Abstract
Background
People living with human immunodeficiency virus (HIV; PLWH) have a markedly elevated anal cancer risk, largely due to loss of immunoregulatory control of oncogenic human papillomavirus infection. To better understand anal cancer development and prevention, we determined whether recent, past, cumulative, or nadir/peak CD4+ T-cell count (CD4) and/or HIV-1 RNA level (HIV RNA) best predict anal cancer risk.
Methods
We studied 102 777 PLWH during 1996–2014 from 21 cohorts participating in the North American AIDS Cohort Collaboration on Research and Design. Using demographics-adjusted, cohort-stratified Cox models, we assessed associations between anal cancer risk and various time-updated CD4 and HIV RNA measures, including cumulative and nadir/peak measures during prespecified moving time windows. We compared models using the Akaike information criterion.
Results
Cumulative and nadir/peak CD4 or HIV RNA measures from approximately 8.5 to 4.5 years in the past were generally better predictors for anal cancer risk than their corresponding more recent measures. However, the best model included CD4 nadir (ie, the lowest CD4) from approximately 8.5 years to 6 months in the past (hazard ratio [HR] for <50 vs ≥500 cells/µL, 13.4; 95% confidence interval [CI], 3.5–51.0) and proportion of time CD4 <200 cells/µL from approximately 8.5 to 4.5 years in the past (a cumulative measure; HR for 100% vs 0%, 3.1; 95% CI, 1.5–6.6).
Conclusions
Our results are consistent with anal cancer promotion by severe, prolonged HIV-induced immunosuppression. Nadir and cumulative CD4 may represent useful markers for identifying PLWH at higher anal cancer risk.
Keywords: HIV infection, CD4+ T-cell count, HIV-1 RNA viral load, anal cancer, risk
Among various CD4 or human immunodeficiency virus (HIV) RNA measures, nadir and cumulative CD4 were the best predictors for anal cancer risk, suggesting a key role for severe, prolonged HIV-induced immunosuppression in anal carcinogenesis. Early antiretroviral therapy may help prevent anal cancer.
Anal cancer risk is markedly elevated among people living with human immunodeficiency virus (HIV; PLWH), especially those with AIDS and men who have sex with men (MSM) [1, 2]. Immunosuppression is associated with persistent oncogenic human papillomavirus (HPV) infection, which promotes development of anal high-grade squamous intraepithelial lesions/anal intraepithelial neoplasia 2-3 (HSIL/AIN2-3) but may play a lesser role in HSIL/AIN2-3 progression to anal cancer [3–6]. Despite improvements in immune function after the introduction of effective antiretroviral therapy (ART) in 1996, most studies of anal cancer incidence trends in PLWH have reported an increase or no change over time [7–15], with an apparent decline in recent years [1, 2]. Among non-AIDS–defining cancers, anal cancer exhibits the highest relative risk in PLWH compared with the general population [1].
Immunosuppression and HIV viremia have been associated with anal cancer risk, but the time courses over which they exert their effects are unclear. Increased anal cancer risk has been associated with lower CD4+ T-cell count (CD4) [8–10, 16–26] and higher HIV-1 RNA level (HIV RNA) [18, 24, 26–28], even after mutual adjustment [18, 24, 26, 27]. It remains unknown whether risk is driven by recent, past, cumulative, or nadir/peak (ie, the lowest CD4 and the highest HIV RNA) CD4 and/or HIV RNA. Evidence suggests that past (ie, 6–7 years before diagnosis) [23], cumulative (ie, duration with CD4 <200 cells/µL) [18], and nadir (ie, the lowest CD4 ever or before ART initiation) [8, 23, 24, 26] CD4 are better risk predictors than recent CD4 and that cumulative HIV RNA (eg, duration with HIV RNA >100 000 copies/mL) is a better predictor than recent HIV RNA [18, 27].
To contribute to our understanding of anal carcinogenesis and help tailor approaches to prevention, we aimed to determine whether recent, past, cumulative, or nadir/peak CD4 and HIV RNA measures best predict anal cancer risk among PLWH in the North American AIDS Cohort Collaboration on Research and Design (NA-ACCORD) [29].
METHODS
Study Sample
Our study population included PLWH (aged ≥18 years) from 21 cohorts in the United States and Canada that contributed data, including cancer diagnoses, to the NA-ACCORD (1996–2014) [15, 29]. Cohorts submitted demographic and clinical data using standardized collection methods. Incident invasive anal cancers were validated using a standardized process of either cancer registry linkage or manual review of medical records and pathology reports [15]. Institutional review board approval was obtained for each cohort.
Follow-up started at the latest of the following dates: 1 January 1996, NA-ACCORD entry (first of 2 HIV primary care visits ≤1 year apart), cohort-specific start for cancer reporting, 18th birthday, or 360 days before the later of the first CD4 or HIV RNA measurement. Follow-up ended at the earliest of anal cancer diagnosis, cohort-specific end for cancer reporting, death, or 540 days after the earlier of the last CD4 or HIV RNA measurement. We excluded persons with <2 CD4 or <2 HIV RNA measurements or with follow-up ≤180 days.
Time-updated CD4 and HIV RNA Measures
We estimated CD4 and HIV RNA values at 30-day intervals using observed laboratory measurements, as previously described (Supplementary Methods) [30]. Based on these estimated values, we constructed time-updated CD4 and HIV RNA measures that were all lagged by at least 180 days to reduce the possibility of reverse causality. Thus, we examined CD4 and HIV RNA lagged by 180 days (approximately 6 months; defined as “recent”) and then by longer lag (lagged by 360 day [approximately 1 year] increments from 540 to 1620 days, and by 720 day [approximately 2 years] increments from 1620 to 3060 days).
We also constructed lagged cumulative and nadir/peak measures during prespecified moving time windows. Cumulative measures were CD4 or HIV RNA average (ie, normalized by time) and proportion of time CD4 <200 cells/µL or HIV RNA >500 copies/mL. Nadir/peak measures were lowest CD4 and highest HIV RNA. The duration of the “early” (more distant past) and “late” (more recent past) windows was 1440 days (approximately 4 years). The early window was lagged by 1620 days (ie, covered approximately 8.5 to approximately 4.5 years in the past), and the late window was lagged by 180 days (ie, covered approximately 4.5 years to approximately 6 months in the past). The “overall” (early and late combined) window (2880 days duration; approximately 8 years) covered 3060 to 180 days (approximately 8.5 years to approximately 6 months) in the past.
Analyses of these various measures were restricted to persons with follow-up duration greater than the examined lag or window start (eg, the late window [starting at 1620 days in the past] can only be assessed among persons with follow-up >1620 days). Thus, the analyzed sample size decreased with increasing lag or amount of time in the past of the window start.
Statistical Analyses
We assessed associations between CD4 and HIV RNA measures and anal cancer risk using cohort-stratified Cox regression models with follow-up time as the time scale using counting process syntax [31] to account for time-varying covariates updated at 30-day intervals. We modeled these measures as categorical variables (calculating likelihood ratio global P values) and as continuous variables to test for trends (Ptrend).
To identify a final model with the most robust independent CD4 and/or HIV RNA predictors, we compared models using the Akaike information criterion (AIC), which accounts for both model fit and parsimony. A smaller AIC indicates a better model; a difference in AIC that is >10 between models is considered meaningful [32]. Models being compared must use the same participant set. To make valid comparisons across all models, we examined AICs among persons with follow-up >3060 days (N = 34 625; anal cancer cases = 170). To choose CD4 measures for further testing, we compared separate models including each CD4 measure individually. Then, to select the best CD4 measure(s), we compared models that included the measures chosen for further testing, 2 at a time in the same model. We did the same for HIV RNA measures. To develop our final model, we compared models including combinations of the best CD4 and/or HIV RNA measures in the same model.
All models were adjusted for sex, race/ethnicity, and baseline age and calendar period. We did not adjust for history of AIDS due to its expected high collinearity with CD4 nadir and other CD4 or HIV RNA measures. In the primary analysis, we did not adjust for HIV risk group or smoking due to a large number of unknowns, and we did not adjust for ART use because ART initiation during the study period was affected by prior CD4 [33]. In separate sensitivity analyses, we adjusted for HIV risk group, smoking, and cumulative ART use in our final model and assessed whether HIV risk group modified the association between the predictors and anal cancer risk in our final model (Supplementary Methods).
