Skip to main content
NCBI home page
PLOS One logoLink to PLOS One
. 2021 Apr 26;16(4):e0250426. doi: 10.1371/journal.pone.0250426

Self-collected and clinician-collected anal swabs show modest agreement for HPV genotyping

Racheal S Dube Mandishora 1,2,*, Trine B Rounge 3,4, Megan Fitzpatrick 5, Irene Kraus Christiansen 6,7, Ole Herman Ambur 8, Sonja Lagström 3,6, Babill Stray-Pedersen 9, Massimo Tommasino 2, Joel Palefsky 10,‡, Zvavahera M Chirenje 11,‡
Editor: Michael Scheurer12
PMCID: PMC8075200  PMID: 33901223

Abstract

Background & aim

Women with HIV/HPV coinfection and cervical lesions are at increased risk of developing HPV related anal cancer. Self-collection of anal swabs may facilitate HPV molecular testing in anal cancer screening, especially in high-risk groups, and yet it is not adequately studied. We evaluated level of agreement between self-collected anal swabs (SCAS) and clinician-collected anal swabs (CCAS) when used for HPV genotyping. We also described the anal HPV genotype distribution and HIV/HPV coinfection.

Methods

We performed a cross sectional study with participants from a visual-inspection-with-acetic-acid and cervicography (VIAC) clinic, in Harare, Zimbabwe. In a clinic setting, the women aged ≥18 years provided anal swabs in duplicate; first CCAS and then SCAS immediately after. HPV detection and genotyping were performed using next generation amplicon sequencing of a 450bp region of the HPV L1 gene. Level of agreement of HPV genotypes between CCAS and SCAS was calculated using the kappa statistic. McNemar tests were used to evaluate agreement in the proportion of genotypes detected by either method.

Results

Three-hundred women provided 600 samples for HPV genotyping. HPV genotypes were detected in 25% of SCAS and in 22% of CCAS. The most common genotypes with CCAS were HPV52, HPV62 and HPV70 and with SCAS were HPV62, HPV44, HPV52, HPV53 and HPV68. Total HPV genotypes detected in CCAS were more than those detected in SCAS, 32 versus 27. The agreement of HPV genotypes between the two methods was 0.55 in kappa value (k). The test of proportions using McNemar gave a Chi-square value of 0.75 (p = 0.39). Multiple HPV infections were detected in 28/75 and 29/67 women for CCAS and SCAS respectively.

Conclusions

SCAS and CCAS anal swabs showed moderate agreement, with no statistically significant difference in the proportion of genotypes detected by either methods. Although the differences between the two methods were not statistically significant, CCAS detected more HPV genotypes than SCAS and more HPV infections were detected in SCAS than in CCAS. Our data suggest that self-collected anal swabs can be used as an alternative to clinician-collected anal swabs for HPV genotyping.

1. Introduction

Human papillomaviruses (HPVs) are the most common viral infections in the genital tract and are the main etiologic agent of anogenital cancers, such as cervical, anal, vulval, penile and vaginal malignancies [1]. Although development of cancer only occurs in a small percentage of individuals who are infected with HPV, the high prevalence of the virus and its severe consequences makes HPV-related cancers a high-priority research area [2]. HPV accounts for almost all new cervical and anal cancer cases, as reported in 2018, with most of the cases disproportionately occurring in low- and middle-income countries. Zimbabwe reports an age-standardised cervical cancer incidence rate of 62.3/100000, ranking it as the 4th highest globally [3]. HPV-related cancers have been widely studied. Women who acquire cervical HPV infections may shed the virus at anatomical sites of vagina, vulva and anus, thus one anogenital site may act as a reservoir for HPV infection in other sites [4]. Women with a history of cervical cancer, especially those with HPV/HIV co-infection, are at increased risk of developing anal cancers [5]. Screening for anal cancer is thus recommended for women with cervical lesions and/or who are HIV infected [6, 7].

Although anal cancers are relatively rare, data from USA indicates a significant increase in the incidence of anal cancers [8]. The increase is particularly high among HIV-infected individuals despite the availability and widespread use of anti-retroviral-therapy (ART), which improves individuals’ immunogenic profile. Conversely, ART extends life among HIV infected individuals, giving time for cancer to develop [9]. Zimbabwe has an HIV prevalence of almost 13% [10] and cervical cancer contributes a third of all cancers among black Zimbabwean women [11]. In 2016, the Zimbabwe Cancer Registry reported a gradual increase in anal cancers recorded in the past decade. Despite the increased prevalence of risk factors in the population and evidence of increasing incidence, anal cancer screening guidelines are not routinely followed in Zimbabwe.

Utility of self-sample cervical swabs for HPV genotyping has been demonstrated [12]. Furthermore, self-collection of cervical swabs has shown to be highly acceptable [13, 14]. For anal screening, there is evidence that self-sampling is acceptable and patients tend to be more comfortable and less embarrassed when an anal swab is self-collected [15]. Its clinical utility has been demonstrated for non-HPV studies [16, 17] and for anal cytology [18–20]. However, there is still scarce data with regards to self-collected sample for HPV DNA testing. Improvements in the performance of molecular based tests permit consideration of self-collection as a promising approach to increase anal HPV screening [21].

Prioritising research on HPV-related anal cancers in women cannot be over emphasised. Most published data on HPV and anal malignancies is on men-who-have-sex-with men (MSM) [18, 22–24]. Heterosexual women have been neglected and they are more vulnerable to infection with oncogenic HPV. This is because of their anatomy, which allows the inoculation of HPV from the vaginal region to the anus even without practicing anal intercourse [25]. Furthermore, understanding the different collection methods will help researchers and clinicians to prepare for future application of anal cancer screening.

Against this background, our study primarily aimed at determining the level of agreement of self-collected anal swabs (SCAS) and clinician-collected anal swabs (CCAS) for anal HPV genotyping, in women reporting for routine cervical cancer screening in a clinical setting. We also reported the anal HPV genotype distribution and HIV co-infection in this group of women who are at increased risk of anogenital cancers, and as additional epidemiologic data. The description of HPV genotype distribution is important for clinical recommendations, for surveillance before and after introducing HPV vaccination and for comparison of the HPV in circulation in both the cervix and anus [26].

2. Methods

2.1 Study design and population

The study was cross-sectional and recruited women reporting for routine cervical cancer screening from the general population. The cervical cancer screening was at a Visual Inspection with Acetic-acid and cervicography (VIAC) clinic within a tertiary hospital, Parirenyatwa Hospital, situated in the capital city of Zimbabwe, with the sample set representing a populations in the city and in the rural and urban peripheries of the city. The questionnaire was available in both English and Shona (S1 File).

The estimated sample size was 300 women, accounting for an assumed 7% attrition rate. The samples were collected for a primary objective to obtain HPV genotype specific population parameters, such as prevalence and genetic evolution of the virus. The A. J Dobson’s formula for prevalence was used, where the estimated prevalence was 24% based on known prevalence of HPV genotype detection from a previous study done in Zimbabwe [27]. The confidence interval was set at 95% with a desired width of 5%, giving a total of 280 women. Twenty women were included to cover any attrition. All the samples were then used to address the secondary objective of comparing self-collected and clinician-collected anal swabs for HPV genotyping.

Ethics approval was obtained from The Joint Parirenyatwa hospital and College of Health Sciences Research Ethics Committee (reference: JREC210/14), Medical Research Council of Zimbabwe (reference: MRCZ/A/1911) and Research Council of Zimbabwe RCZ (Permit number 03064).

2.2 Recruitment and laboratory methods

Written informed consent, in English or Shona, was obtained from the women who were ≥18 years, sexually active and had no history of a total abdominal hysterectomy. Before consenting, they were taken through the study protocol and were made aware of future publications, which would be written with non-identifiers or with no information that could compromise the participant’s identity.

Enrolment was from February to April 2015 and the recruitment process is previously described [26]. Anal DacronTM swabs [28, 29] were collected and VIAC, Pap-smear and HIV testing were performed on these women, as previously described [30]. Briefly, sexually active women aged at least 18 years old, reporting for routine cervical cancer screening at the VIAC clinic were consecutively recruited. Women who did not consent to providing duplicate anal swabs were excluded. Two anal swabs were requested from each woman; one SCAS and one CCAS. The nurse collected the CCAS in the examination room, by gently inserting the swab into the anal canal until the shaft could not move further and rotated it for 10–30 seconds. The nurse explained the self-collection procedure to the women, who then immediately proceeded to the toilet within the clinic facility for the SCAS collection. All swabs had DacronTM tips and a firm plastic shaft. The swabs were immediately broken into a cryotube soon after collection; SCAS were broken by the participant and the CCAS by the collecting clinician. All swabs were stored in 500μl lysis buffer from bioMerieux (containing guanadine thiocyanate) at -80°C until analysed.

DNA was extracted from the anal swabs using the ZR Genomic DNATM Tissue MiniPrep kit, ZYMO Research, USA, as per manufacturer’s instructions [31]. The eluted DNA was stored at -20°C, awaiting next-generation-sequencing. For quality control, betaglobin testing was performed to verify DNA quality using conventional PCR in Zimbabwe before samples were sent for next-generation sequencing in Norway.

2.2.1 Next-generation-sequencing (NGS)

The detection and genotyping of HPV was done by amplification of a 450 bp amplicon from the HPV L1 region, using the Illumina MiSeq platform, as previously described [26]. Briefly, using PGMY primers (Table 1) [32], amplicons with Illumina-tail were generated in 20 μl volumes using Phusion Master Mix (Thermo Fischer Scientific, MA, USA), 0.1 μM of each primer and 5 μl sample under the following conditions; 98°C for 30 seconds, 40 cycles of 98°C for 10 seconds, 56°C for 30 seconds and 72°C for 15 seconds, before a final extension 72°C for 10 minutes. Amplicons were then used as templates in indexing PCR in 20 μl volumes using Phusion Master Mix, 0.375 μM each index [33] and 1 μl template under the following conditions; 98°C for 30 seconds, 8 cycles of 98°C for 10 seconds, 65°C for 30 seconds and 72°C for 20 seconds before a final extension 72°C for 5 min. The resulting amplicon libraries were pooled together in equal volumes, cleaned up twice using 0.7 × AMPure XP (Agencourt Beckman Coulter, CA, USA).Library quality control and quantitation was performed on the Agilent 2100 Bioanalyzer using Agilent High Sensitivity DNA Kit (Agilent Technologies, CA, USA) and by qPCR using KAPA DNA library quantification kit (Kapa Biosystems, Wilmington, MA). For HPV detection and genotyping, the libraries containing the 450 bp PGMY amplicons from the HPV L1 region were sequenced on the MiSeq platform (Illumina, CA, USA) using V2 chemistry and 2 × 250 bp reads.

Table 1. List of PGMY primers used for Illumina next-generation-sequencing.
Primer name Locus specific sequence Sequence with Illumina adapter
PGMY11-A GCACAGGGACATAACAATGG ACACTCTTTCCCTACACGACGCTCTTCCGATCTGCACAGGGACATAACAATGG
PGMY11-B GCGCAGGGCCACAATAATGG ACACTCTTTCCCTACACGACGCTCTTCCGATCTGCGCAGGGCCACAATAATGG
PGMY11-C GCACAGGGACATAATAATGG ACACTCTTTCCCTACACGACGCTCTTCCGATCTGCACAGGGACATAATAATGG
PGMY11-D GCCCAGGGCCACAACAATGG ACACTCTTTCCCTACACGACGCTCTTCCGATCTGCCCAGGGCCACAACAATGG
PGMY11-E GCTCAGGGTTTAAACAATGG ACACTCTTTCCCTACACGACGCTCTTCCGATCTGCTCAGGGTTTAAACAATGG
PGMY09-F CGTCCCAAAGGAAACTGATC AGACGTGTGCTCTTCCGATCTCGTCCCAAAGGAAACTGATC
PGMY09-G CGACCTAAAGGAAACTGATC AGACGTGTGCTCTTCCGATCTCGACCTAAAGGAAACTGATC
PGMY09-H CGTCCAAAAGGAAACTGATC AGACGTGTGCTCTTCCGATCTCGTCCAAAAGGAAACTGATC
PGMY09-I GCCAAGGGGAAACTGATC AGACGTGTGCTCTTCCGATCTGCCAAGGGGAAACTGATC
PGMY09-J CGTCCCAAAGGATACTGATC AGACGTGTGCTCTTCCGATCTCGTCCCAAAGGATACTGATC
PGMY09-K CGTCCAAGGGGATACTGATC AGACGTGTGCTCTTCCGATCTCGTCCAAGGGGATACTGATC
PGMY09-L CGACCTAAAGGGAATTGATC AGACGTGTGCTCTTCCGATCTCGACCTAAAGGGAATTGATC
PGMY09-M CGACCTAGTGGAAATTGATC AGACGTGTGCTCTTCCGATCTCGACCTAGTGGAAATTGATC
PGMY09-N CGACCAAGGGGATATTGATC AGACGTGTGCTCTTCCGATCTCGACCAAGGGGATATTGATC
PGMY09-P GCCCAACGGAAACTGATC AGACGTGTGCTCTTCCGATCTGCCCAACGGAAACTGATC
PGMY09-Q CGACCCAAGGGAAACTGGTC AGACGTGTGCTCTTCCGATCTCGACCCAAGGGAAACTGGTC
PGMY09-R CGTCCTAAAGGAAACTGGTC AGACGTGTGCTCTTCCGATCTCGTCCTAAAGGAAACTGGTC
RSMY09-L CGTCCTAATGGGAATTGGTC AGACGTGTGCTCTTCCGATCTCGTCCTAATGGGAATTGGTC
HMB01 GCGACCCAATGCAAATTGGT AGACGTGTGCTCTTCCGATCTGCGACCCAATGCAAATTGGT
Open in a new tab

