Objective
Management of equivocal Papanicolaou smear result remains to be challenging even with the aid of human papillomavirus test. Recently, 3 novel methylation-silenced genes, PAX1 , WT1 , and PCDH10 , have been found to be specifically associated with cervical cancer. We compared the performances of methylation test of these genes with human papillomavirus tests in triage of equivocal Papanicolaou smear result.
Study Design
Two hundred twenty-two women with Papanicolaou smear results of atypical cells of undetermined significance nested to a multicenter, nation-wide cohort (the T1899 cohort) were studied. Status of cervical neoplasm was diagnosed with colposcopic biopsy. Status of gene methylation was determined by methylation-specific polymerase chain reaction. High-risk human papillomavirus DNA was detected by polymerase chain reaction-reverse line blot hybridization and Hybrid Capture 2.
Results
Cervical intraepithelial neoplasm 1, cervical intraepithelial neoplasm 2, cervical intraepithelial neoplasm 3, carcinoma in situ, carcinoma, and normal cervix were diagnosed in 58, 17, 14, 10, 1, and 120 women, respectively. Methylation of PCDH10 , WT1 , and PAX1 was highly associated with the severity of cervical neoplasm ( P < 10 −9 , < 10 −7 , and < 10 −5 , respectively). In comparison with a negative test result, the odds ratio (95% confidence intervals) for cervical intraepithelial neoplasm 3 or more severe neoplasms for women tested positive for methylation of these 3 genes were 26.4 (9.0–77.3), 18.1 (6.9–47.2), and 10.3 (4.1–25.9), respectively; whereas those positive for human papillomavirus polymerase chain reaction and Hybrid Capture 2 were 10.5 (3.5–31.9) and 5.6 (2.3–21.4). In triage for atypical cells of undetermined significance, each methylation test had less colposcopy referral and false-positive rates, but higher false-negative rate than the human papillomavirus tests. With a combination test of PCDH10 or WT1 methylation, a comparable false-negative rate ( P = .62) but much less false-positive rate ( P = .002) and colposcopy referral rate ( P < 10 −6 ) were achieved.
Conclusion
In triage of atypical cells of undetermined significance Papanicolaou smear results, methylation test of WT1 and PCDH10 is superior to human papillomavirus test in this multicenter cohort. Comparing to current human papillomavirus triage, the new test has only one third of false positivity and half of colposcopy referral, with no compromise of the sensitivity in diagnosis of cervical intraepithelial neoplasm 3 or more severe neoplasms.
As many as 1-5% of Papanicolaou (Pap) smears end with equivocal results. The majority of them are atypical cells of undetermined significance (ASC-US). Around 10-20% and 3-5% of ASC-US are eventually diagnosed with cervical intraepithelial neoplasm 1 (CIN1) and CIN2/3, respectively, leaving a great stress to both patients and physicians, and a high cost for management.
See Journal Club, page 87
Immediate colposcopy is assumed to be the highest standard in detecting CIN2/3 lesions from ASC-US, but problems of overinterpretation of the colposcopy and biopsy findings, and limited resource of qualified colposcopist also exist. Currently, ASC-US women are usually triaged with repeat Pap smear or human papillomavirus (HPV) test before colposcopy referral. Several large scale studies have established HPV test for ASC-US as the preferred management option that decreases the colposcopy referral and keeps a high sensitivity of detecting CIN2/3 lesions. However, the predictive power of HPV test is still unsatisfactory, because as many as 40-60% of women with equivocal smear results are HPV positive but only <1% of them are eventually diagnosed with cancer. A more precise test is needed.
Hypermethylation of CpG sites of promoter of tumor suppressor gene is an early and crucial event in cancer development. A growing number of methylated genes have been found to be associated with different grades of cervical neoplasia. However, there is yet no methylation marker that can be readily translated for use in the screening or triage settings for cervical cancer. In our recent studies, 3 proven or putative tumor suppressor genes were found to be specifically hypermethylated in cervical cancers. Methylation of Protocadherin 10 ( PCDH10 ), WT1 , and PAX1 were found in 86% (24/28), 78% (84/108), and 94% (101/107) of cervical scrapings of cancers and 0% (0/17), 11% (5/45), and 0% (0/41) of that of normal cervix, respectively.
PCDH10 , a member of the cadherin family, is essential for brain development. It has been shown to be a functional tumor suppressor and be methylation-silenced in cancers of the esophagus, stomach, and cervix, and in multiple hematologic malignancies. WT1 , first identified as deleted or rearranged in hereditary cases of Wilms’ tumor, encodes a zinc-finger transcription factor, and plays complex roles in the development and carcinogenesis. It is involved in the development of several organ systems and is both a tumor suppressor and oncogene. PAX1 encodes a transcription factor regulating cellular differentiation, proliferation, migration, and survival, and is expressed during the development of the axial skeleton, the thymus, and the parathyroid glands. Other than our recent findings of a frequently methylation-silencing in cervical and ovarian cancers, the functional role of PAX1 in carcinogenesis is unknown.
