Ultrasound helped reveal high-risk thyroid nodules

Thyroid nodules that include microcalcifications, are larger than 2 cm in size, and have an entirely solid composition on ultrasound imaging are the most likely to be cancerous, according to a report published online Aug. 26 in JAMA Internal Medicine.

A patient’s risk of having primary thyroid cancer ranges from 0.2% when a nodule’s ultrasound image has none of the three characteristics to 1.8% if a nodule has one of the characteristics, 6.2% if a nodule has two of the characteristics, and 96% if a nodule has all three characteristics, said Dr. Rebecca Smith-Bindman of the department of radiology and biomedical imaging, University of California, San Francisco, and her associates.

"Ours is the first study, to our knowledge, that permits estimating this risk," they noted (JAMA Intern. Med. 2013 Aug. 26 [doi:10.1001/jamainternmed.2013.9245]).

Thyroid nodules are extremely common, but almost all of them are benign, so it is crucial to identify which ones may be malignant and to biopsy them, and to avoid subjecting low-risk patients to unnecessary biopsy.

Cindy Chew/UCSF

To determine which features on ultrasound correspond with cancer risk, Dr. Smith-Bindman and her colleagues performed a retrospective case-control study involving 8,806 consecutive patients who underwent 11,618 thyroid ultrasound examinations at the university from January 2000 through March 2005.

The patients who were found to have primary thyroid cancer during up to 7 years of follow-up were identified using data in a comprehensive cancer registry. Those 96 patients were matched for age, sex, and year of ultrasound examination to 369 control subjects who did not have primary thyroid cancer.

Two of the researchers who were blinded to the subjects’ cancer status independently assessed the number and size of all the thyroid nodules that were imaged in the study population and performed a detailed analysis of numerous ultrasound characteristics that might possibly correlate with the presence of malignancy. The agreement between those two reviewers in the categorization of the specific ultrasound image characteristics was "good to outstanding."

"We considered many nodule characteristics endorsed by other authors, but when put into the multiple-predictor models, most of the characteristics were not significantly associated with cancer risk," the researchers said.

In the initial univariable analysis of the data, several ultrasonographic traits were significantly associated with the likelihood that a thyroid nodule harbored cancer.

Microcalcifications had the strongest association with malignancy: They were found in 38.2% of cancerous nodules, compared with only 5.4% of benign nodules. Thus, if microcalcifications were present in a nodule, the chances were seven times greater that the nodule was cancerous, Dr. Smith-Bindman and her coworkers reported.

In the study analysis, the size of the nodule on ultrasound also correlated with cancer risk, with the odds of malignancy increasing as nodule size increased. A size of 2 cm appeared to be a good cutoff point, because nodules larger than 2 cm were much more likely than nodules smaller than 1 cm to be cancerous, with an odds ratio of 3.1.

Several other characteristics correlated with cancer risk in the analysis, but did so to a lesser degree. Coarse calcifications, solid (vs. cystic or mixed) nodule composition, hyperechoic nodule echogenicity, central vascularity, ill-defined or lobulated nodule margins, and taller-than-wide nodule shape all raised the risk of malignancy, with odds ratios ranging from 1.6 to 2.9.

Traits not associated with cancer risk included rim calcifications, comet-tail artifacts, peripheral vascularity, and the presence of a "halo," they said.

In a multivariable analysis, only three nodule characteristics remained significantly associated with cancer risk. The presence of microcalcifications had an odds ratio of 8.1, size larger than 2 cm had an odds ratio of 3.6, and an entirely solid composition had an odds ratio of 4.0, Dr. Smith-Bindman and her associates said.

The findings remained robust through a series of sensitivity analyses.

The investigators found that performing a biopsy only if two of the three characteristics were present would yield a much greater diagnostic sensitivity and specificity than the current practice of biopsying all thyroid nodules larger than 5 mm.

"Compared with existing guidelines ... adoption of this more stringent rule requiring two abnormal [ultrasound] characteristics to prompt biopsy would reduce unnecessary biopsies by 90% while maintaining a low risk of cancer [0.5%] in patients in whom biopsy is deferred," they wrote.

In contrast, requiring all three nodule characteristics to be present before performing a biopsy would detect only a small proportion of thyroid cancers, they added.

In the study, simple cysts never indicated the presence of cancer. Such thyroid cysts should be considered "essentially never malignant" and should not be biopsied, Dr. Smith-Bindman and her associates said.

"The main strength of our study is the large sample size and the linkage of the cohort with data from a comprehensive cancer registry, which allows accurate assessment of the true underlying prevalence of cancer," the researchers noted.

The overall prevalence of primary thyroid cancer was 1.5% in the study population. In contrast, most previous studies have cited a prevalence rate closer to 20%. "All previous studies have inflated the association between nodule characteristics and cancer risk because they limited their analyses to nodules that underwent biopsy," the investigators said.

The National Cancer Institute and the University of California, San Francisco, supported the study. No financial conflicts of interest were reported.

