Content uploaded by Naomi Jay
Author content
All content in this area was uploaded by Naomi Jay on Jun 17, 2019
Content may be subject to copyright.
361
Prevalence and Risk Factors for Human Papillomavirus Infection of the Anal
Canal in Human Immunodeficiency Virus (HIV)–Positive and HIV-Negative
Homosexual Men
Joel M. Palefsky, Elizabeth A. Holly, Mary L. Ralston,
Departments of Laboratory Medicine, Stomatology, and Epidemiology
and Biostatistics, University of California, San Francisco
and Naomi Jay
One of the groups at highest risk of anal cancer is homosexual and bisexual men. Like cervical
cancer, anal cancer is associated with human papillomavirus (HPV) infection. Anal HPV infection
was characterized in a study of 346 human immunodeficiency virus (HIV)– positive and 262 HIV-
negative homosexual and bisexual men. Anal HPV DNA was detected in 93% of HIV-positive and
61% of HIV-negative men by polymerase chain reaction. The spectrum of HPV types was similar
in HIV-positive and HIV-negative men, with HPV-16 the most common type. Infection with multiple
HPV types was found in 73% of HIV-positive and 23% of HIV-negative men. Among HIV-positive
men who were positive by hybrid capture for group B HPV types (16/18/31/33/35/39/45/51/52/56/
58) or group A types (6/11/42/43/44), lower CD4 cell levels were associated with higher levels of
group B DNA (PÅ.004) but not group A DNA. These data suggest increased replication of the
more oncogenic HPV types with more advanced immunosuppression.
Several studies have shown the risk of anal cancer to be of cases of anal cancer with increasing proximity to a diagnosis
of AIDS.
elevated among men with a history of homosexual or bisexual
Several lines of evidence point to an important role for hu-
activity (hereafter called homosexual) [1–3]. The incidence of
man papillomavirus (HPV) infection in the pathogenesis of
anal cancer among homosexual men has been estimated to be
anal cancer and ASIL. HPV DNA has been detected in anal
Ç35/100,000 [2], which is several times higher than current
cancer and ASIL tissues [13–15]. Similar to their oncogenic
rates of cervical cancer in women in the United States [4] and
risk in the cervix [16], most cases of anal high-grade squamous
similar to rates of cervical cancer prior to the introduction of
intraepithelial lesions and anal cancer contained HPV-16,
routine cervical cytology screening.
whereas anal low-grade squamous intraepithelial lesions were
Recent reports have shown that human immunodeficiency
more likely to have HPV types known to have a low risk of
virus (HIV)–positive homosexual men have higher rates of
oncogenicity in the cervix, such as HPV-6 or -11. An under-
anal squamous intraepithelial lesions (ASIL), putative anal can-
standing of anal HPV infection and risk factors for detection
cer precursors, than do HIV-negative men [5 –10]. This sug-
of anal HPV in HIV-positive and HIV-negative homosexual
gests that HIV-positive men may be at higher risk of developing
men is important, given the role of HPV infection in the patho-
anal cancer than HIV-negative men. So far, data on the degree
genesis of ASIL and anal cancer.
of excess incidence of anal cancer in HIV-positive homosexual
Previous studies have reported HIV-positive homosexual men
compared with HIV-negative homosexual men are conflicting.
to have a higher prevalence of anal HPV infection than do HIV-
Rabkin and Yellin [11] reported little increased risk of anal
negative homosexual men [6 –10, 17]. However, most of these
cancer among single, never-married men aged 25–44 years in
studies were small and had limited information on risk factors
the San Francisco Bay Area since the onset of the HIV epi-
for anal HPV infection. No studies to date have described the
demic. However, in a study linking HIV registries to the Sur-
spectrum of HPV types detectable in the anal canal using a
veillance, Epidemiology, and End Results database, Melbye et
polymerase chain reaction (PCR)–based DNA detection method
al. [12] reported an increase in the observed-to-expected ratio
capable of detecting a wide variety of genital and nongenital
HPV types, nor have they characterized the levels of HPV DNA
in anal specimens and their relationship to HIV-related immuno-
Received 13 March 1997; revised 13 August 1997.
suppression. The goal of our analyses was to characterize the
The study protocol was approved by the institution’s Committee on Human
prevalence of 29 different HPV types and a mixture of 10 addi-
Research and informed consent was obtained from all men.
Grant support: NIH (CA-54053). Clinical examinations were performed in
tional types in the anal canal among HIV-positive and HIV-
the General Clinical Research Center, University of California, San Francisco,
negative men. We examined the relationship between HPV DNA
with funds provided by the Division of Research Resources (RR-00079), US
levels and HIV-related immunosuppression and characterized
Public Health Service.
Reprints or correspondence: Dr. Joel Palefsky, Box 0100, Dept. of Labora-
risk factors for detection of HPV DNA in these groups.
tory Medicine, University of California, San Francisco, San Francisco, CA
94143 (joelp@labmed.ucsf.edu).
Materials and Methods
The Journal of Infectious Diseases 1998;177:361 –7
In total, 346 HIV-positive and 262 HIV-negative men were
q1998 by The University of Chicago. All rights reserved.
