Abstract
It is unknown whether reactivation of human papillomavirus (HPV) after latency occurs in the anus. We measured incidence and predictors of incident anal HPV in sexually inactive gay and bisexual men (GBM) as a surrogate of HPV reactivation.
The Study of the Prevention of Anal Cancer collected data on sexual behavior, anal cytology, HPV DNA, histology and HPV serology. HPV incidence during periods when zero sexual partners were reported in the last six months at both the current and previous annual visit (“no sexual activity”) was analyzed by Cox regression using the Wei-Lin-Weissfeld method to determine univariable predictors.
Of 617 men enrolled, 525 had results for ≥2 visits, of whom 58 (11%) had ≥ one period of “no sexual activity”. During sexually inactive periods, there were 29 incident high risk HPV infections in 20 men, which occurred more commonly in older men (Ptrend = 0.010), HIV-positive men (HR = 3.12; 95% CI, 0.91–16.65), longer duration of HIV (Ptrend = 0.028), history of AIDS defining illness (P = 0.010), lower current (P = 0.010) and nadir CD4 count (P = 0.014). For 18 of 29 infections with available results, 12 men remained type-specific HRHPV L1 seronegative. None were consistently seropositive. A new diagnosis of HSIL occurred in only two men, caused by an HPV type other than the incident type.
Our findings suggest that in sexually inactive GBM, anal HRHPV incidence is relatively common, and is associated with increasing age and immune dysfunction, a pattern consistent with HPV reactivation.
Reactivation of anal HPV may occur.
Introduction
Like cervical squamous cell cancer (SCC), anal SCC is preceded by persistent high-risk human papillomavirus (HRHPV) infection associated high-grade squamous intraepithelial lesions (HSIL; refs. 1, 2). Natural history studies of cervical HRHPV infection show that approximately 90% of infections become undetectable within 2 years (3, 4). Women who test HRHPV negative are at low risk of cervical HSIL and even lower risk of cervical cancer (5, 6). In a U.S. study of over one million women, at three years following a negative HPV, the cumulative risk of cervical HSIL-CIN3 was 0.069% and was 0.011% for cervical cancer (7).
Testing HPV negative does not necessarily equate with complete absence of virus. Some viruses have a state of latency in which there is no disease present and no viral particles are produced or detectable by standard assays, but in which there is a possibility of later reactivation (8–12). Reactivation of latent HRHPV in the cervix has been proposed to cause some cases of cervical SCC (13). Reactivation of latent infections often depends upon declines in the host immune function. A virus may remain in a latent state due to a strong, cell medicated immune response, until immune suppression, for instance due to HIV infection, or immunosenescence due to aging, facilitates reactivation (11).
The evidence for HPV reactivation is strongest for low-risk HPV in animals (10, 14) and in the human upper respiratory tract (15, 16). In humans, studies of immune compromised, sexually abstinent (17, 18) and older, less sexually active female populations (9, 19, 20) have reported incident cervical HRHPV detection suggestive of HRHPV reactivation.
It is unclear whether HPV reactivation occurs in the anus, with only one previously published study identified (21). Using data from the Study of the Prevention of Anal Cancer (SPANC), a cohort study of 617 older gay and bisexual men (GBM) (22), we measured incidence and predictors of incident anal HRHPV detection in a sub-cohort of men who had periods of sexual inactivity.
Materials and Methods
Study population
The SPANC study was a 3-year cohort study of GBM, which concluded follow-up in 2018 (23–25). Men aged 35 years and above were invited to attend baseline and three annual visits. At each visit, participants completed a detailed risk factor questionnaire and had an anal swab. A Dacron swab moistened with tap water was used by a study clinician to sample the anal canal. Immediately after sampling, the swab was rotated in a vial containing 20mL of PreservCyt (Hologic, Inc.) fixative medium. The sample was then sent for cytology and HPV DNA testing. HPV genotyping was performed at the Royal Hospital for Women, Melbourne, a Western Pacific WHO reference laboratory for HPV for the HPV Labnet, using Linear Array (Roche Molecular Systems) and Anyplex II HPV HR Detection (Seegene). We considered specimens positive for any HRHPV type on either assay as positive. Blood was collected for HPV L1 antibodies, to eight HRHPV types (HPV16, 18, 31, 33, 35, 45, 52 and 58) using multiplex serology methods (Luminex Corp.; ref. 26). The clinician performed a visual inspection of the peri-anal region and a digital anal rectal examination. A high resolution anoscopy (HRA) was performed and biopsies of any lesions suspicious of HPV-related disease were sent for histology. Participants were asked detailed questions about lifetime and recent (last 6 months) sexual exposures. HIV-positive participants were asked about duration of HIV infection, history of AIDS defining illness, most recent CD4 count and CD4 nadir.