Thyroid nodules that include microcalcifications, are larger than 2 cm in size, and have an entirely solid composition on ultrasound imaging are the most likely to be cancerous, according to a report published online Aug. 26 in JAMA Internal Medicine.

A patient’s risk of having primary thyroid cancer ranges from 0.2% when a nodule’s ultrasound image has none of the three characteristics to 1.8% if a nodule has one of the characteristics, 6.2% if a nodule has two of the characteristics, and 96% if a nodule has all three characteristics, said Dr. Rebecca Smith-Bindman of the department of radiology and biomedical imaging, University of California, San Francisco, and her associates.

"Ours is the first study, to our knowledge, that permits estimating this risk," they noted (JAMA Intern. Med. 2013 Aug. 26 [doi:10.1001/jamainternmed.2013.9245]).

Thyroid nodules are extremely common, but almost all of them are benign, so it is crucial to identify which ones may be malignant and to biopsy them, and to avoid subjecting low-risk patients to unnecessary biopsy.

Cindy Chew/UCSF

To determine which features on ultrasound correspond with cancer risk, Dr. Smith-Bindman and her colleagues performed a retrospective case-control study involving 8,806 consecutive patients who underwent 11,618 thyroid ultrasound examinations at the university from January 2000 through March 2005.

The patients who were found to have primary thyroid cancer during up to 7 years of follow-up were identified using data in a comprehensive cancer registry. Those 96 patients were matched for age, sex, and year of ultrasound examination to 369 control subjects who did not have primary thyroid cancer.

Two of the researchers who were blinded to the subjects’ cancer status independently assessed the number and size of all the thyroid nodules that were imaged in the study population and performed a detailed analysis of numerous ultrasound characteristics that might possibly correlate with the presence of malignancy. The agreement between those two reviewers in the categorization of the specific ultrasound image characteristics was "good to outstanding."

"We considered many nodule characteristics endorsed by other authors, but when put into the multiple-predictor models, most of the characteristics were not significantly associated with cancer risk," the researchers said.

In the initial univariable analysis of the data, several ultrasonographic traits were significantly associated with the likelihood that a thyroid nodule harbored cancer.

Microcalcifications had the strongest association with malignancy: They were found in 38.2% of cancerous nodules, compared with only 5.4% of benign nodules. Thus, if microcalcifications were present in a nodule, the chances were seven times greater that the nodule was cancerous, Dr. Smith-Bindman and her coworkers reported.

In the study analysis, the size of the nodule on ultrasound also correlated with cancer risk, with the odds of malignancy increasing as nodule size increased. A size of 2 cm appeared to be a good cutoff point, because nodules larger than 2 cm were much more likely than nodules smaller than 1 cm to be cancerous, with an odds ratio of 3.1.

Several other characteristics correlated with cancer risk in the analysis, but did so to a lesser degree. Coarse calcifications, solid (vs. cystic or mixed) nodule composition, hyperechoic nodule echogenicity, central vascularity, ill-defined or lobulated nodule margins, and taller-than-wide nodule shape all raised the risk of malignancy, with odds ratios ranging from 1.6 to 2.9.

Traits not associated with cancer risk included rim calcifications, comet-tail artifacts, peripheral vascularity, and the presence of a "halo," they said.

In a multivariable analysis, only three nodule characteristics remained significantly associated with cancer risk. The presence of microcalcifications had an odds ratio of 8.1, size larger than 2 cm had an odds ratio of 3.6, and an entirely solid composition had an odds ratio of 4.0, Dr. Smith-Bindman and her associates said.

The findings remained robust through a series of sensitivity analyses.

The investigators found that performing a biopsy only if two of the three characteristics were present would yield a much greater diagnostic sensitivity and specificity than the current practice of biopsying all thyroid nodules larger than 5 mm.

"Compared with existing guidelines ... adoption of this more stringent rule requiring two abnormal [ultrasound] characteristics to prompt biopsy would reduce unnecessary biopsies by 90% while maintaining a low risk of cancer [0.5%] in patients in whom biopsy is deferred," they wrote.

In contrast, requiring all three nodule characteristics to be present before performing a biopsy would detect only a small proportion of thyroid cancers, they added.

In the study, simple cysts never indicated the presence of cancer. Such thyroid cysts should be considered "essentially never malignant" and should not be biopsied, Dr. Smith-Bindman and her associates said.

"The main strength of our study is the large sample size and the linkage of the cohort with data from a comprehensive cancer registry, which allows accurate assessment of the true underlying prevalence of cancer," the researchers noted.

The overall prevalence of primary thyroid cancer was 1.5% in the study population. In contrast, most previous studies have cited a prevalence rate closer to 20%. "All previous studies have inflated the association between nodule characteristics and cancer risk because they limited their analyses to nodules that underwent biopsy," the investigators said.

The National Cancer Institute and the University of California, San Francisco, supported the study. No financial conflicts of interest were reported.

