- Email: [email protected]
Clinical sensitivity of HPV assays for the detection of high grade cervical disease in cervical samples treated with glacial acetic acid
Journal of Clinical Virology 79 (2016) 32–35
Contents lists available at ScienceDirect
Journal of Clinical Virology journal homepage: www.elsevier.com/locate/jcv
Clinical sensitivity of HPV assays for the detection of high grade cervical disease in cervical samples treated with glacial acetic acid Catherine Moore a , Edward Duvall b , Ellen Braby b , Graham Reid c , Eileen Docherty c , Logan Grieve a , Heather Cubie e , Cat Graham d , Kate Cuschieri a,∗ a
Scottish HPV Reference Laboratory, Royal Infirmary of Edinburgh, Edinburgh, UK Department of Pathology, Royal Infirmary of Edinburgh, Edinburgh, UK c Department of Pathology, Queen Elizabeth University Hospital, Glasgow, UK d Epidemiology and Statistics Core, Wellcome Trust Clinical Research Facility, Edinburgh, UK e HPV Research Group, Queens Medical Research Institute, University of Edinburgh, Edinburgh, UK b
a r t i c l e
i n f o
Article history: Received 22 December 2015 Received in revised form 17 March 2016 Accepted 21 March 2016 Keywords: HPV Molecular detection Cervical Glacial acetic acid
a b s t r a c t Background: Lysis of bloody liquid based cytology (LBC) specimens with glacial acetic acid (GAA) is performed to aid cytological interpretation. However, the influence of GAA treatment on HPV detection is not fully understood and in studies designed to assess this, few cases of high-grade disease have been included. Objectives: To assess the sensitivity of HPV molecular tests for the detection of high grade cervical disease in GAA treated samples Study design: A total of 207 specimens associated with high grade dyskaryosis and treated with GAA were collated prospectively. Overall 140 specimens had underlying CIN2+, including 88 CIN3. All specimens were tested with the Abbott RealTime High Risk HPV test (rtHPV) and the Qiagen Hybrid Capture 2High Risk HPV DNA test (HC2). Specimens associated with a CIN2+ that were negative by either assay were genotyped. Results: The sensitivity of rtHPV for CIN2+ and CIN3+ was 92.8% (87.2, 96.5) and 94.3% (87.2, 98.1) respectively. Sensitivity of the HC2 for CIN2+ and CIN3+ was 97.2% (92.8, 99.2) and 96.6% (90.3, 99.2) respectively. The sensitivity of both assays in GAA treated specimens was thus consistent with the level required for clinical application. HPV negative, CIN2+ specimens were generally attributable to HPV types outside the explicit analytical range of the assays. Conclusions: The data indicate that GAA treatment has little impact on the detection of CIN2+ by HPV testing in LBC specimens. © 2016 Published by Elsevier B.V.
1. Background Cervical LBC specimens with high red-blood cell content are often treated with glacial acetic acid (GAA) to lyse erthythrocytes as an aid to cytological interpretation [1]. While the proportion of specimens treated with GAA is difficult to quantify globally due to a lack of consistent reporting, a recent Scotland-wide survey of the cytology laboratories serving the national cervical screening programme showed the proportion of lysed specimens to be 4.8% overall with a range of 1.6–10% (data not shown).
∗ Corresponding author. E-mail address: [email protected] (K. Cuschieri). http://dx.doi.org/10.1016/j.jcv.2016.03.020 1386-6532/© 2016 Published by Elsevier B.V.
As an increasing component of cervical screening and associated disease management relies on HPV testing as a reflex or co-test with cytology [2], it is important to assess whether GAA has an impact on molecular HPV assays and if this impact is differentially exerted across the different assay chemistries and platforms. Of the small number of studies where the impact of GAA has been assessed, the evidence would indicate that the COBAS 4800HPV Assay (Roche Molecular Systems, Pleasanton, CA, USA) and APTIMA HPV Test (Gen-Probe Inc., San Diego, CA) are unaffected by GAA treatment − as opposed to the Cervista Test (Hologic, Inc., Bedford, MA) where it has a deleterious influence [3–6]. Furthermore, in an earlier splitspecimen study of untreated vs treated specimens undertaken in our laboratory, while the impact of GAA on HPV detection was insignificant at the qualitative level, GAA treatment was associated
C. Moore et al. / Journal of Clinical Virology 79 (2016) 32–35
with lower read-outs according to the semi-quantitative measures of the assays chosen: HC2 and rtHPV [7]. An important limitation of the studies described above is that the assessment of GAA influence on HPV testing has largely been determined at the qualitative level in relatively small analytical studies with, crucially, a lack of knowledge of the underlying pathology or, if pathology was known, only a few cases of highgrade disease. For example, in Munson et al. no pathology (cytology or histology) status was provided, whereas in the studies of McMenamin et al. [3,6] and Moore et al. [7] while cytology was reported, only 17/121 (14.0%) and 4/150 (2.6%) cases respectively were associated with high grade cytology—and no histological data was presented. The present article builds on existing data by focussing on the assessment of a key clinical-performance measure of HPV assays (sensitivity for the detection of CIN2+) in GAA treated samples, rather than focussing solely on analytical performance. 