2.2.2 Bioinformatics and statistics

We counted HPV sequences per genotype for each sample using a custom data analysis pipeline. Adapters, primers, short reads (<50 bases) and low-quality bases (<15) removed with Cutadapt (version 1.8.3). The sequences were mapped to an HPV genome database using Bowtie2 v 2.2.9 in an “end-to-end” mode and pre-set “sensitive” setting and only the best alignments were reported [30]. The HPV genome database consisted of 183 reference HPV genomes obtained from PaVe [34]. Number of read pairs mapped to each genome were counted by FeatureCount (version 1.4.6) from the Subread package. Further data cleaning and statistical analyses were performed on R (v3.6.0). A final cut-off of 500 mapped read pairs or more were considered as positive [35–37]. Choice of cut-off was based on the mean reads of CCAS and SCAS. Frequencies were calculated as a percentage of the total. Plots of HPV genotype distribution and agreement of the two methods were made using frequencies of each HPV genotype. Sequence read counts for genotypes detected as one of the most frequent, on both CCAS and SCAS, were tabulated for comparison. This was to check if the amount of HPV collected by the first swab differed from that collected by the second swab in the same individual.

Using R, data were subset into CCAS and SCAS. Categorical variables were created, of either ‘positive’ for all samples with a read count >499 or ‘negative’ for those with read counts <499. Bar plots were drawn to illustrate frequencies of individual HPV genotypes. To evaluate agreement based on proportions the two collection methods, the McNemar non-parametric test was performed. A kappa test was used to evaluate agreement of HPV genotypes across the two collection methods. The McNemar’s non-parametric test was performed to evaluated agreement in the proportion of genotypes detected by either method [38–40]. Agreement was either when both SCAS and CCAS, from the same woman, were positive for a specific HPV genotype or when both SCAS and CCAS were negative for any HPV genotype.

3. Results

3.1 Summary demographics

A total of 300 women were enrolled into the study. These women provided 600 samples for HPV genotyping. Half (50.3%) of the women were HIV infected (Table 2). The age range of these women was 18–83 years old, with a mean of 39.7 years, with no significant differences among the HIV infected and HIV un-infected women (p = 0.13). VIAC positivity was detected in 10% of the women with more HIV-infected women testing positive at a higher rate as compared to the HIV-uninfected (p = 0.036). Similarly, of the 217 women who had normal cytology outcomes, over half (122/217) were HIV-uninfected.

Table 2. Demographic and surveillance data of participants, grouped by their HIV status.

Category All participants
n = 300
*a/n is total detected over total women with available data		 HIV Positive
n = 151 n(%)except*		 HIV Negative n = 149
n (%)
*except		 P-value
Age in years
Mean (SD)
*39.7 (10.7)
*38.7
(8.1)
*40.6
(12.8)
0.13
Parity, number of children
Median (IQR)
*3.0 (0–10)
*3.0 (2–4)
*3.0 (2–4)
0.12
Sexual debut in years
Median (IQR)
*19 (18–22)
*19(17–21.5)
*20 (18–22)
0.03
Use of contraception
a/n (%)
Yes
Oral tablets
168/298 (56%)
79/168 (46.7)
84/151 (55.6%)
29/84 (34.5)
84/147 (57.1%)
50/84 (58.8)
0.79

History of STI a/n (%)
98/298 (32.7%)
62/151 (41.1%)
36/147 (24.2%)
<0.01
Multiple sexual partners at recruitment a/n (%)
Participant
Partner
25/299 (8.4)
38/240 (15.8)
16/151 (10.6)
21/115
9/148 (6.1)
17/125 (1)
0.16
0.31
Partner circumcised a/n (%) 52/227 (22.9) 24/106 (22.6) 28/121 (23.1) 0.93
VIAC positive a/n (%) 30/299 (10.0) 21/151 (13.9) 9/148 (6.0) 0.0361
Cytology normal a/n (%) 217/283 (76.6) *95/145 (65.5) *122/137 (89.1) <0.0001
Open in a new tab

This table gives a summary of demographics and survey data of 300 women, compiled from a structured questionnaire completed by a research nurse who interviewed each participant. VIAC = visual inspection with acetic acid and cervicography. IQR = Interquartile range. SD = Standard Deviation. STI = Sexually transmitted infection.

3.2 Anal HPV detection and genotyping

All samples were positive for betaglobin. SCAS generated 5 694 724 sequence read pairs whilst CCAS generated 7 843 298 (Table 3). HPV genotypes distribution showed that HPV52 were most frequent in CCAS and HPV62 most frequent in SCAS.

Table 3. Summary of sequence read pairs generated and number of positive clinician-collected (CCAS) and self-collected (SCAS) specimens.

Collection method of anal swab Total read pairs (n = 300) Mean read pairs Max read pairs Min read pairs Number of HPV positive women
Clinician-collected (CCAS) 7 843 298 26 232 104 980 144 67
Self-collected (SCAS) 5 694 724 19 110 73 905 50 75
Open in a new tab

HPV genotypes were detected in SCAS from 75/300 women (25%) and in CCAS from 67/300 women (22%). The most common genotypes in women with positive SCAS were HPV 62 (15%) and 11% for each of HPV 44, HPV 52, HPV 53 and HPV 68, whilst in women with positive CCAS HPV 52 (13%), HPV 62 (12%) and HPV 70 (10%) were the most prevalent types (Fig 1). The agreement of HPV genotypes between the two methods was 0.55 in kappa value (k). The test of proportions using McNemar gave a Chi-square value of 0.75 (p = 0.39), when comparing all the HPV genotypes detected. The total number of HPV genotypes detected in CCAS was more than those detected in SCAS, 32 versus 27 (Fig 1). HPV 31, HPV 42, HPV 54, HPV 84, HPV 86 and HPV 87 were only detected in CCAS. Genotypes detected in both CCAS and SCAS in perfectly equal frequencies are HPV 26, HPV 30 and HPV 32 (Fig 1).

Fig 1.

Fig 1

Open in a new tab

A) HPV genotypes detected in self-collected (SCAS) and clinician-collected (CCAS) anal swabs from women reporting for routine cervical cancer screening. B) Genotype concordance between SCAS and CCAS.

Multiple HPV infections were detected amongst 29/67 (43%) and 28/75 (37%) of CCAS and SCAS respectively.

Sequence read counts for the genotypes that were amongst the most frequently detected, HPV 52 and HPV 62, were compared (S1 Table). A total of 12/20 (60%) CCAS samples and/or SCAS showed lower SCAS than CCAS reads.

3.3 HIV stratification: Anal HPV genotypes detected in clinician-collected swabs

A total of 151 (50.3%) women were HIV-infected, of which 47 (31%) were positive on HPV genotyping. A total of 32 HPV genotypes were detected in this group of women. The most common genotypes were HPV 52 (17%), HPV 16 (13%) and HPV 62 (13%), with 24 (51%) women having multiple HPV infections. A total of 149 women were HIV-uninfected and 7 (5%) were positive for any HPV genotype. Eight different HPV genotypes were detected in these HIV-uninfected women (HPV 6, HPV 39, HPV 45, HPV 52, HPV 58, HPV 66, HPV 81 and HPV 89) with HPV 66 detected in two women. Only one HIV-uninfected woman had multiple HPV infections, with three HPV genotypes (HPV 58, HPV 66 and HPV 89).

4. Discussion

To the best of our knowledge this is one of the few studies globally and the first from Zimbabwe to compare the level of agreement of SCAS and CCAS anal swabs for HPV genotyping. We also describe the distribution of CCAS HPV genotypes, stratified by HIV status, in the anal canal of women reporting for routine cervical cancer screening in a clinical setting in Harare, Zimbabwe. The value of our study is to provide evidence on alternative collection methods for anal samples that can be used for anal cancer screening in high-risk groups, such as HPV/HIV co-infected women and/or with a history of cervical lesions. No comparable literature could be retrieved for anal swab collection methods, specifically on SCAS and CCAS for HPV genotyping, as most studies have focused on acceptability of anorectal swabs for other STI tests [16, 17, 41]. Very few studies compared anal self- and clinician-collection for cytology and not specifically for HPV genotyping [18, 19]. Our data addresses this critical gap of knowledge.

SCAS and CCAS show moderate agreement between the specific HPV genotype detected in these duplicate samples. The difference in detection rates of any HPV genotype in SCAS versus CCAS (25% vs 22%) could be attributed to the complex topology of the anal canal; with such complex anatomy, reproducibility may not be easily achieved, even in a scenario where clinician-collected duplicate samples are compared. Based on the moderate level of agreement of HPV genotypes between the two methods, indicated by 0.55 in kappa value (k) and a non-significant test of proportions, our data suggest that researchers can use either methods for collection. It is important for researchers to note one limitation, the use of the Kappa test alone does not warrant clinical recommendations for SCAS to be used as an alternative to the CCAS, because a kappa value (k) of at least 0.80 is usually required to provide strong support [38]. We however recommend a bigger study that also evaluates fewer, clinically relevant HPV genotypes to strengthen our inferences. For the purposes of our current aim, the analyses of all HPV genotypes were useful although it gives a limited picture for those seeking to visualise agreement of only the clinically relevant genotypes. As is, our levels of agreement are only moderate and not excellent, but are sufficient enough to give perspective of instances where NGS is used for HPV genotyping especially in research.

Self-collection is a good alternative because it is the most practical and less resource demanding method, which participants are less reluctant to perform. Although our current study was performed in a clinical setting, we also recommend future study designs that ensure that patients also collect samples from home and the use of routine laboratory genotyping tools. NGS is a highly sensitive method but is not widely used for routine diagnosis.

Distribution of HPV genotypes in the anal canal in the group of women studied here was diverse; with a total of 32 different HPV genotypes detected. Low-risk (LR) HPV genotypes were the most frequently detected, using both collection methods. HPV 52 was the only oncogenic high-risk (HR) HPV genotype among those commonly detected, regardless of HIV status. HPV 16 was noted as a common genotype only in the HIV-infected subgroup. In a previous study on a smaller number from the same group of women [26], HPV 52 was detected as one of the most common genotypes on both the cervico-vaginal and anal canal. LR-HPV genotypes such as HPV 6 and HPV 11 have been widely detected in individuals with anal or genital warts [42, 43]. The presence of HPV 53, a probable high-risk type [44], as one of the frequently detected anal HPV genotypes, agrees with previous findings from another Zimbabwean study that described HPV genotypes in the anus of men and women [43].

The total number HPV genotypes detected in CCAS was more than those detected in SCAS, 32 versus 27 (Fig 1). Furthermore, the most common HPV genotypes detected on CCAS and SCAS were slightly different. This may have occurred because the CCAS swabs were collected before the SCAS. We have several hypotheses to this effect, which are speculative and should be further investigated in future studies. The first one is that, given the decrease in read counts for over half of the HPV 52 and HPV 62 samples (S1 Table), the differences in collection methods maybe due to the fact that CCAS collects most of the exfoliated cells thus reducing the HPV yield for the SCAS that only collects the remaining bits. In addition, the SCAS may be touching different parts of the anal canal where there are different HPV genotypes from the parts touched by the CCAS. We recommend a different design for future studies, to address the possibility of HPV yield decreasing after the first swab has been collected. An ideal design would be to alternate the order of collection, for a proportion of SCAS and CCAS.

Secondly, in terms of numbers of samples with HPV positivity, SCAS detected more HPV than CCAS. We postulate that on collection of the first swabs, HPV from the surface epithelium may have been harvested, exposing more basal cells with HPV, which were easily harvested by the second swab, SCAS. The likelihood of this theory is reduced by the fact that swab collection is not invasive enough to cause aggressive abrasions on the surface epithelium. Our third, but most unlikely postulate is that contamination occurred during sample collection, accounting for the differences observed. Participants might have touched the cervico-vaginal area with the tip of the anal self-collection swab. Arbyn et al had similar observations, reporting HPV detection rates 2.28 times higher in self-collected versus clinician-collected vaginal swabs [45].