We used SAS version 9.4 (SAS Institute Inc., Cary, NC) to perform analyses, and a 2-sided P value of .05 to determine statistical significance.
RESULTS
Among 122 840 PLWH included in the 21 NA-ACCORD cohorts, 102 777 persons were eligible for this study (Table 1). Excluded persons had <2 CD4 or <2 HIV RNA measurements (n = 13 701) or no follow-up time (n = 6362). Most eligible persons were male (85%), of white (44%) or black race/ethnicity (40%), and started follow-up during the period 1996–2003 (57%). At baseline, most were aged ≥40 years (57%) and ART naive (67%), with CD4 ≥200 cells/µL (69%) and HIV RNA >500 copies/mL (70%). Persons diagnosed with anal cancer (n = 492) included a larger proportion of males (96%), persons of white race/ethnicity (61%), and persons who started follow-up during the period 1996–1999 (55% vs 32%). At baseline, they were more likely to be aged ≥40 years (64%) but less likely to be ART naive (57%) or to have a CD4 ≥200 cells/µL (49%).
Table 1.
Baseline Characteristics of Study Sample, North American AIDS Cohort Collaboration on Research and Design, 1996–2014
| Characteristic | All (N = 102 777) | Anal Cancer Cases (n = 492) |
|---|---|---|
| n (%) | n (%) | |
| Sex | ||
| Male | 87 059 (84.7) | 470 (95.5) |
| Female | 15 718 (15.3) | 22 (4.5) |
| Race/ethnicity | ||
| Black | 41 331 (40.2) | 156 (31.7) |
| White | 44 741 (43.5) | 302 (61.4) |
| Hispanic | 6934 (6.8) | 13 (2.6) |
| Other | 4100 (4.0) | 9 (1.8) |
| Unknown, imputed | 5671 (5.5) | 12 (2.4) |
| Age, y | ||
| 18–29 | 13 107 (12.8) | 18 (3.7) |
| 30–39 | 30 890 (30.1) | 160 (32.5) |
| 40–49 | 35 930 (35.0) | 188 (38.2) |
| ≥50 | 22 850 (22.2) | 126 (25.6) |
| Calendar period | ||
| 1996–1999 | 32 426 (31.6) | 272 (55.3) |
| 2000–2003 | 25 949 (25.2) | 133 (27.0) |
| 2004–2007 | 19 870 (19.3) | 55 (11.2) |
| 2008–2011 | 18 287 (17.8) | 28 (5.7) |
| 2012–2014 | 6245 (6.1) | 4 (0.8) |
| Combination antiretroviral therapy naive | ||
| No | 34 022 (33.1) | 210 (42.7) |
| Yes | 68 755 (66.9) | 282 (57.3) |
| CD4 count, cells/µL | ||
| <50 | 11 259 (11.0) | 87 (17.7) |
| 50 to <100 | 6856 (6.7) | 62 (12.6) |
| 100 to <200 | 13 596 (13.2) | 104 (21.1) |
| 200 to <350 | 22 252 (21.7) | 120 (24.4) |
| 350 to <500 | 20 058 (19.5) | 59 (12.0) |
| ≥500 | 28 756 (28.0) | 60 (12.2) |
| HIV RNA level, copies/mL | ||
| ≤500 | 30 987 (30.1) | 133 (27.0) |
| >500 to <10 000 | 20 738 (20.2) | 113 (23.0) |
| 10 000 to <100 000 | 30 394 (29.6) | 138 (28.0) |
| ≥100 000 | 20 658 (20.1) | 108 (22.0) |
| HIV risk group | ||
| Injection drug use | 12 695 (12.4) | 33 (6.7) |
| Men who have sex with men | 32 669 (31.8) | 222 (45.1) |
| Heterosexual | 16 134 (15.7) | 29 (5.9) |
| Other | 2089 (2.0) | 4 (0.8) |
| Unknown, imputeda | 5636 (5.5) | 14 (2.9) |
| Unknown, not imputeda | 33 554 (32.6) | 190 (38.6) |
| Smoking status | ||
| Ever | 54 994 (53.5) | 350 (71.1) |
| Never | 19 491 (19.0) | 102 (20.7) |
| Unknown, imputeda | 15 711 (15.3) | 19 (3.9) |
| Unknown, not imputeda | 12 581 (12.2) | 21 (4.3) |
Abbreviation: HIV, human immunodeficiency virus.
aWe imputed HIV risk group and smoking status for persons with unknown values, except for cohorts with a high proportion of unknowns, or, for smoking, with all the knowns being smokers.
Selection of Individual CD4 or HIV RNA Measures for Further Testing
In separate models, anal cancer risk was significantly associated with each CD4 and HIV RNA measure, except HIV RNA lagged by 180 and 540 days (Tables 2–4). Models with stronger associations (ie, higher hazard ratios [HRs]) and lower AICs were largely concentrated in single measures lagged by ≥1260 days (ie, occurring approximately 8.5 to approximately 3.5 years in the past) and in early cumulative and nadir/peak measures (ie, from 3060 to 1620 days [approximately 8.5 to approximately 4.5 years] in the past). Based on the results from separate models and models that included 2 cumulative or nadir/peak measures of the same type for CD4 (eg, early and late CD4 average in the same model) and HIV RNA, respectively (Supplementary Results; Supplementary Tables 1 and 2), we chose several CD4 measures (1620 day lag, early and overall average, early and overall proportion of time CD4 <200 cells/ µL, and early and overall nadir) and HIV RNA measures (3060 day lag, early average, and early peak) for further consideration.
Table 2.
Recent and Past CD4 Count Measures and Anal Cancer Risk, North American AIDS Cohort Collaboration on Research and Design, 1996–2014
| CD4 Count Measure (cells/µL) | Anal Cancer Cases | Hazard Ratio (95% Confidence Interval)a | Global P Value (Akaike Information Criterion)b |
|---|---|---|---|
| 180 d (approximately 6 mo) lag (ie, “recent”) | |||
| <50 | 27 | 4.1 (2.7, 6.3) | <.0001 (2603) |
| 50 to <100 | 34 | 4.6 (3.1, 6.7) | … |
| 100 to <200 | 73 | 3.1 (2.3, 4.2) | … |
| 200 to <350 | 135 | 2.6 (2.0, 3.3) | … |
| 350 to <500 | 106 | 1.8 (1.4, 2.3) | … |
| ≥500 | 117 | 1.0 (ref) | … |
| Per 50 cells/µLc | … | 0.91 (0.89, 0.93) | … |
| 540 d (approximately 1.5 y) lag | |||
| <50 | 34 | 5.8 (3.9, 8.7) | <.0001 (2591) |
| 50 to <100 | 27 | 3.7 (2.4, 5.7) | … |
| 100 to <200 | 88 | 3.8 (2.9, 5.1) | … |
| 200 to <350 | 113 | 2.2 (1.7, 2.9) | … |
| 350 to <500 | 92 | 1.6 (1.2, 2.1) | … |
| ≥500 | 106 | 1.0 (ref) | … |
| Per 50 cells/µLc | … | 0.90 (0.88, 0.92) | … |
| 900 d (approximately 2.5 y) lag | |||