2. Objectives To compliment and consolidate existing data, the present study was designed to assess the clinical sensitivity of two clinically validated HPV assays for histologically confirmed, disease in cervical samples treated with GAA. 2.1. Study design 2.1.1. Specimen collection and annotation The two largest cytology laboratories in Scotland which serve the Scottish Cervical Screening Programme (SCSP); Greater Glasgow and Clyde (GGC) and Lothian participated in the study. Both laboratories prospectively collated LBC specimens that (1) indicated “routine” GAA treatment, (2) were graded as high-grade dyskaryosis according to the British Society for Clinical Cytology (BSCC) criteria [8]. Specimens with underlying high-grade dyskaryosis were selected to enrich for CIN or worse (CIN2+) lesions and a total of 214 specimens (ThinPrep® PreservCyt® ) were collated over a 12 month period. In Scotland a result of high-grade dyskaryosis triggers a referral to colposcopy. Histology results associated with these cases were captured via the Scottish Cytology Call Recall System (SCCRS): an integrated IT system that serves the Scottish Cervical Screening Programme (SCSP) and contains cytology, colposcopy and histology information. Of the 214 specimens collected, 7 were excluded from analysis owing to insufficient volume for HPV testing, leaving a total of 207 for assessment. The underlying pathology of the 207 evaluable specimens was as follows −5 (cases of) histology negative, 34 CIN 1 lesions and 140 CIN2+ lesions, which incorporated 78 CIN3 lesions and 10 carcinomas. Histology was not available for 19 cases and no biopsy was taken in a further 9. Permission to deliver the work as a service development project was provided by the South East Scotland Research Ethics Service (NR/1101AB4). 2.1.2. GAA treatment protocol Both cytology laboratories, Greater Glasgow and Clyde (GGC) and Lothian use the Hologic protocol for the processing of ThinPrep Papanicolaou specimens (Hologic, Malborough, MA, USA) with slight variations described previously [7]
33
Table 1 Agreement between HC2 and rtHPV in the overall specimen set (3a), the 140 specimens associated with CIN2+ (3b) and the 88 specimens associated with CIN3+ (3c). 3a
HC2 result
rtHPV result Positive Negative Total
Negative 2 8 10
3b
HC2 result
rtHPV result Positive Negative Total
Negative 1 3 4
3c
HC2 result
rtHPV result Positive Negative Total
Negative 1 2 3
Positive 189 8 197
Total 191 16 207
Positive 129 7 136
Total 130 10 140
Positive 82 3 85
Total 83 5 88
types, whereas the HC2 test is a signal amplification assay which detects 13 high-risk types in aggregate. These assays were those used in the split-sample study described earlier [7].
2.1.4. HPV genotyping Specimens associated with CIN2+ which tested HPV negative by either or both the HC2 or rtHPV tests were also subject to HPV genotyping using the Linear Array HPV Genotyping Test − LA, (Roche Molecular Systems, CA, USA) and the Optiplex HPV Genotyping Assay (Diamex, Heidelberg, Germany). Both genotyping assays were performed according to manufacturers’ instructions. Briefly, the LA assay involves reverse line blot hybridisation of PCR products and can delineate 37HPV types (6, 11, 16, 18, 26, 31, 33, 35, 39, 40, 42, 45, 51, 52, 53, 54, 55, 56, 58, 59, 61, 62, 64, 66, 67, 68, 69, 70, 71, 72, 73 (MM9), 81, 82 (MM4), 83 (MM7), 84 (MM8), IS39 and CP6108) whereas the Optiplex assay uses luminex fluorescent bead array (FBA) technology and detects 24 genotypes: 6,11,16,18, 26, 31, 33, 35, 39, 42, 43, 44, 45, 51, 52, 53, 56, 58, 59, 66, 68, 70, 73, 82.
2.1.5. Analysis The design of the study precluded the capture of a pre-treatment aliquot as cytology status could only be confirmed in real-time after GAA treatment. However, for an HPV assay to perform according to accepted performance standards, sensitivity for CIN3 or worse (CIN3 + ) of at least 90% (+/−) 3% should be achieved [9]. Sensitivity comparisons of the HC2 and rtHPV tests were performed using Fisher’s exact test to assess the non-random associations between two variables. HPV genotype status of HPV negative CIN2+ specimens was analysed descriptively given the small number of cases (n = 11).
3. Results 3.1. HPV positivity and agreement between HC2 and rtHPV
2.1.3. HR-HPV detection The 207 samples with sufficient volume were tested by the HC2 test and rtHPV. To mimic a routine/service testing environment samples were tested only once by each assay. Testing was according to manufacturers’ instructions—briefly, the rtHPV is a target amplification assay designed to detect 14 high-risk types, with identification of HPV 16, HPV 18 and “Other” high-risk HPV
As expected, HPV prevalence was high in this study population with 197/207 (95.2%) and 191/207 (92.3%) testing positive by the HC2 and rtHPV assays respectively. Agreement between the assays was high (Table 1), being 95.2% overall (n = 207), 94.3% in the 140 cases associated with CIN2+ and 95.4% in the 88 cases associated with CIN3+.
34
C. Moore et al. / Journal of Clinical Virology 79 (2016) 32–35
Table 2 Sensitivity of the rtHPV and HC2 assays for the detection of CIN2+ and CIN3+ in GAA treated specimens.