Ethical approval was granted by the Human Research Ethics Committees at St Vincent's Hospital, Sydney (HREC/09/SVH/168) and the University of New South Wales.
Statistical analysis
The main analyses focused on the subcohort of men who reported at least one period of sexual inactivity. A period of sexual inactivity was defined if a man reported no sexual activity with any partner in the current and the previous annual visit at the first, second and/or third annual follow-up visits. We report results for 13 HRHPV types (HPV types 16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59 and 68) and two low-risk HPV types (HPV6 and 11). Type-specific incident HPV infection was defined as testing negative to a specific HPV type at study baseline, followed by a positive result during study follow-up. For each of the 13 HRHPV, person-years (PY) were calculated from study baseline to the date of the first positive test, or to the last study visit in participants who remained negative throughout the study. The any-type HRHPV incidence rate was calculated as the total number of incident HRHPV infections of any type, divided by the total PY cumulated. We calculated HRHPV incidence during a period of sexual inactivity as a new HRHPV detection at a follow-up visit when zero sexual partners were reported in the last six months at the current and previous visit. New all-type HRHPV incidence rate in the sexually inactive period was compared with that in all periods when there was sexual activity.
Potential predictors of incident HRHPV during periods of sexual inactivity included age, sexual position preference during anal intercourse and HIV status. For HIV-positive participants, we considered potential markers of immune suppression, including duration of being HIV-positive, a history of an AIDS defining illness, and current and nadir CD4 cell counts. Univariable Cox regression was used to determine the associations with incident anal HRHPV in periods of sexual inactivity using the Wei-Lin-Weissfeld (WLW) method (27), with HR and corresponding 95% confidence intervals (CI) presented. The WLW method is a Cox model with a baseline hazard function stratified by HPV type. It assumes that HPV types have different incidence rates, and that the exposure has a common relative effect on the incidence rate across types. We chose this method to account for multiple observations per participant with different HPV types. The WLW robust sandwich estimate was applied in our Cox regression models, with each participant's study identification number specified as clusters.
We performed sensitivity analyses using different definitions of sexual inactivity including (i) all visits at which there was no sexual activity in the previous 6 months and (ii) only men who reported no sexual partners at all study visits. Statistical analyses were performed using Stata 16 (StataCorp).
Results
A total of 617 GBM were enrolled. Over a third were HIV positive (220, 35.7%). Median age was 49 years (interquartile range: 43–56), unrelated to HIV status (P = 0.79). Among a total of 2021 study visits, a total of 1978 samples tested using LA, 1920 using Anyplex II and 1878 using both. Over three-quarters of study participants had at least one HRHPV type detected (n = 469, 76.5%) at baseline and HPV16 was detected in over one-third (n = 204, 33.3%; ref. 25). Of the 617 men, 525 had HPV results for at least two visits, and among these 525 men, of median age of 49 years (IQR: 43–56), 39 (7.4%) reported no sexual partners in the last 6 months at baseline. Overall, 58 men (11.0%) reported no sexual partners in the last 6 months on at least two consecutive annual visits. At baseline, these 58 men were significantly older (54.8 vs. 49.8 years, P < 0.001) than the remaining 467 and were more likely to be HIV positive (65.5% vs. 32.1%, P < 0.001). The crude overall incidence rate of any HRHPV was non-significantly lower in sexually inactive periods (25.3 per 100 PY compared with 35.1 per 100PY in the sexually active periods, P = 0.077, incidence rate ratio 0.72 (95% CI, 0.48–1.04)). Using the WLW method, the crude HR associated with sexual inactivity was 0.74 (95% CI, 0.45–1.22; P = 0.236). After adjustment for age and HIV status, the association approached statistical significance (adjusted HR, 0.62; 95% CI, 0.38–1.01; P = 0.055). The crude incidence of the low-risk HPV types HPV6 and HPV11 was 1.82 per 100PY in the sexually inactive periods, as compared with 3.45 per 100PY in the sexually active periods (incidence rate ratio 0.53; 95% CI, 0.06–2.02; P = 0.395). Using WLW method, the crude HR was 0.55 (95% CI, 0.13–2.37, P = 0.425). After adjustment for age and HIV status, the adjusted HR was 0.49 (95% CI, 0.12–2.12, P = 0.342).
