Abstract
Background
Valid diagnostic tests for human papillomavirus (HPV) detection are crucial to identify individuals at high risk of cervical cancer. We assessed and compared the validity of Mehrviru HPV genotyping and Sacace (HPV Genotypes 14 Real-TM Quant) for molecular detection of 14 high-risk human papillomaviruses.
Methods
We used three HPV test results to identify HPV-positive individuals (14 high-risk genotypes) in a specialist gynecology clinic. The HPV test results were collected using Mehrviru®, Sacace®, and a third kit from the clinical diagnostic laboratory. We used Latent class analysis to determine the actual status of HPV infection in study participants. The sensitivity, specificity, positive and negative predictive values, positive and negative likelihood ratios, Youden, and area under the ROC curve indices of each diagnostic kit and their 95% confidence intervals were calculated. The agreement between the Mehrviru and Sacace kits was determined using the Kappa statistic.
Results
We examined 117 women at high risk of HPV infection. The mean age (SD) was 37.2 (9.1). According to LCA, 28.6% of participants had an HPV infection. The sensitivity and specificity (95% CI) of the Mehrviru were 90.8% (73.7-97.2%) and 90.9% (82.1-95.6%), and corresponding figures for Sacace were 92.0% (72.3-98.1%) and 97.4% (90.2-99.3%). The kappa index between the Mehrviru and Sacace kits was 69.7% (55.6-83.9%). The area under the ROC curve for Mehrviru and Sacace test were 91.4% (85.7-97.1%) and 94.4% (89.3-99.5%).
Conclusions
There is an excellent agreement between Mehrviru and Sacace test results, and the diagnostic accuracy indices were similar.
Supplementary Information
The online version contains supplementary material available at 10.1186/s12879-024-10121-9.
Keywords: HPV, Sensitivity, Specificity, Latent class analysis
Background
Human papillomavirus is known as one of the important causes of cervical cancer and the most common sexually transmitted infection worldwide [1, 2]. In 2020, an estimated 604,000 new cases of cervical cancer and 342,000 cervical cancer deaths were reported worldwide [3]. The high-risk genotypes of human papillomavirus, closely related to more than 90% of cervical cancers, include types 16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, 66, 68 [4–7]. 70% of cervical cancer are associated with only two HPV types, 16 and 18 [8–11]. Early diagnosis of these high-risk HPV genotypes and timely treatments are now available through molecular diagnostic methods.
The Mehrviru HPV test is an in vitro real-time PCR-based assay for qualitatively detecting of 14 high-risk HPV genotypes including 16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, 66, 68 in cervical exfoliated cells, urogenital tract secretion, and FFPE (Formalin Fixed Paraffin Embedded) tissue samples. The list of HPV genotypes detectable by Mehrviru HPV test includes the 12 high-risk ones declared by IARC [4] plus HPV 68 and 66. The Mehrviru HPV test’s proficiency was confirmed by the results of the 2021 HPV LabNet international proficiency study [12]. There are more than 125 tests (and more than 80 variants of the main tests) for diagnosing human papillomavirus, but evidence of clinical utility has been demonstrated for relatively few of them [13–15].
Latent Class Analysis (LCA) is a statistical procedure used to identify qualitatively different subgroups (classes) within populations that share certain observed characteristics [16]. LCA has been proposed for evaluating the performance of diagnostic tests in the absence of a gold standard [17–20]. In the current study, we compared the sensitivity and specificity of the new (MehrViru HPV Genotyping) kit and the brand (HPV Genotypes 14 Real-TM Quant) kit for the molecular detection of human papillomavirus in the cervical mucosal specimens using the results from Latent Class Analysis (LCA) as gold standard (for more details please see the Materials and Methods).