CIN2+ CIN3+
rtHPV +ve/n, sensitivity% (95%CI)
HC2 +ve/n sensitivity% (95% CI)
130/140, 92.8 (87.2, 96.5) 83/88, 94.3 (87.2, 98.1)
136/140, 97.1 (92.8, 99.2) 85/88, 96.6 (90.3, 99.2)
3.2. Sensitivity of the HC2 and rtHPV for detection of CIN2+ Sensitivity for the detection of high grade disease in GAA treated samples by HC2 was 97.1% (95% CI: 92.8, 99.2) for CIN2+ and 96.6% (95% CI: 90.3, 99.2) for CIN3+. Comparatively, the sensitivity of the rtHPV for CIN2+ and CIN3+ was 92.8% (95% CI 87.2, 96.5) and 94.3% (95% CI 87.2, 98.1) respectively (Table 2). No statistically significant difference in assay sensitivity was observed between the HC2 and rtHPV for CIN2+ (p = 0.168) and CIN3+ (p = 0.469) although HC2 picked up 6 more cases of CIN2+. Of the 10 specimens associated with invasive carcinoma, 8/10 were detected by the HC2 assay and 9/10 by the rtHPV test; of the two HPV negative cancer cases, 1 case was negative by both assays and 1 by the HC2 only. 3.3. HPV genotyping of specimens associated with CIN2+ negative by HC2 and/or rtHPV assay Genotyping results associated with the 11 CIN2+ specimens that tested negative by the rtHPV and/or HC2 are shown in Table 3. A total of 3/11 specimens were HPV negative by both genotyping assays (one of which had been negative by both the HC2 and rtHPV test) and 6/11 were positive exclusively for types outside the explicit range of the assays (HPV 53, HPV 73 or HPV 82). The remaining 2 specimens were HPV positive for multiple types, both outside and within the ranges of the assays, one (ID 02) contained HPV 16 and HPV 73 according to both assays and HPV 51 according to LA, while the other (ID 10) contained HPV 82 according to both genotyping assays and HPV 16 according to the Optiplex test. 4. Discussion While lysis of bloody specimens with GAA to aid cytological interpretation is a widespread practice, there are few studies which assess effect of this procedure on the performance of HPV assays, particularly in specimen sets where histological confirmation of disease (CIN2+) has been obtained. Exclusion of GAA treated specimens from further molecular testing, or specific manipulation of these specimens to allow testing presents logistical challenges, so further data that informs their suitability for HPV testing is warranted. This study indicates that the sensitivity of HPV testing in GAA treated specimens by two clinically validated HPV assays exceeds the sensitivity threshold required for routine testing [9]. In a previous study of 1366 women in Scotland attending for routine colposcopy services, sensitivity of the HC2 for CIN2+ was 93.9% (95%
CI: 91.9–96.1) and in another UK-based colposcopy study (Predictors 2) the sensitivity of the rtHPV assay for CIN2+ was 93.3% (95% CI: 90.1–95.6) [10,11]. These values are in line with the sensitivities of the HC2 and rtHPV observed in the present analysis, which were 97.1% (92.8, 99.2) and 92.8% (87.2, 96.5) respectively. There are caveats to the analysis; there are protocols for GAA treatment other than those described above. Another limitation of this study was that it was not a ‘split specimen’ study, as knowledge of pathology status was only available after cytology and associated biopsy. However, an earlier split-specimen study undertaken in our laboratory which assessed analytical, rather than clinical, performance again showed that GAA did not affect qualitative HPV detection by rtHPV and HC2 and the present study was designed to be complimentary to those earlier data [7]. Finally only two assays were assessed in this study while an increasing number of clinically validated tests are now available [12]. A total of 11 cases of CIN2+ were not detected by either the rtHPV and/or the HC2, specifically, 4 were negative by HC2 and 10 by rtHPV. It was of interest that 6/11 were infected with “possibly carcinogenic” [13] types outside the defined range of the assays and no specimen was exclusively infected by a high-risk HPV type within the defined range of the assays. Given the relatively narrow study inclusion criteria (high grade dyskaryosis and GAA treated) it is feasible that the CIN2+ attributable to infection with relatively rare types should be found in a collection of specimens which spanned a 1 year collection time frame. It is also tempting to speculate that the observed differences between the assays for the detection of CIN2+ may be attributable to differences in assay cross-reactivity with the HC2 having a higher propensity for cross-reactivity (as has been reported elsewhere) [14]. In conclusion, sensitivity of the HC2 and rtHPV for detection of CIN2+ and CIN3+ in GAA treated specimens was comparable to sensitivity measurements derived from previous studies of colposcopy referral populations and was above the threshold required for clinical application. No significant difference in assay sensitivity was observed between the HC2 and rtHPV for CIN2+ although HC2 detected more CIN2+ associated with cross-reactive types. These data indicate that GAA treated specimens are suitable for clinical HPV testing by the assays described above; however further data on the performance of other assays in this context would be welcome. Funding The project was funded as part of a service development/QC project within NHS Lothian. Contributions CM was principal investigator of the study, EDu, EB, GR and EDo, coordinated the collation of pathology specimens and data and contributed to manuscript versions, LG supported with delivery of HPV