On average, 40% of the HPV infected women had multiple HPV infections, using either sample collection methods. This is a relatively high proportion of multiple HPV infections and is mostly observed for women with HPV/HIV co-infection, similar to studies from South-Africa and Zambia [43, 46]. The prevalence of HIV in the sample population was relatively high (50.3%). This is mainly attributed to the fact that a huge proportion of the women reporting for routine cervical cancer screening, at this tertiary hospital, would be coming from the HIV opportunistic clinic at the same premises. HIV-infected women are likely more informed of their increased risk of developing cervical cancer, thus they follow the screening guidelines prescribed for them.

In the context of the secondary data on the background of study participants, gathered on HIV status and cervical cancer screening, that gives a background of the study participants, HPV was more prevalent in HIV-infected women than in the HIV-un-infected group (31% vs 5%). This finding aligns with literature, describing how immunosuppression promotes persistent HPV infection and poorer clearance of the virus [47–50]. Important to note is that the oncogenic type, HPV 16, was only detected in the HIV-infected women and HPV 52 was detected in both groups. Given the small number of HIV-uninfected women with HPV, the proportion of HPV 52 cannot be elucidated to significance. On the other hand, HPV 52 and HPV 16 each make-up proportions of 17% and 13% of all the genotypes detected, indicating an increased risk of developing cancer for these women.

In addition to this, about half of these HPV/HIV co-infected women 21/47 (Table 2) also had positive VIAC results, signifying possible presence of pre-cancer or cancer lesions. Furthermore, the proportion of normal cervical cytology was higher in HIV-uninfected than the HIV-infected (Table 2). To further highlight risk of disease development, we report multiple HPV infections in half of the co-infected women versus only one woman amongst the HIV-uninfected. The HIV-infected group also exhibited significantly higher risk factors (Table 2), such as sexual debut (p = 0.03) and history of STIs (p <0.01).

To highlight additional limitations of our study, we have used both the Kappa and McNemar tests which are commonly used together to complement each other. Indeed, as mentioned by Ranganathan et al, a non-significant McNemar test alone is not sufficient to indicate good agreement. However, it is still a useful statistic, if interpreted correctly [51]. In our case, the McNemar test indicated that the proportion of discordant pairs (ie +/- vs -/+) was not statistically significant, providing useful information for evaluating (dis)agreement of HPV genotype detection.

We also could not properly measure whether HPV yield decreased when CCAS was collected before SCAS. Therefore, we recommend future studies that alternate the order of collection. Another possible limitation is that a few of the samples may have been reported as false negatives. Due to limitation of funds, a different test could not be used to verify this, but considering the mean and maximum reads recorded for both CCAS and SCAS (Table 3), and the positive betaglobin PCR test that verified DNA quality we are confident that a negligible number of women may have fallen within the false negative range.

5. Conclusion

SCAS and CCAS anal swabs showed moderate HPV genotype agreement, with non-statistically significant difference. SCAS also detected more HPV infections than CCAS, although CCAS detected a higher number of different HPV genotypes than SCAS. Our data suggest that self-collection of anal swabs can be used as an alternative to clinician collection for HPV genotyping in a clinical setting, especially where NGS is being used for research purposes. The diverse anal HPV genotype distribution detected, mostly in the HIV-infected who also had positive VIAC results, supports the recommendation for introducing routine anal cancer screening among this high-risk group.

Supporting information

S1 File

(PDF)

Click here for additional data file. (1.4MB, pdf)
S1 Table. Comparison of sequence read counts for HPV 52 and HPV 62 in both CCAS and SCAS.

(DOCX)

Click here for additional data file. (14.2KB, docx)

Acknowledgments

We would like to acknowledge Maria Da Costa (UCSF), who consistently gave counsel during sample collection and DNA extraction, Vasco Chikwasha for his statistical expertise, which he contributed during the proposal writing stages and Eugénie Lohmann for her contribution with R-studio commands. We also express our gratitude to Sister Mucheche, the nurse who recruited the women, the personnel at UZCHS-CTRC laboratory, the Norwegian HPV reference laboratory at Akershus University Hospital for their molecular expertise and technical support and the Parirenyatwa Hospital VIAC clinic staff for assisting during the recruitment phase. We are also very grateful to Prof. Jane R. Montealegre and all the other reviewers for their constructive critique and suggestions.

Where authors are identified as personnel of the International Agency for Research on Cancer/World Health Organization, the authors alone are responsible for the views expressed in this article and they do not necessarily represent the decisions, policy or views of the International Agency for Research on Cancer /World Health Organization.

Data Availability

All relevant data are within the manuscript and its Supporting Information files.

Funding Statement

Two organisations funded the research reported in this publication; Letten Foundation Norway and The Fogarty International Center of the National Institutes of Health under Award Number D43TW010137 within the University of Zimbabwe PERFECT program.

References

  • 1.zur Hausen H. Papillomaviruses in the causation of human cancers—a brief historical account. Virology. 2009;384:260–5. 10.1016/j.virol.2008.11.046 [DOI] [PubMed] [Google Scholar]
  • 2.Wagner M, Bennetts L, Patel H, Welner S, de Sanjose S, Weiss TW. Global availability of data on HPV genotype-distribution in cervical, vulvar and vaginal disease and genotype-specific prevalence and incidence of HPV infection in females. Infectious Agents and Cancer. 2015;10:13. 10.1186/s13027-015-0008-y [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Cervical cancer statistics. World Cancer Research Fund. 2018. https://www.wcrf.org/dietandcancer/cancer-trends/cervical-cancer-statistics. Accessed 20 Aug 2020.
  • 4.Goodman MT, Shvetsov YB, McDuffie K, Wilkens LR, Zhu X, Thompson PJ, et al. Sequential Acquisition of Human Papillomavirus Infection of the Anus and Cervix: The Hawaii HPV Cohort Study. J Infect Dis. 2010;201:1331–9. 10.1086/651620 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Park IU, Ogilvie JW Jr., Anderson KE, Li Z, Darrah L, Madoff R, et al. Anal human papillomavirus infection and abnormal anal cytology in women with genital neoplasia. Gynecologic Oncology. 2009;114:399–403. 10.1016/j.ygyno.2009.05.008 [DOI] [PubMed] [Google Scholar]
  • 6.Gosens KCM, Richel O, Prins JM. Human papillomavirus as a cause of anal cancer and the role of screening. Curr Opin Infect Dis. 2017;30:87–92. 10.1097/QCO.0000000000000337 [DOI] [PubMed] [Google Scholar]
  • 7.Wieland U, Kreuter A. Anal cancer risk: HPV-based cervical screening programmes. The Lancet Infectious Diseases. 2019;19:799–800. 10.1016/S1473-3099(19)30296-8 [DOI] [PubMed] [Google Scholar]
  • 8.Deshmukh AA, Suk R, Shiels MS, Sonawane K, Nyitray AG, Liu Y, et al. Recent Trends in Squamous Cell Carcinoma of the Anus Incidence and Mortality in the United States, 2001–2015. J Natl Cancer Inst. 2020;112:829–38. 10.1093/jnci/djz219 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Palefsky JM. HPV-associated anal and cervical cancers in HIV-infected individuals: Incidence and prevention in the antiretroviral therapy era. Curr Opin HIV AIDS. 2017;12:26–30. 10.1097/COH.0000000000000336 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.ZWE_2018_countryreport.pdf. http://www.unaids.org/sites/default/files/country/documents/ZWE_2018_countryreport.pdf. Accessed 4 Mar 2019.
  • 11.Zimbabwe National cancer registry. Pattern of cancer in Zimbabwe. National. Zimbabwe: ZNCR; 2016.
  • 12.Lissouba P, Perre PV de, Auvert B. Association of genital human papillomavirus infection with HIV acquisition: a systematic review and meta-analysis. Sex Transm Infect. 2013;89:350–6. 10.1136/sextrans-2011-050346 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Yi X, Zou J, Xu J, Liu T, Liu T, Hua S, et al. Development and Validation of a New HPV Genotyping Assay Based on Next-Generation Sequencing. American Journal of Clinical Pathology. 2014;141:796–804. 10.1309/AJCP9P2KJSXEKCJB [DOI] [PubMed] [Google Scholar]
  • 14.Waller J, McCaffery K, Forrest S, Szarewski A, Cadman L, Austin J, et al. Acceptability of unsupervised HPV self-sampling using written instructions. J Med Screen. 2006;13:208–13. http://msc.sagepub.com/content/13/4/208. Accessed 24 Apr 2015. 10.1177/096914130601300409 [DOI] [PubMed] [Google Scholar]
  • 15.Ortiz AP, Alejandro N, Pérez CM, Otero Y, Soto-Salgado M, Palefsky JM, et al. Acceptability of Cervical and Anal HPV Self-sampling in a Sample of Hispanic Women in Puerto Rico. P R Health Sci J. 2012;31:205–12. http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3954748/. [PMC free article] [PubMed] [Google Scholar]
  • 16.Wayal S, Llewellyn C, Smith H, Hankins M, Phillips A, Richardson D, et al. Self-sampling for oropharyngeal and rectal specimens to screen for sexually transmitted infections: acceptability among men who have sex with men. Sexually Transmitted Infections. 2009;85:60–4. 10.1136/sti.2008.032193 [DOI] [PubMed] [Google Scholar]
  • 17.Bell S, Wapling J, Ase S, Boli-Neo R, Vallely AJ, Kaldor JM, et al. Acceptability of testing for anorectal sexually transmitted infections and self-collected anal swabs in female sex workers, men who have sex with men and transgender women in Papua New Guinea. BMC Public Health. 2018;18:776. 10.1186/s12889-018-5684-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Lampinen TM, Miller ML, Chan K, Anema A, van Niekerk D, Schilder AJ, et al. Randomized clinical evaluation of self-screening for anal cancer precursors in men who have sex with men. Cytojournal. 2006;3:4. 10.1186/1742-6413-3-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Cranston RD, Darragh TM, Holly EA, Jay N, Berry JM, Da Costa M, et al. Self-collected versus clinician-collected anal cytology specimens to diagnose anal intraepithelial neoplasia in HIV-positive men. J Acquir Immune Defic Syndr. 2004;36:915–20. 10.1097/00126334-200408010-00004 [DOI] [PubMed] [Google Scholar]
  • 20.Chin-Hong PV, Berry JM, Cheng S-C, Catania JA, Costa MD, Darragh TM, et al. Comparison of Patient- and Clinician-Collected Anal Cytology Samples to Screen for Human Papillomavirus–Associated Anal Intraepithelial Neoplasia in Men Who Have Sex with Men. Annals of Internal Medicine. 2008;149:300. 10.7326/0003-4819-149-5-200809020-00004 [DOI] [PubMed] [Google Scholar]
  • 21.Vahabi M, Lofters A. HPV self-sampling: A promising approach to reduce cervical cancer screening disparities in Canada. Curr Oncol. 2018;25:13–8. 10.3747/co.25.3845 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Brown B, Davtyan M, Galea J, Chow E, Leon S, Klausner JD. The Role of Human Papillomavirus in Human Immunodeficiency Virus Acquisition in Men who Have Sex with Men: A Review of the Literature. Viruses. 2012;4:3851–8. 10.3390/v4123851 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Chin-Hong PV, Vittinghoff E, Cranston RD, Buchbinder S, Cohen D, Colfax G, et al. Age-Specific Prevalence of Anal Human Papillomavirus Infection in HIV-Negative Sexually Active Men Who Have Sex with Men: The EXPLORE Study. J Infect Dis. 2004;190:2070–6. 10.1086/425906 [DOI] [PubMed] [Google Scholar]
  • 24.Machalek DA, Poynten M, Jin F, Fairley CK, Farnsworth A, Garland SM, et al. Anal human papillomavirus infection and associated neoplastic lesions in men who have sex with men: a systematic review and meta-analysis. Lancet Oncol. 2012;13:487–500. 10.1016/S1470-2045(12)70080-3 [DOI] [PubMed] [Google Scholar]
  • 25.Castro FA, Quint W, Gonzalez P, Katki HA, Herrero R, van Doorn L-J, et al. Prevalence of and Risk Factors for Anal Human Papillomavirus Infection Among Young Healthy Women in Costa Rica. J Infect Dis. 2012;206:1103–10. 10.1093/infdis/jis458 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Dube Mandishora RS, Christiansen IK, Chin’ombe N, Duri K, Ngara B, Rounge TB, et al. Genotypic diversity of anogenital human papillomavirus in women attending cervical cancer screening in Harare, Zimbabwe. J Med Virol. 2017;:1671–7. 10.1002/jmv.24825 [DOI] [PubMed] [Google Scholar]
  • 27.Smith-McCune KK, Shiboski S, Chirenje MZ, Magure T, Tuveson J, Ma Y, et al. Type-Specific Cervico-Vaginal Human Papillomavirus Infection Increases Risk of HIV Acquisition Independent of Other Sexually Transmitted Infections. PLoS ONE. 2010;5:1–7. 10.1371/journal.pone.0010094 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Mbatha JN, Galapaththi-Arachchige HN, Mtshali A, Taylor M, Ndhlovu PD, Kjetland EF, et al. Self-sampling for human papillomavirus testing among rural young women of KwaZulu-Natal, South Africa. BMC Res Notes. 2017;10. 10.1186/s13104-016-2357-z [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Gage JC, Ghosh A, Borgonovo S, Follansbee S, Wentzensen N, Gravitt PE, et al. A Comparison of Dacron versus Flocked Nylon Swabs for Anal Cytology Specimen Collection. ACY. 2011;55:364–7. 10.1159/000329488 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Dube Mandishora RS, Gjøtterud KS, Lagström S, Stray-Pedersen B, Duri K, Chin’ombe N, et al. Intra-host sequence variability in human papillomavirus. Papillomavirus Res. 2018;5:180–91. 10.1016/j.pvr.2018.04.006 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Zymo-research. ZR Genomic DNATM -Tissue MiniPrep—Solid & FFPE Tissue DNA—Genomic DNA—DNA. Zymo-Research. 2015. http://www.zymoresearch.com/dna/genomic-dna/solid-ffpe-tissue-dna/zr-genomic-dna-tissue-miniprep. Accessed 18 Jan 2017.
  • 32.Gravitt PE, Peyton CL, Alessi TQ, Wheeler CM, Coutlée F, Hildesheim A, et al. Improved Amplification of Genital Human Papillomaviruses. J Clin Microbiol. 2000;38:357–61. http://jcm.asm.org/content/38/1/357. Accessed 20 Mar 2017. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Kozich JJ, Westcott SL, Baxter NT, Highlander SK, Schloss PD. Development of a Dual-Index Sequencing Strategy and Curation Pipeline for Analyzing Amplicon Sequence Data on the MiSeq Illumina Sequencing Platform. Appl Environ Microbiol. 2013;79:5112–20. 10.1128/AEM.01043-13 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Van Doorslaer K, Li Z, Xirasagar S, Maes P, Kaminsky D, Liou D, et al. The Papillomavirus Episteme: a major update to the papillomavirus sequence database. Nucleic Acids Res. 2017;45:D499–506. 10.1093/nar/gkw879 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Petrackova A, Vasinek M, Sedlarikova L, Dyskova T, Schneiderova P, Novosad T, et al. Standardization of Sequencing Coverage Depth in NGS: Recommendation for Detection of Clonal and Subclonal Mutations in Cancer Diagnostics. Front Oncol. 2019;9. 10.3389/fonc.2019.00009 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Meisal R, Rounge TB, Christiansen IK, Eieland AK, Worren MM, Molden TF, et al. HPV Genotyping of Modified General Primer-Amplicons Is More Analytically Sensitive and Specific by Sequencing than by Hybridization. PLoS One. 2017;12:1–14. 10.1371/journal.pone.0169074 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Cullen M, Boland JF, Schiffman M, Zhang X, Wentzensen N, Yang Q, et al. Deep sequencing of HPV16 genomes: A new high-throughput tool for exploring the carcinogenicity and natural history of HPV16 infection. Papillomavirus Research. 2015;1:3–11. 10.1016/j.pvr.2015.05.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.McHugh ML. Interrater reliability: the kappa statistic. Biochem Med (Zagreb). 2012;22:276–82. http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3900052/. Accessed 23 Apr 2017. [PMC free article] [PubMed] [Google Scholar]
  • 39.Krummenauer F, Kalden P, Kreitner KF. [Cohen’s kappa or McNemar’s test? A comparison of binary repeated measurements]. Rofo. 1999;171:226–31. [PubMed] [Google Scholar]
  • 40.Wirtz M, Kutschmann M. Analyse der Beurteilerübereinstimmung für kategoriale Daten mittels Cohens Kappa und alternativer Maße. Rehabilitation (Stuttg). 2007;46:370–7. 10.1055/s-2007-976535 [DOI] [PubMed] [Google Scholar]
  • 41.Tamalet C, Ravaux I, Dhiver C, Menard A, Colson P, Stein A. Feasibility and Acceptability of Anal Self-Sampling for Human Papillomavirus Screening in HIV-Infected Patients. INT. 2016;59:118–22. 10.1159/000452710 [DOI] [PubMed] [Google Scholar]
  • 42.Flores-Díaz E, Sereday KA, Ferreira S, Sirak B, Sobrinho JS, Baggio ML, et al. HPV-6 Molecular Variants Association With the Development of Genital Warts in Men: The HIM Study. J Infect Dis. 2017;215:559–65. 10.1093/infdis/jiw600 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Chinyowa S, Palefsky JM, Chirenje ZM, Makunike-Mutasa R, Munjoma M, Muguti GI. Anal human papillomavirus infection in HIV-positive men and women at two opportunistic infections clinics in Harare, Zimbabwe. BMC Public Health. 2018;18. 10.1186/s12889-018-6170-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Meyer T, Arndt R, Beckmann ER, Padberg B, Christophers E, Stockfleth E. Distribution of HPV 53, HPV 73 and CP8304 in genital epithelial lesions with different grades of dysplasia. Int J Gynecol Cancer. 2001;11:198–204. 10.1046/j.1525-1438.2001.01009.x [DOI] [PubMed] [Google Scholar]
  • 45.Arbyn M, Smith SB, Temin S, Sultana F, Castle P, Collaboration on Self-Sampling and HPV Testing. Detecting cervical precancer and reaching underscreened women by using HPV testing on self samples: updated meta-analyses. BMJ. 2018;363:k4823. 10.1136/bmj.k4823 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Bello BD, Spinillo A, Alberizzi P, Cesari S, Gardella B, D’Ambrosio G, et al. Cervical infections by multiple human papillomavirus (HPV) genotypes: Prevalence and impact on the risk of precancerous epithelial lesions. J Med Virol. 2009;81:703–12. 10.1002/jmv.21429 [DOI] [PubMed] [Google Scholar]
  • 47.Palefsky JM, Holly EA, Ralston ML, Da Costa M, Greenblatt RM. Prevalence and risk factors for anal human papillomavirus infection in human immunodeficiency virus (HIV)-positive and high-risk HIV-negative women. J Infect Dis. 2001;183:383–91. 10.1086/318071 [DOI] [PubMed] [Google Scholar]
  • 48.Mbulawa ZZA, Coetzee D, Williamson A-L. Human papillomavirus prevalence in South African women and men according to age and human immunodeficiency virus status. BMC Infect Dis. 2015;15:459. 10.1186/s12879-015-1181-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 49.Mbulawa ZZA, Marais DJ, Johnson LF, Coetzee D, Williamson A-L. Impact of human immunodeficiency virus on the natural history of human papillomavirus genital infection in South African men and women. J Infect Dis. 2012;206:15–27. 10.1093/infdis/jis299 [DOI] [PubMed] [Google Scholar]
  • 50.Meiring TL, Mbulawa ZZA, Lesosky M, Coetzee D, Williamson A-L. High diversity of alpha, beta and gamma human papillomaviruses in genital samples from HIV-negative and HIV-positive heterosexual South African men. Papillomavirus Res. 2017;3:160–7. 10.1016/j.pvr.2017.05.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 51.Ranganathan P, Pramesh CS, Aggarwal R. Common pitfalls in statistical analysis: Measures of agreement. Perspect Clin Res. 2017;8:187–91. 10.4103/picr.PICR_123_17 [DOI] [PMC free article] [PubMed] [Google Scholar]