0022–1899/98/7702– 0012$02.00
recruited between November 1991 and March 1994 from among
/ 9d3f$$fe10 12-01-97 13:08:24 jinfa UC: J Infect
Downloaded from https://academic.oup.com/jid/article-abstract/177/2/361/925365 by guest on 17 June 2019
362 Palefsky et al. JID 1998;177 (February)
homosexual and bisexual men in the San Francisco Men’s Health A if the RLU ratio for group A mix, HPV-6/11, or HPV-42/43/
44 was ú1.0. Specimens with an RLU ratio ú0.65 with probeStudy (SFMHS) [18], in the San Francisco General Hospital Co-
hort Study (SFGH) [19], and at the University of California, San group B mix were separately reprobed for HPV-16 alone, HPV-
18/31/33/45/58, and HPV-35/39/51/52/56, subgroups that were se-Francisco (UCSF), responding to newspaper advertisements. Sub-
jects were interviewed using a comprehensive questionnaire re- lected according to their frequency of detection in cervical cancer
tissues (HPV-16 had the highest frequency, HPV-18/31/33/45/58garding their medical history, including prescription drug use, x-
ray exposure, hospitalizations, transfusions, history of sexually were found in ú2% of cervical cancers, and HPV-35/39/51/52/56
were found in õ2% of cervical cancers [16]). Specimens weretransmitted diseases, intestinal parasites, hepatitis, anal conditions
such as fissures and fistulas, and symptoms related to HIV. Subjects considered positive for group B if the RLU ratio for group B mix,
HPV-16/18/31/33/45/58, or HPV-35/39/51/52/56 was ú1.0. Allwere queried about sex practices, cigarette smoking, drinking, and
recreational drug use. experiments were performed in triplicate, and the average of the
three results was used to compute the RLU ratio.HPV testing. To obtain a sample for HPV testing, a Dacron
swab was inserted into the anal canal [20] and the swab placed Measurement of HIV status and CD4 cell levels. ELISA
screening was used to determine HIV status in all men, and allinto a tube containing 1 mL of Sample Transport Medium (Digene
Diagnostics, Silver Spring, MD). The tubes were frozen at 0707C positive ELISAs were confirmed with a Western blot assay. Differ-
ent techniques were used to measure CD4 cell levels, and appro-until processed for analysis. After defrosting, the tubes were heated
to 567Cfor1htoinactivate HIV. priate adjustments were made to correct for differences in these
techniques, using calibration curves that compared the results ofBecause of its high sensitivity, PCR was used in this study to
detect low-level HPV infection. Since a positive PCR result does the different methods using the same samples. For HIV-positive
men, CD4 cell levels were classified as ú500/mm
3
, 200–500/mm
3
,not discriminate between low-level and high-level infection, we
also used the hybrid capture (HC) test (Digene), a non–amplifica- and õ200/mm
3
. After adjusting for differences in the techniques, 1
subject was moved from the õ200/mm
3
category into the 200–tion-based test that is not as sensitive as PCR. The quantity of
HPV DNA in the specimen can be determined by measuring the 500/mm
3
category.
Statistical analysis. Potential risk factors for anal HPV infec-HC relative light unit (RLU) ratio, computed by dividing the che-
miluminescent signal of the specimen by that obtained with a tion that were examined included history of anal conditions (hem-
orrhoids, anal fissures or fistulas, anal infection or discharge, rectalcontrol sample containing 10 pg/mL HPV-16 DNA.
Part (200
m
L) of the specimen was used for PCR and the remain- itching or burning, constipation, and use of enemas or laxatives),
history of sexually transmitted diseases (genital warts, syphilis,der was used for HC. PCR was performed using MY09/MY11
consensus HPV-L1 primers as well as primers for amplification Chlamydia trachomatis, Neisseria gonorrhea, and herpes simplex
virus), lifetime exposure to sexual activities (receptive anal inter-of the human
b
-globin gene [21]. After 30 amplification cycles,
6
m
L of amplification mixture was applied to a nylon membrane course, insertive anal intercourse, rimming, receptive fisting, and
rectal use of recreational drugs), and substance use (lifetime smok-and probed with a biotin-labeled HPV-L1 consensus probe mixture
consisting of full-length HPV-16, -18, and -51 genomic DNA [21]. ing, lifetime alcohol consumption, and use in the previous 10 years
of marijuana, cocaine, hallucinogens, ‘‘ecstasy,’’ ‘‘poppers,’’ ‘‘up-A separate membrane was probed with a biotin-labeled probe to
the human
b
-globin gene. Each specimen was also studied for pers,’’ and ‘‘downers’’).
