant options that include chemotherapy and radiation.^2,6 Due to the high rate of lymph node metastases at presentation (40%–50%), SLNB is recommended. A positive SLNB should be followed with complete axillary lymphadenectomy^4,6; however, there is a lack of consensus regarding the role of SLNB and lymph node dissection in detecting subclinical lymph node disease, which might improve local recurrence rate and prognosis.⁶ Similarly, research shows variable results with adjunctive treatment such as chemotherapy or radiation therapy.^6,9,13 Adjuvant treatment with chemotherapy or radiation therapy should be considered in cases with large tumor size; perineural, lymphatic, or vascular invasion; or when complete removal of the tumor is not possible due to location or size.^2,6 However, neither the role nor the efficacy of such treatments in AA is well established.^6,9,13

There is little information in the literature regarding the prognosis of AA. Although no specific or well-documented prognostic criteria exist, it is generally believed that patients with well-differentiated AA will have higher cure rates or lower rates of local recurrence and lymph node metastasis than patients with poorly differentiated neoplasms.^3,6,10 A few small case series with long-term follow-up of patients ranging from 2 to 10 years have shown that prognosis may be favorable for AA patients despite local recurrence and regional lymph node metastasis.^1,5

Conclusion

Primary AA is a rare cutaneous neoplasm that most commonly occurs in the axillae and the anogenital region. Apocrine adenocarcinoma presents with highly variable clinical and histopathological findings that make diagnosis a challenge. Clinicians should keep this entity in their differential diagnosis for patients who present with nodules arising in apocrine gland–bearing skin. Ultimately, histopathology is critical to diagnosis, and special stains are often required. To make the diagnosis, a tissue biopsy demonstrating apocrine differentiation and anaplastic features to suggest a malignant process are required. Additionally, a careful workup to rule out other diagnoses should be performed. Testing modalities that detect the presence of useful markers such as apocrine or epithelial origin should be used, and the presence of positive findings should support the diagnosis of AA. However, immunohistochemical findings should be used in the context of the patient’s clinical presentation and other available data. Treatment includes wide local excision, and lymphadenectomy is recommended in the setting of nodal spread. For aggressive tumors or metastases, excision may be followed by radiation therapy and chemotherapy.

Primary apocrine adenocarcinoma (AA) is a rare cutaneous malignancy, with most of the available information about this disease consolidated from anecdotal evidence of single case reports and small case series with fewer than 30 patients.^1-11 Although certain histologic and immunohistochemical features have been suggested to be useful in the diagnosis of AA, there is no clear consensus on the required pathologic criteria.^{1,5,6,9,10,12,13} Additionally, the clinical presentation of AA is highly variable, which further adds to the challenge of making an accurate diagnosis.^{1-3,5,9,10,13}

Apocrine adenocarcinoma usually arises in areas of high apocrine gland density such as the axillae or anogenital region.^2,4,6 It also has been reported in areas such as the scalp, ear canal, eyelids, chest, nipples, arms, wrists, and fingers.^4,8,10,14-16 Apocrine adenocarcinoma in unusual locations such as the eyelid and ear canal is thought to arise from modified apocrine glands such as the Moll glands of the eyelid and the ceruminous glands of the ear canal.^9,10 The presence of ectopic apocrine glands may lead to AA in atypical sites such as the wrists and fingers.^5,16 The areola is an apocrine-dense area; therefore, AA may present on the nipples or within supernumerary nipples anywhere along the milk lines.⁴

Apocrine adenocarcinoma clinically presents as an asymptomatic to slightly painful, slowly growing, and erythematous to violaceous nodule or tumor.^4,6,9 However, in a minority of cases the initial presentation consists of a cystic or ulcerated mass with overlying granulation tissue and purulent discharge.^6,9,11 A wide time frame from the onset of symptoms to diagnosis has been reported, ranging from weeks to decades.^4,6-8 The conventional treatment of AA is wide local excision.^2,4,6,9 Although AA often presents with local lymph node metastasis at the time of diagnosis, there is no consensus on the use of sentinel lymph node biopsy (SLNB), nodal dissection, or adjuvant chemoradiation therapy.^1,3,8,9

We report the case of a 49-year-old man with primary AA of the left axilla; the clinical and histologic features of AA as well as the appropriate diagnostic and treatment modalities also are provided.