Among the cohort of 58 men who had a period of sexual inactivity, using the WLW method there were a total of 1225.8 PY of follow up during the periods in which they had no sexual partners. Among these men, none tested positive for other sexually transmitted infections, including urethral Neisseria gonorrhoea (NG) or Chlamydia trachomatis (CT), anorectal NG/CT, pharyngeal NG or syphilis at the time of their incident HRHPV infection. At each visit at which a new HRHPV was detected only one HRHPV type was detected in 14 participants and more than one type (range 2–4) in six participants. There were 29 incident HRHPV infections in 20 men, an average incidence of 2.37 per 100PY (95% CI, 1.64–3.40; Table 1). Of these 29 incident infections, 18 (62.1%) were detected by AnyPlex II only, three (10.3%) were detected by LA only and eight (27.6%) were detected by both assays. Thirteen of the 20 men had persistent infection with at least one other HRHPV, the most common types being HPV16 (seven men, 53.8%) and HPV68 (five men, 38.5%). There were only two cases (one each for HPV6 and HPV11) of low-risk HPV incidence in intervals of sexual inactivity.
. | N . | Person-years . | Incidence (per 100PY) . | HR (95% CIb) . | P . |
---|---|---|---|---|---|
Age | 0.010 | ||||
35–44 | 1 | 85.5 | 1.2 | 1 (—) | |
45–54 | 4 | 386.8 | 1.0 | 0.57 (0.07–4.83) | |
55–64 | 11 | 413.4 | 2.7 | 2.13 (0.30–15.06) | |
>65 | 13 | 340.1 | 3.8 | 3.39 (0.46–25.00) | |
Sexual position preference | 0.268 | ||||
Mostly top | 15 | 386.4 | 3.9 | 1 (—) | |
Versatile | 8 | 468.7 | 1.7 | 0.53 (0.18–1.51) | |
Mostly bottom | 6 | 370.8 | 1.6 | 0.56 (0.19–1.69) | |
HIV status | |||||
Negative | 6 | 481.5 | 1.2 | 1 (—) | 0.069 |
Positive | 23 | 744.3 | 3.1 | 3.12 (0.91–10.65) | |
Duration of HIV | 0.028 | ||||
HIV negative | 6 | 481.5 | 1.2 | 1 (—) | |
<10 years | 5 | 143.0 | 3.5 | 2.58 (0.50–13.42) | |
10–20 years | 6 | 263.5 | 2.3 | 2.46 (0.51–11.90) | |
>20 years | 12 | 337.8 | 3.6 | 4.13 (1.13–15.07) | |
History of AIDS defining Illness | 0.010 | ||||
HIV negative | 6 | 481.5 | 1.2 | 1 (—) | |
No | 5 | 373.0 | 1.3 | 1.33 (0.29–6.22) | |
Yes | 17 | 351.7 | 4.8 | 4.94 (1.43–17.05) | |
Last CD4 count (µmol/L) | 0.010 | ||||
HIV negative | 6 | 481.5 | 1.2 | 1 (—) | |
>500 | 14 | 480.5 | 2.9 | 2.92 (0.78–10.98) | |
351–500 | 3 | 157.9 | 1.9 | 2.05 (0.45–9.31) | |
≤350 | 6 | 80.2 | 7.5 | 7.04 (1.64–30.30) | |
CD4 Nadir (µmol/L) | 0.014 | ||||
HIV negative | 6 | 481.5 | 1.2 | 1 (—) | |
≥200 | 6 | 397.5 | 1.5 | 1.55 (0.35–6.83) | |
<200 | 15 | 316.2 | 4.7 | 4.6 (1.33–15.88) | |
Current CD4 count in those with a nadir CD4 count of <200 | |||||
Over 350 | 11 | 263.6 | 4.2 | 1 (—) | 0.613 |