Decision Letter 0

Michael Scheurer

10 Aug 2020

PONE-D-20-19224

Self-collected and clinician-collected anal swabs show modest agreement for HPV genotyping

PLOS ONE

Dear Dr. Dube Mandishora,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

In particular, please ensure that your statistical methods are fully described and appropriately support your study design. Address any inconsistencies noted between your conclusions and the existing body of literature. Finally, be sure to correct any noted spelling or grammatical errors in the manuscript.

Please submit your revised manuscript by Sep 24 2020 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

  • A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

  • A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

  • An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

We look forward to receiving your revised manuscript.

Kind regards,

Michael Scheurer, Ph.D.

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. Please include additional information regarding the survey or questionnaire used in the study and ensure that you have provided sufficient details that others could replicate the analyses. For instance, if you developed a questionnaire as part of this study and it is not under a copyright more restrictive than CC-BY, please include a copy, in both the original language and English, as Supporting Information.

3. In your Methods section, please provide additional information about the participant recruitment method and the demographic details of your participants. Please ensure you have provided sufficient details to replicate the analyses such as:  a) a description of any inclusion/exclusion criteria that were applied to participant recruitment, b) a statement as to whether your sample can be considered representative of a larger population, c) a description of how participants were recruited, and d) descriptions of where participants were recruited and where the research took place.

4. Please provide a sample size and power calculation in the Methods, or discuss the reasons for not performing one before study initiation.

5. To comply with PLOS ONE submission guidelines, in your Methods section, please provide additional information regarding your statistical analyses, including the threshold set for statistical significance for the analyses. For more information on PLOS ONE's expectations for statistical reporting, please see https://journals.plos.org/plosone/s/submission-guidelines.#loc-statistical-reporting.

6. In the Methods section, please provide the specific sequences of the primers used in the next-generation sequencing analysis.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer #1: This is an interesting manuscript describing concordance between self- and clinician-collected anal swabs for HPV DNA testing. Given increasing incidence of anal cancer worldwide and new guidelines for anal HPV screening, self-sampling may help screening programs achieve greater coverage. Thus, there is a need for research regarding utility of self-sampling vis a vis clinician sampling. I have the following comments and suggestions:

Abstract

I suggest changing the last sentence to, “Our data suggest that self-collected anal swabs can be used as an alternative to clinician-collected anal swabs for HPV DNA testing in a clinical setting.”

Intro

Paragraph 1 would benefit from mention of burden of disease. I suggest including data on burden of cervical cancer, narrowing in on burden is sub-Saharan Africa.

Paragraph 2 “Although anal cancers are relatively rare, data from USA reports a gradual increase in the incidence of anal cancers especially in HIV-infected individuals, regardless of the anti

retroviral (ART) era [5]. Furthermore, a cumulative incidence of 0.1% by age 75 years was

reported in HIV uninfected individuals as compared to 1.5% in HIV infected individuals.

The increase in anal cancer seems to be broader than among HIV-infected individuals. I suggest revising to: Although anal cancers are relatively rare, data from USA indicates a significant increase in the incidence of anal cancers [REF: https://doi.org/10.1093/jnci/djz219]. The increase is particularly high among HIV-infected individuals in the anti-retroviral era[include data on increasing incidence].

Sidenote: “The increase is particularly high among HIV-infected individuals, regardless of the antiretroviral (ART) era [5].” I’m not sure that it’s “regardless” of the anti-retroviral era, rather than “in” the antiretroviral era. Isn’t the extension of life due to ART driving increasing cancer incidence among HIV+ individuals?

Paragraph 1 and 2 jump around quite a bit. Also, since the paper doesn’t seek to fill the gap in data regarding burden of anal cancer in Zimbabwe, I don’t think it makes sense to set this up as an argument in the Intro. How about ending paragraph 1 with “Women with a history of cervical cancer, especially those with HPV/HIV co-infection, are at increased risk of developing anal cancers [4]. Screening for anal cancer is thus recommended for women with cervical lesions and/or who are HIV infected [9, 10].” And starting paragraph 2 with “Although anal cancers are relatively rare, data from USA indicates a significant increase in the incidence of anal cancers. The increase is particularly high among HIV-infected individuals in the anti-retroviral era. Zimbabwe has an HIV prevalence of almost 13% and cervical cancer contributes a third of all cancers among black Zimbabwean women. In 2016, the Zimbabwe Cancer Registry reported a gradual increase in anal cancers recorded in the past decade. Despite the increased prevalence of risk factors in the population and evidence of increasing incidence, anal cancer screening guidelines are not routinely followed in Zimbabwe.”

Given focus of paper on HPV genotypes and multiple infections, it would be helpful to reader to understand why different genotypes matter. Could this be woven into the intro? This is mentioned in Discussion, but I’d move to Intro

Paragraph 3 jumps between accuracy and acceptability and between anal and cervical swabs. I suggest walking the reader through the arguments to better set up the rationale for the study. Eg. “Utility of self-sample cervical swabs for HPV DNA testing has been demonstrated. Furthermore, self-collection of cervical swabs has shown to be highly acceptable. For anal screening, there is evidence that self-sampling is acceptable and patients tend to be more comfortable and less embarrassed when an anal swab is self-collected. Its clinical utility has been demonstrated for non-HPV studies [sidenote: which tests?] and for cytology. However, here is still scarce data with regards to self-collected sample for HPV DNA testing.”

