Analysis of the Peninsula Health Colposcopy Service Data of One Year Since Transitioning to the New Cervical Screening Program

Article Information

Stephanie Sii*, Amarendra Trivedi, Wei Chun Wang

Peninsula Health, Frankston, Victoria, Australia

*Corresponding Author: Stephanie Sii, Peninsula Health, Frankston Hospital, 2 Hastings Road, Frankston 3199, Melbourne, Victoria, Australia

Received: 11 November 2019; Accepted: 18 November 2019; Published: 19 November 2019

Citation:

Stephanie Sii, Amarendra Trivedi, Wei Chun Wang. Analysis of the Peninsula Health Colposcopy Service Data of One Year Since Transitioning to the New Cervical Screening Program. Obstetrics and Gynecology Research 2 (2019): 088-099.

View / Download Pdf Share at Facebook

Abstract

Aims: We aim to analyze the data from our colposcopy clinic since its transition to the new cervical screening program to determine its effectiveness in clinical practice.

Materials and Methods: We extracted data from hospital electronic records, which was analyzed by a bio-statistician. Associations and concordance between referring (Human Papilloma Virus) HPV types, Liquid Based Cytology (LBC), colposcopic impression and cervical biopsy were calculated and analyzed.

Results: Data from 262 women who attended our colposcopy clinic from December 2017 to November 2018 was analyzed. Our study showed that the volume of referrals had increased by 6-fold since the implementation of the new cervical screening program in Australia. HPV Other (non 16/18) had the highest prevalence among women referred and accounted for the highest number of high-grade referrals to our unit.

Conclusion: 83.3% of referrals with HPV Other (non 16/18) had abnormal cytology. HPV Other appears to be the most pathogenic of all HPV subtypes. The concordance between colposcopic impression and cervical biopsy was 61.9%.

Keywords

Cytology and GYN Pathology, HPV Infection and CIN

Cytology and GYN Pathology articles, HPV Infection and CIN articles

Article Details

1. Introduction

The new risk-based cervical screening program was implemented in Australia in December 2017. It was based on partial genotyping to classify the type of HPV and then reflex LBC on the same sample to determine if a cervical cytological abnormality was present. This was to determine the patient’s risk rating and triaging for colposcopy. This new screening program was supposed to have equivalent or better sensitivity, specificity and positive predictive value to that of the Pap smear. Using HPV as a primary screening tool was expected to reduce the number of false negatives and the reflex LBC testing to reduce the number of false positives, thereby decreasing the number of referrals to the colposcopy clinic and unnecessary follow ups. Comparisons of HPV testing with cytology in two randomized controlled trials (RCT) showed increased detection of Cervical Intraepithelial Neoplasia 2 (CIN2) or worse over subsequent screening rounds, compared with women who were tested with cytology only [1]. HPV 16 or 18 positive had been associated with increased incidence of CIN3 or worse up to 10 years later when cervical abnormalities were left untreated [2].  HPV 16, 18 and 45 accounted for 94% of HPV-related cervical cancers [3]. Studies had shown that HPV 16 and 18 were associated with CIN that were more likely to progress than regress [4-5].

1.1 Aim

The aim of this study was to analyze the data of the colposcopy clinic of Peninsula Health from December 2017 to November 2018 to streamline the clinical practice. The analysis focussed on the prevalence of different types of HPV in the local population, their association with referring LBC  (LBC as mentioned in the referral letter of the referring doctor), concordance between the referring LBC and colposcopic impression, and concordance between the colposcopic impression and histological confirmation of the concurrent cervical biopsy. This colposcopy clinic is the only public colposcopy clinic in the Peninsula area of Melbourne and serves a population of approximately 400, 000. The clinic is held once a week and is run 5 gynaecologists, who take turns supervising registrars allocated to the clinic.

2. Materials and Methods

Data were extracted from the hospital’s electronic medical records and analyzed by the bio-statistician.   Different associations and concordance were calculated and analyzed. The analyses were conducted in the R program using epi.R package [6]. Two by two tables, chi-squared tests, odds ratio with 95% confidence interval were computed from the observed count data and presented.