| <50 | 25 | 4.6 (2.9, 7.2) | <.0001 (2596) |
| 50 to <100 | 32 | 4.6 (3.1, 6.9) | … |
| 100 to <200 | 79 | 3.6 (2.6, 4.8) | … |
| 200 to <350 | 98 | 2.0 (1.5, 2.7) | … |
| 350 to <500 | 74 | 1.4 (1.0, 1.8) | … |
| ≥500 | 98 | 1.0 (ref) | … |
| Per 50 cells/µLc | … | 0.90 (0.88, 0.92) | … |
| 1260 d (approximately 3.5 y) lag | |||
| <50 | 36 | 8.0 (5.4, 12.0) | <.0001 (2584) |
| 50 to <100 | 25 | 4.2 (2.7, 6.6) | … |
| 100 to <200 | 66 | 3.4 (2.4, 4.7) | … |
| 200 to <350 | 80 | 1.9 (1.4, 2.5) | … |
| 350 to <500 | 69 | 1.5 (1.1, 2.0) | … |
| ≥500 | 83 | 1.0 (ref) | … |
| Per 50 cells/µLc | … | 0.89 (0.86, 0.91) | … |
| 1620 d (approximately 4.5 y) lag | |||
| <50 | 33 | 9.5 (6.2, 14.7) | <.0001 (2575) |
| 50 to <100 | 28 | 6.0 (3.8, 9.5) | … |
| 100 to <200 | 58 | 3.8 (2.6, 5.4) | … |
| 200 to <350 | 65 | 1.9 (1.4, 2.8) | … |
| 350 to <500 | 74 | 2.1 (1.5, 2.9) | … |
| ≥500 | 62 | 1.0 (ref) | … |
| Per 50 cells/µLc | … | 0.88 (0.86, 0.91) | … |
| 2340 d (approximately 6.5 y) lag | |||
| <50 | 27 | 9.0 (5.6, 14.6) | <.0001 (2581) |
| 50 to <100 | 14 | 3.6 (2.0, 6.6) | … |
| 100 to <200 | 47 | 3.6 (2.4, 5.5) | … |
| 200 to <350 | 53 | 1.9 (1.3, 2.8) | … |
| 350 to <500 | 48 | 1.6 (1.1, 2.4) | … |
| ≥500 | 50 | 1.0 (ref) | … |
| Per 50 cells/µLc | … | 0.88 (0.86, 0.91) | … |
| 3060 d (approximately 8.5 y) lag | |||
| <50 | 13 | 6.3 (3.3, 12.2) | <.0001 (2580) |
| 50 to <100 | 18 | 6.6 (3.7, 11.9) | … |
| 100 to <200 | 37 | 4.2 (2.6, 6.8) | … |
| 200 to <350 | 47 | 2.5 (1.6, 3.9) | … |
| 350 to <500 | 23 | 1.2 (0.7, 2.0) | … |
| ≥500 | 32 | 1.0 (ref) | … |
| Per 50 cells/µLc | … | 0.87 (0.84, 0.90) | … |
aEach measure was individually included in a cohort-stratified Cox model adjusted for sex, race/ethnicity, and baseline age and calendar period. The N and number of anal cancer cases used for the model of each measure was: 180 d lag (N = 102 777; number of anal cancer cases = 492), 540 d lag (94 458; 460), 900 d lag (79 882; 406), 1260 d lag (68 953; 359), 1620 d lag (60 006; 320), 2340 d lag (45 343; 239), and 3060 d lag (34 625; 170).
bGlobal P value and Akaike information criterion from models among persons with follow-up >3060 days (N = 34 625; number of anal cancer cases = 170).
c P trend < .0001.
Table 4.
Cumulative and Nadir/Peak CD4 Count and Human Immunodeficiency Virus RNA Level Measures During Moving Time Windows and Anal Cancer Risk, North American AIDS Cohort Collaboration on Research and Design, 1996–2014
| Measurea | Moving Time Window | |||||
|---|---|---|---|---|---|---|
| Early | Late | Overall | ||||
| (3060 to 1620 d in the Past) | (1620 to 180 d in the Past) | (3060 to 180 d in the Past) | ||||
| Anal Cancer Cases | HR (95% CI) | Anal Cancer Cases | HR (95% CI) | Anal Cancer Cases | HR (95% CI) | |
| CD4 count average, cells/µL | ||||||
| <50 | 10 | 12.2 (6.0, 24.8) | 15 | 7.6 (4.3, 13.3) | 8 | 20.7 (9.5, 45.0) |
| 50 to <100 | 18 | 9.1 (5.1, 16.2) | 24 | 5.4 (3.4, 8.6) | 14 | 9.3 (5.0, 17.3) |
| 100 to <200 | 28 | 3.7 (2.2, 6.1) | 60 | 3.8 (2.7, 5.3) | 30 | 4.4 (2.7, 7.2) |
| 200 to <350 | 43 | 2.4 (1.5, 3.7) | 74 | 2.0 (1.5, 2.8) | 45 | 2.7 (1.7, 4.1) |
| 350 to <500 | 36 | 1.7 (1.1, 2.7) | 66 | 1.5 (1.1, 2.0) | 35 | 1.6 (1.0, 2.6) |
| ≥500 | 35 | 1.0 (ref) | 81 | 1.0 (ref) | 38 | 1.0 (ref) |
| Per 50 cells/µLb | … | 0.86 (0.83, 0.90) | … | 0.88 (0.86, 0.91) | … | 0.86 (0.83, 0.89) |
| Global P value (AIC)c | … | <.0001 (2577) | … | <.0001 (2594) | … | <.0001 (2569) |
| Proportion of time CD4 count <200 cells/µL | ||||||
| 0.00 | 73 | 1.0 (ref) | 148 | 1.0 (ref) | 61 | 1.0 (ref) |
| >0.00 to 0.25 | 29 | 2.1 (1.4, 3.3) | 44 | 1.7 (1.2, 2.5) | 36 | 1.8 (1.2, 2.7) |
| >0.25 to 0.50 | 12 | 2.1 (1.1, 3.8) | 24 | 2.3 (1.5, 3.5) | 19 | 2.7 (1.6, 4.6) |
| >0.50 to 0.75 | 10 | 2.2 (1.2, 4.3) | 26 | 3.2 (2.1, 4.8) | 18 | 3.6 (2.1, 6.2) |
| >0.75 to <1.00 | 9 | 2.1 (1.1, 4.3) | 32 | 4.1 (2.8, 6.0) | 16 | 4.0 (2.3, 7.1) |
| 1.00 | 37 | 7.9 (5.3, 11.9) | 46 | 4.7 (3.4, 6.6) | 20 | 10.3 (6.1, 17.1) |
| Per 20% of time CD4 <200 cells/µLb | … | 1.39 (1.29, 1.49) | … | 1.36 (1.28, 1.43) | … | 1.46 (1.35, 1.58) |
| Global P value (AIC)c | … | <.0001 (2572) | … | <.0001 (2591) | … | <.0001 (2575) |
| CD4 count nadir, cells/µL | ||||||
| <50 | 45 | 12.6 (6.1, 26.0) | 61 | 7.0 (4.4, 11.2) | 54 | 22.9 (7.1, 73.9) |
| 50 to <100 | 17 | 5.5 (2.4, 12.5) | 34 | 4.5 (2.7, 7.6) | 20 | 11.2 (3.3, 37.7) |
| 100 to <200 | 35 | 4.1 (1.9, 8.5) | 77 | 3.8 (2.4, 6.0) | 35 | 7.6 (2.3, 24.8) |
| 200 to <350 | 49 | 3.2 (1.6, 6.6) | 80 | 2.2 (1.4, 3.5) | 47 | 6.5 (2.0, 20.9) |
| 350 to <500 | 15 | 1.5 (0.6, 3.4) | 42 | 1.6 (1.0, 2.6) | 11 | 2.6 (0.7, 9.4) |
| ≥500 | 9 | 1.0 (ref) | 26 | 1.0 (ref) | 3 | 1.0 (ref) |
| Per 50 cells/µLb | … | 0.82 (0.78, 0.86) | … | 0.85 (0.82, 0.88) | … | 0.79 (0.75, 0.84) |
| Global P value (AIC)c | … | <.0001 (2564) | … | <.0001 (2583) | … | <.0001 (2562) |
| HIV RNA level average, copies/mL | ||||||
| ≤500 | 47 | 1.0 (ref) | 122 | 1.0 (ref) | 44 | 1.0 (ref) |
| >500 to <10 000 | 41 | 1.2 (0.8, 1.9) | 81 | 1.3 (1.0, 1.8) | 51 | 1.4 (0.9, 2.1) |
| 10 000 to <100 000 | 61 | 2.4 (1.6, 3.5) | 88 | 1.9 (1.4, 2.5) | 59 | 2.2 (1.5, 3.3) |
| ≥100 000 | 21 | 5.3 (3.1, 8.9) | 29 | 3.6 (2.4, 5.5) | 16 | 5.4 (3.0, 9.8) |
| Per log10 copies/mLb | … | 1.58 (1.36, 1.84) | … | 1.36 (1.23, 1.52) | … | 1.56 (1.33, 1.83) |
| Global P value (AIC)c | … | <.0001 (2607) | … | .0001 (2628) | … | <.0001 (2617) |