Thyroid nodules that include microcalcifications, are larger than 2 cm in size, and have an entirely solid composition on ultrasound imaging are the most likely to be cancerous, according to a report published online Aug. 26 in JAMA Internal Medicine.

A patient’s risk of having primary thyroid cancer ranges from 0.2% when a nodule’s ultrasound image has none of the three characteristics to 1.8% if a nodule has one of the characteristics, 6.2% if a nodule has two of the characteristics, and 96% if a nodule has all three characteristics, said Dr. Rebecca Smith-Bindman of the department of radiology and biomedical imaging, University of California, San Francisco, and her associates.

"Ours is the first study, to our knowledge, that permits estimating this risk," they noted (JAMA Intern. Med. 2013 Aug. 26 [doi:10.1001/jamainternmed.2013.9245]).

Thyroid nodules are extremely common, but almost all of them are benign, so it is crucial to identify which ones may be malignant and to biopsy them, and to avoid subjecting low-risk patients to unnecessary biopsy.

Cindy Chew/UCSF

To determine which features on ultrasound correspond with cancer risk, Dr. Smith-Bindman and her colleagues performed a retrospective case-control study involving 8,806 consecutive patients who underwent 11,618 thyroid ultrasound examinations at the university from January 2000 through March 2005.

The patients who were found to have primary thyroid cancer during up to 7 years of follow-up were identified using data in a comprehensive cancer registry. Those 96 patients were matched for age, sex, and year of ultrasound examination to 369 control subjects who did not have primary thyroid cancer.

Two of the researchers who were blinded to the subjects’ cancer status independently assessed the number and size of all the thyroid nodules that were imaged in the study population and performed a detailed analysis of numerous ultrasound characteristics that might possibly correlate with the presence of malignancy. The agreement between those two reviewers in the categorization of the specific ultrasound image characteristics was "good to outstanding."

"We considered many nodule characteristics endorsed by other authors, but when put into the multiple-predictor models, most of the characteristics were not significantly associated with cancer risk," the researchers said.

In the initial univariable analysis of the data, several ultrasonographic traits were significantly associated with the likelihood that a thyroid nodule harbored cancer.

Microcalcifications had the strongest association with malignancy: They were found in 38.2% of cancerous nodules, compared with only 5.4% of benign nodules. Thus, if microcalcifications were present in a nodule, the chances were seven times greater that the nodule was cancerous, Dr. Smith-Bindman and her coworkers reported.

In the study analysis, the size of the nodule on ultrasound also correlated with cancer risk, with the odds of malignancy increasing as nodule size increased. A size of 2 cm appeared to be a good cutoff point, because nodules larger than 2 cm were much more likely than nodules smaller than 1 cm to be cancerous, with an odds ratio of 3.1.

Several other characteristics correlated with cancer risk in the analysis, but did so to a lesser degree. Coarse calcifications, solid (vs. cystic or mixed) nodule composition, hyperechoic nodule echogenicity, central vascularity, ill-defined or lobulated nodule margins, and taller-than-wide nodule shape all raised the risk of malignancy, with odds ratios ranging from 1.6 to 2.9.

Traits not associated with cancer risk included rim calcifications, comet-tail artifacts, peripheral vascularity, and the presence of a "halo," they said.

In a multivariable analysis, only three nodule characteristics remained significantly associated with cancer risk. The presence of microcalcifications had an odds ratio of 8.1, size larger than 2 cm had an odds ratio of 3.6, and an entirely solid composition had an odds ratio of 4.0, Dr. Smith-Bindman and her associates said.

The findings remained robust through a series of sensitivity analyses.

The investigators found that performing a biopsy only if two of the three characteristics were present would yield a much greater diagnostic sensitivity and specificity than the current practice of biopsying all thyroid nodules larger than 5 mm.

"Compared with existing guidelines ... adoption of this more stringent rule requiring two abnormal [ultrasound] characteristics to prompt biopsy would reduce unnecessary biopsies by 90% while maintaining a low risk of cancer [0.5%] in patients in whom biopsy is deferred," they wrote.

In contrast, requiring all three nodule characteristics to be present before performing a biopsy would detect only a small proportion of thyroid cancers, they added.

In the study, simple cysts never indicated the presence of cancer. Such thyroid cysts should be considered "essentially never malignant" and should not be biopsied, Dr. Smith-Bindman and her associates said.

"The main strength of our study is the large sample size and the linkage of the cohort with data from a comprehensive cancer registry, which allows accurate assessment of the true underlying prevalence of cancer," the researchers noted.

The overall prevalence of primary thyroid cancer was 1.5% in the study population. In contrast, most previous studies have cited a prevalence rate closer to 20%. "All previous studies have inflated the association between nodule characteristics and cancer risk because they limited their analyses to nodules that underwent biopsy," the investigators said.

The National Cancer Institute and the University of California, San Francisco, supported the study. No financial conflicts of interest were reported.