Materials and methods
We conducted the study in a specialist obstetrics and gynecology clinic in 2022. According to WHO guidelines [21], patients who required a pap smear test for cervical cancer screening were invited and specimens were taken by specialist gynecologist physician. A speculum was inserted into the vagina to visualize the cervix. Lubricants were not used, as it may interfere with results. Cells were collected using a cytobrush. The collected cells were placed into a vial of liquid preservative (CYToFAST collection medium). We asked each participant willing to enter the study to fill out an individual self-administered one-page questionnaire developed for this study containing a written statement of consent at the top (See supplement file). Demographic, lifestyle, reproductive, and sexual health information were recorded for each participant.
Two cervical mucosal specimens were collected for molecular detection of HPV. One of the samples was sent to the routine clinical laboratory, and the result was used in the clinical management decision-making process. The other sample was divided into two separate samples, deidentified, coded, and sent to the research laboratory for testing with Mehrviru or Sacace molecular diagnostic kits. We used batches of three linked unique codes to label the specimens, two of them were concealed behind the scratch labels and used on the tubes, and the third one was put on the questionnaire. Therefore the research laboratory team members were kept unaware of the identity of the specimens.
We performed real-time PCR on clinical specimens and identified those positive for one of the 14 high-risk HPV genotypes. Total DNA was extracted using the High Pure PCR Template Preparation Kit (Roche) following the manufacturer’s instructions. Two aliquots of the extracted DNA were used with the MehrViru HPV Genotyping and Sacace HPV Genotypes 14 Real-TM Quant diagnostic kits. For real-time PCR using MehrViru HPV Genotyping Kit (Ref. 02.0701.24), a 25 µL PCR final reaction volume contained 20 µL of Reaction Mix and 5 µL of sample DNA. The thermal cycles applied sequentially were 95 °C for 10 min, 40 cycles at 95 °C for 10 s, 58 °C for 30 s, and 72 °C for 15 s. Fluorescence was read in the FAM, Hex, ROX, and Cy5 channels. For real-time PCR using Sacace HPV Genotypes 14 Real-TM Quant (Ref. V67-100FRT), a 25 µL PCR final reaction volume contained 15 µL of Reaction Mix, and 10 µL of sample DNA. The thermal cycles applied sequentially were 95 °C for 15 min, 5 cycles at 95 °C for 5 s, 60 °C for 20 s, 72 °C for 15 s, 40 cycles at 95 °C for 5 s, 60 °C for 30 s, and 72 °C for 15 s. Fluorescence was read in the FAM, Hex, ROX, and Cy5 channels. For quality control and quality assurance, negative, positive, and internal controls for each assay were used.
We used latent class analysis (LCA) to determine a positive or negative cervical mucosal sample HPV test result for each participant based on the results from the three MehrViru HPV Genotyping, HPV Genotypes 14 Real-TM Quant (Sacace), and Routine clinical laboratory tests. The sensitivity and specificity of each diagnostic kit (validity indices) were estimated and compared. Other diagnostic accuracy indices were also estimated, including positive and negative predictive values, positive and negative likelihood ratios, youden, and area under the ROC curve. The agreement between the Mehrviru and Sacace kits was determined using the Kappa statistic.
Results
In total, 117 women with a mean age of 37.2 years (SD = 9.1) entered the study. 76.1% of participants were married, and 79.8% had a university education. Study participants’ characteristics have been summarised in Table 1. 96 samples (82.1%) had the same test result for all three diagnostic tests, either positive (n = 28; 23.9%) or negative (n = 68; 58.1%) (Fig. 1). HPV test results were positive in 32.5%, 28.2%, and 34.2%of subjects tested with MehrViru HPV Genotyping, HPV Genotypes 14 Real-TM Quant (Sacace), and Routine clinical laboratory kits, respectively (Table 2).
Table 1.