3. Result

Women referred to our colposcopy clinics had their cervical screening in different laboratories using HPV-type specific DNA tests.  Before December 2017, the average number of new referrals to our colposcopy clinic was 4 per week. Under the new system, there has been a 6-fold increase in this number. A total of 262 women were referred to the colposcopy clinic in the time frame of the study. 34 of 262 women referred to the Peninsula Health colposcopy clinic were under 25 years of age: 4 were aged 18-20 years and 30 were 21-25 years. The new cervical screening program recommends screening from 25 years onwards, but these women had HPV testing earlier either due to risk factors or at clinician discretion. 60 women were aged 26-30 years, 82 aged 31-40 years, 51 aged 41- 50 years, 20 aged 51-60 years, 13 aged 61-70 years, 1 aged 71-80 and 1 aged 81-85 years (see Table 1). HPV Other (non 16/18) accounted for the highest proportion of referrals, 41.2%, followed by HPV 16, 23.6%. Other referred HPV subtypes include HPV 18, 6.1%, HPV 16 & Other, 6.9%, HPV 18 & Other, 3.4%, HPV 16 & 18, 0.76%, HPV 16 &18 & Other, 0.38% and HPV hasn't detected, 0.38%. HPV testing was not done in 17.1% women and their referral was based on LBC (see Table 2). The breakdown of the referring LBC included normal, high grade (HSIL), possible high grade (pHSIL), low grade (LSIL), possible low grade (pLSIL), unsatisfactory and not done. There were 98 high grade or possible high grade LBC in the referrals, 86, low grade or possible low grade LBC, 2 unsatisfactory, 74 negative LBC and 2 where LBC was not done (see Table 3). When evaluating the relationship between the different HPV subtypes and abnormal reflex LBC, HPV Other (non 16/18) appeared to be the most pathogenic of the HPV subtypes. 90 out of 108 women (83.3%) referred to HPV Other (non 16/18) had abnormal LBC. Of these women, 62 had high grade or possible high grade reflex LBC (68.8%), 28 had low grade or possible low grade LBC (31.1%), 17 had normal LBC and 1 was unsatisfactory. A total of 62 out of 98 of all the patients (63.2%) referred to the highest grade or possible high grade reflex LBC were HPV Other (non 16/18) positive. When compared to other HPV types, HPV Other accounted for the highest proportion of abnormal referred LBC, both high and low grades in our study (see Table 4 and 5). Of the 262 women who attended the Peninsula Health colposcopy clinic, 257 underwent colposcopy. Five patients did not undergo colposcopy due to lack of consent or difficulty performing a speculum examination. We evaluated the relationship between referred LBC subtypes and colposcopic impression. 51 women had a high-grade colposcopy and of them, 42 had high-grade or possible high-grade referred LBC, 5 had low-grade or possible low-grade referred LBC, three had negative LBC and one did not have LBC done. 111 women had low-grade abnormal colposcopic impression. Of these women, 45 had low-grade or possible low-grade referred LBC, 47 high-grade or possible high-grade referred LBC, 18 negative referred LBC and one was unsatisfactory. Of the 88 patients with normal impression at colposcopy, 47 had negative referred LBC, 32 low-grade or possible low-grade referred LBC, seven high-grade or possible high-grade referred LBC, one unsatisfactory LBC and 1 in whom LBC was not done (see Table 6).