Methods

Laboratory Methods—DNA extraction. Is this per manufacturers’ instruction? If so, I ‘m not sure that this level of detail is needed.

Results

The results and (discussion) seem to jump back and forth between describing HPV genotypes and presence of HPV DNA. As written, I can’t follow where kappa statistic is for presence of HPV DNA or for specific genotypes.

Discussion

Postulate 3 seems very unlikely.

Paragraphs 7 and 8 (lines 338-353) seem to go beyond the data and focus of the paper. Given this, and that Discussion gets quite lengthy, I’d suggest omitting these paragraphs.

Conclusion: “Cautiously recommend…” should be replaced with “Our data suggest that…” (see comments on abstract).

Reviewer #2: The authors of this generally well-written manuscript were interested in assessing potential anal cancer screening methods that focus on exfoliated cell collection from the anal canal. They compared clinician-collected vs self-collected anal swabs (CCAS vs SCAS) from women attending a cervical cancer screening program in Zimbabwe. The analysis found higher HPV prevalence in SCAS but a higher number of genotypes in CCAS. The method for HPV detection was next generation sequencing and this method leads to a number of questions about the results. At times, the attention to differing options for cut-off read numbers (100 vs 500 vs 1000) obscures the HPV results and calls into question the method of detecting HPV. And if the paper doesn’t have a sound method for detecting HPV, then the HPV results that depend on that detection method, e.g., comparisons of CCAS and SCAS, are called into question. Overall recommendation is to decide which cut-off level is best, defend that decision in the methods with literature, and then report results only for that cut-off level. Until that cut-off level is understood to be clearly the right one, then recommendations for anal cancer screening based on these data may be inappropriate. In addition, since anal cancer screening methods is the reason this study was done, the analysis needs to pay more attention to reporting results of high-risk HPV and results within HIV-positive people.

Abstract:

L53 – focus these results on high-risk types and difference by HIV status.

Methods:

L134 – why were Dacron swabs chosen over other options like flocked swab?

L135 – who did the breaking of the swab into the cryovial? Participant or staff?

L177 – why is 500 better than 100 or 1000? It seems to me that the results can have very different interpretations if, say 100, is chosen instead of 500. There is speculation about false-positive and false-negative results in the discussion, but it seems this should be clearly laid out by the authors in the methods.

L180 – it would be nice to see simple agreement in addition to kappa (agreement beyond chance).

L181 – 183 – this sentence is difficult to understand.

L_ - most assays assess whether or not a sample is adequate for testing. How is this assessed with NGS? Is it possible that a participant could have waived the swab in the air and this swab was judged to be HPV-negative by the authors?

Results:

Table 1 – what kind of contraception for the Yes group? Does the multiple sexual partners variable mean the participant acknowledged having >1 concurrent partner at recruitment?

Supp table 2 I don’t understand the heading: “Number of positive CCAS and SCAS GENOTYPES with cut-off……” Do you mean, “Number of HPV positive CCAS and SCAS specimens with cut-off……”?

Figure 1 A CCAS - HPV16 is typically the most prevalent genotype in other studies as it almost is for CCAS at the 100 cut-off. And it certainly is the most important type for anal cancer. Why not give preference to the most sensitive scenario for detecting HPV16? i.e., choose 100 as the cut-off. The authors state there may be too many false-positives; but the paper doesn’t seem to have a convincing argument that these HPV16 reads are mostly false-positive. Also, since you’re reporting data for three cut-offs, and this is not a methods paper about cut-offs, the message of this figure is that reasonable persons might believe that any one of these 3 cut-offs is closer to the truth.

Fig 2 A and B – the grid lines fall in odd intervals making it tough to assess the frequency for any given column. B – This figure makes no sense to me. It does not communicate the level of concordance in any way that I can understand.

L207 – “type distributions were similar regardless of …cutoffs”….this doesn’t seem to be true since HPV16 – the most important type for anal cancer - has a prevalence of ~13% in CCAS with 100 reads. Prevalence of HPV 16 is otherwise 5 or 6 percent.

L209 – reason for choosing 500 over 100 cutoff is very briefly addressed but reason for choosing 500 over 1000 is not addressed. Again, it seems this should be addressed in the methods.

L221 – overall positivity for each swabbing group is central to this paper and should appear in a table or figure. There are too many data in the results narrative that are not included in a table or figure. In this area I would also expect to see results that are specific to high-risk types since it’s a cancer outcome that the authors care about most.

L227 – why calculate kappa for all genotypes when only a few are responsible for almost all anal cancer?

L252 – It would be nice to see these data in a table stratified by HIV status.

Discussion:

L275 – Other studies have also compared CCAS and SCAS for cytology – Chin-Hong, 2008 and Cranston, 2004

L302 – Add citation for claim that HPV53 is possibly carcinogenic.

L312-313. If the observations are dependent as you suggest (which makes sense), then the kappa statistic is not appropriate since it requires independent observations. Add this to limitations.

L328 – do you mean 40% of women with HPV had multiple infections?

L360 – this range is troubling, because it implies that the McNemar test was almost significant at one of the cut-off thresholds (I presume 100), i.e., at one of the cut-off levels, the p-value was 0.06 for a difference between CCAS and SCAS.

L361 – the authors indicate here that they have adjusted the methods of the study (i.e., chosen a preferred cut-off threshold) in order to maintain statistical significance in their analysis. This is not appropriate. It brings me back to my central concern: that the authors need to justify which cut-off is most appropriate with literature. And if that’s not possible, then maybe a methods paper (that does not make recommendations on anal cancer screening) is most appropriate for these data.

L366. Other than limitation of funds, there is not a clear listing of limitations of this study. One of these might be that kappa assumes the two raters are equally competent which is not the case here.

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

PLoS One. 2021 Apr 26;16(4):e0250426. doi: 10.1371/journal.pone.0250426.r002

Author response to Decision Letter 0

17 Nov 2020

Dear Dr Michael Scheurer

We would like to thank you and the reviewers for the thorough assessment of our work and for the constructive feedback. In the following, we respond to all concerns and comments point-by-point.

Editor comments

Journal Requirements

When submitting your revision, we need you to address these additional requirements.

Editor point 1: Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

Reply: We have now formatted and renamed our documents to ensure that they meet the journal’s style.

Editor point 2: Please include additional information regarding the survey or questionnaire used in the study and ensure that you have provided sufficient details that others could replicate the analyses. For instance, if you developed a questionnaire as part of this study and it is not under a copyright more restrictive than CC-BY, please include a copy, in both the original language and English, as Supporting Information.

Reply: The questionnaire used in our study, written in both English and Shona, is now included as supplementary document 1. We have also indicated the inclusion of the document under section 2.1 L127-128.

Editor point 3: In your Methods section, please provide additional information about the participant recruitment method and the demographic details of your participants. Please ensure you have provided sufficient details to replicate the analyses such as: a) a description of any inclusion/exclusion criteria that were applied to participant recruitment. b) a statement as to whether your sample can be considered representative of a larger population, c) a description of how participants were recruited, and d) descriptions of where participants were recruited and where the research took place.

Reply: a-d) A line that briefly states that sexually active women aged at least 18 years old were reporting for routine cervical cancer screening were included is now captured in section 2.2 L145-150, whilst L124-129 inform the reader of how and where the study was carried out and to what extent the sample set was representative of a bigger population.

Editor point 4: Please provide a sample size and power calculation in the Methods, or discuss the reasons for not performing one before study initiation.

Reply: The sample size calculation is now detailed in section 2.1 L130-138. The formular accounted for a power of 0.80.

Editor point 5: To comply with PLOS ONE submission guidelines, in your Methods section, please provide additional information regarding your statistical analyses, including the threshold set for statistical significance for the analyses. For more information on PLOS ONE's expectations for statistical reporting, please see https://journals.plos.org/plosone/s/submission-guidelines.#loc-statistical-reporting.

Reply: The authors have a working R-script that can be made available upon the readers’ requests directly from the corresponding author. However a summary of the analyses performed on R-Studio is now included on section 2.2.2 L192-213.

Editor point 6: In the Methods section, please provide the specific sequences of the primers used in the next-generation sequencing analysis.

Reply: We have now included specific sequences of the primers as Table 1. L188

[Note: HTML markup is below. Please do not edit.]

Reviewer #1 comments

Reviewer #1: This is an interesting manuscript describing concordance between self- and clinician-collected anal swabs for HPV DNA testing. Given increasing incidence of anal cancer worldwide and new guidelines for anal HPV screening, self-sampling may help screening programs achieve greater coverage. Thus, there is a need for research regarding utility of self-sampling vis a vis clinician sampling. I have the following comments and suggestions:

Abstract

Reviewer #1: I suggest changing the last sentence to, “Our data suggest that self-collected anal swabs can be used as an alternative to clinician-collected anal swabs for HPV DNA testing in a clinical setting.”

Reply: We have now incorporated the suggested wording on L61-63

Introduction

Reviewer #1: Paragraph 1 would benefit from mention of burden of disease. I suggest including data on burden of cervical cancer, narrowing in on burden is sub-Saharan Africa.

Reply: We have now included facts on the burden of disease in Low-middle income countries and in Zimbabwe. This is captured on L75-78.

Reviewer #1: Paragraph 2 “Although anal cancers are relatively rare, data from USA reports a gradual increase in the incidence of anal cancers especially in HIV-infected individuals, regardless of the antiretroviral (ART) era [5]. Furthermore, a cumulative incidence of 0.1% by age 75 years was reported in HIV uninfected individuals as compared to 1.5% in HIV infected individuals.

The increase in anal cancer seems to be broader than among HIV-infected individuals. I suggest revising to: Although anal cancers are relatively rare, data from USA indicates a significant increase in the incidence of anal cancers [REF: https://doi.org/10.1093/jnci/djz219]. The increase is particularly high among HIV-infected individuals in the anti-retroviral era[include data on increasing incidence].

Side note: “The increase is particularly high among HIV-infected individuals, regardless of the antiretroviral (ART) era [5].” I’m not sure that it’s “regardless” of the anti-retroviral era, rather than “in” the antiretroviral era. Isn’t the extension of life due to ART driving increasing cancer incidence among HIV+ individuals?

Reply: We are grateful for the suggestions and we have now incorporated them on L84-86. We have also cited the paper by Deshmukh et al. Regarding the side note, we used ‘regardless of the ART era’ to indicate that in as much as ART is meant to improve the immunological profiles of an individual, we still have consistent HPV infections leading to cancer. The reviewer points out a different school of thought that describes the extension of life, due to ART, leading to cancer development. Both points are now integrated and captured in L85-88.

Reviewer #1: Paragraph 1 and 2 jump around quite a bit. Also, since the paper doesn’t seek to fill the gap in data regarding burden of anal cancer in Zimbabwe, I don’t think it makes sense to set this up as an argument in the Intro. How about ending paragraph 1 with “Women with a history of cervical cancer, especially those with HPV/HIV co-infection, are at increased risk of developing anal cancers [4]. Screening for anal cancer is thus recommended for women with cervical lesions and/or who are HIV infected [9, 10].” And starting paragraph 2 with “Although anal cancers are relatively rare, data from USA indicates a significant increase in the incidence of anal cancers. The increase is particularly high among HIV-infected individuals in the anti-retroviral era. Zimbabwe has an HIV prevalence of almost 13% and cervical cancer contributes a third of all cancers among black Zimbabwean women. In 2016, the Zimbabwe Cancer Registry reported a gradual increase in anal cancers recorded in the past decade. Despite the increased prevalence of risk factors in the population and evidence of increasing incidence, anal cancer screening guidelines are not routinely followed in Zimbabwe.”

Reply: Paragraph 1 now ends with the sentence on screening for anal cancer and paragraph 2 begins with the incidence of anal cancer in USA. L82-86.

Reviewer #1: Given focus of paper on HPV genotypes and multiple infections, it would be helpful to reader to understand why different genotypes matter. Could this be woven into the intro? This is mentioned in Discussion, but I’d move to Intro

Reply: We agree with the reviewer, for the reader to grasp the importance of reporting HPV genotypes right from the beginning, we have now moved the lines from discussion to the last paragraph of the introduction. L112-119.

Reviewer #1: Paragraph 3 jumps between accuracy and acceptability and between anal and cervical swabs. I suggest walking the reader through the arguments to better set up the rationale for the study. Eg. “Utility of self-sample cervical swabs for HPV DNA testing has been demonstrated. Furthermore, self-collection of cervical swabs has shown to be highly acceptable. For anal screening, there is evidence that self-sampling is acceptable and patients tend to be more comfortable and less embarrassed when an anal swab is self-collected. Its clinical utility has been demonstrated for non-HPV studies [sidenote: which tests?] and for cytology. However, here is still scarce data with regards to self-collected sample for HPV DNA testing.”