Characteristics of the study population
| Characteristics | Population | HPV+ (n/%)a |
|---|---|---|
| Total number of study participants | 117 | 34 (29.1) |
| Age years (mean/SD) | 37.2 (9.12) | - |
| Age groups (n/%) | ||
| 20–29 | 20 (17.54) | 6 (30) |
| 30–39 | 53 (46.49) | 15 (28.3) |
| 40–49 | 30 (26.32) | 9 (30) |
| 50 ≤ | 11 (9.65) | 3 (27.3) |
| Age at first sex (mean/SD) | 22.9 (5.01) | - |
| Partner age at first sex (mean/SD) | 28.9 (22.54) | - |
| BMI (mean/SD) | 25.2 (3.7) | - |
| BMI groups (n/%) | ||
| Normal/under-weight (BMI < 25) | 56 (53.84) | 19 (33.9) |
| Over-weight (25 ≤ BMI < 30) | 38 (36.54) | 11 (28.9) |
| Obese (BMI ≥ 30) | 10 (9.61) | 2 (20) |
| Marital status (n/%) | ||
| Single | 15 (12.82) | 7 (46.6) |
| Married | 89 (76.07) | 22 (24.7) |
| Divorced or widowed | 13 (11.11) | 5 (38.4) |
| Education (n/%) | ||
| Diploma or lower | 23 (20.18) | 6 (26) |
| Bachelor degree | 59 (51.75) | 14 (23.7) |
| Master degree or higher | 32 (28.07) | 13 (40.6) |
| Smoking status (n/%) | ||
| Never-smoker | 66 (67.35) | 16 (24.2) |
| Current smoker | 20 (20.41) | 10 (50) |
| former smokers | 12 (12.24) | 3 (25) |
| HPV vaccination (n/%) | ||
| Yes | 14 (13.59) | 5 (35.7) |
| No | 89 (86.41) | 26 (29.2) |
| OCP use (n/%) | ||
| Yes | 13 (12.15) | 3 (23) |
| No | 94 (87.85) | 29 (30.8) |
| Multiple partners (n/%) | ||
| Yes | 10 (8.55) | 4 (40) |
| No | 107 (91.45) | 30 (28) |
| Pregnancies (n/%) | ||
| No history of pregnancy | 42 (40) | 13 (30.9) |
| History of one pregnancy | 27 (25.71) | 5 (18.5) |
| History of two or more pregnancies | 36 (34.29) | 11 (30.5) |
| Abortions (n/%) | ||
| No history of abortion | 71 (68.27) | 17 (23.9) |
| History of abortion | 33 (31.73) | 11 (33.3) |
a HPV status determined based on latent class analysis (LCA) results
Fig. 1.
Venn diagram to show the overlap among positive (A) and negative (B) test results
Table 2.
Diagnostic acuracy estimates and their 95% confidence intervals. Diagnostic acuracy estimates for MehrViru HPV Genotyping, HPV Genotypes 14 Real-TM Quant (Sacace), and routine clinical laboratory test results have been calculated using latent class analysis
| Test | HPV+(%)a N = 117 |
Sensitivity | Specificity | PPVb | NPVc | Pos. likelihood ratio | Neg. likelihood ratio | Youden | Area under the ROCd curve |
|---|---|---|---|---|---|---|---|---|---|
| MehrViru HPV Genotyping | 32.5% | 90.8% (73.7-97.2%) | 90.9% (82.1-95.6%) | 81.6% (65.7-92.3%) | 96.2% (89.3-99.2%) | 10.8 (5.28–22.1) | 0.0964 (0.0323–0.284) | 82.7% (67.2-91.3%) | 91.4% (85.7-97.1%) |
| HPV Genotypes 14 Real-TM Quant | 28.2% | 92.0% (72.3-98.1%) | 97.4% (90.2-99.3%) | 93.9% (79.8-99.3%) | 96.4% (89.9-99.3%) | 37.8 (9.59–149) | 0.0904 (0.0307–0.267) | 88.8% (72.8-95.6%) | 94.4% (89.3-99.5%) |
| Routine clinical lab test result | 34.2% | 100% (89.7-100%) | 92.2% (83.1-96.6%) | %85 (70.2-94.3%) | 100% (95.3-100%) | 13.8 (6.4–29.9) | 0 (0–0) | 92.8% (84.3-96.7%) | 96.4% (93.6-99.2%) |
a Propotion of HPV positive patients; bPositive Predictive Value; c Negative Predictive Value; d receiver operating characteristic curv
Based on LCA results, 28.6% (95%CI: 22 -37.8) of participants had a positive test for human papillomavirus. The sensitivity and specificity (95% CI) of the Mehrviru were 90.8% (73.7–97.2) and 90.9% (82.1-95.6%), and corresponding figures for Sacace were 92.0% (72.3-98.0%) and 97.4% (90.2-99.3%). The observed agreement between the Mehrviru and Sacace kits was 87.2%, and the kappa index was 69.7% (95%CI: 55.6–83.9) (Table 2).