The presence or absence of the condition or behavior was usedthe presence of specific HPV types by preparing membranes as
described above with 6
m
L of specimen. The membranes were when practical to evaluate the association between HPV infection
and potential risk factors. However, many of the behaviors werestudied with probes to 29 different HPV types: 6, 11, 16, 18, 26,
31, 32, 33, 35, 39, 40, 45, 51, 52, 53, 54, 55, 56, 58, 59, 61, 66, practiced by almost all of the men. Therefore, summary variables
quantifying exposure to these factors were created to characterize68, 69, 70, 73, AE2, Pap 155, and Pap 291, as well as the following
10 HPV types together in a probe mixture: 2, 13, 34, 42, 57, 62, lifetime exposure to the activity [22]. Different measures of dura-
tion and quantity of activity were used depending on the activity.64, 67, 72, and W13B. A sample was considered HPV-positive
when it was positive with the consensus probes or with §1 probes To obtain exposure measures comparable for the different cohorts,
cut points were chosen to create the summary variables so that thefor specific HPV types (only 1 specimen that was
b
-globin–posi-
tive and consensus-negative was positive for a specific HPV type). distributions of the exposure measures were similar for the HIV-
positive subjects in each cohort. The derived variables were con-Specimens negative for
b
-globin gene amplification were excluded
from analysis. Negative controls for each experiment consisted of verted to scores of 0, 1, 2, and 3, with larger numbers representing
higher levels of exposure. The scores were averaged to create aamplification of solution containing all the above components ex-
cept for sample DNA. Positive controls included amplification of summary variable for each behavior for each individual.
The summary risk factor exposure categories thus created werecloned HPV DNA.
HC testing was conducted according to the manufacturer’s rec- used to divide study subjects into the following exposure groups:
(1) never/rare: men who were exposed to the risk factor over aommendations. Two separate tests were performed on each anal
specimen, consisting of 50
m
L for the HPV types associated with small portion of their lifetime and never or rarely engaged in the
activity, (2) low exposure: men who exceeded the limits of thea low risk of cervical cancer in probe group A (6/11/42/43/44)
and 50
m
L for detection of the HPV types associated with interme- never/rare category in at least one aspect of the activity, (3) moder-
ate exposure: men who engaged in the activity most of their livesdiate and high risk of cervical cancer in probe group B (16/18/31/
33/35/39/45/51/52/56/58). Specimens with an RLU ratio ú0.65 for a moderate amount of time and may have occasionally engaged
in it extensively, and (4) high exposure: men who engaged in thewith probe group A mix were reprobed separately for HPV-6/11
and HPV-42/43/44. Specimens were considered positive for group activity most of their lives and usually engaged in it extensively.
/ 9d3f$$fe10 12-01-97 13:08:24 jinfa UC: J Infect
Downloaded from https://academic.oup.com/jid/article-abstract/177/2/361/925365 by guest on 17 June 2019
363JID 1998;177 (February) Anal HPV in HIV
/
and HIV
0
Homosexual Men
Because there were few HIV-negative men in the highest exposure
HIV-positive men, 269 (93%) tested positive with the consen-
category for most behaviors, the medium- and high-risk categories
sus HPV probes or §1 of the type-specific probes. Among the
were grouped together to evaluate risk factors for HPV infection
specific HPV types, HPV-16 was detected in 109 samples
in HIV-negative men.
(38%) and was the single most common HPV type. Other
The SAS statistical package was used for data analysis. Analyses
common HPV types included 6, 18, 31, 53, and 70, each of
were conducted separately for HIV-positive and HIV-negative
which was found in at least 20% of the HIV-positive men.
men. Among HIV-negative men, overall positivity for HPV by
Among HIV-negative men, 122 (61%) tested positive with
PCR and HC was used as the outcome variable. Relative risks
the consensus HPV probes or §1 of the type-specific probes.
(RR) and 95% confidence interval (CI) were calculated [23]. The
As in the HIV-positive men, the most common HPV type in
Mantel-Haenszel procedure was used to compute adjusted RRs.
the HIV-negative men was HPV-16, which was found in 37
For multilevel categoric variables, the Mantel-Haenszel test for
trend was computed. Mean numbers of HPV types and mean RLU
samples (19%). The next most common HPV types were 6,
ratios were compared for various strata of potential risk factors by
31, 53, and 58, each of which was found in 7% – 9% of the
using analysis of variance. For dichotomous risk factors, CIs for
HIV-negative men. Of the HIV-negative men, 18 (9%) were
differences between means were based on the tstatistic. For multi-
positive with the consensus probes but were negative for the
level risk factors, the least-significant difference was used to deter-
29 specific and 10 grouped HPV types. Table 2 shows that the
mine which means were different from each other [24]. The loga-
proportion of men positive for HPV by PCR was higher among
rithmic transformation of the RLU ratio was done before analysis,
HIV-positive men than among HIV-negative men. Among
and the means, differences in means, and CIs for the differences
HIV-positive men, a high proportion were HPV-positive by
were transformed back to the original scale to obtain a normal
PCR at all CD4 cell levels.
distribution. This resulted in the expression of differences between
The number of HPV types detected by PCR was used as
groups as a ratio of the two means. For number of HPV types, a
another measure of the amount of HPV infection in HIV-positive
transformation was not necessary.