≤350 | 4 | 44.2 | 9.1 | 1.43 (0.36–5.65) |
. | N . | Person-years . | Incidence (per 100PY) . | HR (95% CIb) . | P . |
---|---|---|---|---|---|
Age | 0.010 | ||||
35–44 | 1 | 85.5 | 1.2 | 1 (—) | |
45–54 | 4 | 386.8 | 1.0 | 0.57 (0.07–4.83) | |
55–64 | 11 | 413.4 | 2.7 | 2.13 (0.30–15.06) | |
>65 | 13 | 340.1 | 3.8 | 3.39 (0.46–25.00) | |
Sexual position preference | 0.268 | ||||
Mostly top | 15 | 386.4 | 3.9 | 1 (—) | |
Versatile | 8 | 468.7 | 1.7 | 0.53 (0.18–1.51) | |
Mostly bottom | 6 | 370.8 | 1.6 | 0.56 (0.19–1.69) | |
HIV status | |||||
Negative | 6 | 481.5 | 1.2 | 1 (—) | 0.069 |
Positive | 23 | 744.3 | 3.1 | 3.12 (0.91–10.65) | |
Duration of HIV | 0.028 | ||||
HIV negative | 6 | 481.5 | 1.2 | 1 (—) | |
<10 years | 5 | 143.0 | 3.5 | 2.58 (0.50–13.42) | |
10–20 years | 6 | 263.5 | 2.3 | 2.46 (0.51–11.90) | |
>20 years | 12 | 337.8 | 3.6 | 4.13 (1.13–15.07) | |
History of AIDS defining Illness | 0.010 | ||||
HIV negative | 6 | 481.5 | 1.2 | 1 (—) | |
No | 5 | 373.0 | 1.3 | 1.33 (0.29–6.22) | |
Yes | 17 | 351.7 | 4.8 | 4.94 (1.43–17.05) | |
Last CD4 count (µmol/L) | 0.010 | ||||
HIV negative | 6 | 481.5 | 1.2 | 1 (—) | |
>500 | 14 | 480.5 | 2.9 | 2.92 (0.78–10.98) | |
351–500 | 3 | 157.9 | 1.9 | 2.05 (0.45–9.31) | |
≤350 | 6 | 80.2 | 7.5 | 7.04 (1.64–30.30) | |
CD4 Nadir (µmol/L) | 0.014 | ||||
HIV negative | 6 | 481.5 | 1.2 | 1 (—) | |
≥200 | 6 | 397.5 | 1.5 | 1.55 (0.35–6.83) | |
<200 | 15 | 316.2 | 4.7 | 4.6 (1.33–15.88) | |
Current CD4 count in those with a nadir CD4 count of <200 | |||||
Over 350 | 11 | 263.6 | 4.2 | 1 (—) | 0.613 |
≤350 | 4 | 44.2 | 9.1 | 1.43 (0.36–5.65) |
Among men who reported a period of sexual inactivity, HRHPV incidence was more common in those who were HIV-positive (HR, 3.12; 95% CI, 0.91–16.65) and older than 55 years (Ptrend = 0.010). When stratified by HIV status, older age was borderline significantly related to higher HRHPV incidence in HIV-negative men (Ptrend = 0.070) and significant in HIV-positive men (Ptrend = 0.002.) Among the 16 men who were HIV-positive, incidence was significantly related to longer duration of HIV infection (Ptrend = 0.028), having a history of AIDS defining illness (P = 0.010), and a lower current (P = 0.010) and nadir CD4 count (P = 0.014; Table 1).
HRHPV L1 serology results in relation to the specific incident HRHPV type were available for 18 of 29 incident infections. The other 11 infections were with HRHPV types not included in our assay (HPV39, 51, 56, 68). In relation to the incident HRHPV type, 12 of these 18 men were type-specific L1 seronegative to the relevant HRHPV type at baseline and remained so after incident HRHPV infection, and none were consistently L1 seropositive (Table 2).