| Proportion of time HIV RNA level >500 copies/mL | ||||||
| 0.00 | 31 | 1.0 (ref) | 95 | 1.0 (ref) | 25 | 1.0 (ref) |
| >0.00 to 0.25 | 29 | 1.6 (0.9, 2.6) | 51 | 1.2 (0.8, 1.7) | 42 | 1.6 (1.0, 2.7) |
| >0.25 to 0.50 | 19 | 1.6 (0.9, 2.8) | 49 | 1.8 (1.3, 2.6) | 24 | 1.5 (0.9, 2.7) |
| >0.50 to 0.75 | 21 | 1.8 (1.0, 3.2) | 47 | 1.9 (1.3, 2.8) | 32 | 2.4 (1.4, 4.1) |
| >0.75 to <1.00 | 40 | 3.1 (1.9, 5.1) | 57 | 2.2 (1.6, 3.2) | 41 | 3.7 (2.2, 6.4) |
| 1.00 | 30 | 2.5 (1.5, 4.3) | 21 | 0.9 (0.6, 1.5) | 6 | 1.8 (0.7, 4.5) |
| Per 20% of time HIV RNA >500 copies/mLb | … | 1.21 (1.11, 1.31) | … | 1.11 (1.04, 1.18) | … | 1.25 (1.14, 1.37) |
| Global P value (AIC)c | … | .0002 (2629) | … | .0001 (2627) | … | <.0001 (2625) |
| HIV RNA level peak, copies/mL | ||||||
| ≤500 | 31 | 1.0 (ref) | 95 | 1.0 (ref) | 25 | 1.0 (ref) |
| >500 to <10 000 | 24 | 1.2 (0.7, 2.0) | 44 | 1.0 (0.7, 1.4) | 23 | 1.3 (0.7, 2.2) |
| 10 000 to <100 000 | 54 | 1.9 (1.2, 3.0) | 92 | 1.6 (1.2, 2.1) | 54 | 1.9 (1.1, 3.0) |
| ≥100 000 | 61 | 3.2 (2.1, 5.1) | 89 | 2.3 (1.7, 3.1) | 68 | 2.8 (1.7, 4.5) |
| Per log10 copies/mLb | … | 1.49 (1.30, 1.70) | … | 1.30 (1.19, 1.43) | … | 1.43 (1.25, 1.64) |
| Global P value (AIC)c | … | <.0001 (2615) | … | .0001 (2628) | … | <.0001 (2626) |
Abbreviations: AIC, Akaike information criterion; CI, confidence interval; HIV, human immunodeficiency virus; HR, hazard ratio.
aEach measure was individually included in a cohort-stratified Cox model adjusted for sex, race/ethnicity, and baseline age and calendar period. The N and number of anal cancer cases for each moving window was: early (N = 34 625; number of anal cancer cases = 170), late (60 006; 320), and overall (34 625; 170).
b P trend < .0001, except for early proportion of time HIV RNA level >500 copies/mL (Ptrend = .0010).
cGlobal P value and AIC from models among persons with follow-up >3060 days (N = 34 625; number of anal cancer cases = 170).
Table 3.
Recent and Past Human Immunodeficiency Virus RNA Level Measures and Anal Cancer Risk, North American AIDS Cohort Collaboration on Research and Design, 1996–2014
| Human Immunodeficiency Virus RNA Level Measure (copies/mL) | Anal Cancer Cases | Hazard Ratio (95% Confidence Interval)a | Global P Value (Akaike Information Criterion)b |
|---|---|---|---|
| 180 d (approximately 6 mo) lag (ie, “recent”) | |||
| ≤500 | 341 | 1.0 (ref) | .14 (2643) |
| >500 to <10 000 | 66 | 0.9 (0.7, 1.2) | … |
| 10 000 to <100 000 | 59 | 1.0 (0.8, 1.3) | … |
| ≥100 000 | 26 | 1.3 (0.9, 1.9) | … |
| Per log10 copies/mLc | … | 1.01 (0.93, 1.11) | … |
| 540 d (approximately 1.5 y) lag | |||
| ≤500 | 288 | 1.0 (ref) | .086 (2642) |
| >500 to <10 000 | 77 | 1.1 (0.8, 1.4) | … |
| 10 000 to <100 000 | 63 | 1.2 (0.9, 1.5) | … |
| ≥100 000 | 32 | 1.8 (1.2, 2.7) | … |
| Per log10 copies/mLc | … | 1.13 (1.03, 1.23) | … |
| 900 d (approximately 2.5 y) lag | |||
| ≤500 | 233 | 1.0 (ref) | .013 (2638) |
| >500 to <10 000 | 74 | 1.2 (0.9, 1.6) | … |
| 10 000 to <100 000 | 70 | 1.5 (1.1, 1.9) | … |
| ≥100 000 | 29 | 2.0 (1.3, 2.9) | … |
| Per log10 copies/mLc | … | 1.22 (1.11, 1.33) | … |
| 1260 d (approximately 3.5 y) lag | |||
| ≤500 | 189 | 1.0 (ref) | <.0001 (2624) |
| >500 to <10 000 | 70 | 1.3 (1.0, 1.7) | … |
| 10 000 to <100 000 | 70 | 1.7 (1.3, 2.3) | … |
| ≥100 000 | 30 | 2.5 (1.7, 3.7) | … |
| Per log10 copies/mLc | … | 1.27 (1.16, 1.39) | … |
| 1620 d (approximately 4.5 y) lag | |||
| ≤500 | 165 | 1.0 (ref) | <.0001 (2627) |
| >500 to <10 000 | 61 | 1.1 (0.8, 1.5) | … |
| 10 000 to <100 000 | 62 | 1.6 (1.2, 2.1) | … |
| ≥100 000 | 32 | 2.8 (1.9, 4.2) | … |
| Per log10 copies/mLc | … | 1.31 (1.19, 1.45) | … |
| 2340 d (approximately 6.5 y) lag | |||
| ≤500 | 105 | 1.0 (ref) | .0002 (2629) |
| >500 to <10 000 | 59 | 1.5 (1.1, 2.0) | … |
| 10 000 to <100 000 | 53 | 1.8 (1.3, 2.6) | … |
| ≥100 000 | 22 | 2.7 (1.7, 4.3) | … |
| Per log10 copies/mLc | … | 1.32 (1.18, 1.48) | … |
| 3060 d (approximately 8.5 y) lag | |||
| ≤500 | 66 | 1.0 (ref) | <.0001 (2619) |
| >500 to <10 000 | 37 | 1.3 (0.9, 1.9) | … |
| 10 000 to <100 000 | 50 | 2.5 (1.7, 3.7) | … |
| ≥100 000 | 17 | 3.1 (1.8, 5.3) | … |
| Per log10 copies/mLc | … | 1.49 (1.30, 1.70) | … |
aEach measure was individually included in a cohort-stratified Cox model adjusted for sex, race/ethnicity, and baseline age and calendar period. The N and number of anal cancer cases used for the model of each measure was: 180 d lag (N = 102 777; number of anal cancer cases = 492), 540 d lag (94 458; 460), 900 d lag (79 882; 406), 1260 d lag (68 953; 359), 1620 d lag (60 006; 320), 2340 d lag (45 343; 239), and 3060 d lag (34 625; 170).
bGlobal P value and Akaike information criterion from models among persons with follow-up >3060 days (N = 34 625; number of anal cancer cases = 170).
c P trend < .0001, except for 180 d lag (Ptrend = .75) and 540 d lag (Ptrend = .0086).
Selection of Best CD4 Predictor(s)
We fit models including the CD4 measures chosen for further consideration above, 2 at a time. The model with both overall CD4 nadir and early proportion of time CD4 <200 cells/µL showed an AIC that was meaningfully lower than any other model containing 1 or 2 CD4 measures (data not shown). These 2 measures were highly correlated (Supplementary Table 3; Spearman correlation coefficient, –0.77; P < .0001), as were most cumulative and nadir CD4 measures.