There was a moderately strong correlation between the referred LBC result and colposcopic impression (Cramer’ s V= 0.29). The odds ratio for a high-grade colposcopic impression with a high-grade referring LBC was 4.24 (95% CI, 2.17- 8.29%, P value <0.001).  There was a positive correlation between low-grade colposcopy impression with a low-grade referring LBC with an odds ratio of 1.87 (95% CI, 1.02- 3.44%, P value 0.042). The odds ratio of having a normal colposcopy with a negative referring LBC is 6.24 (95% CI, 3.47-11.22%, P value <0.001). The odds ratio of possible high- grade referring LBC correlates with a high-grade colposcopy was 4.80 (95% CI, 2.38- 9.67%, P value <0.001). We also evaluated the correlation between referred HPV and colposcopic impression. The odds ratio of HPV 16 correlating with low-grade colposcopy was 2.59 (95% CI, 1.07- 6.26%, P value 0.03) and of HPV Other with high-grade colposcopy was 1.96 (95% CI, 1.01- 3.79%, P value 0.044). The odds ratio of HPV Other and normal colposcopy was 0.47 (95% CI, 0.24-0.91%, P value 0.022). See Table 7. The women aged below 25 were analyzed separately. In our study, HPV Other (non 16/18) had the highest prevalence in these younger women, accounting for 18 in 34 referrals (52%). 19 women had low-grade or possible low-grade referring LBC, twelve had high-grade or possible high-grade referring LBC, two had negative LBC and one had unsatisfactory LBC. Of the 19 women with low-grade referring LBC, four had high-grade changes on concurrent cervical biopsy at colposcopy. Of them, three were positive for HPV Other and underwent Large Loop Excision of the Transformation Zone (LLETZ), all of them confirming CIN2 or higher on histology.  Of the twelve with high grade referring LBC, four had a high-grade proven cervical biopsy at colposcopy. Two of them underwent LLETZ and their histology confirmed CIN 2 or 3. The other two chose to repeat colposcopy in 6 months. This implied that abnormal cytology, if detected, should have close follow up or treatment even in those below 25 years of age. There was no consensus in our unit to repeat colposcopy or perform LLETZ for those with biopsy-proven high-grade changes and the decision seemed to be individualized to clinician choice. Given that 8 of 12 with high-grade referring LBC were low-grade or normal on the initial cervical biopsy, it seemed reasonable to repeat colposcopy in 6 months. Preliminary data from the PRINCess trial suggests there is a 43% likelihood of regression of CIN 2 in women below age of 25 [7]. A study evaluating cervical screening in 4767 women under 25 years old showed that only 63 women (2.5%) had CIN 2-3 confirmed on LLETZ.  This study concluded that screening women under the age of 25 may cause unnecessary referral to colposcopy and subjecting women to anxiety and psychosocial morbidity [8]. Till more evidence is available, the management of high-grade cervical abnormalities for women aged 25 is controversial, largely due to limited data around safety and benefit in this group. A total number of 168 patients were biopsied in our colposcopy clinic. For the purposes of data comparison, we grouped cervicitis, inflammation, HPV effect only, CIN1 and LSIL as low-grade biopsy, whereas CIN 2-3 and HSIL were grouped as high grade biopsy. Of the 51 women with high-grade colposcopic impression, 24 had a high-grade proven biopsy (47%), 21 had low grade, three had normal biopsies and three were not biopsied.  In other words, all women with biopsy proven high-grade abnormal changes had abnormal colposcopic impression (sensitivity 100%). Of the 111 women with low-grade colposcopic impression, 80 women had a low-grade proven biopsy (72%), 20 had a high-grade proven biopsy, 5 had normal biopsies and 6 were not biopsied. Of 88 women with a normal colposcopy impression, 11 women had a low-grade biopsy, 1 had a normal biopsy and 76 women were not biopsied. The number of patients who had the low-grade colposcopy impression with concordant low-grade cervical biopsy result was 80/111 (72%) and 24/51 (47%) in the high-grade group. Together, this correlated to 61.9% concordance (104 in 168 biopsies). The odds ratio of high-grade biopsy with a high-grade colposcopic impression was 8.49 (95% CI, 4.14- 17.39%, P value <0.001) and of low-grade biopsy with the low-grade colposcopic impression was 8.55 (95% CI, 4.88-14.99%, P value <0.001). See Table 8.

Table 1: number of referrals in different age groups.

