Melvin Gerbie, MD

THE QUESTION:
Is self-sampling effective in cervical cancer screening?

Past studies

Prior research has shown selfcollected cervical cell samples to be of variable concurrence to clinician-collected samples.

This study

Researchers evaluated 1 selfand 1 clinician-collected sample from 253 women, ages 16 to 88, randomly selected from a population at high risk for cervical neoplasia. Participants collected self-samples by rotating a cotton swab against the vaginal epithelium, and possibly, the cervix. Next, a colposcopist collected samples with an Ayre’s spatula and endocervical brush, then conducted a colposcopy. All specimens were frozen. Polymerase chain reaction (PCR) studies tested for both low-risk (LR) and high-risk (HR) viral types.

Positive HPV results from self-sampling and physician sampling were 23% and 29%, respectively. While there was no significant difference in the overall results from self-collected samples compared with physician-collected specimens, the prevalence of HR HPV among self-collected samples was 17%, as opposed to 26% in physician-collected samples. Further, testing for HPV detection on self-collected samples resulted in 50% more missed diagnoses of cervical cancer than did samples collected with a spatula and endocervical brush, and in 33.3% more missed diagnoses of high-grade cervical intraepithelial neoplasia. Hence, the researchers concluded that a sample for cervical HPV detection collected with a spatula and endocervical brush provides better results for primary cervical cancer screening than does self-sampling with a cotton-tipped swab.

Find this study

Lorenzato FR, Singer A, Ho L, et al. May 2002 issue of the American Journal of Obstetrics and Gynecology; abstract online at www.us.elsevierhealth.com/ajog.

Who may be affected by these findings?

All women screened for cervical cytology.

Expert commentary

There has been a marked decrease in invasive cervical cancer prevalence and mortality in countries where there are organized cytologic screening programs. Even so, an estimated 30% of US women with invasive cervical cancer had not had a Pap smear in at least 5 years, and 50% had never been screened. Moreover, developing countries have both health-care infrastructure deficiencies and cultural taboos that have thus far precluded universal screening.

Many women may find self-sampling to be a more convenient and comfortable way of screening for cervical cancer. Thus, if self-sampling could be substituted for clinician examinations, the detection of preinvasive cervical neoplasia cases could be substantially increased. This study, however, demonstrated a poor correlation with colposcopic and histologic diagnoses of high-grade squamous intraepithelial lesions and invasive cancer. Of note, this study’s conclusions differed from a similar project conducted in South Africa in which researchers found a high sensitivity but poor specificity in self-collected specimens.¹

Bottom line

Although self-sampling is promising, the present self-collection technique may provide a false sense of security for patients with high-grade lesions. Since these patients need more close monitoring than those with low-grade lesions, this method is not ready for widespread implementation. Further research aimed at improving the safety and efficacy of self-sampling is still needed.

THE QUESTION:
Is self-sampling effective in cervical cancer screening?

Past studies

Prior research has shown selfcollected cervical cell samples to be of variable concurrence to clinician-collected samples.

This study

Researchers evaluated 1 selfand 1 clinician-collected sample from 253 women, ages 16 to 88, randomly selected from a population at high risk for cervical neoplasia. Participants collected self-samples by rotating a cotton swab against the vaginal epithelium, and possibly, the cervix. Next, a colposcopist collected samples with an Ayre’s spatula and endocervical brush, then conducted a colposcopy. All specimens were frozen. Polymerase chain reaction (PCR) studies tested for both low-risk (LR) and high-risk (HR) viral types.

Positive HPV results from self-sampling and physician sampling were 23% and 29%, respectively. While there was no significant difference in the overall results from self-collected samples compared with physician-collected specimens, the prevalence of HR HPV among self-collected samples was 17%, as opposed to 26% in physician-collected samples. Further, testing for HPV detection on self-collected samples resulted in 50% more missed diagnoses of cervical cancer than did samples collected with a spatula and endocervical brush, and in 33.3% more missed diagnoses of high-grade cervical intraepithelial neoplasia. Hence, the researchers concluded that a sample for cervical HPV detection collected with a spatula and endocervical brush provides better results for primary cervical cancer screening than does self-sampling with a cotton-tipped swab.