Selection of the Best HIV RNA Predictor(s)
We fit models including the HIV RNA measures chosen for further consideration above, 2 at a time. None of these models showed an AIC that was meaningfully lower than that for early HIV RNA average alone, the single HIV RNA measure with the lowest AIC (data not shown). Although the AIC difference between separate models for early HIV RNA average and early HIV RNA peak was not meaningful (Table 4), when they were included in the same model, only early HIV RNA average remained significant (global P = .0061).
Final Model
We compared models with combinations of the best CD4 predictors (overall nadir and early proportion of time CD4 <200 cells/µL) and HIV RNA predictors (early average; Table 5). The model with the 2 CD4 measures prevailed as the model with the lowest AIC. Furthermore, early HIV RNA average lost its significance when added to this model. Thus, our final model included overall CD4 nadir (HR for <50 vs ≥500 cells/µL, 13.4; 95% confidence interval [CI], 3.5–51.0; Ptrend < .0001) and early proportion of time CD4 <200 cells/µL (HR for entire vs no time, 3.1; 95% CI, 1.5–6.6; Ptrend = .0030). In sensitivity analyses (data not shown), the HRs for these 2 measures did not meaningfully change after our final model was adjusted for HIV risk group, smoking, or cumulative ART use and were not significantly different between MSM and other/unknown HIV risk groups (P = .69 for overall CD4 nadir and P = .065 for early proportion of time CD4 <200 cells/µL).
Table 5.
Best CD4 Count and Human Immunodeficiency Virus RNA Level Predictors and Final Model for Anal Cancer Risk, North American AIDS Cohort Collaboration on Research and Design, 1996–2014
| Best Predictor | Anal Cancer Cases | Separate Modelsa | Combined Modelsb | |||
|---|---|---|---|---|---|---|
| Model 1 (Final Model) | Model 2 | Model 3 | Model 4 | |||
| HR (95% CI) | HR (95% CI) | HR (95% CI) | HR (95% CI) | HR (95% CI) | ||
| Overall CD4 count nadir, cells/µLc | ||||||
| <50 | 54 | 22.9 (7.1, 73.9) | 13.4 (3.5, 51.0) | 18.4 (5.5, 61.7) | … | 11.7 (3.0, 46.1) |
| 50 to <100 | 20 | 11.2 (3.3, 37.7) | 8.1 (2.1, 31.2) | 10.1 (2.9, 35.0) | … | 7.9 (2.0, 31.0) |
| 100 to <200 | 35 | 7.6 (2.3, 24.8) | 6.2 (1.7, 22.2) | 7.3 (2.2, 24.0) | … | 6.1 (1.7, 22.2) |
| 200 to <350 | 47 | 6.5 (2.0, 20.9) | 6.5 (2.0, 20.9) | 6.4 (2.0, 20.8) | … | 6.4 (2.0, 20.9) |
| 350 to <500 | 11 | 2.6 (0.7, 9.4) | 2.6 (0.7, 9.4) | 2.6 (0.7, 9.4) | … | 2.6 (0.7, 9.4) |
| ≥500 | 3 | 1.0 (ref) | 1.0 (ref) | 1.0 (ref) | … | 1.0 (ref) |
| Global P value | … | <.0001 | <.0001 | <.0001 | … | <.0001 |
| Per 50 cells/µL | … | 0.79 (0.75, 0.84) | 0.85 (0.79, 0.92) | 0.82 (0.77, 0.87) | … | 0.87 (0.80, 0.93) |
| Ptrend | … | <.0001 | <.0001 | <.0001 | … | .0002 |
| AIC | … | 2562 | … | … | … | … |
| Early proportion of time CD4 count <200 cells/µLd |
||||||
| 0.00 | 73 | 1.0 (ref) | 1.0 (ref) | … | 1.0 (ref) | 1.0 (ref) |
| >0.00 to 0.25 | 29 | 2.1 (1.4, 3.3) | 1.4 (0.7, 2.7) | … | 1.9 (1.2, 2.9) | 1.3 (0.7, 2.6) |
| >0.25 to 0.50 | 12 | 2.1 (1.1, 3.8) | 1.1 (0.5, 2.5) | … | 1.7 (0.9, 3.2) | 1.1 (0.5, 2.4) |
| >0.50 to 0.75 | 10 | 2.2 (1.2, 4.3) | 1.1 (0.4, 2.6) | … | 1.8 (0.9, 3.5) | 1.0 (0.4, 2.4) |
| >0.75 to <1.00 | 9 | 2.1 (1.1, 4.3) | 0.9 (0.4, 2.3) | … | 1.6 (0.8, 3.3) | 0.9 (0.3, 2.2) |
| 1.00 | 37 | 7.9 (5.3, 11.9) | 3.1 (1.5, 6.6) | … | 5.8 (3.7, 9.1) | 2.9 (1.4, 6.1) |
| Global P value | … | <.0001 | .0009 | … | <.0001 | .0015 |
| Per 20% of time CD4 <200 cells/µL |
… | 1.39 (1.29, 1.49) | 1.17 (1.05, 1.29) | … | 1.31 (1.20, 1.42) | 1.15 (1.04, 1.28) |
| Ptrend | … | <.0001 | .0030 | … | <.0001 | .0065 |
| AIC | … | 2572 | … | … | … | … |
| Early HIV RNA level average, copies/mLd |
||||||
| ≤500 | 47 | 1.0 (ref) | … | 1.0 (ref) | 1.0 (ref) | 1.0 (ref) |
| >500 to <10 000 | 41 | 1.2 (0.8, 1.9) | … | 0.9 (0.6, 1.3) | 1.1 (0.7, 1.6) | 0.9 (0.6, 1.3) |
| 10 000 to <100 000 | 61 | 2.4 (1.6, 3.5) | … | 1.2 (0.8, 1.8) | 1.6 (1.0, 2.4) | 1.2 (0.7, 1.8) |
| ≥100 000 | 21 | 5.3 (3.1, 8.9) | … | 1.9 (1.0, 3.4) | 2.5 (1.4, 4.5) | 1.7 (0.9, 3.2) |
| Global P value | … | <.0001 | … | .089 | .014 | .15 |
| Per log10 copies/mL | … | 1.58 (1.36, 1.84) | … | 1.17 (0.99, 1.39) | 1.28 (1.09, 1.51) | 1.14 (0.96, 1.35) |
| Ptrend | … | <.0001 | … | .064 | .0030 | .15 |
| AIC | … | 2607 | … | … | … | … |
| Combined models’ AIC | … | 2551 | 2561 | 2567 | 2551 |
Abbreviations: AIC, Akaike information criterion; CI, confidence interval; HR, hazard ratio; HIV, human immunodeficiency virus.
aEach measure was individually included in a cohort-stratified Cox model adjusted for sex, race/ethnicity, and baseline age and calendar period among persons with follow-up >3060 days (N = 34 625; number of anal cancer cases = 170).
bFrom cohort-stratified Cox model with combinations of key predictors as covariates and adjusted for sex, race/ethnicity, and baseline age and calendar period among persons with follow-up >3060 days.
cCD4 count nadir from 3060 to 180 days (approximately 8.5 years to approximately 6 months) in the past.
dProportion of time CD4 count <200 cells/µL and HIV RNA level average from 3060 to 1620 days (approximately 8.5 to approximately 4.5 years) in the past.
The median overall CD4 nadir at the end of follow-up was 137 cells/µL (interquartile range [IQR], 38–256) among anal cancer cases compared with 259 cells/µL (IQR, 134–401) among noncases. The median time from CD4 nadir to anal cancer diagnosis was 4.9 years (IQR, 3.0–7.4).
DISCUSSION
Among PLWH in the United States and Canada, the most robust independent predictors for anal cancer risk were overall CD4 nadir (ie, lowest CD4 from approximately 8.5 years to approximately 6 months in the past) and early cumulative CD4 (ie, proportion of time CD4 <200 cells/µL from approximately 8.5 to approximately 4.5 years in the past). HIV RNA measures did not further improve anal cancer risk prediction. Our results confirmed the association between HIV-related immunosuppression and anal cancer risk [8–10, 16–26]; suggested a key role for severe, prolonged immunosuppression in anal carcinogenesis; and favored a greater role of immunosuppression at earlier compared with late phases of anal carcinogenesis.