Age

Number of Referrals

Percentage of referrals

Below 21

4

1.5%

21-25

30

11.4%

26-30

60

22.9%

31-40

82

31.3%

41-50

51

19.4%

51-60

20

7.6%

61-70

13

4.9%

71-80

1

0.3%

81-85

1

0.3%

Total

262

100%

Table 2: number of referrals with different types of HPV.

HPV Types

Number of Referrals

Percentage of Referrals

HPV Other

108

41.2%

HPV 16

62

23.6%

HPV 18

16

6.1%

HPV 16 & Other

18

6.9%

HPV 18 & Other

9

3.4%

HPV 16 & 18

2

0.76%

HPV 16, 18 & Other

1

0.38%

HPV Not Detected

1

0.38%

HPV not done

45

17.1%

Total

262

100%

Referring LBC

Numbers

Percentage of Referrals

High grade (HSIL)

54

20.6%

Possible high grade (pHSIL)

44

17.2%

Low grade (LSIL)

53

20.2%

Possible low grade (pLSIL)

33

12.2%

Unsatisfactory

2

0.8%

Negative

74

28.2%

Not done

2

0.8%

Total

262

100%

Table 3: Number of referrals with different referring LBC.

HPV Subtype

HSIL or pHSIL Numbers

Percentage of Referrals

HPV Other

62

63.2%

HPV 16

9

9.1%

HPV 18

2

2.0%

HPV 16+ Other

6

6.1%

HPV 18 + Other

2

2.0%

Not done

17

17.3%

Total

98

100%

Total

262

100%

Table 4: Number of referrals and High Grade Change (HSIL or pHSIL) by HPV Subtypel.

Table icon

Table 5: Referring HPV and Referring LBC Crosstabulation.

Table icon

Table 6: Referring LBC and Colposcopy Impression.

Table icon

Table 7: Referring HPV and Colposcopy Impression Crosstabulation.

Table icon

(unsatisfactory and glandular have been removed so n=255)

Table 8: Colposcopy Impression and Biopsy Result.

4. Discussion

One of the expectations of the HPV-based cervical screening program was that it would reduce the number of referrals to colposcopy clinics in the long run because of its low false positive and negative rates. However, in the past year since the implementation of the new cervical screening test, there had been a major increase in the average number of referrals to our Peninsula Health colposcopy clinic. The average number of referrals of 4 per week under the previous Pap-smear based screening system rose to 24 per week, a 6 fold rise.  Since there had been no increase in the resources to the colposcopy clinic, it had resulted in increased waiting times for women to be seen in our colposcopy clinic to about 12 months. In the long term, this new system of screening was expected to give major cost savings. However, major fluctuations in rates of colposcopy referrals and follow up were also expected [9]. It is vital that in the short term, data from individual colposcopy clinics is audited to review the impact of these changes, while this new program is still getting established. According to these authors [9], referrals to colposcopy clinics are expected to decline in future, when mostly the incident disease will be detected. The advantage of partial HPV genotyping is that it allows identification of most oncogenic HPV types, which are then referred for colposcopy, not all HPV types. Although HPV 16 and 18 are the commonest oncogenic HPV types and account for 70% of cervical cancers and pre-cancerous cervical lesions, HPV Other (non16/18) also account for a high volume of referrals and a significant cause of disease. In our study, HPV Other (non 16/18) had the most numbers of referrals to our clinic (41.2%), followed by HPV 16 (23.6%). Other large studies around the world also have reported a high prevalence of HPV Other (non 16/18). A study of 38,000 women in China showed that HPV 52 is the most common HPV genotype in patients with cervical pre-cancerous changes, followed by 16, 58, 39, 18, and 56 [10]. In another study, it is estimated that the prevalence of HPV Other (non 16/18) is three times higher than HPV 16/18 [11]. In a Sydney-based study, the incidence of HPV 16 and 18 combined was 2.2% (95% CI, 2.1-2.3%) and the incidence of HPV Other (non 16/18) was more than double, 5.9% (95% CI, 5.8-6.0%) in primary screening [12]. This study also showed that in women who are HPV 16 or 18 positive, 36.4% (95% CI, 34.8-38.1%) had cytological cervical abnormalities; 21.1% (95% CI, 19.7-22.5%) of whom had low grade and 15.3% (95% CI, 14.2-16.6%) had high grade changes. In our study, HPV Other had the highest rate of cytological abnormalities, accounting for 63.2% of all our high-grade referrals, which is higher than the rate by the Sydney study of 35.2% (95% CI, 34.2-36.2%). HPV infection is necessary for the development of cervical neoplasia, but given only a small percentage of women infected with HPV developed HSIL, this indicates that HPV alone is insufficient for cervical cancer oncogenesis. Other co-variables, such as sexual, environmental, behavioural and socioeconomic factors may come into play. The distribution of HPV subtypes varies to some degree according to the geographical region [13]. Furthermore, the oncogenic potential within the same HPV type also varies. Sequential infection with different HPV types and simultaneous infection with more than one HPV types is not an uncommon phenomenon [14]. Infection with one HPV type and simultaneous clearance of another type happen independently.