Table 3 Type specific results of 11 CIN2+ specimens which tested negative by HC2 and/or rtHPV. Specimen ID
Linear Array result
Optiplex result
Histology
HC2 result
rtHPV result
01 02 03 04 05 06 07 08 09 10 11
No type detected 16, 51, 73 No type detected 70 53, 73 No type detected 82 82 82 82 73
No type detected 16,73 No type detected 70 53, 73 No type detected 82 82 82 16,82 73
CIN2 Cancer CIN3 CIN2 Cancer CIN3 CIN2 CIN2 CIN2 CIN3 CIN3
Neg Neg Pos Pos Neg Pos Pos Pos Pos Pos Neg
Neg Pos Neg Neg Neg Neg Neg Neg Neg Neg Neg
C. Moore et al. / Journal of Clinical Virology 79 (2016) 32–35
testing and database creation, HAC supported with original concept of interference and with manuscript drafts, CG performed statistical analysis and KC advised on study design and supported with manuscript drafts. Competing interests CM, EDu, EB, GR, EDo, LG, HAC and CG have no conflict of interest to declare. KC has received institutional project funding and/or consumables to carry out assay evaluations from HOLOGIC, Qiagen, Roche, NorChip, Cepheid, Becton Dickinson, GSK & Abbott. Acknowledgement We thank staff at the Pathology Departments, Royal Infirmary of Edinburgh, Edinburgh and Queen Elizabeth University Hospital, Glasgow for support in the collection of specimens. References [1] S.N. Agoff, T. Dean, B.K. Nixon, K. Ingalls-Severn, L. Rinker, V.S. Grieco, The efficacy of reprocessing unsatisfactory cervicovaginal thinprep specimens with and without glacial acetic acid: effect of hybrid capture II human papillomavirus testing and clinical follow-up, Am. J. Clin. Pathol. 2002 (118) (2002) 727–732. [2] M. Arbyn, G. Ronco, A. Anttila, C.J. Meijer, M. Poljak, G. Ogilvie, G. Koliopoulos, P. Naucler, R. Sankaranarayanan, J. Peto, Evidence regarding human papillomavirus testing in secondary prevention of cervical cancer, Vaccine 30 (Suppl. 5) (2012) F88–F89. [3] M. McMenamin, M. McKenna, Effect of glacial acetic acid treatment of cervical ThinPrep specimens on HPV DNA detection with the cobas 4800HPV test, Cytopathology 24 (2013) 321–326. [4] E. Munson, E.R. Schroeder, K.C. Ross, C. Yauck, T. Bieganski, R.D. Amrhein, M. Napierala, A.L. Harkins, Effect of preanalytical processing of ThinPrep specimens on detection of high-risk human papillomavirus by the Aptima HPV assay, J. Clin. Microbiol. 52 (2014) 1448–1452.
35
[5] E. Munson, B.K. Du Chateau, B.E. Nelson, J. Griep, J. Czarnecka, R.D. Amrhein, E.R. Schroeder, Effect of glacial acetic acid treatment of liquid-based cytology collections on performance of cervista HPV HR for detection of high-risk human papillomavirus, J. Clin. Microbiol. 50 (2012) 2129–2131. [6] M. McMenamin, M. McKenna, Stability of human papillomavirus (HPV) in cervical ThinPrep specimens previously lysed with glacial acetic acid: effect on cobas 4800HPV test performance, Cancer Cytopathol. 122 (2014) 250–256. [7] C. Moore, K. Cuschieri, F. McQueen, E. Duvall, C. Graham, H.A. Cubie, Effect of glacial acetic acid pre-treatment of cervical liquid-based cytology specimens on the molecular detection of human papillomavirus, Cytopathology 24 (2013) 314–320. [8] http://www.cancerscreening.nhs.uk/cervical/publications/nhscsp01.html; . [9] M.H. Stoler, P.E. Castle, D. Solomon, M. Schiffman, American Society for Colposcopy and Cervical Pathology. The expanded use of HPV testing in gynecologic practice per ASCCP-guided management requires the use of well-validated assays, Am. J. Clin. Pathol. 127 (2007) 335–337. [10] K. Cuschieri, H. Cubie, C. Graham, J. Rowan, A. Hardie, A. Horne, C.B. Earle, A. Bailey, E.J. Crosbie, H. Kitchener, Clinical performance of RNA and DNA based HPV testing in a colposcopy setting: influence of assay target, cut off and age, J. Clin. Virol. 59 (2014) 104–108. [11] A. Szarewski, D. Mesher, L. Cadman, J. Austin, L. Ashdown-Barr, L. Ho, G. Terry, S. Liddle, M. Young, M. Stoler, J. McCarthy, C. Wright, C. Bergeron, W.P. Soutter, D. Lyons, J. Cuzick, Comparison of seven tests for high-grade cervical intraepithelial neoplasia in women with abnormal smears: the predictors 2 study, J. Clin. Microbiol. 50 (2012) 1867–1873. [12] M. Arbyn, P.J. Snijders, C.J. Meijer, H. Berkhof, K. Cuschieri, B.J. Kocjan, M. Poljak, Which high-risk HPV assays fulfil criteria for use in primary cervical cancer screening? Clin. Microbiol. Infect. (2015) 00426–00427, pii: S1198-743X(15). [13] V. Bouvard, R. Baan, K. Straif, Y. Grosse, B. Secretan, F. El Ghissassi, L. Benbrahim-Tallaa, N. Guha, C. Freeman, L. Galichet, V. Cogliano, W.H.O. International Agency for Research on Cancer Monograph Working Group, A review of human carcinogens-Part B: biological agents, Lancet Oncol. 10 (2009) 321–322. [14] M. Poljak, A. Kovanda, B.J. Kocjan, K. Seme, N. Jancar, E. Vrtacnik-Bokal, The Abbott RealTime High Risk HPV test: comparative evaluation of analytical specificity and clinical sensitivity for cervical carcinoma and CIN 3 lesions with the Hybrid Capture 2HPV DNA test, Acta Dermatovenerol Alp Pannonica Adriat 18 (2009) 94–103.