. | HIV . | Age . | HRHPV type detected . | Visit at which HRHPV detecteda . | Nadir CD4 count . | CD4 count at HRHPV detection . | Incident HSIL . | HPV L1 serology at each visitb . |
---|---|---|---|---|---|---|---|---|
1 | Neg | 72 | 33 | Year 1 | No | -/-/-/- | ||
68 | Year 1 | No | Nd | |||||
18 | Year 2 | No | -/-/-/- | |||||
2 | Neg | 55 | 56 | Year 3 | No | Nd | ||
3 | Neg | 58 | 58 | Year 2 | No | -/na/-/- | ||
4 | Neg | 76 | 52 | Year 3 | No | -/-/± | ||
5 | Pos | 55 | 31 | Year 1 | unknown | >750 | No | -/-/na/- |
51 | Year 1 | No | Nd | |||||
6 | Pos | 56 | 52 | Year 1 | 200–500 | 501–750 | No | -/-/-/- |
7 | Pos | 68 | 39 | Year 2 | 200–500 | 501–750 | No | Nd |
16 | Year 3 | No | -/+/+/+ | |||||
8 | Pos | 67 | 45 | Year 1 | >500 | 501–750 | No | -/-/-/- |
9 | Pos | 55 | 51 | Year 2 | 200–500 | >750 | No | Nd |
10 | Pos | 64 | 45 | Year 1 | 200–500 | 501–750 | No | +/±/- |
11 | Pos | 46 | 18 | Year 1 | <50 | <50 | No | -/-/na/- |
12 | Pos | 60 | 52 | Year 1 | <50 | 201–350 | Yesc | -/-/-/- |
13 | Pos | 64 | 51 | Year 2 | 50–199 | 351–500 | No | Nd |
14 | Pos | 52 | 18 | Year 1 | <50 | <50 | No | -/+/± |
15 | Pos | 50 | 58 | Year 1 | <50 | 351–500 | Yesd | +/±/- |
16 | Pos | 69 | 68 | Year 1 | <50 | 501–750 | No | Nd |
35 | Year 2 | No | -/-/-/- | |||||
17 | Pos | 67 | 51 | Year 1 | 50–199 | >750 | No | Nd |
58 | Year 1 | No | -/-/na/- | |||||
68 | Year 1 | No | Nd | |||||
33 | Year 3 | No | -/-/na/- | |||||
18 | Pos | 43 | 35 | Year 1 | < 50 | 351–500 | No | -/+/na/na |
19 | Pos | 61 | 52 | Year 1 | 50–199 | ≤350 | No | -/-/na/na |
68 | Year 1 | Nd | ||||||
20 | Pos | 46 | 68 | Year 2 | < 50 | 201–350 | No | Nd |
. | HIV . | Age . | HRHPV type detected . | Visit at which HRHPV detecteda . | Nadir CD4 count . | CD4 count at HRHPV detection . | Incident HSIL . | HPV L1 serology at each visitb . |
---|---|---|---|---|---|---|---|---|
1 | Neg | 72 | 33 | Year 1 | No | -/-/-/- | ||
68 | Year 1 | No | Nd | |||||
18 | Year 2 | No | -/-/-/- | |||||
2 | Neg | 55 | 56 | Year 3 | No | Nd | ||
3 | Neg | 58 | 58 | Year 2 | No | -/na/-/- | ||
4 | Neg | 76 | 52 | Year 3 | No | -/-/± | ||
5 | Pos | 55 | 31 | Year 1 | unknown | >750 | No | -/-/na/- |
51 | Year 1 | No | Nd | |||||
6 | Pos | 56 | 52 | Year 1 | 200–500 | 501–750 | No | -/-/-/- |
7 | Pos | 68 | 39 | Year 2 | 200–500 | 501–750 | No | Nd |
16 | Year 3 | No | -/+/+/+ | |||||
8 | Pos | 67 | 45 | Year 1 | >500 | 501–750 | No | -/-/-/- |
9 | Pos | 55 | 51 | Year 2 | 200–500 | >750 | No | Nd |
10 | Pos | 64 | 45 | Year 1 | 200–500 | 501–750 | No | +/±/- |
11 | Pos | 46 | 18 | Year 1 | <50 | <50 | No | -/-/na/- |
12 | Pos | 60 | 52 | Year 1 | <50 | 201–350 | Yesc | -/-/-/- |
13 | Pos | 64 | 51 | Year 2 | 50–199 | 351–500 | No | Nd |
14 | Pos | 52 | 18 | Year 1 | <50 | <50 | No | -/+/± |
15 | Pos | 50 | 58 | Year 1 | <50 | 351–500 | Yesd | +/±/- |
16 | Pos | 69 | 68 | Year 1 | <50 | 501–750 | No | Nd |
35 | Year 2 | No | -/-/-/- | |||||
17 | Pos | 67 | 51 | Year 1 | 50–199 | >750 | No | Nd |
58 | Year 1 | No | -/-/na/- | |||||
68 | Year 1 | No | Nd | |||||
33 | Year 3 | No | -/-/na/- | |||||
18 | Pos | 43 | 35 | Year 1 | < 50 | 351–500 | No | -/+/na/na |
19 | Pos | 61 | 52 | Year 1 | 50–199 | ≤350 | No | -/-/na/na |