Case Report

A 49-year-old man with a slowly growing tender mass of the left axilla of 1 year’s duration was referred to our dermatology clinic for evaluation. A review of systems revealed loss of appetite, fatigue, and a 4-month history of unintentional weight loss (15–20 lb). The patient had a history of hepatitis C virus, intravenous drug use, alcohol abuse, and cigarette smoking (1 pack daily) for many years. Additionally, the patient reported a paternal family history of numerous visceral malignancies. Examination of the left axilla revealed a 1.5×5-cm ulcerated tumor that produced serosanguineous discharge and was tender to palpation (Figure 1). Two 1-cm, firm, freely mobile subcutaneous nodules with no overlying skin changes were palpable at the medial border of the ulcerated nodule. There was no additional cervical or axillary lymphadenopathy, and a breast examination was normal.

Figure 1. A 1.5×5-cm ulcerated tumor on the left axilla with firm palpable lymph nodes on the medial border.

The differential diagnosis included primary squamous cell carcinoma or adnexal neoplasm, primary breast carcinoma, lymphoma, scrofuloderma, atypical mycobacterial infection, and cutaneous metastasis from an internal malignancy. Two 4-mm punch biopsies were performed and sent for routine histopathology and bacterial, fungal, and mycobacterial tissue cultures. To exclude a primary visceral malignancy or metastasis, computed tomography of the chest, abdomen, and pelvis; positron emission tomography (PET) from the base of the skull to the thighs; colonoscopy; magnetic resonance imaging of the brain; esophagogastroduodenoscopy; and mammography were conducted. Prominent left axillary lymphadenopathy was noted on computed tomography. Additionally, PET identified extranodal spread in the left axilla, left lateral chest wall, and the left sternocleidomastoid region. Furthermore, a 1-cm hypermetabolic nodule involving the right rectus abdominus muscle was noted on the PET scan. Based on their appearance, the nodules most likely represented metastasis from a primary skin malignancy. The rest of the studies were unremarkable. Serum tumor markers including prostate-specific antigen, cancer antigen 19-9, and carcinoembryonic antigen were within reference range. Immunostaining for estrogen receptor, progesterone receptor, and ERBB2 (formerly HER2/neu) was negative. The only abnormalities noted on serum chemistries were slight elevations in aspartate aminotransferase, alanine aminotransferase, alkaline phosphatase, and the a-fetoprotein tumor marker, which was attributed to chronic hepatitis C infection. Bacterial, fungal, and mycobacterial tissue cultures also were negative. These results ruled out infection and suggested against a primary visceral malignancy with cutaneous metastasis.

Histopathology revealed a moderately differentiated adenocarcinoma adjacent to healthy-appearing apocrine glands (Figure 2A). The normal glands were composed of cuboidal cells with abundant eosinophilic cytoplasm and prominent nuclei. The cells were arranged in a single layer in a glandular formation with prominent decapitation secretion. Adjacent to the normal apocrine glandular tissue was a focus of malignant epithelioid cells that extended to the lateral and inferior margins. The neoplastic cells were cuboidal to angulated in appearance with prominent nuclei and seemed to form ill-defined tubular or glandular structures that partially resembled apocrine glands (Figure 2B). Decapitation secretion is a feature of apocrine differentiation. Examination of additional tissue sections of the tumor did not reveal remarkable decapitation secretion in contrast to the adjacent healthy apocrine glands. Rather, a solid sheet arrangement was primarily noted in several sections (Figure 2B). Neither frequent mitoses nor prominent cellular atypia were seen, and there was no evidence of lymphatic, perineural, or vascular invasion.

Figure 2. A lesional punch biopsy revealed normal apocrine glands adjacent to poorly differentiated glandular and cordlike structures (A)(H&E, original magnification ×10). The neoplastic cells were cuboidal to angulated in appearance with prominent nuclei and abundant cytoplasm demonstrating poorly differentiated apocrine glands (B)(H&E, original magnification ×40).