and HIV-negative men. Of the HIV-positive men, 212 (73%)
had ú1 specific HPV type. In contrast, 46 (23%) of the HIV-
negative men had ú1 HPV type. The mean number of HPV
Results types among HIV-positive men who were positive for §1HPV
type did not differ significantly among those with different CD4In total, 608 men were eligible for this study; 346 were HIV-
positive and 262 were HIV-negative. There were no differences cell counts, with a mean of 3.4 types among those with counts
ú500/mm
3
, 3.3 types in those with counts of 200–500/mm
3
,andbetween the HIV-positive and HIV-negative men or between the
SFGH, UCSF, and SFMHS cohorts by race, ethnicity, or educa- 3.9 types in those with counts õ200/mm
3
. The mean number of
types among HIV-negative men who were positive for §1HPVtion. Overall, 91% of the subjects were Caucasian (not Hispanic),
7% were Hispanic, 2% were African-American, and 1% were type was 1.9, significantly fewer than among the HIV-positive
men (Põ.001). We also determined whether there was a rela-other. Nine percent reported £12 years of education, 54% re-
ported 13– 16 years, and 36% reported ú16 years of education. tionship between detection of particular HPV types in HIV-posi-
tive men and more advanced immunosuppression as measuredThe mean age of the HIV-positive men was 42 years (range,
24–64) and the mean age of the HIV-negative men was 45 years by CD4 cell levels. In an examination of the association between
CD4 cell counts and positivity for the 20 individual HPV types(range, 26–73). The UCSF cohort was slightly younger (mean
age, 42; range, 24 –73) and had a higher proportion of HIV- or groups detected in at least 20 HIV-positive men, types 18, 45,
53, and 59 showed an increasing prevalence with decreasingpositive men (70%) than the SFMHS (mean age, 43; range, 31–
63, 44% HIV-positive) and SFGH cohorts (mean age, 45; range, CD4 cell count, but none was statistically significant.
HC data were available for 600 of the 608 men in the study;28– 68, 56% HIV-positive). Behavioral variables, such as smok-
ing, alcohol consumption, recreational drug use, and sexual prac- 297 (87%) of the HIV-positive men and 97 (37%) of the HIV-
negative men were positive for HPV by HC. Table 2 showstices, were similar among the cohorts when stratified by HIV
status. HIV-positive men reported significantly more receptive that the proportion of men positive for HPV by HC was higher
among HIV-positive men than HIV-negative men. Similar toanal intercourse than the HIV-negative men (Põ.001).
Results of HPV testing. Of the 601 men for whom a sample the results of PCR, among HIV-positive men, a high proportion
were HPV-positive by HC at all CD4 cell levels, but the propor-was available for analysis by PCR, 112 (19%) tested negative
for
b
-globin and were excluded from analysis. The
b
-globin tion was highest among those with the lowest CD4 cell level.
Among the 268 HIV-positive men who were also positive fornegativity rate varied little between the HIV-negative men
(23%) and the HIV-positive men who had CD4 cell counts HPV DNA by PCR and who had HC results, 256 (96%) were
HPV-positive by HC. Significantly fewer (Põ.001) of theú500/mm
3
(17%) or between 200 and 500/mm
3
(19%). How-
ever, HIV-positive men with CD4 cell counts õ200/mm
3
were HIV-negative men who were positive for HPV DNA by PCR
were also positive for HPV by HC (74/122, 61%).less likely (7%) to be
b
-globin negative.
Table 1 shows the number of men stratified by HIV status Of the 340 HIV-positive men for whom HC results were
available, 272 (80%) were positive for §1 group B types andwho were positive for specific HPV types by PCR. Among
/ 9d3f$$fe10 12-01-97 13:08:24 jinfa UC: J Infect
Downloaded from https://academic.oup.com/jid/article-abstract/177/2/361/925365 by guest on 17 June 2019
364 Palefsky et al. JID 1998;177 (February)
Table 1. No. of HIV-positive and HIV-negative homosexual men with anal infection by HPV type
as measured by polymerase chain reaction (PCR).
HIV positive HIV negative
(nÅ289) (nÅ200)
% HPV
/
of
HPV type % HPV
/
of all HPV
/
% HPV
/
of % HPV
/
of all
using PCR No. total subjects* subjects
†
No. total subjects* HPV
/
subjects
†
HPV-positive 269 93 100 122 61 100
6582022147 11
11 43 15 16 12 6 10
16 109 38 41 37 19 30
18 80 28 30 5 3 4
26 7 2 3 0 0 0
31 72 25 27 16 8 13
32 15 5 6 3 2 2
33 34 12 13 9 5 7
35 13 5 5 2 1 2
39 19 7 7 3 2 2
40 4 1 1 1 1 1
45 33 11 12 7 4 6
51 12 4 4 4 2 3
52 40 14 15 7 4 6
53 75 26 28 16 8 13
54 21 7 8 5 3 4
55 15 5 6 6 3 5
56 18 6 7 6 3 5
58 52 18 19 17 9 14
59 23 8 9 2 1 2
61 42 15 16 12 6 10
66 27 9 10 2 1 2
68 26 9 10 6 3 5
69 23 8 9 6 3 5
70 63 22 23 8 4 7
73 18 6 7 3 2 2
AE2 7 2 3 1 1 1
AP155 28 10 10 6 3 5
AP291 18 6 7 4 2 3
Mix
‡
36 12 13 5 3 4
Other
§
23 8 9 18 9 15
* % given for each specific type represents proportion of all subjects.