All the tests had an outstanding ability to separate HPV-positive and HPV-negative subjects. The areas under the ROC curve for Mehrviru and Sacace tests were 91.4% (85.7-97.1%) and 94.4% (89.3-99.5%) (Table 2).
Discussion
We found 28.63% of participants had an HPV infection based on LCA. Both Mehrviru and Sacace HPV genotyping diagnostic kits had very high sensitivity and specificity indices, and the agreement between the two was excellent. They also had an outstanding ability to separate HPV-positive and HPV-negative subjects.
Mehrviru and Sacace have demonstrated low false positive and false negative rates, two essential features of HPV diagnostic tests. Failing to diagnose HPV infection (false negatives) with certain serotypes in an individual could have serious long-term implications, such as dissemination of infection and increased risk of cervical cancer in those infected. Besides, the diagnosis of HPV infection in an individual could be associated with stigma, anxiety, and self-blame [22–24] and requires costly additional medical investigation [25]. Therefore, wrong diagnoses of HPV infection (false positives) should be avoided as much as possible. The new Mehrviru HPV genotyping kit performance in this regard was similar to the internationally-known brand of Sacace (HPV Genotypes 14 Real-TM Quant).
We performed latent class analysis using the test results of Mehrviru and Sacace and the routine clinical laboratory test to identify individuals infected with one of the high-risk HPV genotypes because there was no gold standard method available for the diagnosis. The latent class method enabled us to use all the available information to diagnose high-risk HPV infection, although it was not a definitive indicator of the infection. As the number of classes was known in advance, we ran the LCA model with two latent classes of HPV-positive and HPV-negative. Despite the efforts made to obtain information about the manufacturer, the identity of the third routine clinical laboratory test was not disclosed to the research team. However, based on the genotypes reported, we believe it was a separate diagnostic kit.
We kept the research laboratory team members blind to the identity of all the specimens tested by Mehrviru and Sacace diagnostic kits. This is likely to prevent potential biases when performing the RT-PCR tests on the samples. We could not confirm the brand type of the third routine clinical laboratory diagnostic kit and, therefore, could not guarantee the independence of the method used. However, the reported genotypes on two occasions (not in the list of declared genotypes identifiable by Sacace or Mehrviru) indicated that a separate diagnostic kit had been used. The three diagnostic kits had identified the same genotypes in all who were tested positive except in 4 cases. The information on genotypes were not used for the comparison of the tests. Nevertheless, the clinical and prognostic interpretation remains the same because a high-risk HPV genotype had been diagnosed. We made appropriate use of the latent class analysis method in this study; however, the LCA classifies individuals based on their probability of being in classes, which does not always guarantee proper class assignments [26].
Conclusions
Based on the findings of the present study, the MehrViru HPV Genotyping kit has acceptable diagnostic accuracy, and given its considerably lower price compared to most internationally known brands, we recommend its use as a reliable alternative for the diagnosis of human papillomavirus infection.
Electronic Supplementary Material
Below is the link to the electronic supplementary material.
Acknowledgements
Not applicable.
Abbreviations
- HPV
Human Papillomavirus
- LCA
Latent Class Analysis
- ROC
Receiver Operating Characteristic curve
- FFPE
Formalin Fixed Paraffin Embedded
Author contributions
HN, FRT, and MSD contributed to the design, conduct, data collection, analysis, reporting and had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. MS and DH provided administrative, technical, and material support for the study. DH facilitated the conduct of RT-PCR tests. All authors helped with the writing and critical revision of the manuscript.