Our study confirmed that colposcopy alone is not an effective screening tool. The efficacy of colposcopy to detect CIN largely depends upon the experience and training of the colposcopist, which can be highly variable. To increase the sensitivity of colposcopy, biopsies have to be from the abnormal areas of the transformation zone. Random biopsies do not increase sensitivity, however multiple biopsies do [15]. Using 2011 International Federation of Cervical Pathology and Colposcopic criteria, it is possible to distinguish normal cervix from CIN/ cancer with a sensitivity of 86% and specificity of 30% and to distinguish normal cervix/ CIN 1 from HSIL with sensitivity of 61% and specificity of 94% [16]. In our study, the concordance between low-grade colposcopy and biopsy was 72%, which is comparable to other studies. One study showed a 78.5% correlation in the CIN 1 category and 84-88.6% in the CIN 2-3 category [17]. The colposcopy- biopsy concordance in our high-grade abnormal group is a mere 47%, which is lower than quoted other literature. Only 24 of the 48 biopsies in the high-grade colposcopic impression group had high-grade biopsy results, with the others being low-grade or normal. A possible reason for this may be the subjectivity around colposcopic features of HSIL.  Biopsies will be misrepresentative of the pathology if not sampled from the abnormal areas of the transformation zone.  Although trainees work in the clinic under the direct supervision of a gynaecologist, the accuracy of the biopsy site may be affected by the level of experience of the trainee. Our study showed that the volume of referrals had increased by 6-fold since the implementation of the new cervical screening program in Australia. HPV Other (non 16/18) had the highest prevalence among women referred and accounted for the highest number of high-grade referrals to our unit. On the basis of this, it is suggested that women with HPV Other (non 16/18) should be prioritized in the appointment schedule. All colposcopists must be credentialed by the Royal Australian and New Zealand College of Obstetricians and Gynaecologists (RANZCOG) to minimise the subjectivity of recognizing the colposcopic features of different types of abnormalities. Colposcopy clinic capacity should be increased at least 6 fold to keep pace with the increased number of referrals. Moving forward, the number of HPV Other (non16/18) may continue to be on the rise given that women were previously vaccinated with the Cervarix and quadrivalent Gardasil vaccine, which do not cover for the other oncogenic HPV types.

Acknowledgement

We gratefully acknowledge the contribution of Ms. Marea Fitzsimons, senior  colposcopy nurse, who diligently helped with the data collection.

Disclosure

All authors have no conflict of interest.