†
% given for each specific type represents proportion of all HPV-positive subjects.
‡
Mix contains probes for HPV-2/13/34/42/57/62/64/67/72/W13B.
§
Specimens positive with consensus probes but negative for 39 specific types.
213 (63%) were positive for infection with §1 group A types. and was found in 15 subjects (6%). HPV-16 sought as a single
type was found in 23 (9%) HIV-negative men.When group B was separated into 2 groups containing types
18/31/33/45/58 and types 35/39/51/52/56 (based on their fre- The HC RLU ratio was used to measure the amount of HPV
DNA in the specimens. Table 3 compares the levels of HPVquency of detection in cervical cancer), these were equally
common with 194 subjects (57%) in each group. The least DNA among HIV-positive and HIV-negative subjects who
were HPV-positive by HC. Among those positive for group A,common group consisted of types 6/11 and was found in 126
subjects (37%). HPV-16 sought as a single type was found in the mean group A RLU ratio was higher for the HIV-positive
men than the HIV-negative men, but there was little difference119 (35%) HIV-positive men. Of the 260 HIV-negative men
for whom HC results were available, 75 (29%) were positive by CD4 cell level among the HIV-positive men. Among men
positive for group B, the mean group B ratio was higher amongfor §1 group B HPV types and 42 (16%) were positive for
§1 group A HPV types. The most common specific group HIV-positive men than HIV-negative men, and among the
HIV-positive men, there was an association between higherconsisted of types 18/31/33/45/58 and was found in 39 subjects
(15%). The least common group consisted of types 42/43/44 RLU ratios and lower CD4 cell levels (PÅ.004).
/ 9d3f$$fe10 12-01-97 13:08:24 jinfa UC: J Infect
Downloaded from https://academic.oup.com/jid/article-abstract/177/2/361/925365 by guest on 17 June 2019
365JID 1998;177 (February) Anal HPV in HIV
/
and HIV
0
Homosexual Men
Table 2. No. of men positive for HPV and relative risks for HPV positivity using polymerase chain
reaction (PCR) and hybrid capture (HC), stratified by HIV status and CD4 cell level.
HIV status and PCR
/
,HC
/
,
CD4 cells/mm
3
no. (%) PCR
0
, no. RR (95% CI) no. (%) HC
0
, no. RR (95% CI)
HIV-negative 122 (61) 78 1.0 97 (37) 163 1.0
HIV-positive* 269 (93) 20 1.5 (1.4 –1.7) 297 (87) 43 2.3 (2.0 –2.8)
CD4 ú500 60 (86) 10 1.4 (1.2 –1.6) 65 (77) 19 2.2 (1.8– 2.7)
CD4 200–500 110 (96) 4 1.6 (1.4– 1.8) 124 (89) 16 2.6 (2.1 –3.1)
CD4 õ200 99 (94) 6 1.6 (1.4– 1.8) 106 (94) 7 2.7 (2.3– 3.2)
Pfor trend
†
.05 .001
NOTE. RR, relative risk; CI, confidence interval.
* CD4 cell counts were not available for 3 HIV-positive men.
†
Tests for trend were calculated on data for HIV-positive subjects only.
Risk factors for detection of HPV. Lifetime exposure to to HPV infection. The RRs for HPV DNA positivity by HC
were similar to those by PCR.potential risk factors for detection of anal HPV DNA was
evaluated for HIV-negative men. For HIV-positive men, risk
factors for the presence or absence of HPV DNA could not be
Discussion
assessed because most were positive for HPV DNA by PCR
or HC. Among the risk factors examined in univariate analyses Compared with studies on cervical cancer [16], the number
for the HIV-negative men, the RRs for HPV infection detected of studies of anal cancer and the number of anal cancer cases
by PCR were lifetime rectal drug use (RR, 1.4; 95% CI, 1.1– analyzed have been limited [15]. Furthermore, no studies have
1.7), lifetime history of rectal discharge (RR, 1.3; 95% CI, subjected anal cancer to techniques capable of detecting the
1.0–1.7), and lifetime level of receptive anal intercourse when wide number of HPV types reported in cervical cancer. There-
compared with no receptive anal intercourse (low, RR, 1.3; fore, the true range of HPV types associated with anal cancer
95% CI, 0.97–1.7; medium or high, RR, 1.5; 95% CI, 1.1 – is not known. Likewise, no studies have characterized the range
2.1; Pfor trend Å.03). There was no confounding of the of HPV types detectable in the anal canal of men without anal
univariate relative risks by other factors and little change in lesions or with precancerous lesions. In this study, we used
the risk estimates. Other risk factors examined were not related PCR to detect 29 individual HPV types as well as 10 grouped
types and compared the distribution of HPV types in HIV-
positive and HIV-negative men. Anal HPV infection detected
Table 3. Mean RLU among men positive for hybrid capture group
by PCR was nearly universal (93%) among HIV-positive ho-
A or group B HPV infection, stratified by HIV status and CD4 cell
mosexual men, consistent with the results of an earlier study
levels.