Find this study

Lorenzato FR, Singer A, Ho L, et al. May 2002 issue of the American Journal of Obstetrics and Gynecology; abstract online at www.us.elsevierhealth.com/ajog.

Who may be affected by these findings?

All women screened for cervical cytology.

Expert commentary

There has been a marked decrease in invasive cervical cancer prevalence and mortality in countries where there are organized cytologic screening programs. Even so, an estimated 30% of US women with invasive cervical cancer had not had a Pap smear in at least 5 years, and 50% had never been screened. Moreover, developing countries have both health-care infrastructure deficiencies and cultural taboos that have thus far precluded universal screening.

Many women may find self-sampling to be a more convenient and comfortable way of screening for cervical cancer. Thus, if self-sampling could be substituted for clinician examinations, the detection of preinvasive cervical neoplasia cases could be substantially increased. This study, however, demonstrated a poor correlation with colposcopic and histologic diagnoses of high-grade squamous intraepithelial lesions and invasive cancer. Of note, this study’s conclusions differed from a similar project conducted in South Africa in which researchers found a high sensitivity but poor specificity in self-collected specimens.¹

Bottom line

Although self-sampling is promising, the present self-collection technique may provide a false sense of security for patients with high-grade lesions. Since these patients need more close monitoring than those with low-grade lesions, this method is not ready for widespread implementation. Further research aimed at improving the safety and efficacy of self-sampling is still needed.

THE QUESTION:
Is self-sampling effective in cervical cancer screening?

Past studies

Prior research has shown selfcollected cervical cell samples to be of variable concurrence to clinician-collected samples.

This study

Researchers evaluated 1 selfand 1 clinician-collected sample from 253 women, ages 16 to 88, randomly selected from a population at high risk for cervical neoplasia. Participants collected self-samples by rotating a cotton swab against the vaginal epithelium, and possibly, the cervix. Next, a colposcopist collected samples with an Ayre’s spatula and endocervical brush, then conducted a colposcopy. All specimens were frozen. Polymerase chain reaction (PCR) studies tested for both low-risk (LR) and high-risk (HR) viral types.

Positive HPV results from self-sampling and physician sampling were 23% and 29%, respectively. While there was no significant difference in the overall results from self-collected samples compared with physician-collected specimens, the prevalence of HR HPV among self-collected samples was 17%, as opposed to 26% in physician-collected samples. Further, testing for HPV detection on self-collected samples resulted in 50% more missed diagnoses of cervical cancer than did samples collected with a spatula and endocervical brush, and in 33.3% more missed diagnoses of high-grade cervical intraepithelial neoplasia. Hence, the researchers concluded that a sample for cervical HPV detection collected with a spatula and endocervical brush provides better results for primary cervical cancer screening than does self-sampling with a cotton-tipped swab.

Find this study

Lorenzato FR, Singer A, Ho L, et al. May 2002 issue of the American Journal of Obstetrics and Gynecology; abstract online at www.us.elsevierhealth.com/ajog.

Who may be affected by these findings?

All women screened for cervical cytology.

Expert commentary

There has been a marked decrease in invasive cervical cancer prevalence and mortality in countries where there are organized cytologic screening programs. Even so, an estimated 30% of US women with invasive cervical cancer had not had a Pap smear in at least 5 years, and 50% had never been screened. Moreover, developing countries have both health-care infrastructure deficiencies and cultural taboos that have thus far precluded universal screening.

Many women may find self-sampling to be a more convenient and comfortable way of screening for cervical cancer. Thus, if self-sampling could be substituted for clinician examinations, the detection of preinvasive cervical neoplasia cases could be substantially increased. This study, however, demonstrated a poor correlation with colposcopic and histologic diagnoses of high-grade squamous intraepithelial lesions and invasive cancer. Of note, this study’s conclusions differed from a similar project conducted in South Africa in which researchers found a high sensitivity but poor specificity in self-collected specimens.¹

Bottom line

Although self-sampling is promising, the present self-collection technique may provide a false sense of security for patients with high-grade lesions. Since these patients need more close monitoring than those with low-grade lesions, this method is not ready for widespread implementation. Further research aimed at improving the safety and efficacy of self-sampling is still needed.