Immunohistochemically, tumor cells reacted strongly to cytokeratin AE1/AE3 and CAM5.², stains used to identify various cytokeratins present in epithelial tissue. Staining for epithelial membrane antigen and carcinoembryonic antigen revealed focal glandular differentiation, which further supported the epithelial origin of the neoplastic cells. Gross cystic disease fluid protein 15 (GCDFP-15) is a marker of apocrine differentiation and may indicate a carcinoma of apocrine or eccrine origin. In our case, staining for GCDFP-15 was negative in the cutaneous sections but highlighted tumor cells in 6 of 13 ipsilateral lymph nodes from locoregional metastasis. The cellular and structural morphology, immunohistochemistry, and absence of an alternative primary visceral malignancy supported the diagnosis of primary AA.

Initially the patient was not considered to be a candidate for surgery due to the rapid growth of the tumor with metastases, fatigue, weight loss, and pain. Therefore, radiation therapy was started. The patient responded well to treatment with controlled pain and resolution of the palpable mass of the left axilla. Moreover, a follow-up PET scan revealed no residual tumor and persistent, albeit decreased, axillary lymphadenopathy. As the patient’s clinical status had improved, excision of the left axillary tumor with lymph node dissection was performed 10 months after initial presentation.

In this case, the differential diagnosis consisted of various cutaneous neoplasms, primary mammary carcinoma, cutaneous metastasis, and infection. Diagnostic imaging and laboratory testing failed to identify any primary internal malignancies. Similarly, the negative cultures ruled out an infectious process. Furthermore, the axillary mass was noted to be separate from the breast tissue on physical examination and mammography. Histologically, the tumor showed features that were suggestive of an anaplastic process as well as decapitation secretion and glandular formation that clearly resembled apocrine differentiation.

Comment

Apocrine adenocarcinoma arises from apocrine sweat glands and therefore is mostly reported in areas of high apocrine gland density such as the axillae and the anogenital region.^2,4,6 However, AA also has been reported in unusual locations,^1,5,10,14-16 and they may arise from a pre-existing nevus sebaceous or from supernumerary nipples, which can occur anywhere along the milk lines.^4,15 Apocrine adenocarcinoma most commonly arises in individuals aged 40 to 50 years.^3,17 A slight male predominance has been reported but no racial predilection.^1,4-6 Although a few reports have described the development of AAs within pre-existing benign tumors such as apocrine adenomas, apocrine hyperplasias, cylindromas, and nevi sebaceous, they usually are thought to arise de novo.^4-6

Clinical Presentation

Apocrine adenocarcinoma is highly variable in its clinical manifestation.^1,6 Most cases arise as erythematous to violaceous, firm, solitary nodules. Nonetheless, AA also can present as erythematous patches of skin resembling erysipelas and ulcerated nodules with overlying granulation tissue and purulent exudate.^4,6,9,11 Although AA typically is slow growing and indolent, the time frame reported from onset to diagnosis ranges from weeks to decades.^1,6,7 Most cases present asymptomatically; when symptoms do occur, the most common ones are tenderness, purulent discharge, and restricted range of motion from extremely large tumors.^3,9 Incidence of lymph node metastasis is reported at 40% to 50% at the time of presentation.^4,6 Additionally, AA has a high rate of local recurrence, but extranodal metastasis rarely is seen.^2,6 When metastasis does occur, it is via lymphatic and hematogenous spread.^6,9 Metastatic dissemination of AA may occur in the liver, lungs, bone, brain, and parotid glands, as well as the skin via intraepidermal pagetoid spread.^4,6,9,13

Histopathology

The histologic characteristics essential to the diagnosis of primary AA are anaplastic differentiation and apocrine origin.^1,2,9,10,17 Apocrine units include coiled secretory glands that reside in the deep dermis connecting to a straight duct that empties into the isthmus of the hair follicle.^9,13 These secretory glands have a single row of cuboidal secretory cells lining the tubular component and stratified squamous epithelium lining the straight intradermal component that opens onto the hair follicle.⁹ Contractile myoepithelial cells surround the secretory cell layer of the gland.^9,13

The cuboidal secretory cells of the apocrine gland have abundant eosinophilic cytoplasm^1,4,9 and are further characterized by glandular arrangement and decapitation secretion, 2 features that are strongly suggestive of apocrine differentiation.^4-6 In contrast, the tumor cells of AA can be characterized by hyperchromatic nuclei, nuclear pleomorphism, mitotic figures, and a lack of decapitation secretion.^1,2,6 In malignancy, erratic or poorly differentiated ductal structures may be seen,^1,3-6 including papillary, cordlike, solid, or complex glandular patterns that can potentially invade the adjacent tissue without a clearly recognizable myoepithelial layer that contains them.^1,3,4,6 Moreover, AA may progress with lymphatic, vascular, or neural invasion.^1,13