Contents lists available at ScienceDirect
Journal of Clinical Virology journal homepage: www.elsevier.com/locate/jcv
Clinical sensitivity of HPV assays for the detection of high grade cervical disease in cervical samples treated with glacial acetic acid Catherine Moore a , Edward Duvall b , Ellen Braby b , Graham Reid c , Eileen Docherty c , Logan Grieve a , Heather Cubie e , Cat Graham d , Kate Cuschieri a,∗ a
Scottish HPV Reference Laboratory, Royal Infirmary of Edinburgh, Edinburgh, UK Department of Pathology, Royal Infirmary of Edinburgh, Edinburgh, UK c Department of Pathology, Queen Elizabeth University Hospital, Glasgow, UK d Epidemiology and Statistics Core, Wellcome Trust Clinical Research Facility, Edinburgh, UK e HPV Research Group, Queens Medical Research Institute, University of Edinburgh, Edinburgh, UK b
a r t i c l e
i n f o
Article history: Received 22 December 2015 Received in revised form 17 March 2016 Accepted 21 March 2016 Keywords: HPV Molecular detection Cervical Glacial acetic acid
a b s t r a c t Background: Lysis of bloody liquid based cytology (LBC) specimens with glacial acetic acid (GAA) is performed to aid cytological interpretation. However, the influence of GAA treatment on HPV detection is not fully understood and in studies designed to assess this, few cases of high-grade disease have been included. Objectives: To assess the sensitivity of HPV molecular tests for the detection of high grade cervical disease in GAA treated samples Study design: A total of 207 specimens associated with high grade dyskaryosis and treated with GAA were collated prospectively. Overall 140 specimens had underlying CIN2+, including 88 CIN3. All specimens were tested with the Abbott RealTime High Risk HPV test (rtHPV) and the Qiagen Hybrid Capture 2High Risk HPV DNA test (HC2). Specimens associated with a CIN2+ that were negative by either assay were genotyped. Results: The sensitivity of rtHPV for CIN2+ and CIN3+ was 92.8% (87.2, 96.5) and 94.3% (87.2, 98.1) respectively. Sensitivity of the HC2 for CIN2+ and CIN3+ was 97.2% (92.8, 99.2) and 96.6% (90.3, 99.2) respectively. The sensitivity of both assays in GAA treated specimens was thus consistent with the level required for clinical application. HPV negative, CIN2+ specimens were generally attributable to HPV types outside the explicit analytical range of the assays. Conclusions: The data indicate that GAA treatment has little impact on the detection of CIN2+ by HPV testing in LBC specimens. © 2016 Published by Elsevier B.V.
1. Background Cervical LBC specimens with high red-blood cell content are often treated with glacial acetic acid (GAA) to lyse erthythrocytes as an aid to cytological interpretation [1]. While the proportion of specimens treated with GAA is difficult to quantify globally due to a lack of consistent reporting, a recent Scotland-wide survey of the cytology laboratories serving the national cervical screening programme showed the proportion of lysed specimens to be 4.8% overall with a range of 1.6–10% (data not shown).
∗ Corresponding author. E-mail address: [email protected] (K. Cuschieri). http://dx.doi.org/10.1016/j.jcv.2016.03.020 1386-6532/© 2016 Published by Elsevier B.V.
As an increasing component of cervical screening and associated disease management relies on HPV testing as a reflex or co-test with cytology [2], it is important to assess whether GAA has an impact on molecular HPV assays and if this impact is differentially exerted across the different assay chemistries and platforms. Of the small number of studies where the impact of GAA has been assessed, the evidence would indicate that the COBAS 4800HPV Assay (Roche Molecular Systems, Pleasanton, CA, USA) and APTIMA HPV Test (Gen-Probe Inc., San Diego, CA) are unaffected by GAA treatment − as opposed to the Cervista Test (Hologic, Inc., Bedford, MA) where it has a deleterious influence [3–6]. Furthermore, in an earlier splitspecimen study of untreated vs treated specimens undertaken in our laboratory, while the impact of GAA on HPV detection was insignificant at the qualitative level, GAA treatment was associated
C. Moore et al. / Journal of Clinical Virology 79 (2016) 32–35
with lower read-outs according to the semi-quantitative measures of the assays chosen: HC2 and rtHPV [7]. An important limitation of the studies described above is that the assessment of GAA influence on HPV testing has largely been determined at the qualitative level in relatively small analytical studies with, crucially, a lack of knowledge of the underlying pathology or, if pathology was known, only a few cases of highgrade disease. For example, in Munson et al. no pathology (cytology or histology) status was provided, whereas in the studies of McMenamin et al. [3,6] and Moore et al. [7] while cytology was reported, only 17/121 (14.0%) and 4/150 (2.6%) cases respectively were associated with high grade cytology—and no histological data was presented. The present article builds on existing data by focussing on the assessment of a key clinical-performance measure of HPV assays (sensitivity for the detection of CIN2+) in GAA treated samples, rather than focussing solely on analytical performance. 