[8]. A large proportion of HIV-negative homosexual men were
Mean RLU of
also HPV-positive by PCR (61%). The range of HPV types
HIV status and HIV
/
/mean
detected in the anal canal was wide and did not differ apprecia-
CD4 cells/mm
3
No. Mean RLU RLU of HIV
0
95% CI
bly between HIV-positive and HIV-negative men. In both
groups, the single most frequently detected type was HPV-16,
Group A*
the type most commonly detected in both cervical cancer [16]
HIV-negative 42 5.6 1.0 —
HIV-positive
†
213 13.3 2.4 1.4 –4.0
and anal cancer [15].
CD4 ú500 39 12.7 2.3 1.1 –4.6
A large proportion of the men had HPV types typically
CD4 200–500 94 13.8 2.4 1.4 –4.4
categorized as medium to high risk, on the basis of the HPV
CD4 õ200 79 12.9 2.3 1.3 –4.2
types associated with cervical cancer [16]. It is not known if
Group B
‡
these HPV types confer a similar risk for anal cancer. Among
HIV-negative 75 4.5 1.0 —
HIV-positive
†
272 17.5 4.0 2.7 –5.4
the HPV types found relatively infrequently in cervical cancer
CD4 ú500 56 10.8 2.4 1.5 –3.8
but commonly in the anal canal in our study were types 53,
CD4 200–500 108 17.6 3.9 2.6–5.8
58, 61, and 70. HPV-70 is most closely related phylogenetically
CD4 õ200 106 23.4 5.1 3.4– 7.7
to HPV 39, which is considered an intermediate-risk HPV
NOTE. RLU, ratio of chemiluminescent signal in test sample divided by
type [25]. Similarly, types 53, 58, and 61 are phylogenetically
control sample that contains 10 pg/mL viral DNA. CI, confidence interval.
related to other intermediate-risk HPV types.
* HPV-6/11/42/43/44.
We also report detection of HPV-32 outside the oral cavity
†
Some HIV-positive subjects did not have CD4 cell results available.
‡
HPV-16/18/31/33/35/39/45/51/52/56/58.
in a small proportion of both HIV-positive and HIV-negative
/ 9d3f$$fe10 12-01-97 13:08:24 jinfa UC: J Infect
Downloaded from https://academic.oup.com/jid/article-abstract/177/2/361/925365 by guest on 17 June 2019
366 Palefsky et al. JID 1998;177 (February)
men. HPV-32 is typically associated with an oral lesion known with decreasing CD4 cell levels are not known. Cell-mediated
immune response may play a role, as demonstrated by loweras focal epithelial hyperplasia (Heck’s disease), a benign lesion
of the oral cavity characterized by multiple nodular lesions rates of immune response to HPV antigens among women who
were positive for HPV-16 with cervical lesions compared withhistologically similar to flat warts. Detection of HPV-32 in
anal specimens is consistent with the possibility of oral-anal women without cervical lesions [26, 27]. Therefore, higher
levels of anal HPV DNA or replication of antigenically relatedtransmission of HPV infection, perhaps through the sexual
practice of oral-anal intercourse. HPV types may partially be due to loss of HPV-specific immu-
nity subsequent to HIV-mediated immunosuppression. DirectOne of the most striking findings in this study was the fre-
quent detection of multiple HPV types in a single sample. Of interactions between HIV and HPV may also play a role, given
evidence of transactivation of HPV early-region gene expres-the HIV-positive men, 73% had ú1 HPV type, and on average
these men had ú3 types. Detection of higher numbers of HPV sion by the HIV-1 tat protein [28].
types among the HIV-positive men than among HIV-negative Among HIV-negative men, detection of anal HPV DNA was
men may reflect more reporting of receptive anal intercourse associated with anal activity such as receptive anal intercourse
by HIV-positive men. Since it is not known when HPV was and recreational use of rectal drugs. These data suggest that
acquired by the men in this study, we do not know if differences among HIV-negative men, anal HPV infection is sexually ac-
between HIV-positive and HIV-negative men reflect in part quired, similar to data showing that cervical HPV infection in
the length of time infected with HPV prior to study entry. The women is sexually acquired [29 –31]. Although risk factors for
differences between the groups may also reflect immunosup- HPV positivity could not be assessed among HIV-positive men,
pression-mediated HPV DNA replication, since this may facili- it is likely that sexual activity plays a role in this group as it
tate detection of HPV types that would otherwise be missed if does in the HIV-negative men, given the generally high levels
present at levels below the threshold of sensitivity of PCR. of sexual activity in HIV-positive men.