Various stains may be used in immunohistochemical analysis to aid in the diagnosis of AA.^1,5 Cytokeratin AE1/AE3, CAM5.2, epithelial membrane antigen, smooth muscle antigen, periodic acid–Schiff positivity with diastase resistance, and GCDFP-15 are useful in supporting the diagnosis of AA.^2,6,10,17 Cytokeratin AE1/AE3 and CAM5.2 stain various cytokeratins to confirm the epithelial origin of the tissue.² Epithelial membrane antigen is an antigen present on the apical surface of glandular epithelial cells that also has been used to identify epithelial cells in AA.² Additionally, smooth muscle actin may be used to detect the myoepithelial layer of cells surrounding the apocrine glands.¹⁷ The lack of a continuous layer surrounding the secretory cells suggests invasion into the adjacent tissue.^1,9,17 Periodic acid–Schiff staining with diastase resistance can be used to identify the mucin stored in the intracytoplasmic granules of apocrine cells and the lumen.³ Some stains such as GCDFP-15 may highlight cells of multiple origins (eg, apocrine and eccrine).¹⁰ However, there is the possibility that poorly differentiated AAs would fail to be identified as such even with well-established apocrine markers, which may explain the differential GCDFP-15 staining patterns in our patient’s skin and lymph node sections.^1,5 Therefore, there is not a single perfect set of immunohistological criteria to aid in the diagnosis of AA.^6,10,12 Fundamentally, diagnosis requires detection of primary apocrine differentiation with features such as invasion or spread to adjacent tissue to suggest malignancy and rule out an alternate primary malignant process.^1,2,9,10,17

Treatment and Prognosis

Primary treatment of AA consists of wide local excision with adjuvant options that include chemotherapy and radiation.^2,6 Due to the high rate of lymph node metastases at presentation (40%–50%), SLNB is recommended. A positive SLNB should be followed with complete axillary lymphadenectomy^4,6; however, there is a lack of consensus regarding the role of SLNB and lymph node dissection in detecting subclinical lymph node disease, which might improve local recurrence rate and prognosis.⁶ Similarly, research shows variable results with adjunctive treatment such as chemotherapy or radiation therapy.^6,9,13 Adjuvant treatment with chemotherapy or radiation therapy should be considered in cases with large tumor size; perineural, lymphatic, or vascular invasion; or when complete removal of the tumor is not possible due to location or size.^2,6 However, neither the role nor the efficacy of such treatments in AA is well established.^6,9,13

There is little information in the literature regarding the prognosis of AA. Although no specific or well-documented prognostic criteria exist, it is generally believed that patients with well-differentiated AA will have higher cure rates or lower rates of local recurrence and lymph node metastasis than patients with poorly differentiated neoplasms.^3,6,10 A few small case series with long-term follow-up of patients ranging from 2 to 10 years have shown that prognosis may be favorable for AA patients despite local recurrence and regional lymph node metastasis.^1,5

Conclusion

Primary AA is a rare cutaneous neoplasm that most commonly occurs in the axillae and the anogenital region. Apocrine adenocarcinoma presents with highly variable clinical and histopathological findings that make diagnosis a challenge. Clinicians should keep this entity in their differential diagnosis for patients who present with nodules arising in apocrine gland–bearing skin. Ultimately, histopathology is critical to diagnosis, and special stains are often required. To make the diagnosis, a tissue biopsy demonstrating apocrine differentiation and anaplastic features to suggest a malignant process are required. Additionally, a careful workup to rule out other diagnoses should be performed. Testing modalities that detect the presence of useful markers such as apocrine or epithelial origin should be used, and the presence of positive findings should support the diagnosis of AA. However, immunohistochemical findings should be used in the context of the patient’s clinical presentation and other available data. Treatment includes wide local excision, and lymphadenectomy is recommended in the setting of nodal spread. For aggressive tumors or metastases, excision may be followed by radiation therapy and chemotherapy.