68 | Year 1 | Nd | ||||||
20 | Pos | 46 | 68 | Year 2 | < 50 | 201–350 | No | Nd |
aVisit at which incident HPV infection occurred.
bSerology at baseline and up to 3 annual follow-up visits for HPV types = HPV16, 18, 31, 33, 35, 45, 52, and 58. “na” is when results were not available, “nd” is when serology for this HPV type was not performed.
cIncident HSIL at visit after incident HRHPV detection.
dIncident HSIL at the same visit as incident HRHPV visit.
A new diagnosis of HSIL (HSIL-AIN3) was made in only one of the 20 sexually inactive participants during the interval that incident HRHPV infection developed, an HSIL incidence of 7.32 per 100PY, compared with 11.35 per 100PY in the rest of the SPANC cohort (HR, 0.75; 95% CI, 0.10–5.39, P = 0.775). An additional one HSIL (HSIL-AIN3) diagnosis was made in the visit subsequent to incident HRHPV detection, an HSIL incidence of 9.80 per 100 PY, which was not significantly different to the rest of cohort (HR, 0.74; 95% CI, 0.10–5.27; P = 0.760; Table 2). Both of these HSIL diagnoses were in participants who had persistent infection with HPV16 throughout the study, and both had incident infection with a different type (HPV58 and HPV52). In both men, HSIL persisted for the rest of study follow-up. Whole-tissue section HPV typing of all HSIL biopsies were negative for the incident HRHPV type for both participants.
Similar results for HRHPV incidence were seen in sensitivity analyses where the cohort characteristics were broadened to include men who were sexually inactive at one visit only (n = 107) or restricted to include only men who were sexually inactive throughout the entire study (n = 22). There were nine men (1.7% of the cohort with HPV results for at least two visits) who reported no sexual exposure throughout the entire study period who developed incident HRHPV and eight (88.9%) were HIV-positive, as compared with 36.5% (106/290) of the remaining cohort who developed incident HRHPV (P = 0.001).
Discussion
In this cohort study of older GBM, 11% (58) reported no sexual activity in at least two consecutive annual visits over three years of follow-up. In sexually inactive periods, HRHPV incidence was lower than in the remainder of the cohort, although this did not reach statistical significance (P = 0.074). During periods of sexual inactivity, HRHPV incidence was significantly associated with older age and being HIV-positive, and incident HRHPV was not associated with increased HSIL occurrence during follow-up. During periods of sexual inactivity in HIV-positive men, anal HRHPV incidence was associated with a history of AIDS defining illness and a lower current and nadir CD4 count. The association of anal HRHPV incidence with increased age and markers of HIV-associated immune dysfunction in men without recent sexual exposure, strongly supports the hypothesis that HRHPV detection in these men represented HRHPV reactivation after latency, modulated by impaired immune surveillance related to increasing age and/or HIV-related immune dysfunction. Only 2 of the 20 men who had HRHPV reactivation developed HSIL at the same or the subsequent visit, and in both individuals the HSIL persisted for at least 12 months.