2. Objectives To compliment and consolidate existing data, the present study was designed to assess the clinical sensitivity of two clinically validated HPV assays for histologically confirmed, disease in cervical samples treated with GAA. 2.1. Study design 2.1.1. Specimen collection and annotation The two largest cytology laboratories in Scotland which serve the Scottish Cervical Screening Programme (SCSP); Greater Glasgow and Clyde (GGC) and Lothian participated in the study. Both laboratories prospectively collated LBC specimens that (1) indicated “routine” GAA treatment, (2) were graded as high-grade dyskaryosis according to the British Society for Clinical Cytology (BSCC) criteria [8]. Specimens with underlying high-grade dyskaryosis were selected to enrich for CIN or worse (CIN2+) lesions and a total of 214 specimens (ThinPrep® PreservCyt® ) were collated over a 12 month period. In Scotland a result of high-grade dyskaryosis triggers a referral to colposcopy. Histology results associated with these cases were captured via the Scottish Cytology Call Recall System (SCCRS): an integrated IT system that serves the Scottish Cervical Screening Programme (SCSP) and contains cytology, colposcopy and histology information. Of the 214 specimens collected, 7 were excluded from analysis owing to insufficient volume for HPV testing, leaving a total of 207 for assessment. The underlying pathology of the 207 evaluable specimens was as follows −5 (cases of) histology negative, 34 CIN 1 lesions and 140 CIN2+ lesions, which incorporated 78 CIN3 lesions and 10 carcinomas. Histology was not available for 19 cases and no biopsy was taken in a further 9. Permission to deliver the work as a service development project was provided by the South East Scotland Research Ethics Service (NR/1101AB4). 2.1.2. GAA treatment protocol Both cytology laboratories, Greater Glasgow and Clyde (GGC) and Lothian use the Hologic protocol for the processing of ThinPrep Papanicolaou specimens (Hologic, Malborough, MA, USA) with slight variations described previously [7]
33
Table 1 Agreement between HC2 and rtHPV in the overall specimen set (3a), the 140 specimens associated with CIN2+ (3b) and the 88 specimens associated with CIN3+ (3c). 3a
HC2 result
rtHPV result Positive Negative Total
Negative 2 8 10
3b
HC2 result
rtHPV result Positive Negative Total
Negative 1 3 4
3c
HC2 result
rtHPV result Positive Negative Total
Negative 1 2 3
Positive 189 8 197
Total 191 16 207
Positive 129 7 136
Total 130 10 140
Positive 82 3 85
Total 83 5 88
types, whereas the HC2 test is a signal amplification assay which detects 13 high-risk types in aggregate. These assays were those used in the split-sample study described earlier [7].
2.1.4. HPV genotyping Specimens associated with CIN2+ which tested HPV negative by either or both the HC2 or rtHPV tests were also subject to HPV genotyping using the Linear Array HPV Genotyping Test − LA, (Roche Molecular Systems, CA, USA) and the Optiplex HPV Genotyping Assay (Diamex, Heidelberg, Germany). Both genotyping assays were performed according to manufacturers’ instructions. Briefly, the LA assay involves reverse line blot hybridisation of PCR products and can delineate 37HPV types (6, 11, 16, 18, 26, 31, 33, 35, 39, 40, 42, 45, 51, 52, 53, 54, 55, 56, 58, 59, 61, 62, 64, 66, 67, 68, 69, 70, 71, 72, 73 (MM9), 81, 82 (MM4), 83 (MM7), 84 (MM8), IS39 and CP6108) whereas the Optiplex assay uses luminex fluorescent bead array (FBA) technology and detects 24 genotypes: 6,11,16,18, 26, 31, 33, 35, 39, 42, 43, 44, 45, 51, 52, 53, 56, 58, 59, 66, 68, 70, 73, 82.
2.1.5. Analysis The design of the study precluded the capture of a pre-treatment aliquot as cytology status could only be confirmed in real-time after GAA treatment. However, for an HPV assay to perform according to accepted performance standards, sensitivity for CIN3 or worse (CIN3 + ) of at least 90% (+/−) 3% should be achieved [9]. Sensitivity comparisons of the HC2 and rtHPV tests were performed using Fisher’s exact test to assess the non-random associations between two variables. HPV genotype status of HPV negative CIN2+ specimens was analysed descriptively given the small number of cases (n = 11).
3. Results 3.1. HPV positivity and agreement between HC2 and rtHPV
2.1.3. HR-HPV detection The 207 samples with sufficient volume were tested by the HC2 test and rtHPV. To mimic a routine/service testing environment samples were tested only once by each assay. Testing was according to manufacturers’ instructions—briefly, the rtHPV is a target amplification assay designed to detect 14 high-risk types, with identification of HPV 16, HPV 18 and “Other” high-risk HPV
As expected, HPV prevalence was high in this study population with 197/207 (95.2%) and 191/207 (92.3%) testing positive by the HC2 and rtHPV assays respectively. Agreement between the assays was high (Table 1), being 95.2% overall (n = 207), 94.3% in the 140 cases associated with CIN2+ and 95.4% in the 88 cases associated with CIN3+.
34
C. Moore et al. / Journal of Clinical Virology 79 (2016) 32–35
Table 2 Sensitivity of the rtHPV and HC2 assays for the detection of CIN2+ and CIN3+ in GAA treated specimens.