Reply: With the reviewer 1’s suggested phrasing, we have now re-written paragraph 3 and improved on the general flow. L95-100.

Methods

Reviewer #1: Laboratory Methods—DNA extraction. Is this per manufacturers’ instruction? If so, I ‘m not sure that this level of detail is needed.

Reply: Yes, the extraction was as per manufacturer’s instruction. The extra detail on DNA extraction has now been deleted L164-166

Results

Reviewer #1: The results and (discussion) seem to jump back and forth between describing HPV genotypes and presence of HPV DNA. As written, I can’t follow where kappa statistic is for presence of HPV DNA or for specific genotypes.

Reply: To avoid losing the reader and to improve on the flow we have now used consistent terminology, “HPV genotypes” in all sections. The Kappa test is for presence of any HPV genotype.

Discussion

Reviewer #1: Postulate 3 seems very unlikely.

Reply: We have kept postulate 3 in the text and added the word “unlikely” to guide the readers. L332

Reviewer #1: Paragraphs 7 and 8 (lines 338-353) seem to go beyond the data and focus of the paper. Given this, and that Discussion gets quite lengthy, I’d suggest omitting these paragraphs.

Reply: Paragraphs 7 describes the distribution of HPV genotypes by HIV status and paragraph 8 briefly describes cervical disease (as detected by Visual-inspection-with-acetic acid). Inasmuch as these two paragraphs mainly discuss secondary data, they are important for the reader to have a full perspective of the sample population. Given that women who have a history of cervical lesions and/or HIV co-infection are at increased risk of anal HPV infection. We have re-written the beginning of paragraph 8, L342, to capture the relevance of the paragraphs.

Reviewer #1: Conclusion: “Cautiously recommend…” should be replaced with “Our data suggest that…” (see comments on abstract).

Reply: We agree with the reviewer, therefore the conclusion is now phrased as ‘our data suggest that….’ L61 and L375.

Reviewer #2 comments

Reviewer #2: The authors of this generally well-written manuscript were interested in assessing potential anal cancer screening methods that focus on exfoliated cell collection from the anal canal. They compared clinician-collected vs self-collected anal swabs (CCAS vs SCAS) from women attending a cervical cancer screening program in Zimbabwe. The analysis found higher HPV prevalence in SCAS but a higher number of genotypes in CCAS. The method for HPV detection was next generation sequencing and this method leads to a number of questions about the results. At times, the attention to differing options for cut-off read numbers (100 vs 500 vs 1000) obscures the HPV results and calls into question the method of detecting HPV. And if the paper doesn’t have a sound method for detecting HPV, then the HPV results that depend on that detection method, e.g., comparisons of CCAS and SCAS, are called into question. Overall recommendation is to decide which cut-off level is best, defend that decision in the methods with literature, and then report results only for that cut-off level. Until that cut-off level is understood to be clearly the right one, then recommendations for anal cancer screening based on these data may be inappropriate. In addition, since anal cancer screening methods is the reason this study was done, the analysis needs to pay more attention to reporting results of high-risk HPV and results within HIV-positive people.

Abstract

Reviewer #2: L53 – focus these results on high-risk types and difference by HIV status.

Reply: We summarise the detection of high-risk HPV genotypes in the discussion section, L302-311 and L348-351. The primary aim and scope of this manuscript was to determine whether the same HPV genotypes were detected in self-collected and clinician-collected swabs, therefore we reported all the genotypes, regardless of whether they are high risk or low-risk. In addition, the most detected genotypes did not include a large number of high-risk genotypes. This is probably because these were mostly ‘disease free’ women reporting for routine screening, a good sample population for the purposes of primarily comparing CCAS to SCAS. We agree that in future studies, where we will focus on HPV genotype distribution other than comparison of collection methods, we can explore in-depth the anal high-risk HPV genotypes and HIV status.

Methods

Reviewer #2: L134 – why were Dacron swabs chosen over other options like flocked swab?

Reply: Our choice was based on literature that reported dacron and flocked swabs having harvested similar amounts of cells. Although in some scenarios flocked swabs are deemed superior, the difference with dacron swabs is minimal. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5717820/ and https://pubmed.ncbi.nlm.nih.gov/21791907/. In our setting dacron swabs are the most readily available synthetic swabs, therefore given their proven use, we chose them to avoid inconsistency. The likelihood of running out of flocked swabs and having to wait for very long periods for new procurements would have been high. The citations used for justifying the choice of dacron swabs are now included on L145.

Reviewer #2: L150 – who did the breaking of the swab into the cryovial? Participant or staff?

Reply: The collection procedure was explained to all participants by the recruiting nurse. Self-collected swabs were broken into the tube by the participant, immediately after swabbing. Clinician-collected swabs were broken into the tube by the clinician as soon as they completed the swabbing. This is now indicated under section 2.2 L155.

Reviewer #2: L177 – why is 500 better than 100 or 1000? It seems to me that the results can have very different interpretations if, say 100, is chosen instead of 500. There is speculation about false-positive and false-negative results in the discussion, but it seems this should be clearly laid out by the authors in the methods.

Reply: We agree with the reviewer that presenting more than 1 cut-off can be confusing to the readers, thus we have carefully chosen to report all our findings based on a 500 read pairs cut-off. Our choice is based on the analyses of our read pairs which had a mean of 26232 for CCAS and 19110 for SCAS. Due to the difference in means, instead of normalising we calculated for a slightly higher cut-off. Assuming limit of detection of 5%, our 500 cut-off allows us to filter any false positive results and aligns with clinical recommendations of 300-500 published by Petrackova and colleagues https://pubmed.ncbi.nlm.nih.gov/31552176/. Although, there are a few papers that suggest varied cut-offs, Meisal et al DOI:10.1371/journal.pone.0169074 and Cullen et al doi: 10.1016/j.pvr.2015.05.004 , there is generally no consensus on the minimum required read counts using deep targeted sequencing therefore each laboratory sets its own parameters based on the depth of coverage, the average read pairs and the statistical analyses methods employed. We have now clearly presented our justification (L199-202) and consistently only reported results using a cut-off that’s at least 500 read pairs. Fortunately, one of the co-authors is now preparing a methods paper to suggest the best cut-offs for our laboratory.

Reviewer #2: L180 – it would be nice to see simple agreement in addition to kappa (agreement beyond chance).

Reply: We present both the Cohen’s Kappa and the McNemer because they complement each other. Kappa assumes the two raters are equally competent whilst McNemar allows for unequal raters. We agree that the Kappa assumes the observed agreement to be due to chance. Therefore we included a McNemar test to enhance the agreement analysis and to ensure a more valid interpretation, L210. The McNemar assumes that HPV genotypes detected from self-collected swabs are equal to those detected by clinician-collected swabs. To the best of our knowledge, the complementary effect of these two tests is sufficient to describe the agreement of the collection methods https://pubmed.ncbi.nlm.nih.gov/18188809/ and https://pubmed.ncbi.nlm.nih.gov/10520333/.

Reviewer #2: L181 – 183 – this sentence is difficult to understand.

Reply: The sentence has now been rephrased. We have now specifically described how the plots were made using the HPV genotype frequencies. L201-203.

Reviewer #2: L_ - most assays assess whether or not a sample is adequate for testing. How is this assessed with NGS?

Reply: NGS detects betaglobin as part of a control step to check if the DNA is adequate for testing. However, our samples were treated differently because they had already undergone betaglobin testing as an upstream procedure for a different assay (dot-blot-hybridisation) in Zimbabwe, for a different research scope. This was not repeated before NGS. We have now indicated this in the methods section. L166-168

Reviewer #2: Is it possible that a participant could have waived the swab in the air and this swab was judged to be HPV-negative by the authors?

Reply: This is highly unlikely, because the participants had the procedure explained to them by the nurse, who also closely monitored any women who requested supervision. As soon as the swab was collected it was transferred into the cryotube.

Results

Reviewer #2: Table 1 – what kind of contraception for the Yes group?

Reply: This question referred to any type of contraception. A follow-up question also captured the specific contraception; however, this detailed information is not relevant for the scope of this manuscript. Please see question 14 of the questionnaire, attached as supplementary document 1.

Reviewer #2: Does the multiple sexual partners variable mean the participant acknowledged having >1 concurrent partner at recruitment?

Reply: Yes, this is correct. The participants self-reported having more than 1 sexual partners at the time of recruitment. Please see the questionnaire attached as supplementary material on question numbers 9 to 11.

Reviewer #2: Supp table 2 I don’t understand the heading: “Number of positive CCAS and SCAS GENOTYPES with cut-off……” Do you mean, “Number of HPV positive CCAS and SCAS specimens with cut-off……”?

Reply: Thank you for this note, this was a typographic error. The caption for supplementary table 2 (which is now table 3) has now been corrected to “specimens” instead of “genotypes”. L226

Reviewer #2: Figure 1 A CCAS - HPV16 is typically the most prevalent genotype in other studies as it almost is for CCAS at the 100 cut-off. And it certainly is the most important type for anal cancer. Why not give preference to the most sensitive scenario for detecting HPV16? i.e., choose 100 as the cut-off. The authors state there may be too many false-positives; but the paper doesn’t seem to have a convincing argument that these HPV16 reads are mostly false-positive. Also, since you’re reporting data for three cut-offs, and this is not a methods paper about cut-offs, the message of this figure is that reasonable persons might believe that any one of these 3 cut-offs is closer to the truth.

Reply: We have considered your suggestions and settled for a cut-off of 500. This is based on the explanation given above, for your question on L177. Concerning HPV 16 prevalence, indeed there seems to be a reduction in the cases but it was not statistically significant. Furthermore, with NGS we expect the sensitivity to be standard for all genotypes. We are also of the view that presenting an optimum cut-off will be better regardless of the loss of seemingly ‘positive’ samples because if they do not meet the optimum cut-off, we consider them to be negative. We also agree that presenting more than one cut-off is not suited for the scope of this paper. These will now be presented in a different methods manuscript that some of the co-authors are working on. To avoid misleading the readers, we have removed Figure 1. This will be best suited for a methods manuscript.

Reviewer #2: Fig 2 A and B – the grid lines fall in odd intervals making it tough to assess the frequency for any given column. B – This figure makes no sense to me. It does not communicate the level of concordance in any way that I can understand.

Reply: The authors have taken a closer look at figure 2 (which is now renamed to Figure 1) and we agree that the way the frequencies were presented may have been confusing. We have now relabelled the y-axis to present frequencies as whole numbers as opposed to the 0.5 intervals. The main message conveyed by the figure is the difference of frequencies of the HPV genotypes detected by each method (A), whilst (B) further gives a pictorial illustration of instances when an HPV genotype was detected by either of the collection methods or by both.

Reviewer #2: L207 – “type distributions were similar regardless of …cutoffs”….this doesn’t seem to be true since HPV16 – the most important type for anal cancer - has a prevalence of ~13% in CCAS with 100 reads. Prevalence of HPV 16 is otherwise 5 or 6 percent. L209 – reason for choosing 500 over 100 cutoff is very briefly addressed but reason for choosing 500 over 1000 is not addressed. Again, it seems this should be addressed in the methods.

Reply: Similar to the response to Figure 1 above. We have now justified the use of 500 as a cut-off, therefore we have now discarded the need to compare HPV genotypes based on cut-offs, as we agree that this is not a methods paper.

Reviewer #2: L221 – overall positivity for each swabbing group is central to this paper and should appear in a table or figure. There are too many data in the results narrative that are not included in a table or figure. In this area I would also expect to see results that are specific to high-risk types since it’s a cancer outcome that the authors care about most.

Reply: Regarding the overall positivity for each swabbing group, we have now presented these in table 3 L229, previously labelled as supplementary table 1 and 2. We agree with the reviewer, cancer outcome is a critical end point which we wish to prevent by providing data that can offer more screening options. In the context of cancer development, high-risk genotypes have the most clinical relevance. However, the primary aim and scope of this manuscript was to determine whether the same HPV genotypes were detected in self-collected and clinician-collected swabs, therefore we reported all the genotypes including those with less clinical relevance. Having determined and described the agreement using a broad array of HPV genotypes, we envision that when applications are being made in a clinical setting, the clinician will receive a report specifying the clinically relevant HPV genotypes.

Reviewer #2: L227 – why calculate kappa for all genotypes when only a few are responsible for almost all anal cancer?

Reply: The reviewer has rightfully highlighted that high-risk HPV genotypes have the most clinical relevance. We used Illumina next-generation-sequencing which detects a broad array of HPV genotypes, thus our overall HPV anal tests should report good accuracy for all HPV genotypes. The primary aim and scope of this manuscript was to determine whether the same HPV genotypes were detected in self-collected and clinician-collected swabs, therefore we reported all the genotypes including those with less clinical relevance.