However, since there was no association between number of The results of this study should be interpreted with caution,
HPV types and CD4 cell level among those who were HPV- since they were performed in highly sexually active urban
positive, these data suggest that any influence of immunosup- populations with a mean age ú40 years, and it is not known
pression manifests relatively early in the course of HIV disease. if such results are applicable to populations of homosexual
The role of multiple HPV types in the pathogenesis of ASIL men with different levels of sexual activity or at different ends
and anal cancer is unclear, and it is unknown whether different of the age spectrum. In addition, many men in this study had
HPV types interact to potentiate pathogenesis. However, detec- multiple HPV types, and the number of HPV types was proba-
tion of multiple HPV types was associated with concurrent bly underestimated because probes were not available for all
anal lesions and progression of anal lesions to a higher grade types. Finally, compared with many other studies, a high pro-
over a 2-year follow-up period in both HIV-positive and HIV- portion of the samples were negative for
b
-globin. This proba-
negative men compared with detection of a single type or no bly represents the presence of unknown inhibitors in the anal
type (Palefsky J, unpublished data). specimens for the majority of these cases. The reason for the
HIV-positive men had higher mean RLU ratios for both low
b
-globin negativity rate among the HIV-positive men with
group A and group B HPV types than did HIV-negative men, the lowest CD4 cell counts is not clear, but the high rate of
and this may reflect the higher average number of HPV types negativity most likely did not affect our conclusions, since
detected in the HIV-positive men. However, among HIV-posi- the specimens negative for
b
-globin were removed from the
tive men, lower CD4 cell counts were associated with higher analyses and the remaining PCR data were stratified by HIV
group B DNA RLU ratios but not higher group A HPV DNA status and CD4 cell group.
RLU ratios. Since the mean number of HPV types varied little Several studies have shown that anal HPV infection is an
by CD4 cell level, these data suggest that there may be selective important risk factor for ASIL [7 –9, 20], and our data are
replication of group B types as immunosuppression increases. consistent with the higher rates of ASIL found among HIV-
Further, our data showed a nonsignificant trend toward increas- positive men [5 –10]. The present study suggests that a substan-
ing detection of 4 specific HPV types among HIV-positive men tial number of HIV-positive and HIV-negative homosexual
with lower CD4 cell levels: 18, 45, 53, and 59. This is of men may be at risk for ASIL. Studies of change in HPV infec-
interest for two reasons: (1) In a study of cervicovaginal HPV tion and levels of HPV DNA and their relationship to incident
infection in 1781 HIV-positive women enrolled in the Wom- ASIL are in progress in this cohort.
en’s Interagency HIV Study, the same 4 types were signifi-
cantly more common among the women with lower CD4 cell
levels (Palefsky J, unpublished data), and (2) 3 of these types
(18, 45, and 59) are closely related phylogenetically [25].
Acknowledgments
The mechanisms underlying the association between HIV-
associated immunosuppression and higher levels of group B
We thank Rosanna Botts and Maria daCosta for their expert
technical assistance.
HPV DNA or increasing detection of particular HPV types
/ 9d3f$$fe10 12-01-97 13:08:24 jinfa UC: J Infect
Downloaded from https://academic.oup.com/jid/article-abstract/177/2/361/925365 by guest on 17 June 2019
367JID 1998;177 (February) Anal HPV in HIV
/
and HIV
0
Homosexual Men
References 16. Bosch FX, Manos MM, Munoz N, et al. Prevalence of human papillomavi-
rus in cervical cancer: a worldwide perspective. International biological
1. Daling JR, Weiss NS, Klopfenstein LL, Cochran LE, Chow WH, Daifuku study on cervical cancer (IBSCC) Study Group. J Natl Cancer Inst 1995;
R. Correlates of homosexual behavior and the incidence of anal cancer. 87:796– 802.
JAMA 1982;247:1988– 90. 17. Kiviat N, Rompalo A, Bowden R, et al. Anal human papillomavirus infec-
2. Daling JR, Weiss NS, Hislop TG, et al. Sexual practices, sexually transmit- tion among human immunodeficiency virus –seropositive and – seroneg-
ted diseases, and the incidence of anal cancer. N Engl J Med 1987; 317: ative men. J Infect Dis 1990;162:358 –61.
973–7. 18. Winkelstein W Jr, Lyman DM, Padian N, et al. Sexual practices and risk
3. Holly EA, Whittemore AS, Aston DA, Ahn DK, Nickoloff BJ, Kristiansen of infection by the human immunodeficiency virus. The San Francisco
JJ. Anal cancer incidence: genital warts, anal fissure or fistula, hemor- Men’s Health Study. JAMA 1987;257:321 –5.
rhoids, and smoking. J Natl Cancer Inst 1989;81:1726– 31. 19. Moss AR, Bacchetti P, Osmond D, et al. Seropositivity for HIV and the
4. Qualters JR, Lee NC, Smith RA, Aubert RE. Breast and cervical cancer development of AIDS or AIDS related condition: three year follow up
surveillance, United States, 1973– 1987. MMWR Morb Mortal Wkly of the San Francisco General Hospital cohort. Br Med J (Clin Res Ed)
Rep 1992;41:1– 15. 1988;296:745 –50.