Funding
This study was jointly funded by Iran University of Medical Sciences (IUMS), Tehran, Iran (Reference number: 21791) and Mehr Clinical Diagnostic Laboratory (MCDL), Hashtgerd, Iran.
Data availability
Deidentified Individual Participant Data (IPD) from this study could be made available for researchers from academic institutions for up to two years after publication to be used for educational and not for profit purposes. Please contact the coresponding author at solaymani.m@iums.ac.ir email address.
Declarations
Ethics approval and consent to participate
This study was approved by Iran University of Medical Sciences’ Ethics Committee, reference: IR.IUMS.REC.1400.924, and written informed consent was obtained from all the participants.
Consent for publication
Not applicable.
Competing interests
MS is the technical manager and DH is the head of department of Pathology and Molecular Medicine at Mehr Clinical Diagnostic Laboratory. MSD advises MCDL on research project. HN and FRT declare no conflict of interest.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Walboomers JM, Jacobs MV, Manos MM, Bosch FX, Kummer JA, Shah KV, et al. <ArticleTitle Language=“En”>Human papillomavirus is a necessary cause of invasive cervical cancer worldwide. J Pathol. 1999;189(1):12–9. [DOI] [PubMed] [Google Scholar]
- 2.Tolstov Y, Hadaschik B, Pahernik S, Hohenfellner M, Duensing S, editors. Human papillomaviruses in urological malignancies: a critical assessment. Urologic Oncology: Seminars and Original Investigations; 2014. [DOI] [PubMed] [Google Scholar]
- 3.Rao H-L, Chen J-W, Li M, Xiao Y-B, Fu J, Zeng Y-X, et al. Increased intratumoral neutrophil in colorectal carcinomas correlates closely with malignant phenotype and predicts patients’ adverse prognosis. PLoS ONE. 2012;7(1):e30806. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Humans IWGotEoCRt. Biological agents. 100 B. A review of human carcinogens: Human papillomaviruses. IARC Monogr Eval Carcinog Risks Hum. 2012;100:255–313. [Google Scholar]
- 5.Bzhalava D, Guan P, Franceschi S, Dillner J, Clifford G. A systematic review of the prevalence of mucosal and cutaneous human papillomavirus types. Virology. 2013;445(1–2):224–31. [DOI] [PubMed] [Google Scholar]
- 6.Chacón J, Sanz I, Rubio MD, de la Morena ML, Díaz E, Mateos ML, et al. Detection and genotyping of high-risk human papillomavirus in cervical specimens. Enferm Infecc Microbiol Clin. 2007;25(5):311–6. [DOI] [PubMed] [Google Scholar]
- 7.Del Mistro A, Salamanca HF, Trevisan R, Bertorelle R, Parenti A, Bonoldi E, et al. Human papillomavirus typing of invasive cervical cancers in Italy. Infect Agents Cancer. 2006;1(1):1–3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Munoz N, Bosch FX, Castellsagué X, Díaz M, De Sanjose S, Hammouda D, et al. Against which human papillomavirus types shall we vaccinate and screen? The international perspective. Int J Cancer. 2004;111(2):278–85. [DOI] [PubMed] [Google Scholar]
- 9.Zeng Z, Yang H, Li Z, He X, Griffith CC, Chen X, et al. Prevalence and genotype distribution of HPV infection in China: analysis of 51,345 HPV genotyping results from China’s largest CAP certified laboratory. J Cancer. 2016;7(9):1037. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Muñoz N, Bosch FX, De Sanjosé S, Herrero R, Castellsagué X, Shah KV, et al. Epidemiologic classification of human papillomavirus types associated with cervical cancer. N Engl J Med. 2003;348(6):518–27. [DOI] [PubMed] [Google Scholar]
- 11.Castellsagué X, Díaz M, De Sanjosé S, Muñoz N, Herrero R, Franceschi S, et al. Worldwide human papillomavirus etiology of cervical adenocarcinoma and its cofactors: implications for screening and prevention. J Natl Cancer Inst. 2006;98(5):303–15. [DOI] [PubMed] [Google Scholar]
- 12.Arroyo Mühr LS, Eklund C, Lagheden C, Forslund O, Robertsson KD, Dillner J. Improving human papillomavirus (HPV) testing in the cervical cancer elimination era: The 2021 HPV LabNet international proficiency study. J Clin Virol. 2022;154:105237. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Arbyn M, Ronco G, Anttila A, Meijer CJ, Poljak M, Ogilvie G, et al. Evidence regarding human papillomavirus testing in secondary prevention of cervical cancer. Vaccine. 2012;30:F88–99. [DOI] [PubMed] [Google Scholar]
- 14.Poljak M, Cuzick J, Kocjan BJ, Iftner T, Dillner J, Arbyn M. Nucleic acid tests for the detection of alpha human papillomaviruses. Vaccine. 2012;30:F100–6. [DOI] [PubMed] [Google Scholar]
- 15.Poljak M, Kocjan BJ. Commercially available assays for multiplex detection of alpha human papillomaviruses. Expert Rev anti-infective therapy. 2010;8(10):1139–62. [DOI] [PubMed] [Google Scholar]
- 16.Hagenaars JA, McCutcheon AL. Applied latent class analysis. Cambridge University Press; 2002.