Reviewer #2: L252 – It would be nice to see these data in a table stratified by HIV status.

Reply: The scope of this manuscript is to compare two methods and we do not expect the HIV status of the women to affect the accuracy of the test, therefore we do not present the HPV genotype agreement of CCAS and SCAS by HIV status. However, we give a summary description of the HIV status of the sample population based on the gold-standard, which is the clinician-collected in this case. This gives the reader a good indication of the HPV/HIV co-infection in this group of women. Sections 3.3 L260-270.

Discussion

Reviewer #2: L275 – Other studies have also compared CCAS and SCAS for cytology – Chin-Hong, 2008 and Cranston, 2004

Reply: Thank you for highlighting these important studies to us. The line has now been rephrased and the citations are now updated L98 and L282-283.

Reviewer #2: L302 – Add citation for claim that HPV53 is possibly carcinogenic.

Reply: A citation (Meyer et al 2001) has now been added and the phrase is also slightly modified to “probable high-risk”. L309-310.

Reviewer #2: L312-313. If the observations are dependent as you suggest (which makes sense), then the kappa statistic is not appropriate since it requires independent observations. Add this to limitations.

Reply: We have now highlighted this limitation on L363-370.

Reviewer #2: L328 – do you mean 40% of women with HPV had multiple infections?

Reply: Yes, 40% of women with HPV had multiple infections. We have added the phrase “of HPV infected” to clearly capture this L335.

Reviewer #2: L360 – this range is troubling, because it implies that the McNemar test was almost significant at one of the cut-off thresholds (I presume 100), i.e., at one of the cut-off levels, the p-value was 0.06 for a difference between CCAS and SCAS.

Reply: Thank you for this note. We have chosen our best cut-off to be 500, as described in the earlier responses in this document. We have now removed the comparisons that may confuse the reader. We agree that at one of the cut-offs the p-value of 0.06 was boarder line significant, therefore it should be confirmed using a larger sample size in other future studies.

Reviewer #2: L361 – the authors indicate here that they have adjusted the methods of the study (i.e., chosen a preferred cut-off threshold) in order to maintain statistical significance in their analysis. This is not appropriate. It brings me back to my central concern: that the authors need to justify which cut-off is most appropriate with literature. And if that’s not possible, then maybe a methods paper (that does not make recommendations on anal cancer screening) is most appropriate for these data.

Reply: We agree that our inferences may have not been well spelt out. The comparison of cut-offs was not appropriate for this paper. We will publish such comparisons in a different paper. To avoid any misinterpretations by the reader we have now presented all our data based on the optimum cut-off of 500 as justified in the response to similar comments above.

Reviewer #2: L366. Other than limitation of funds, there is not a clear listing of limitations of this study. One of these might be that kappa assumes the two raters are equally competent which is not the case here.

Reply: We have now collated our limitations and presented them on L360-366 and L363-370. We present the Kappa and McNemer because they complement each other. Kappa assumes the two raters are equally competent whilst McNemar allows for unequal raters.

Attachment

Submitted filename: v2Response to reviewers Plos.docx

Click here for additional data file. (38.6KB, docx)

Decision Letter 1

Michael Scheurer

31 Dec 2020

PONE-D-20-19224R1

Self-collected and clinician-collected anal swabs show modest agreement for HPV genotyping

PLOS ONE

Dear Dr. Dube Mandishora,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Please submit your revised manuscript by Feb 14 2021 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

  • A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

  • A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

  • An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

We look forward to receiving your revised manuscript.

Kind regards,

Michael Scheurer, Ph.D.

Academic Editor

PLOS ONE

Additional Editor Comments (if provided):

In addressing the remaining reviewers' comments below, please also ensure that you have cited the correct references to support the use of the statistical tests that you used.

Please confirm that your manuscript meets the PLOS One data submission/availability policy. In particular, the policy states that "data points used to create, for example, means and medians, need to be available."

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: (No Response)

Reviewer #2: (No Response)

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Partly

Reviewer #2: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: No

Reviewer #2: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: (No Response)

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: The authors have satisfactorily addressed all of my previous comments. However, I have some concerns that I had overlooked regarding the statistical analysis, specifically the use of McNamar’s test as a test of agreement (see below). The authors conclude that there is moderate agreement (based on kappa=0.55), but state that this was not statistically significant based on McNamar’s test. However, McNamar’s is a test of proportions, rather than agreement. Thus, the p-value used to support the conclusion is only able to conclude that the proportion of a particular genotype was not significantly different between CCAS and SCAS, rather than indicating agreement. My other comments relate to minor issues that should be easy to address.

Abstract: There’s a small typo in conclusions: “with no statistically-significant difference.”

Intro:

I appreciate the authors’ clarification of the increasing incidence of anal cancer among HIV-infected individuals in the era of ART: “The increase is particularly high among HIV-infected

individuals regardless of a repaired immune system due to ART, and conversely, extension

to life due to ART tends to give time for cancer to develop [9].” Just to clarify, is the null association at the individual level (HIV+ individuals on ART are as likely than those not on ART to develop anal cancer) or at the ecologic level (despite advances and coverage in ART, anal cancer incidence continues to rise among HIV+)? If the latter, I think it may be better to say something like, “The increase is particularly high among HIV-infected individual despite the availability and widespread (?)/ increasing (?) use of ART, which improves individuals’ immunogenic profile. Conversely, ART extends life among HIV infected individuals, giving time for cancer to develop.”

Also, please define ART on first use

101-103: The last sentence of this paragraph is repeated from above: “Furthermore,

patients tend to be more comfortable and less embarrassed when an anal swab is self

collected [21].

112-13: please define SCAS and CCAS on first use

Methods:

144-148 The information in this paragraph seems redundant with the paragraph below. I suggest omitting this paragraph and integrating any new information provided into the paragraph below.

Agreement was tested using kappa and McNamar’s test. However, McNamar’s test compares overall proportions, not agreement (see: (see: Ranganathan P, Pramesh CS, Aggarwal R. Common pitfalls in statistical analysis: Measures of agreement. Perspect Clin Res. 2017 Oct-Dec;8(4):187-191.). I’m not confident that this test can be used to support the authors’ conclusions.

Results:

Table 3 says CCAS: 124 HPV + samples, SCAS:114 HPV + samples. But paragraph below says 67 and 75, respectively. What is Table 3, last column referring to?

249-250: This sentence is confusing and I’m not sure of relevance: In addition, 13/29 of these women had multiple infections in both SCAS and CCAS regardless of HPV genotype and number of infections.

Figure 3. I’m confused as to why there’s a single p-value for the McNamar’s test. McNamar’s is used for paired nominal data in a 2x2 contingency table. So, for each genotype CCAS vs. SCAS and genotype present versus genotype absent. It seems like there should be a test statistic and associated p-value for each genotype-comparison, rather than a global McNamar’t test statistic and p-value.

Discussion:

327-334: The same has been observed for self-collected versus clinician-collected cervical swabs [See Arbyn et al 2018-- https://pubmed.ncbi.nlm.nih.gov/30518635/], attributed to vaginal sampling with self-collection versus cervical sampling with CC.

The author’s conclude that there is moderate agreement (based on kappa=0.55), but state that this was not statistically significant, based on McNamar’s test. However, McNamar’s isn’t a test of agreement. Thus, I’m not certain that the data support this conclusion.

Reviewer #2: Thanks for addressing my comments. I appreciate the authors complete and thoughtful replies. I would only point out that the following critique does seem to have been addressed in the revised manuscript.

L312-313. If the observations are dependent as you suggest (which makes sense), then the kappa statistic is not appropriate since it requires independent observations. Add this to limitations.

Reply: We have now highlighted this limitation on L363-370.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

PLoS One. 2021 Apr 26;16(4):e0250426. doi: 10.1371/journal.pone.0250426.r004

Author response to Decision Letter 1

14 Feb 2021

Dear Dr Michael Scheurer

We would like to thank you and the reviewers for the additional comments on our work and for the constructive feedback. In the following, we respond to all concerns and comments point-by-point.

Additional Editor Comments

Comment 1: In addressing the remaining reviewers' comments below, please also ensure that you have cited the correct references to support the use of the statistical tests that you used.

Response: Additional references have now been included, such as Ranganathan P, Pramesh CS, Aggarwal R. Common pitfalls in statistical analysis: Measures of agreement. Perspect Clin Res. 2017 Oct-Dec;8(4):187-191.

Comment 2: Please confirm that your manuscript meets the PLOS One data submission/availability policy. In particular, the policy states that "data points used to create, for example, means and medians, need to be available."

Response: The authors have a working R-script that can be made available upon the readers’ requests directly from the corresponding author.

Reviewer #1 Comments to the Author

Comment 1: The authors have satisfactorily addressed all of my previous comments. However, I have some concerns that I had overlooked regarding the statistical analysis, specifically the use of McNemar’s test as a test of agreement (see below). The authors conclude that there is moderate agreement (based on kappa=0.55), but state that this was not statistically significant based on McNemar’s test. However, McNemar’s is a test of proportions, rather than agreement. Thus, the p-value used to support the conclusion is only able to conclude that the proportion of a particular genotype was not significantly different between CCAS and SCAS, rather than indicating agreement. My other comments relate to minor issues that should be easy to address.

Response: Thank you for highlighting the interpretation of these statistical analyses. To avoid confusion for the readers, the authors have now separated the two statistical analyses to represent (1) measure of agreement (L54-55) and (2) test of proportions (L55). This has also been applied throughout the manuscript in the methods (L223-225), results (L253-255) and discussion (L314-318). The tests are not used as one but rather as two separate complimentary tests.

Comment 2: Abstract: There’s a small typo in conclusions: “with no statistically-significant difference.”

Response:

We have now deleted the “none” and used “no” on L58.

Introduction

Comment 3: I appreciate the authors’ clarification of the increasing incidence of anal cancer among HIV-infected individuals in the era of ART: “The increase is particularly high among HIV-infected individuals regardless of a repaired immune system due to ART, and conversely, extension to life due to ART tends to give time for cancer to develop [9].” Just to clarify, is the null association at the individual level (HIV+ individuals on ART are as likely than those not on ART to develop anal cancer) or at the ecologic level (despite advances and coverage in ART, anal cancer incidence continues to rise among HIV+)? If the latter, I think it may be better to say something like, “The increase is particularly high among HIV-infected individual despite the availability and widespread (?)/ increasing (?) use of ART, which improves individuals’ immunogenic profile. Conversely, ART extends life among HIV infected individuals, giving time for cancer to develop.” Also, please define ART on first use 101-103: The last sentence of this paragraph is repeated from above: “Furthermore, patients tend to be more comfortable and less embarrassed when an anal swab is self-collected [21].

Response: Yes, the intention was to explain the latter scenario. The null association is at an ecological level. We have therefore edited the paragraph to include the reviewer’s useful suggestions, L86-88. The acronym ART has now been introduced in L87. The repeated sentence has now been deleted from line 108.

Comment 4: 112-13: please define SCAS and CCAS on first use

Response: The acronyms SCAS and CCAS are now introduced on L121, in addition to the definition that was provided in the abstract L40.

Methods

Comment 5: 144-148 The information in this paragraph seems redundant with the paragraph below. I suggest omitting this paragraph and integrating any new information provided into the paragraph below.

Agreement was tested using kappa and McNamar’s test. However, McNamar’s test compares overall proportions, not agreement (see: (see: Ranganathan P, Pramesh CS, Aggarwal R. Common pitfalls in statistical analysis: Measures of agreement. Perspect Clin Res. 2017 Oct-Dec;8(4):187-191.). I’m not confident that this test can be used to support the authors’ conclusions.

Response: We agree with the reviewer, stating that agreement was tested using kappa and McNemar can be misleading. To remove this confusion, we have now separated the two statistical analyses to represent (1) measure of agreement (L54-55) and (2) test of proportions (L55). This is also explained in the responses above. The authors believe that both statistical tests are useful. We are also conscious of the fact that there is no perfect test to evaluate agreement and it’s a classical source of debate. Both tests are commonly used together to complement each other https://pubmed.ncbi.nlm.nih.gov/10520333/. McNemar is indeed a test of proportion, but the proportion of discordant pairs (ie +/- vs -/+), which gives useful information when evaluating (dis)agreement based on proportions. Indeed, as mentioned by Ranganathan et al, a non-significant McNemar test is not enough to say the agreement is good. It’s however still useful, if interpreted correctly. We have now further clarified our interpretations and limitations in L399-405 of the Discussion section.

Results

Comment 6: Table 3 says CCAS: 124 HPV + samples, SCAS:114 HPV + samples. But paragraph below says 67 and 75, respectively. What is Table 3, last column referring to?