5. Melbye M, Palefsky J, Gonzales J, et al. Immune status as a determinant of 20. Palefsky JM, Gonzales J, Greenblatt RM, Ahn DK, Hollander H. Anal
human papillomavirus detection and its association with anal epithelial intraepithelial neoplasia and anal papillomavirus infection among homo-
abnormalities. Int J Cancer 1990;46:203– 6. sexual males with group IV HIV disease. JAMA 1990;263:2911 –6.
6. Caussy D, Goedert JJ, Palefsky J, et al. Interaction of human immunodefi- 21. Ting Y, Manos MM. Detection and typing of genital human papillomavi-
ciency and papilloma viruses: association with anal epithelial abnormal- ruses. In: Innis MA, Gelfand DH, Sninsky JJ, White TJ, eds. PCR
ity in homosexual men. Int J Cancer 1990;46:214– 9. protocols and applications. San Diego: Academic Press, 1990:356 – 67.
7. Critchlow CW, Holmes KK, Wood R, et al. Association of human immuno- 22. Palefsky J, Holly E, Ralston M, Arthur S, Hogeboom C, Darragh T. Anal
deficiency virus and anal human papillomavirus infection among homo- cytologic abnormalities and anal HPV infection in men with Centers
sexual men. Arch Intern Med 1992;152:1673– 6. for Disease Control group IV HIV disease. Genitourinary Med 1997;
8. Kiviat N, Critchlow C, Holmes K, et al. Association of anal dysplasia 73:174– 80.
and human papillomavirus with immunosuppression and HIV infection 23. Breslow NE, Day NE. Statistical methods in cancer research. Vol 2— The
among homosexual men. AIDS 1993;7:43–9. design and analysis of cohort studies. IARC Sci Publ 1987;82:1 –406.
9. Palefsky JM, Shiboski S, Moss A. Risk factors for anal human papillomavi- 24. Winer BJ. Statistical principles in experimental design. New York:
rus infection and anal cytologic abnormalities in HIV-positive and HIV- McGraw-Hill, 1971:199– 200.
negative homosexual men. J Acquir Immune Defic Syndr 1994;7:599 – 25. Chan SY, Delius H, Halpern AL, Bernard HU. Analysis of genomic se-
606. quences of 95 papillomavirus types: uniting typing, phylogeny, and
10. Breese PL, Judson FN, Penley KA, Douglas J Jr. Anal human papillomavi- taxonomy. J Virol 1995;69:3074– 83.
rus infection among homosexual and bisexual men: prevalence of type- 26. Nakagawa M, Stites DP, Kong F, Farhat S, Moscicki AB, Palefsky JM.
specific infection and association with human immunodeficiency virus. Cytotoxic T lymphocyte responses to the human papillomavirus type
Sex Transm Dis 1995;22:7– 14. 16 E6 and E7 proteins: relationship to cervical intraepithelial neoplasia.
11. Rabkin CS, Yellin F. Cancer incidence in a population with a high preva- J Infect Dis 1997; 175:927– 31.
lence of infection with human immunodeficiency virus type 1. J Natl 27. Nakagawa M, Stites D, Farhat S, et al. T cell response to human papillo-
Cancer Inst 1994;86:1711– 6. mavirus type 16: relationship to cervical intraepithelial neoplasia. Clin
12. Melbye M, Cote TR, Kessler L, Gail M, Biggar RJ. High incidence of Diagn Lab Immunol 1996; 3:205– 10.
anal cancer among AIDS patients. The AIDS/Cancer Working Group. 28. Vernon SD, Hart CE, Reeves WC, Icenogle JP. The HIV-1 tat protein
Lancet 1994;343:636– 9. enhances E2-dependent human papillomavirus 16. Virus Res 1993;27:
13. Beckmann AM, Daling JR, Sherman KJ, et al. Human papillomavirus 133– 45.
infection and anal cancer. Int J Cancer 1989;43:1042– 9. 29. Schneider A, Koutsky LA. Natural history and epidemiological features
14. Palefsky JM, Holly EA, Gonzales J, Berline J, Ahn DK, Greenspan JS. of genital HPV infection. IARC Sci Publ 1992; 119:25– 52.
Detection of human papillomavirus DNA in anal intraepithelial neopla- 30. Bauer HM, Hildesheim A, Schiffman MH, et al. Determinants of genital
sia and anal cancer. Cancer Res 1991;51:1014– 9. human papillomavirus infection in low-risk women in Portland, Oregon.
15. Zaki SR, Judd R, Coffield LM, Greer P, Rolston F, Evatt BL. Human Sex Transm Dis 1993;20:274 –8.
papillomavirus infection and anal carcinoma. Retrospective analysis by 31. Hildesheim A, Gravitt P, Schiffman MH, et al. Determinants of genital
in situ hybridization and the polymerase chain reaction. Am J Pathol human papillomavirus infection in low-income women in Washington,
D.C. Sex Transm Dis 1993;20:279– 85.1992;140:1345– 55.
/ 9d3f$$fe10 12-01-97 13:08:24 jinfa UC: J Infect
Downloaded from https://academic.oup.com/jid/article-abstract/177/2/361/925365 by guest on 17 June 2019