CIN2+ CIN3+
rtHPV +ve/n, sensitivity% (95%CI)
HC2 +ve/n sensitivity% (95% CI)
130/140, 92.8 (87.2, 96.5) 83/88, 94.3 (87.2, 98.1)
136/140, 97.1 (92.8, 99.2) 85/88, 96.6 (90.3, 99.2)
3.2. Sensitivity of the HC2 and rtHPV for detection of CIN2+ Sensitivity for the detection of high grade disease in GAA treated samples by HC2 was 97.1% (95% CI: 92.8, 99.2) for CIN2+ and 96.6% (95% CI: 90.3, 99.2) for CIN3+. Comparatively, the sensitivity of the rtHPV for CIN2+ and CIN3+ was 92.8% (95% CI 87.2, 96.5) and 94.3% (95% CI 87.2, 98.1) respectively (Table 2). No statistically significant difference in assay sensitivity was observed between the HC2 and rtHPV for CIN2+ (p = 0.168) and CIN3+ (p = 0.469) although HC2 picked up 6 more cases of CIN2+. Of the 10 specimens associated with invasive carcinoma, 8/10 were detected by the HC2 assay and 9/10 by the rtHPV test; of the two HPV negative cancer cases, 1 case was negative by both assays and 1 by the HC2 only. 3.3. HPV genotyping of specimens associated with CIN2+ negative by HC2 and/or rtHPV assay Genotyping results associated with the 11 CIN2+ specimens that tested negative by the rtHPV and/or HC2 are shown in Table 3. A total of 3/11 specimens were HPV negative by both genotyping assays (one of which had been negative by both the HC2 and rtHPV test) and 6/11 were positive exclusively for types outside the explicit range of the assays (HPV 53, HPV 73 or HPV 82). The remaining 2 specimens were HPV positive for multiple types, both outside and within the ranges of the assays, one (ID 02) contained HPV 16 and HPV 73 according to both assays and HPV 51 according to LA, while the other (ID 10) contained HPV 82 according to both genotyping assays and HPV 16 according to the Optiplex test. 4. Discussion While lysis of bloody specimens with GAA to aid cytological interpretation is a widespread practice, there are few studies which assess effect of this procedure on the performance of HPV assays, particularly in specimen sets where histological confirmation of disease (CIN2+) has been obtained. Exclusion of GAA treated specimens from further molecular testing, or specific manipulation of these specimens to allow testing presents logistical challenges, so further data that informs their suitability for HPV testing is warranted. This study indicates that the sensitivity of HPV testing in GAA treated specimens by two clinically validated HPV assays exceeds the sensitivity threshold required for routine testing [9]. In a previous study of 1366 women in Scotland attending for routine colposcopy services, sensitivity of the HC2 for CIN2+ was 93.9% (95%
CI: 91.9–96.1) and in another UK-based colposcopy study (Predictors 2) the sensitivity of the rtHPV assay for CIN2+ was 93.3% (95% CI: 90.1–95.6) [10,11]. These values are in line with the sensitivities of the HC2 and rtHPV observed in the present analysis, which were 97.1% (92.8, 99.2) and 92.8% (87.2, 96.5) respectively. There are caveats to the analysis; there are protocols for GAA treatment other than those described above. Another limitation of this study was that it was not a ‘split specimen’ study, as knowledge of pathology status was only available after cytology and associated biopsy. However, an earlier split-specimen study undertaken in our laboratory which assessed analytical, rather than clinical, performance again showed that GAA did not affect qualitative HPV detection by rtHPV and HC2 and the present study was designed to be complimentary to those earlier data [7]. Finally only two assays were assessed in this study while an increasing number of clinically validated tests are now available [12]. A total of 11 cases of CIN2+ were not detected by either the rtHPV and/or the HC2, specifically, 4 were negative by HC2 and 10 by rtHPV. It was of interest that 6/11 were infected with “possibly carcinogenic” [13] types outside the defined range of the assays and no specimen was exclusively infected by a high-risk HPV type within the defined range of the assays. Given the relatively narrow study inclusion criteria (high grade dyskaryosis and GAA treated) it is feasible that the CIN2+ attributable to infection with relatively rare types should be found in a collection of specimens which spanned a 1 year collection time frame. It is also tempting to speculate that the observed differences between the assays for the detection of CIN2+ may be attributable to differences in assay cross-reactivity with the HC2 having a higher propensity for cross-reactivity (as has been reported elsewhere) [14]. In conclusion, sensitivity of the HC2 and rtHPV for detection of CIN2+ and CIN3+ in GAA treated specimens was comparable to sensitivity measurements derived from previous studies of colposcopy referral populations and was above the threshold required for clinical application. No significant difference in assay sensitivity was observed between the HC2 and rtHPV for CIN2+ although HC2 detected more CIN2+ associated with cross-reactive types. These data indicate that GAA treated specimens are suitable for clinical HPV testing by the assays described above; however further data on the performance of other assays in this context would be welcome. Funding The project was funded as part of a service development/QC project within NHS Lothian. Contributions CM was principal investigator of the study, EDu, EB, GR and EDo, coordinated the collation of pathology specimens and data and contributed to manuscript versions, LG supported with delivery of HPV