References

  1. Leinonen MK, Nieminen P, Lonnberg S, Nea Malila, Arun Pokhrel, Pekka Laurila, et al. Detection rates of precancerous and cancerous cervical lesions within one screening round of primary human papillomavirus DNA testing: prospective randomized trial in Finland. BMJ 345 (2012): e7789.
  2. Khan MJ, Astle PE, Lorincz AT, Sholom Wacholder, Mark Sherman, David R. Scott, et al. The elevated 10 year risk of cervical precancer and cancer in women with HPV type 16 or 18 and the possible utility of type specific HPV testing in clinical practice. J. Cancer Inst 97 (2005): 1072-1079.
  3. Pirog EC, Lloveras B, Molijin A, Sara Tous, Núria Guimerà, Maria Alejo, et al. HPV prevalence and genotypes in different histological subtypes of cervical adenocarcinoma, a worldwide analysis of 760 cases. Mod Pathol 27 (2014): 1559-1567.
  4. Castle PE, Schiffman M, Wheeler CM. Evidence of frequent regression of cervical intraepithelial neoplasia-grade 2. Obstet Gynecol 113 (2009): 18-25.
  5. Kjaer SK, Frederiksen K, Munk C. Long term absolute risk of CIN grade 3 or worse following HPV infection: role of persistence. J Natl Cancer Inst 102 (2010): 1478-1488.
  6. Stevenson M, Telmo Nunes, Cord Heuer, Jonathon Marshall, Javier Sanchez, Ron Thornton, et al. epiR: Tools for the Analysis of Epidemiological Data. R package version 0.9-97 (2018). https://CRAN.R-project.org/package=epiR.
  7. Sykes P, Innes C, Harker D, Whitehead M, van der Griend R, Lawton B, et al. Observational Management of CIN2 in Young Women: A Prospective Multicentre Trial. J Low Geni Tract Dis 20 (2016): 343-347.
  8. Al-Kalbani M. Does Cervical Screening in Young Women Aged 20-25 Years Lead to Unnecessary and Harmful Interventions? Asian Pac J Cancer Prev 16 (2015): 6557-6559.
  9. Hall MT, Simms KT, Lew JB, Megan A Smith, Marion Saville, Karen Canfell. Projected future impact of HPV vaccination and primary HPV screening on cervical cancer rates from 2017-2035:example from Australia. PLoS One 13 (2018).
  10. Xu HH, Lin A, Chen YH, Dong SS, Shi WW, Yu JZ, et al. Prevalence characteristics of cervical human papillomavirus (HPV) genotypes in the Taizhou area, China: a cross-sectional study of 37 967 women from the general population. BMJ Open 7 (2017): e014135.
  11. Robadi I, Pharaon M, Ducatman B. Frequency of high-risk human papillomavirus types in a rural Appalachian population. Mod Pathol 29 (2016): 306A.
  12. Machalek DA, Roberts JM, Garland SM, Julia Thurloe, Adele Richards, Ian Chambers, et al. Routine cervical screening by primary HPV testing: early findings in the renewed national cervical screening program. Med J Aus 211 (2019): 113-119.
  13. Mosciki AB, Hills N, Shiboski S, Powell K, Jay N, Hanson E, et al. Risks of incident HPV infection and LSIL development in young females. JAMA 285 (2001): 2995-3002.
  14. Woodman CB, Collins SI, Young LS. The natural history of cervical HPV infection: unresolved issues. Nat Rev Cancer 7 (2007): 11.
  15. Gage JC, Hanson VW, Abbey K, Dippery S, Gardner S, Kubota J, et al. Number of cervical biopsies and sensitivity of colposcopy. Obstet Gynecol 108 (2006): 264-272.
  16. Bornstein J, Bentley J, Bosze P, Girardi F, Haefner H, Menton M, et al. 2011 colposcopic terminology of the International Federation for Cervical Pathology and Colposcopy. Obstet Gynecol 120 (2012): 166-172.
  17. Boicea A, P?tra?cu A, Surlin V, Iliescu D, Schenker M, Chiu?u L. Correlations between Colposcopy and Histologic Results from Colposcopically Directed Biopsy in Cervical Precancerous Lesions. Rom J Morphol Embryol 53 (2012): 735-41.

© 2016-2022, Copyrights Fortune Journals. All Rights Reserved