We were only able to identify one previously published study on anal HRHPV reactivation, which was also conducted in GBM (21). This Dutch study examined potential HPV reactivation of seven HRHPV types in a cohort of HIV-negative and HIV-positive GBM. The definition of sexually non-exposed in this study was less stringent than in our study, in that it included men who may have had forms of sexual contact other than receptive anal sex, rimming or fisting. In our study, we had a stringent definition of sexually inactive which required men to report no sexual partners at two consecutive annual visits. Incident HPV infection in the Dutch study was less likely to occur in the sexually non-exposed but remained common, suggesting the possibility of HRHPV reactivation. Similar to our findings, baseline HRHPV serology was not associated with incident infection. The paper did not report on the association between CD4 count and HIV status and potential HRHPV reactivation (21).
Most research on HRHPV reactivation in humans has been in cervical HPV. There has been debate around the contribution of HPV reactivation to cervical cancer risk in older women (13, 28) and the role of vaccination in prevention of reactivation of latent infection. HRHPV prevalence peaks in young women, and the more modest second peak in HRHPV prevalence women of menopausal age has been attributed to new sexual partners with consequent new infections. However, reactivation of latent infection has been proposed as an alternate explanation (13, 19, 29). A US study of 700 women aged 35–60 years found that the vast majority (155 of 183, 85%) of incident HPV detections occurred during periods of sexual abstinence or monogamy. However, reported monogamy does not preclude a new exposure from their sexual partner. On the other hand, if the women were truly non-exposed, these data would suggest that most incident cervical HPV infection was attributable to past, not current, sexual behaviour at older ages (20). There are no conclusive data yet available on whether cervical HPV reactivation is associated with the occurrence of HSIL.
Two of the twenty SPANC participants with sexually inactive periods who developed HRHPV reactivation developed persistent anal HSIL. However, whole tissue HPV typing attributed the new HSIL to persistent HRHPV types. Our results are too sparse to make firm conclusions, but it remains possible that reactivation might be important in causing disease. An increased rate of HRHPV-associated disease associated with reactivation in people with immune dysfunction would be consistent with data on HRHPV infection, whereby it has been demonstrated that immunosuppression may elevate HPV titers, promote viral gene expression and the efficiency of lesion formation (30). The disease risks associated with HRHPV reactivation can only be studied in large scale cohorts of older individuals in which sexual behavior, HPV and associated disease data are collected.
Our study of HRHPV in periods of sexual inactivity in GBM had some limitations. Self-reported sexual activity may result in under- or over-reporting bias because of perceived social desirability. We attempted to minimize this by having interviews which were self-administered in a private room. Every 12 months, participants were asked about their past sexual behavior in the 6 months prior to interview meaning it was possible that some sexual activity occurred each year in the half of the year that was not covered by the interview. However, among the 58 men who reported they were sexually inactive, none tested positive for anal chlamydia, anal gonorrhoea, or syphilis. The complete absence of these common conditions in the sexually inactive men, when these conditions are common in sexually active GBM (31), strongly supports the validity of our definition of sexual inactivity. The number of men reporting no sexual activity was small, and this prevented a meaningful multivariable analysis to adjust for potential confounding factors. The specific associations we have described between HRHPV reactivation with increasing age, HIV status, and markers of impaired immune function, strongly support our interpretation of these cases as representing reactivation of latent HRPV related to impaired immune function. We do not believe it is possible to draw strong conclusions about the lack of an association between reactivated HRHPV with HSIL development in our study. Over 90% of HIV-positive SPANC participants were on effective anti-retroviral therapies and had current CD4 counts in the normal range, and it may well be that immune function was sufficient to prevent lesion development (11). Larger studies are needed to assess this in people with currently impaired immune function.
In conclusion, among sexually inactive GBM, incident HRHPV infection, likely reflecting reactivation of latent HRHPV, was associated with older age, HIV-positive status, and measures of past and present immune dysfunction due to HIV. In most cases, HRHPV reactivation was not associated with concurrent development of HSIL. Our results suggest that in the anus, HRHPV reactivation is relatively common, and is related to immune dysfunction caused by aging or HIV infection. Further research is needed to determine whether anal HRHPV reactivation in sexually inactive populations causes significant HRHPV-related disease.