- 17.Garrett ES, Eaton WW, Zeger S. Methods for evaluating the performance of diagnostic tests in the absence of a gold standard: a latent class model approach. Stat Med. 2002;21(9):1289–307. [DOI] [PubMed] [Google Scholar]
- 18.Sundaramurthy R, Dhodapkar R, Kaliaperumal S, Harish BN. Investigational approach to adenoviral conjunctivitis: comparison of three diagnostic tests using a Bayesian latent class model. J Infect Developing Ctries. 2018;12(01):043–51. [DOI] [PubMed] [Google Scholar]
- 19.Fowler JR, Cipolli W, Hanson T. A comparison of three diagnostic tests for carpal tunnel syndrome using latent class analysis. JBJS. 2015;97(23):1958–61. [DOI] [PubMed] [Google Scholar]
- 20.Thaler MA, Bietenbeck A, Yin M-X, Steigerwald U, Holmes AB, Lindhoff-Last E, et al. Evaluation of antiphospholipid antibody assays using latent class analysis to address the lack of a reference standard. Clin Chem Lab Med (CCLM). 2016;54(12):1929–37. [DOI] [PubMed] [Google Scholar]
- 21.WHO guideline for screening. and treatment of cervical pre-cancer lesions for cervical cancer prevention 2021 [ https://www.who.int/publications/i/item/9789240030824 [PubMed]
- 22.Qaderi K, Mirmolaei ST, Geranmayeh M, Sheikh Hasani S, Farnam F. Iranian women’s psychological responses to positive HPV test result: a qualitative study. BMC Womens Health. 2021;21(1):1–11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Korfage IJ, van Ballegooijen M, Huveneers H, Essink-Bot M-L. Anxiety and borderline PAP smear results. Eur J Cancer. 2010;46(1):134–41. [DOI] [PubMed] [Google Scholar]
- 24.Macios A, Nowakowski A. False Negative Results in Cervical Cancer Screening-Risks, Reasons and Implications for Clinical Practice and Public Health. Diagnostics (Basel). 2022;12(6). [DOI] [PMC free article] [PubMed]
- 25.Teoh D, Hultman G, DeKam M, Isaksson Vogel R, Downs LS Jr., Geller MA, et al. Excess Cost of Cervical Cancer Screening Beyond Recommended Screening Ages or After Hysterectomy in a Single Institution. J Low Genit Tract Dis. 2018;22(3):184–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Muthén B, Muthén LK. Integrating person-centered and variable‐centered analyses: Growth mixture modeling with latent trajectory classes. Alcoholism: Clin experimental Res. 2000;24(6):882–91. [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
Deidentified Individual Participant Data (IPD) from this study could be made available for researchers from academic institutions for up to two years after publication to be used for educational and not for profit purposes. Please contact the coresponding author at solaymani.m@iums.ac.ir email address.