We compared the efficacy of testing these 3 new biomarkers with the widely used HPV test in triage of equivocal Pap smear results in a nation-wide, multicenter cohort.
Materials and Methods
Study population
Study subjects were nested to a nation-wide, multicenter cohort of abnormal cervical cytology (the T1899 cohort) conducted by the National Health Research Institute (NHRI), Taiwan. In brief, subjects with abnormal Pap smear findings were enrolled at 11 medical centers around Taiwan, from August 1999 to March 2004. All the studied women received cervical scraping, pelvic examination, and colposcopy with indicated biopsy. The T1899 cohort study was approved by the Joint Institutional Review Board (IRB) of Taiwan and IRB of each participating center, and this ancillary study was approved by IRB of NHRI. Informed consent was obtained from each subject. For a quality control of diagnosis, slides of cytology and histology were reviewed by 2 expert cytologists and pathologists, respectively. In addition, the diagnoses in cytology and pathology were validated with the National Cervical Cytology and Histology Registry and National Cancer Registry of Taiwan. Among 1284 eligible cases, 321 had a Pap smear of ASC-US; 101 of them had a negative colposcopy finding and did not receive cervical biopsy; and the remaining 220 underwent colposcopy-directed biopsy. The results of histology were normal in 120, condyloma or koilocytosis in 11, CIN1 in 47, CIN2 in 17, CIN3 in 14, carcinoma in situ (CIS) in 10 women, and invasive carcinoma in 1 woman ( Table 1 ).
| Age | ||||
|---|---|---|---|---|
| Diagnosis | n | % | Mean ± SE | Median (range) |
| Negative colposcopy, no biopsy | 101 | 31.5 | 47.39 ± 1.17 | 48.0 (19-77) |
| Positive colposcopy with biopsy | 220 | 68.5 | 44.87 ± 0.82 | 44.9 (22-76) |
| Normal | 120 | 54.5 | 44.75 ± 1.08 | 44.0 (22-76) |
| Condyloma or koilocytosis | 11 | 5.0 | 41.64 ± 4.12 | 36.0 (27-67) |
| CIN1 | 47 | 21.4 | 44.43 ± 1.90 | 43.0 (22-76) |
| CIN2 | 17 | 7.7 | 42.82 ± 2.41 | 44.0 (31-69) |
| CIN3 | 14 | 6.4 | 46.00 ± 2.74 | 43.0 (36-71) |
| CIS | 10 | 4.5 | 52.70 ± 4.88 | 55.0 (31-74) |
| Carcinoma | 1 | 0.5 | 56.00 | 56.0 |
| Total | 321 | 100.0 | 45.66 ± 0.68 | 45.0 (19-77) |
Methylation specific-polymerase chain reaction (MSP)
Genomic DNA from cervical scrapings was extracted with QIAmp tissue kit (Qiagen, Hilden, Germany) and was modified with sodium bisulfite using DNA Methylation Gold Kit (Zymo Research, CA). Peripheral blood lymphocyte (PBL) DNA was used as a negative control and PBL DNA treated with Sss1 methylase was used as a positive control. The MSP included an initial incubation at 94°C for 10 minutes, followed by 40 cycle of 94°C for 30 seconds (1 minute for PCDH10 ), annealing at appropriate temperatures for 30 seconds (2 minutes for PCDH10 ), and at extension at 72°C for 30 seconds (1 minute for PCDH10 ). The final extension was done at 72°C for 7 minutes. The primers and conditions for PCR were listed in Table 2 . MSP products were visualized in 2.0% agarose gel containing ethidium bromide and illuminated under ultraviolet (UV) light.
| Gene name | Forward primer | Reverse primer | Annealing temp | Cycle no. | Product size, bp |
|---|---|---|---|---|---|
| PAX1 | M TATTTTGGGTTTGGGGTCGC | CCCGAAAACCGAAAACCG | 64°C | 40 | 153 |
| U GTTTATTTTGGGTTTGGGGTTGTG | CACCCAAAAACCAAAAACCAC | 64°C | 40 | 158 | |
| WT1 | M TGTTGAGTGAATGGAGCGGTC | CGAAAAACCCCCGAATATAAACG | 56°C | 40 | 146 |
| U GTTGTTGAGTGAATGGAGTGGTTG | AATTACAAAAAACCCCCAAATATAAACAC | 61°C | 40 | 153 | |
| PCDH10 | M TCGTTAAATAGATACGTTACGC | TAAAAACTAAAAACTTTCCGCG | 60°C | 40 | 153 |
| U GTTGTTAAATAGATATGTTATGT | CTAAAAACTAAAAACTTTCCACA | 56°C | 40 | 155 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