Response: We apologise for the misleading column. The 124 and 114 refer to every HPV positivity regardless of multiple infections. In other words, for CCAS, the total read pairs amounted to 124 HPV genotypes in 67 women. This is because some of the 67 women had multiple HPV genotypes, such that there were 32 different HPV genotypes repeatedly detected in 67 women. However, to avoid confusing the reader, the column is now titled “number of HPV positive women”.

Comment 7: 249-250: This sentence is confusing and I’m not sure of relevance: In addition, 13/29 of these women had multiple infections in both SCAS and CCAS regardless of HPV genotype and number of infections.

Response: We have now removed the statement.

Comment 8: Figure 3. I’m confused as to why there’s a single p-value for the McNamar’s test. McNamar’s is used for paired nominal data in a 2x2 contingency table. So, for each genotype CCAS vs. SCAS and genotype present versus genotype absent. It seems like there should be a test statistic and associated p-value for each genotype-comparison, rather than a global McNamar’t test statistic and p-value.

Response: The reviewer raises an important question which fortunately we also touched on in the first rebuttal. In our data, each genotype was considered independent because we were most interested in knowing how much HPV was detected in either methods or in comparison. Given the broad array of HPV genotypes that the Illumina sequencer could detect, it was complicated to break down our analyses into each and every HPV genotype. Discordant results were defined as those that were HPV positive on one specimen type while negative on the other. However, we have highlighted this under limitations L294-299 and L371-384. Furthermore, NGS is a highly sensitive assay (DOI:10.1038/s41598-018-36669-6 ) and we do not expect a difference in the capacity for detecting HPV genotypes assuming there are infected cells on the swabs, rather, our research question is whether each collection method harvests enough cells for HPV genotypes to be detected, when comparing two detection methods.

Discussion

Comment 9: 327-334: The same has been observed for self-collected versus clinician-collected cervical swabs [See Arbyn et al 2018-- https://pubmed.ncbi.nlm.nih.gov/30518635/], attributed to vaginal sampling with self-collection versus cervical sampling with CC.

Response: Thank you for this important reference. We have included it in our discussion, L340-342

Comment 10: The authors conclude that there is moderate agreement (based on kappa=0.55), but state that this was not statistically significant, based on McNamar’s test. However, McNamar’s isn’t a test of agreement. Thus, I’m not certain that the data support this conclusion.

Response: We agree with the reviewer. Yes, the McNemar is best referred to as a method to ‘evaluate agreement’, as opposed to a ‘test of agreement’. We have now carefully rephrased the abstract and conclusion to fully portray what both statistical tests do and mean. This has now been applied throughout the manuscript in the methods (L210-212), results (L238-240) and discussion (L314-318). We have also toned down our conclusion to ensure that the readers fully understand what our results mean L287-293.

Reviewer #2 Comments to the Author

Comment 1: Thanks for addressing my comments. I appreciate the authors complete and thoughtful replies. I would only point out that the following critique does seem to have been addressed in the revised manuscript.

L312-313. If the observations are dependent as you suggest (which makes sense), then the kappa statistic is not appropriate since it requires independent observations. Add this to limitations.

Reply: We have now highlighted this limitation on L363-370.

Response: Thank you for the comment, but unfortunately there is probably a mix up on the line numbering and we are not sure if we have fully understood your comment. Nonetheless, as mentioned above, the authors believe that both statistical tests are useful. We are also conscious of the fact that there is no perfect test to evaluate agreement. Both tests are commonly used together to complement each other. However, we have spelt out the limitations in L294-299 and L371-384. If the reviewer is talking about the different hypotheses for higher detection with SC, (because L302 corresponds to L232 on the original pdf and L328 to L259), then the fact that the second swab is influenced by the first one, is just a hypothesis, the observations (SC and CC) can be considered independent. We have reiterated our speculative hypotheses for these differences in L322-323.

Attachment

Submitted filename: V2_2nd Response to reviewers_100221.doc

Click here for additional data file. (41.5KB, doc)

Decision Letter 2

Michael Scheurer

19 Mar 2021

PONE-D-20-19224R2

Self-collected and clinician-collected anal swabs show modest agreement for HPV genotyping

PLOS ONE

Dear Dr. Dube Mandishora,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Please address the reviewer's suggestions for edits to clarify information presented to the reader in the methods and the discussion sections.

Please submit your revised manuscript by May 03 2021 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

  • A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

  • A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

  • An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

We look forward to receiving your revised manuscript.

Kind regards,

Michael Scheurer, Ph.D.

Academic Editor

PLOS ONE

Journal Requirements:

Please review your reference list to ensure that it is complete and correct. If you have cited papers that have been retracted, please include the rationale for doing so in the manuscript text, or remove these references and replace them with relevant current references. Any changes to the reference list should be mentioned in the rebuttal letter that accompanies your revised manuscript. If you need to cite a retracted article, indicate the article’s retracted status in the References list and also include a citation and full reference for the retraction notice.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: (No Response)

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: The authors have fully addressed all of my comments and suggestions. I have a few minor suggestions related to how the kappa and McNemar’s are presented and a few minor edits in the Discussion.

Abstract (and throughout) Thank you for clarifying that kappa and McNamar’s were two separate tests used to evaluate concordance. For the sake of clarity, could the following edits be made to the abstract:

Methods: Currently: Level of agreement was calculated using the kappa and McNemar tests. Suggested: Level of agreement of HPV genotypes between CCAS and SCAS was calculated using the kappa statistic. McNemar tests were used to evaluate agreement in the proportion of genotypes detected by either method.

The same info is available elsewhere in the abstract, but it would improve clarity to include it in the Methods so that readers can better understand the Results.

Methods: Similarly, I’d clarify how the statistics were used in the last paragraph of the Methods. Currently: To evaluate agreement of the two collection methods, the McNemar non-parametric test was performed. A kappa test for agreement was also performed to complement the McNemar [38–40]. Suggested: A kappa test was used to evaluate agreement of HPV genotypes across the two collection methods. The McNemar’s non-parametric test was performed to evaluated agreement in the proportion of genotypes detected by either method.

Discussion, line 294-300: Current: Based on the moderate level of agreement of HPV genotypes between the two methods, signified by 0.55 in kappa value (k), our data suggest that researchers can use either methods for collection. Furthermore, to evaluate agreement of the methods, McNemar gave a Chi-square value of 0.75 (p=0.39), which showed that the difference in detection rates on the two methods was not statistically significant. Suggested: Based on the moderate level of agreement of HPV genotypes between the two methods, indicated by 0.55 in kappa value (k) and a non-significant test of proportions, our data suggest that researchers can use either methods for collection.

Line 351-353 (minor edit): Arbyn et al had similar observations, reporting HPV detection rates 2.28 times higher in self-collected versus clinician-collected vaginal swabs [45].

Line 384-389, Current: Indeed, as mentioned by Ranganathan et al, a non-significant McNemar test is not always enough to say the agreement is good. It’s however still useful, if interpreted correctly [51]. In our case, the McNemar is indeed a test of proportion, but the proportion of discordant pairs (ie +/- vs -/+), which gave useful information for evaluating (dis)agreement of HPV genotype detection. Suggested: Indeed, as mentioned by Ranganathan et al, a non-significant McNemar test alone is not sufficient to indicate good agreement. However, it is still a useful statistic, if interpreted correctly [51]. In our case, the McNemar test indicated that the proportion of discordant pairs (ie +/- vs -/+) was not statistically significant, providing useful information for evaluating (dis)agreement of HPV genotype detection.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: Yes: Jane R Montealegre

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

PLoS One. 2021 Apr 26;16(4):e0250426. doi: 10.1371/journal.pone.0250426.r006

Author response to Decision Letter 2

31 Mar 2021

Dear Dr Michael Scheurer

We would like to thank you and the reviewers for the additional comments on our work and for the constructive feedback. In the following, we respond to all concerns and comments point-by-point.

Review # 1 Comments to the Author

Comment 1: The authors have fully addressed all of my comments and suggestions. I have a few minor suggestions related to how the kappa and McNemar’s are presented and a few minor edits in the Discussion.

Abstract (and throughout) Thank you for clarifying that kappa and McNamar’s were two separate tests used to evaluate concordance. For the sake of clarity, could the following edits be made to the abstract:

Methods: Currently: Level of agreement was calculated using the kappa and McNemar tests. Suggested: Level of agreement of HPV genotypes between CCAS and SCAS was calculated using the kappa statistic. McNemar tests were used to evaluate agreement in the proportion of genotypes detected by either method.

The same info is available elsewhere in the abstract, but it would improve clarity to include it in the Methods so that readers can better understand the Results.

Response: We have now included the suggested phrasing on L48-51.

Comment 2: Methods: Similarly, I’d clarify how the statistics were used in the last paragraph of the Methods. Currently: To evaluate agreement of the two collection methods, the McNemar non-parametric test was performed. A kappa test for agreement was also performed to complement the McNemar [38–40]. Suggested: A kappa test was used to evaluate agreement of HPV genotypes across the two collection methods. The McNemar’s non-parametric test was performed to evaluated agreement in the proportion of genotypes detected by either method.

Response: We have now included the suggested phrasing on L213-216.

Comment 3: Discussion, line 294-300: Current: Based on the moderate level of agreement of HPV genotypes between the two methods, signified by 0.55 in kappa value (k), our data suggest that researchers can use either methods for collection. Furthermore, to evaluate agreement of the methods, McNemar gave a Chi-square value of 0.75 (p=0.39), which showed that the difference in detection rates on the two methods was not statistically significant. Suggested: Based on the moderate level of agreement of HPV genotypes between the two methods, indicated by 0.55 in kappa value (k) and a non-significant test of proportions, our data suggest that researchers can use either methods for collection.

Response: We have now included the suggested phrasing on L294-297.

Comment 4: Line 351-353 (minor edit): Arbyn et al had similar observations, reporting HPV detection rates 2.28 times higher in self-collected versus clinician-collected vaginal swabs [45].

Response: We have now added the word “versus” to the sentence, L343-345.

Comment 5: Line 384-389, Current: Indeed, as mentioned by Ranganathan et al, a non-significant McNemar test is not always enough to say the agreement is good. It’s however still useful, if interpreted correctly [51]. In our case, the McNemar is indeed a test of proportion, but the proportion of discordant pairs (ie +/- vs -/+), which gave useful information for evaluating (dis)agreement of HPV genotype detection. Suggested: Indeed, as mentioned by Ranganathan et al, a non-significant McNemar test alone is not sufficient to indicate good agreement. However, it is still a useful statistic, if interpreted correctly [51]. In our case, the McNemar test indicated that the proportion of discordant pairs (ie +/- vs -/+) was not statistically significant, providing useful information for evaluating (dis)agreement of HPV genotype detection.

Response: We have now rephrased L388-393, to in cooperate the suggestion.

Additional edits by authors: Additional information on author contributions has now been included on L428-431, to indicate the unfortunate loss of one of our co-authors BSP, who passed away before the submission of the final version. We have also taken the opportunity to thank the reviewers for their contributions in remoulding this manuscript, L408 specifically mentions Prof. Jane Montealegre and those who also remained anonymous.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: Yes: Jane R Montealegre

Attachment

Submitted filename: V1_3rd Response to reviewers.docx

Click here for additional data file. (17.7KB, docx)

Decision Letter 3

Michael Scheurer

7 Apr 2021

Self-collected and clinician-collected anal swabs show modest agreement for HPV genotyping

PONE-D-20-19224R3

Dear Dr. Dube Mandishora,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Michael Scheurer, Ph.D.

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

Acceptance letter

Michael Scheurer

12 Apr 2021

PONE-D-20-19224R3

 Self-collected and clinician-collected anal swabs show modest agreement for HPV genotyping

Dear Dr. Dube Mandishora:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Michael Scheurer

Academic Editor

PLOS ONE

Associated Data

    This section collects any data citations, data availability statements, or supplementary materials included in this article.

    Supplementary Materials

    S1 File

    (PDF)

    Click here for additional data file. (1.4MB, pdf)
    S1 Table. Comparison of sequence read counts for HPV 52 and HPV 62 in both CCAS and SCAS.

    (DOCX)

    Click here for additional data file. (14.2KB, docx)
    Attachment

    Submitted filename: v2Response to reviewers Plos.docx

    Click here for additional data file. (38.6KB, docx)
    Attachment

    Submitted filename: V2_2nd Response to reviewers_100221.doc

    Click here for additional data file. (41.5KB, doc)
    Attachment

    Submitted filename: V1_3rd Response to reviewers.docx

    Click here for additional data file. (17.7KB, docx)

    Data Availability Statement

    All relevant data are within the manuscript and its Supporting Information files.

    Articles from PLoS ONE are provided here courtesy of PLOS

    RESOURCES