Authors' Disclosures
F. Jin reports grants from Australian National Health and Medical Research Council and grants from Cancer Council New South Wales during the conduct of the study. M. Molano reports grants from Cancer Council Victoria during the conduct of the study; grants from Cancer Council Victoria outside the submitted work. D.J. Templeton reports grants from NHMRC and grants from Cancer Council NSW during the conduct of the study. T. Waterboer reports personal fees from MSD (Merck) Sharp & Dohme outside the submitted work. A. Farnsworth reports grants from NHMRC during the conduct of the study. C.K. Fairley reports ownership of shares in CSL, which markets an HPV vaccine in Australia. S.M. Garland reports grants from Merck, personal fees from Merck, and personal fees from Merck outside the submitted work; and is a member of the Merck Global Advisory Board as a consultant. A.E. Grulich reports grants from National Health and Medical Research Council, non-financial support from Hologic, and grants from Cancer Council of NSW during the conduct of the study; personal fees from MSD and grants from Seqirus outside the submitted work. No disclosures were reported by the other authors.
Disclaimer
The views expressed in this publication do not necessarily represent the position of the Australian Government.
Authors' Contributions
I.M. Poynten: Conceptualization, formal analysis, funding acquisition, validation, methodology, writing–original draft, project administration, writing–review and editing. F. Jin: Conceptualization, data curation, software, formal analysis, validation, methodology, writing–review and editing. M. Molano: Formal analysis, validation, investigation, methodology, writing–review and editing. J.M. Roberts: Conceptualization, data curation, supervision, validation, methodology, writing–review and editing. R.J. Hillman: Conceptualization, supervision, investigation, methodology, writing–review and editing. D.J. Templeton: Data curation, investigation, methodology, writing–review and editing. C. Law: Data curation, validation, investigation, project administration, writing–review and editing. M.A. Stanley: Methodology, writing–review and editing. T. Waterboer: Data curation, software, formal analysis, supervision, validation, investigation, methodology, writing–review and editing. A. Farnsworth: Resources, supervision, funding acquisition, methodology, writing–review and editing. C.K. Fairley: Conceptualization, resources, funding acquisition, methodology, writing–review and editing. S.M. Garland: Conceptualization, resources, supervision, funding acquisition, methodology, writing–review and editing. A.E. Grulich: Conceptualization, resources, supervision, funding acquisition, methodology, writing–review and editing.
Acknowledgments
We would like to acknowledge and thank the participants of the SPANC study. The SPANC study team includes Brian Acraman, Marjorie Adams, Claire Biro, Andrew Carr, Susan Carroll, Simon Comben, David Cooper, Alyssa Cornall, Leonie Crampton, Deborah Ekman, Amber Ellis, Christopher Fairley, Annabelle Farnsworth, Lance Feeney, Marko Garcia, Suzanne Garland, Andrew Grulich, Richard Hillman, Kirsten Howard, Fengyi Jin, Carmella Law, Matthew Law, Dorothy Machalek, Kirsten McCaffery, Ross McDonald, Patrick McGrath, Robert Mellor, Monica Molano, Kathy Petoumenos, Piero Pezzopane, Samuel Phillips, Mary Poynten, Garrett Prestage, Adele Richards, Jennifer Roberts, Daniel Seeds, Sepehr Tabrizi, David Templeton, Julia Thurloe, Winnie Tong, and Rick Varma. This work was supported by the National Health and Medical Research Council Program Grant (Sexually transmitted infections: causes, consequences and interventions Grant #568971; awarded to authors AE Grulich, SM Garland, and CK Fairley); and a Cancer Council New South Wales Strategic Research Partnership Program Grant (Preventing morbidity and mortality from anal cancer Grant #13–11; awarded to A.E. Grulich, I.M. Poynten, F. Jin, R.J. Hillman, and J.M. Roberts). Cytological testing materials were provided by Hologic (Australia) Pty Ltd. The Kirby Institute is affiliated with the Faculty of Medicine, University of New South Wales and funded by the Australian Government of Health and Ageing.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.