Similar to our study, a previous study found nadir and cumulative CD4 to be independent predictors of anal cancer risk [25]. Other studies have found nadir [8, 9, 16, 19, 23, 24, 26] or cumulative CD4 [18, 21] to be associated with anal cancer risk and found nadir [8, 23, 24, 26], cumulative [18], and past [23] CD4 to be better predictors than recent CD4. Furthermore, PLWH with a prior AIDS diagnosis, which captures a history of severe immunosuppression (ie, a low CD4 nadir), have an elevated anal cancer risk compared to PLWH without AIDS [1].
In our study, based on AICs and strengths of association, single past CD4 measures (ie, lagged by approximately 3.5 to approximately 8.5 years) and early cumulative and nadir CD4 measures (from approximately 8.5 to approximately 4.5 years in the past) were generally better predictors of anal cancer risk than more recent CD4 measures, with early proportion of time CD4 <200 cells/µL included in our final model. Furthermore, only early measures remained significant when both early and late CD4 measures of the same type were modeled together (Supplementary Table 1). Nevertheless, as overall CD4 nadir was highly correlated with both early and late CD4 nadir, it was difficult to definitively tease apart these 3 constructs. Moreover, besides capturing severity of immunosuppression, overall CD4 nadir likely captured cumulative immunosuppression, as it was highly correlated with cumulative CD4 measures.
We and others have found cumulative HIV RNA measures to be significantly associated with anal cancer risk [18, 24, 26, 27] and to be better predictors than recent HIV RNA [18, 27]. However, when we added early HIV RNA average to a model with our 2 key CD4 predictors, it lost statistical significance and the model fit did not improve. These results suggest that the HIV RNA effect was entirely mediated by its effect of lowering CD4 [34], with no independent HIV RNA effect. In contrast, the only other study to comprehensively evaluate different CD4 and HIV RNA measures determined both cumulative CD4 and cumulative HIV RNA to be independent key predictors [18].
Prolonged immunosuppression-driven loss of control of oncogenic HPV infection is considered the primary mechanism that drives anal carcinogenesis in PLWH [3–5]. Lower CD4 is associated with increased anal HPV infection prevalence and persistence, greater amounts of HPV DNA, more HPV types in anal specimens, higher incidence and prevalence of anal intraepithelial neoplasia, and faster progression to HSIL/AIN2-3 [3, 4]. However, HIV-induced immunosuppression may not be critical for HSIL/AIN2-3 progression to anal cancer [3–6], and HPV-associated genetic instability and other factors (eg, smoking) that lead to genetic changes in HSIL/AIN2-3 may have a greater influence on progression [3, 4]. Our results generally supported this model, with early CD4 measures being better predictors for anal cancer risk than late CD4 measures. Nevertheless, our identification of overall CD4 nadir (ie, from approximately 8.5 years to approximately 6 months in the past) as a key predictor suggests a role for severe, prolonged immunosuppression in both HSIL/AIN2-3 development and progression.
HIV-encoded proteins may also contribute to HSIL/AIN2-3 development by facilitating initial HPV infection [35] or by upregulating HPV oncogenes [5, 36]. Although our finding of no independent HIV RNA effect contradicts this hypothesis, HIV RNA may be an imperfect marker for viral protein effects.
ART’s role in preventing anal carcinogenesis is unclear [3–5]. If severe, prolonged immunosuppression in the presence of infection with oncogenic HPV types is a key risk factor in anal carcinogenesis, then ART would have little preventive effect if initiated late, once immunosuppression has already played its main role of promoting HSIL/AIN2-3 [4]. Presumably a high proportion of persons who initiated ART at the start of the ART era already had experienced severe, prolonged immunosuppression, likely explaining the lack of decline in anal cancer incidence at that time [2, 7–15]. The trend toward earlier ART initiation since then [37] might explain the apparent decreasing incidence since 2008 [1, 2].
Nevertheless, anal cancer risk remains markedly elevated in PLWH [1], so optimization of prevention and screening efforts are needed. Early HIV diagnosis and prompt ART initiation [38, 39] could prevent the severe, prolonged immunosuppression associated with increased anal cancer risk. Vaccination against oncogenic HPV infection, especially among young MSM, and smoking cessation could also reduce risk and merit incorporation into primary HIV care [39]. Furthermore, a firm evidence base needs to be developed for HSIL/AIN2-3 screening and management in PLWH [39, 40]. Anal cancer screening may be more effective if targeted to high-risk subgroups such as MSM, especially young MSM with AIDS [2]. Nadir and cumulative CD4, along with having a prior AIDS diagnosis (a proxy for low CD4 nadir), may represent useful markers for identifying patients at higher risk for anal cancer to target for screening and other prevention efforts.
Our study’s limitations included a lack of information on anal HPV infection (including HPV genotype), anal intercourse frequency (a risk factor for HPV infection), presence of anal warts (an indicator of HPV infection), and anal cancer screening. HIV risk group and smoking data were incomplete; nevertheless, adjusting for these variables in sensitivity analyses revealed that our final model was highly robust. Furthermore, our study included a low percentage of females and Hispanics and only US and Canadian populations, limiting generalizability. Finally, none of the examined CD4 or HIV RNA measures fully characterize HIV-associated immune dysfunction (before or after ART).
Our study’s strengths included large sample size and number of events, coverage across 19 years during the ART era among people with varied characteristics and from diverse North American locations, validated anal cancer diagnoses, and comprehensive evaluation of various CD4 and HIV RNA measures, capturing exposures that occurred as far back as approximately 8.5 years. Our approach of examining cumulative and nadir/peak measures during moving windows of fixed duration may be superior to examining these measures during total follow-up time, which varies by person.
In summary, we found CD4 nadir from approximately 8.5 years to approximately 6 months in the past and cumulative CD4 (ie, proportion of time CD4 <200 cells/µL) from approximately 8.5 to approximately 4.5 years in the past, in combination, to be the most robust independent predictors of anal cancer risk. Besides capturing severity of immunosuppression during the approximately 8-year window, CD4 nadir likely also captured cumulative immunosuppression. As evidence suggested that CD4 in the early part of the window had the greatest influence on risk, our results were consistent with severe, prolonged immunosuppression affecting the earlier stages in anal carcinogenesis. Initiating ART promptly upon HIV diagnosis, as currently recommended [38], and monitoring nadir and cumulative CD4 to help target prevention efforts may be important for anal cancer prevention in PLWH.
Supplementary Data
Supplementary materials are available at Clinical Infectious Diseases online. Consisting of data provided by the authors to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the authors, so questions or comments should be addressed to the corresponding author.
Notes
North American AIDS Cohort Collaboration on Research and Design (NA-ACCORD) Collaborating Cohorts and Representatives. AIDS Clinical Trials Group Longitudinal Linked Randomized Trials: Constance A. Benson and Ronald J. Bosch; AIDS Link to the IntraVenous Experience: Gregory D. Kirk; Fenway Health HIV Cohort: Kenneth H. Mayer and Chris Grasso; HAART Observational Medical Evaluation and Research: Robert S. Hogg, P. Richard Harrigan, Julio S. G. Montaner, Benita Yip, Julia Zhu, Kate Salters, and Karyn Gabler; HIV Outpatient Study: Kate Buchacz and Jun Li; HIV Research Network: Kelly A. Gebo and Richard D. Moore; Johns Hopkins HIV Clinical Cohort: Richard D. Moore; John T. Carey Special Immunology Unit Patient Care and Research Database, Case Western Reserve University: Benigno Rodriguez; Kaiser Permanente Mid-Atlantic States: Michael A. Horberg; Kaiser Permanente Northern California: Michael J. Silverberg; Longitudinal Study of Ocular Complications of AIDS: Jennifer E. Thorne; Multicenter Hemophilia Cohort Study–II: Charles Rabkin; Multicenter AIDS Cohort Study: Joseph B. Margolick, Lisa P. Jacobson, and Gypsyamber D’Souza; Montreal Chest Institute Immunodeficiency Service Cohort: Marina B. Klein; Ontario HIV Treatment Network Cohort Study: Abigail Kroch, Ann Burchell, Adrian Betts, and Joanne Lindsay; Retrovirus Research Center, Bayamon Puerto Rico: Robert F. Hunter-Mellado and Angel M. Mayor; Southern Alberta Clinic Cohort: M. John Gill; Study of the Consequences of the Protease Inhibitor Era: Steven G. Deeks and Jeffrey N. Martin; Study to Understand the Natural History of HIV/AIDS in the Era of Effective Therapy: Jun Li and John T. Brooks; University of Alabama at Birmingham 1917 Clinic Cohort: Michael S. Saag, Michael J. Mugavero, and James Willig; University of California at San Diego: William C. Mathews; University of North Carolina at Chapel Hill HIV Clinic Cohort: Joseph J. Eron and Sonia Napravnik; University of Washington HIV Cohort: Mari M. Kitahata, Heidi M. Crane, and Daniel R. Drozd; Vanderbilt Comprehensive Care Clinic HIV Cohort: Timothy R. Sterling, David Haas, Peter Rebeiro, and Megan Turner; Veterans Aging Cohort Study: Amy C. Justice, Robert Dubrow, and David Fiellin; and Women’s Interagency HIV Study: Stephen J. Gange and Kathryn Anastos.