Table 3 Type specific results of 11 CIN2+ specimens which tested negative by HC2 and/or rtHPV. Specimen ID
Linear Array result
Optiplex result
Histology
HC2 result
rtHPV result
01 02 03 04 05 06 07 08 09 10 11
No type detected 16, 51, 73 No type detected 70 53, 73 No type detected 82 82 82 82 73
No type detected 16,73 No type detected 70 53, 73 No type detected 82 82 82 16,82 73
CIN2 Cancer CIN3 CIN2 Cancer CIN3 CIN2 CIN2 CIN2 CIN3 CIN3
Neg Neg Pos Pos Neg Pos Pos Pos Pos Pos Neg
Neg Pos Neg Neg Neg Neg Neg Neg Neg Neg Neg
C. Moore et al. / Journal of Clinical Virology 79 (2016) 32–35
testing and database creation, HAC supported with original concept of interference and with manuscript drafts, CG performed statistical analysis and KC advised on study design and supported with manuscript drafts. Competing interests CM, EDu, EB, GR, EDo, LG, HAC and CG have no conflict of interest to declare. KC has received institutional project funding and/or consumables to carry out assay evaluations from HOLOGIC, Qiagen, Roche, NorChip, Cepheid, Becton Dickinson, GSK & Abbott. Acknowledgement We thank staff at the Pathology Departments, Royal Infirmary of Edinburgh, Edinburgh and Queen Elizabeth University Hospital, Glasgow for support in the collection of specimens. References [1] S.N. Agoff, T. Dean, B.K. Nixon, K. Ingalls-Severn, L. Rinker, V.S. Grieco, The efficacy of reprocessing unsatisfactory cervicovaginal thinprep specimens with and without glacial acetic acid: effect of hybrid capture II human papillomavirus testing and clinical follow-up, Am. J. Clin. Pathol. 2002 (118) (2002) 727–732. [2] M. Arbyn, G. Ronco, A. Anttila, C.J. Meijer, M. Poljak, G. Ogilvie, G. Koliopoulos, P. Naucler, R. Sankaranarayanan, J. Peto, Evidence regarding human papillomavirus testing in secondary prevention of cervical cancer, Vaccine 30 (Suppl. 5) (2012) F88–F89. [3] M. McMenamin, M. McKenna, Effect of glacial acetic acid treatment of cervical ThinPrep specimens on HPV DNA detection with the cobas 4800HPV test, Cytopathology 24 (2013) 321–326. [4] E. Munson, E.R. Schroeder, K.C. Ross, C. Yauck, T. Bieganski, R.D. Amrhein, M. Napierala, A.L. Harkins, Effect of preanalytical processing of ThinPrep specimens on detection of high-risk human papillomavirus by the Aptima HPV assay, J. Clin. Microbiol. 52 (2014) 1448–1452.
35
[5] E. Munson, B.K. Du Chateau, B.E. Nelson, J. Griep, J. Czarnecka, R.D. Amrhein, E.R. Schroeder, Effect of glacial acetic acid treatment of liquid-based cytology collections on performance of cervista HPV HR for detection of high-risk human papillomavirus, J. Clin. Microbiol. 50 (2012) 2129–2131. [6] M. McMenamin, M. McKenna, Stability of human papillomavirus (HPV) in cervical ThinPrep specimens previously lysed with glacial acetic acid: effect on cobas 4800HPV test performance, Cancer Cytopathol. 122 (2014) 250–256. [7] C. Moore, K. Cuschieri, F. McQueen, E. Duvall, C. Graham, H.A. Cubie, Effect of glacial acetic acid pre-treatment of cervical liquid-based cytology specimens on the molecular detection of human papillomavirus, Cytopathology 24 (2013) 314–320. [8] http://www.cancerscreening.nhs.uk/cervical/publications/nhscsp01.html; . [9] M.H. Stoler, P.E. Castle, D. Solomon, M. Schiffman, American Society for Colposcopy and Cervical Pathology. The expanded use of HPV testing in gynecologic practice per ASCCP-guided management requires the use of well-validated assays, Am. J. Clin. Pathol. 127 (2007) 335–337. [10] K. Cuschieri, H. Cubie, C. Graham, J. Rowan, A. Hardie, A. Horne, C.B. Earle, A. Bailey, E.J. Crosbie, H. Kitchener, Clinical performance of RNA and DNA based HPV testing in a colposcopy setting: influence of assay target, cut off and age, J. Clin. Virol. 59 (2014) 104–108. [11] A. Szarewski, D. Mesher, L. Cadman, J. Austin, L. Ashdown-Barr, L. Ho, G. Terry, S. Liddle, M. Young, M. Stoler, J. McCarthy, C. Wright, C. Bergeron, W.P. Soutter, D. Lyons, J. Cuzick, Comparison of seven tests for high-grade cervical intraepithelial neoplasia in women with abnormal smears: the predictors 2 study, J. Clin. Microbiol. 50 (2012) 1867–1873. [12] M. Arbyn, P.J. Snijders, C.J. Meijer, H. Berkhof, K. Cuschieri, B.J. Kocjan, M. Poljak, Which high-risk HPV assays fulfil criteria for use in primary cervical cancer screening? Clin. Microbiol. Infect. (2015) 00426–00427, pii: S1198-743X(15). [13] V. Bouvard, R. Baan, K. Straif, Y. Grosse, B. Secretan, F. El Ghissassi, L. Benbrahim-Tallaa, N. Guha, C. Freeman, L. Galichet, V. Cogliano, W.H.O. International Agency for Research on Cancer Monograph Working Group, A review of human carcinogens-Part B: biological agents, Lancet Oncol. 10 (2009) 321–322. [14] M. Poljak, A. Kovanda, B.J. Kocjan, K. Seme, N. Jancar, E. Vrtacnik-Bokal, The Abbott RealTime High Risk HPV test: comparative evaluation of analytical specificity and clinical sensitivity for cervical carcinoma and CIN 3 lesions with the Hybrid Capture 2HPV DNA test, Acta Dermatovenerol Alp Pannonica Adriat 18 (2009) 94–103.