NA-ACCORD Study Administration. Executive Committee: Richard D. Moore, Michael S. Saag, Stephen J. Gange, Mari M. Kitahata, Keri N. Althoff, Michael A. Horberg, Marina B. Klein, Rosemary G. McKaig, and Aimee M. Freeman; Administrative Core: Richard D. Moore and Aimee M. Freeman; Data Management Core: Mari M. Kitahata, Stephen E. Van Rompaey, Heidi M. Crane, Daniel R. Drozd, Liz Morton, Justin McReynolds, and William B. Lober; and Epidemiology and Biostatistics Core: Stephen J. Gange, Keri N. Althoff, Jennifer S. Lee, Bin You, Brenna Hogan, Jinbing Zhang, Jerry Jing, Elizabeth Humes, and Sally Coburn.
Acknowledgments. Some of the data were collected by cancer registries that participate in the National Program of Cancer Registries of the Centers for Disease Control and Prevention (CDC).
Disclaimer. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health (NIH) or the CDC.
Financial support. This work was supported by the NIH (grant numbers U01AI069918, F31AI124794, F31DA037788, G12MD007583, K01AI093197, K01AI131895, K23EY013707, K24AI065298, K24AI118591, K24DA000432, KL2TR000421, M01RR000052, N01CP001004, N02CP055504, N02CP91027, P30AI027757, P30AI027763, P30AI027767, P30AI036219, P30AI050410, P30AI094189, P30AI110527, P30MH62246, R01AA016893, R01CA165937, R01DA011602, R01DA012568, R01AG053100, R24AI067039, U01AA013566, U01AA020790, U01AI031834, U01AI034989, U01AI034993, U01AI034994, U01AI035004, U01AI035039, U01AI035040, U01AI035041, U01AI035042, U01AI037613, U01AI037984, U01AI038855, U01AI038858, U01AI042590, U01AI068634, U01AI068636, U01AI069432, U01AI069434, U01AI103390, U01AI103397, U01AI103401, U01AI103408, U01DA03629, U01DA036935, U01HD032632, U10EY008057, U10EY008052, U10EY008067, U24AA020794, U54MD007587, UL1RR024131, UL1TR000004, UL1TR000083, UL1TR000454, UM1AI035043, Z01CP010214, and Z01CP010176); the CDC (contracts CDC-200-2006-18797 and CDC-200-2015-63931); the Agency for Healthcare Research and Quality (US; contract 90047713); the Health Resources and Services Administration (US; contract 90051652); the Canadian Institutes of Health Research (grant numbers CBR-86906, CBR-94036, HCP-97105, and TGF-96118); the Ontario Ministry of Health and Long Term Care; and the Government of Alberta. Additional support was provided by the National Cancer Institute, National Institute for Mental Health, and National Institute on Drug Abuse.
Potential conflicts of interest. K. N. A. serves on the scientific advisory board of TrioHealth, Inc. M. J. G. has served as an ad hoc member of Canadian HIV Advisory Boards of Merck, Gilead, and ViiV. A. R. has provided clinical research support to Merck Frosst, ViiV Healthcare, Gilead Sciences, and Janssen. R. D. M. has previously consulted for Medscape. M. J. S. has received grants from Gilead and, formerly, Merck. All other authors report no potential conflicts. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.
Contributor Information
North American AIDS Cohort Collaboration on Research and Design of the International Epidemiologic Databases to Evaluate AIDS:
Constance A Benson, Ronald J Bosch, Gregory D Kirk, Kenneth H Mayer, Chris Grasso, Robert S Hogg, P Richard Harrigan, Julio S G Montaner, Benita Yip, Julia Zhu, Kate Salters, Karyn Gabler, Kate Buchacz, Jun Li, Kelly A Gebo, Richard D Moore, Richard D Moore, Benigno Rodriguez, Michael A Horberg, Michael J Silverberg, Jennifer E Thorne, Charles Rabkin, Joseph B Margolick, Lisa P Jacobson, Gypsyamber D’Souza, Marina B Klein, Abigail Kroch, Ann Burchell, Adrian Betts, Joanne Lindsay, Robert F Hunter-Mellado, Angel M Mayor, M John Gill, Steven G Deeks, Jeffrey N Martin, Jun Li, John T Brooks, Michael S Saag, Michael J Mugavero, James Willig, William C Mathews, Joseph J Eron, Sonia Napravnik, Mari M Kitahata, Heidi M Crane, Daniel R Drozd, Timothy R Sterling, David Haas, Peter Rebeiro, Megan Turner, Amy C Justice, Robert Dubrow, David Fiellin, Stephen J Gange, Kathryn Anastos, Richard D Moore, Michael S Saag, Stephen J Gange, Mari M Kitahata, Keri N Althoff, Michael A Horberg, Marina B Klein, Rosemary G McKaig, Aimee M Freeman, Richard D Moore, Aimee M Freeman, Mari M Kitahata, Stephen E Van Rompaey, Heidi M Crane, Daniel R Drozd, Liz Morton, Justin McReynolds, William B Lober, Stephen J Gange, Keri N Althoff, Jennifer S Lee, Bin You, Brenna Hogan, Jinbing Zhang, Jerry Jing, Elizabeth Humes, and Sally Coburn
References
- 1. Hernández-Ramírez RU, Shiels MS, Dubrow R, Engels EA. Cancer risk in HIV-infected people in the USA from 1996 to 2012: a population-based, registry-linkage study. Lancet HIV 2017; 4:e495–504. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Colón-López V, Shiels MS, Machin M, et al. . Anal cancer risk among people with HIV infection in the United States. J Clin Oncol 2018; 36:68–75. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Palefsky JM, Holly EA. Chapter 6: immunosuppression and co-infection with HIV. J Natl Cancer Inst Monogr 2003;2003:41–6. [DOI] [PubMed] [Google Scholar]
- 4. Palefsky J. Biology of HPV in HIV infection. Adv Dent Res 2006; 19:99–105. [DOI] [PubMed] [Google Scholar]
- 5. Brickman C, Palefsky JM. Human papillomavirus in the HIV-infected host: epidemiology and pathogenesis in the antiretroviral era. Curr HIV/AIDS Rep 2015; 12:6–15. [DOI] [PubMed] [Google Scholar]
- 6. Mathews WC, Agmas W, Cachay ER, Cosman BC, Jackson C. Natural history of anal dysplasia in an HIV-infected clinical care cohort: estimates using multi-state Markov modeling. PLoS One 2014; 9:e104116. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Crum-Cianflone N, Hullsiek KH, Marconi V, et al. . Trends in the incidence of cancers among HIV-infected persons and the impact of antiretroviral therapy: a 20-year cohort study. AIDS 2009; 23:41–50. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Crum-Cianflone NF, Hullsiek KH, Marconi VC, et al. ; Infectious Disease Clinical Research Program HIV Working Group Anal cancers among HIV-infected persons: HAART is not slowing rising incidence. AIDS 2010; 24:535–43. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Piketty C, Selinger-Leneman H, Bouvier AM, et al. . Incidence of HIV-related anal cancer remains increased despite long-term combined antiretroviral treatment: results from the French hospital database on HIV. J Clin Oncol 2012; 30:4360–6. [DOI] [PubMed] [Google Scholar]
- 10. Silverberg MJ, Lau B, Justice AC, et al. ; North American AIDS Cohort Collaboration on Research and Design (NA-ACCORD) of IeDEA Risk of anal cancer in HIV-infected and HIV-uninfected individuals in North America. Clin Infect Dis 2012; 54:1026–34. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Worm SW, Bower M, Reiss P, et al. ; D:A:D Study Group Non-AIDS defining cancers in the D:A:D Study— time trends and predictors of survival: a cohort study. BMC Infect Dis 2013; 13:471. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Robbins HA, Shiels MS, Pfeiffer RM, Engels EA. Epidemiologic contributions to recent cancer trends among HIV-infected people in the United States. AIDS 2014; 28:881–90. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Hleyhel M; Writing Committee of the Cancer Risk Group of the French Hospital Database on HIV (FHDH-ANRS CO4) Risk of non-AIDS-defining cancers among HIV-1-infected individuals in France between 1997 and 2009: results from a French cohort. AIDS 2014; 28:2109–18. [DOI] [PubMed] [Google Scholar]
- 14. Park LS, Tate JP, Sigel K, et al. . Time trends in cancer incidence in persons living with HIV/AIDS in the antiretroviral therapy era: 1997–2012. AIDS 2016; 30:1795–806. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Silverberg MJ, Lau B, Achenbach CJ, et al. ; North American AIDS Cohort Collaboration on Research and Design of the International Epidemiologic Databases to Evaluate AIDS Cumulative incidence of cancer among persons with HIV in North America: a cohort study. Ann Intern Med 2015; 163:507–18. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Patel P, Hanson DL, Sullivan PS, et al. . Incidence of types of cancer among HIV-infected persons compared with the general population in the United States, 1992–2003. Ann Intern Med 2008; 148:728–36. [DOI] [PubMed] [Google Scholar]
- 17. Chaturvedi AK, Madeleine MM, Biggar RJ, Engels EA. Risk of human papillomavirus-associated cancers among persons with AIDS. J Natl Cancer Inst 2009; 101:1120–30. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Guiguet M, Boué F, Cadranel J, Lang JM, Rosenthal E, Costagliola D; Clinical Epidemiology Group of the FHDH-ANRS CO4 cohort Effect of immunodeficiency, HIV viral load, and antiretroviral therapy on the risk of individual malignancies (FHDH-ANRS CO4): a prospective cohort study. Lancet Oncol 2009; 10:1152–9. [DOI] [PubMed] [Google Scholar]
- 19. Powles T, Robinson D, Stebbing J, et al. . Highly active antiretroviral therapy and the incidence of non-AIDS-defining cancers in people with HIV infection. J Clin Oncol 2009; 27:884–90. [DOI] [PubMed] [Google Scholar]
- 20. Reekie J, Kosa C, Engsig F, et al. ; EuroSIDA Study Group Relationship between current level of immunodeficiency and non-acquired immunodeficiency syndrome-defining malignancies. Cancer 2010; 116:5306–15. [DOI] [PubMed] [Google Scholar]
- 21. Kesselring A, Gras L, Smit C, et al. . Immunodeficiency as a risk factor for non-AIDS-defining malignancies in HIV-1-infected patients receiving combination antiretroviral therapy. Clin Infect Dis 2011; 52:1458–65. [DOI] [PubMed] [Google Scholar]
- 22. Silverberg MJ, Chao C, Leyden WA, et al. . HIV infection, immunodeficiency, viral replication, and the risk of cancer. Cancer Epidemiol Biomarkers Prev 2011; 20:2551–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Bertisch B, Franceschi S, Lise M, et al. ; Swiss HIV Cohort Study Investigators Risk factors for anal cancer in persons infected with HIV: a nested case-control study in the Swiss HIV Cohort Study. Am J Epidemiol 2013; 178:877–84. [DOI] [PubMed] [Google Scholar]
- 24. Chiao EY, Hartman CM, El-Serag HB, Giordano TP. The impact of HIV viral control on the incidence of HIV-associated anal cancer. J Acquir Immune Defic Syndr 2013; 63:631–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Duncan KC, Chan KJ, Chiu CG, et al. . HAART slows progression to anal cancer in HIV-infected MSM. AIDS 2015; 29:305–11. [DOI] [PubMed] [Google Scholar]
- 26. Mbang PA, Kowalkowski MA, Amirian ES, et al. . Association between time on protease inhibitors and the incidence of squamous cell carcinoma of the anus among U.S. male veterans. PLoS One 2015; 10:e0142966. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Kowalkowski MA, Day RS, Du XL, Chan W, Chiao EY. Cumulative HIV viremia and non-AIDS-defining malignancies among a sample of HIV-infected male veterans. J Acquir Immune Defic Syndr 2014; 67:204–11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Crum-Cianflone NF, Wang X, Ganesan A, et al. . Short communication: HIV RNA levels predict AIDS-defining and non-AIDS-defining cancers after antiretroviral therapy initiation among HIV-infected adults. AIDS Res Hum Retroviruses 2015; 31:514–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Gange SJ, Kitahata MM, Saag MS, et al. . Cohort profile: the North American AIDS Cohort Collaboration on Research and Design (NA-ACCORD). Int J Epidemiol 2007; 36:294–301. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Dubrow R, Qin L, Lin H, et al. ; North American AIDS Cohort Collaboration on Research and Design of the International Epidemiologic Databases to Evaluate AIDS Association of CD4+ T-cell count, HIV-1 RNA viral load, and antiretroviral therapy with Kaposi sarcoma risk among HIV-infected persons in the United States and Canada. J Acquir Immune Defic Syndr 2017; 75:382–90. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Andersen PK, Gill RD. Cox’s regression model for counting processes: a large sample study. Ann Statist 1982; 10:21. [Google Scholar]
- 32. Burnham KP, Anderson DR. Multimodel inference: understanding AIC and BIC in model selection. Sociol Methods Res 2004; 33:261–304. [Google Scholar]
- 33. Richardson ET, Grant PM, Zolopa AR. Evolution of HIV treatment guidelines in high- and low-income countries: converging recommendations. Antiviral Res 2014; 103:88–93. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Mellors JW, Muñoz A, Giorgi JV, et al. . Plasma viral load and CD4+ lymphocytes as prognostic markers of HIV-1 infection. Ann Intern Med 1997; 126:946–54. [DOI] [PubMed] [Google Scholar]
- 35. Tugizov SM, Herrera R, Chin-Hong P, et al. . HIV-associated disruption of mucosal epithelium facilitates paracellular penetration by human papillomavirus. Virology 2013; 446:378–88. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. Kim RH, Yochim JM, Kang MK, Shin KH, Christensen R, Park NH. HIV-1 Tat enhances replicative potential of human oral keratinocytes harboring HPV-16 genome. Int J Oncol 2008; 33:777–82. [PubMed] [Google Scholar]
- 37. Hanna DB, Buchacz K, Gebo KA, et al. . Trends and disparities in antiretroviral therapy initiation and virologic suppression among newly treatment-eligible HIV-infected individuals in North America, 2001–2009. Clin Infect Dis 2013; 56:1174–82. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38. Panel on Antiretroviral Guidelines for Adults and Adolescents. Guidelines for the use of antiretroviral agents in adults and adolescents living with HIV Available at: https://aidsinfo.nih.gov/contentfiles/lvguidelines/adultandadolescentgl.pdf. Accessed 2 March 2018.
- 39. Goncalves PH, Montezuma-Rusca JM, Yarchoan R, Uldrick TS. Cancer prevention in HIV-infected populations. Semin Oncol 2016; 43:173–88. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40. Hillman RJ, Cuming T, Darragh T, et al. . 2016 IANS international guidelines for practice standards in the detection of anal cancer precursors. J Low Genit Tract Dis 2016; 20:283–91. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.