Abstract
Background:
People living with HIV (PWH) have improved life expectancy due to effective HIV therapy but still experience immune impairment (e.g., altered CD4/CD8 T-cells). We hypothesized that tumors diagnosed in PWH would have distinct molecular features.
Methods:
We utilized whole exome sequencing of paired tumor and germline DNA and RNA sequencing of tumors from 229 patients with cancer enrolled into ORIEN to classify: total tumor mutation burden (TMB), major histocompatibility complex (MHC) class I neoantigen count, and MHC class II neoantigen count.
Results:
Specimens from 229 patients with cancer (110 PWH and 119 without HIV) were evaluated. Average TMB for tumors diagnosed in PWH was 249, compared to 172 for those without HIV. After adjustment for age, sex, race, smoking, and cancer site, the association between HIV and TMB remained statistically significant (OR=1.72; 95% CI: 1.26-2.43). We further observed an association between HIV and higher class I neoantigen count (OR=1.62; 95% CI: 1.10-2.41) but no association with putative class II neoantigens. When considering cancer sites separately in unadjusted analyses, average TMB was elevated in PWH for thyroid (p<0.01) and bladder cancers (p=0.03), and sarcoma (p=0.04). Similarly, putative class I neoantigen count was elevated in PWH for head & neck (p<0.01) and thyroid (p=0.01) cancers, and sarcoma (p=0.04).
Conclusion:
Our findings indicate that tumors diagnosed in PWH harbor a higher TMB and a higher number of putative class I neoantigens.
Impact.
A higher TMB in PWH may portend a more favorable response to cancer treatment modalities such as immune checkpoint inhibitors.
Keywords: HIV and cancer, tumor mutation burden, neoantigens, ORIEN population
INTRODUCTION.
People living with HIV (PWH) now have life expectancies comparable to the general population due to effective antiretroviral therapy (ART),(1, 2) but this population still experiences long-term immune dysregulation.(3–7) Recent work from a large consortium of North American HIV cohorts (NA-ACCORD) described consistently altered CD8 T-cells distributions, even during years when the majority of PWH had access to effective ART to control HIV.(8) The immune system is crucial for controlling cancer development.(9, 10) It is therefore biologically plausible that immune impairment among PWH, including those on ART, impacts various steps in the tumorigenesis process. This could extend to the process of tumor editing, whereby the immune system eliminates cells recognized as mutated or exhibiting non-self-antigens produced by malignant tissue. We therefore hypothesized that tumors developing in PWH would present with distinct molecular features compared to tumors in persons without HIV, specifically an altered mutational profile.
Recent data from the AIDS Cancer Specimen Resource (ACSR) examined tumor mutational burden (TMB) in tissues from twenty-one women with HIV diagnosed with either breast (N=13) or lung (N=8) cancer.(11) Seventy-five percent and 61% of the patients with breast and lung cancer, respectively, were observed to have an elevated TMB, defined as at least 10 mutations per mega-base of sequenced tumor DNA. However, this report did not include a direct comparison to the TMB observed for patients with cancer without HIV.
The goal of our report was to evaluate a well-controlled comparison of tumor mutational features in solid organ tumor patients with and without HIV. In addition, given the dual facts that TMB may have an impact on response to cancer therapy, and that PWH have documented poor clinical outcomes compared to cancer patients without HIV diagnosed with the same cancer site and stage, we also assessed overall survival stratified by both HIV status and TMB count.
MATERIALS AND METHODS.
This study leveraged existing clinical and genomic data obtained by Aster Insights in partnership with ORIEN, an alliance of18 cancer centers in North America established in 2014 to create a centralized repository of harmonized clinical and molecular data to support translational research. All ORIEN collaborative network member institutions have opened the Total Cancer Care (TCC) protocol that consents patients to specimen collection and allows for future testing to support cancer research, as well as ongoing access to clinical outcome data. A subset of the TCC-consented patient population is enrolled into the Avatar program, which includes research use only (RUO) grade whole-exome tumor and germline sequencing, RNA sequencing, and collection of longitudinal clinical data. Preference for the selection of Avatar patients has historically been given to high-risk cases, but this is not a requirement for enrollment. Aster Insights harmonizes all abstracted clinical data elements and molecular sequencing files into a standardized, structured format to enable aggregation of de-identified data for sharing across the network. This research was conducted in accordance with the Belmont Report guidelines, and each ORIEN institution’s ethical review board has approved the TCC protocol. All data utilized were HIPAA-compliant and de-identified, and the study was exempted from full review by the Advarra institutional review board at Moffitt Cancer Center.
Study Eligibility.
This study included clinical and genomic data from participants enrolled into the ORIEN cohort between 2005 and 2022 and diagnosed with a solid organ tumor (i.e., not a hematological malignancy). To be classified as a PWH, patients with cancer had to have evidence in the ORIEN dataset of an HIV/AIDS diagnosis; this was achieved through diagnostic coding (e.g., ICD10 codes) provided by each ORIEN institution. For each identified PWH, one cancer patient without HIV was individually matched on sex, age, race, cancer site, cancer stage, and smoking status.
Tumor Genomic Metrics.
The genomic data included whole exome sequencing (WES) generated from paired tumor and germline DNA, as well as RNA sequencing from tumors (RNASeq). Three tumor-specific metrics were evaluated: total tumor mutation burden (TMB), putative, or predicted, major histocompatibility complex (MHC) class I neoantigen count, and putative MHC class II neoantigen count.
TMB was defined as the number of non-synonymous mutations observed to be specific to DNA generated from the tumor rather than germline (i.e., non-tumor) DNA. WES reads were aligned to the reference human genome with the Burrows-Wheeler Aligner; duplicate identification, insertion/deletion realignment, quality score recalibration, and variant identification were performed with PICARD (http://picard.sourceforge.net/) and the Genome Analysis ToolKit (GATK). Tumor-specific mutations were then identified with Strelka and MuTect and annotated to determine genic context (i.e., non-synonymous, missense, splicing) using ANNOVAR. TMB was calculated by counting the number of protein-altering (i.e., non-synonymous) mutations. WES quality control included Picard and in-house scripts to examine alignment fraction, sequence duplication rates, depth of coverage across target regions, and inherited variant concordance between matched tumor versus germline sample pairs in each participant.
Putative class I and II neoantigens were defined as the subset of non-synonymous, tumor-specific mutations predicted to be able to elicit a tumor-specific immune response. To guide this determination, RNASeq data were used to classify (1) the degree of expression of identified tumor-specific mutations and (2) the human leukocyte antigen (HLA) type of the patient using ArcasHLA(12) to understand if the expressed mutation could elicit an immune response. Specifically, RNA-seq reads were aligned to the human reference genome in a splice-aware fashion using STAR, allowing for accurate alignments of sequences across introns. Sequences were quantified to RefSeq genes using the HTseq package. RNAseq quality control includes in-house scripts and RSeqC to examine read count metrics, alignment fraction, chromosomal alignment counts, expression distribution measures, as well as principal components analysis and hierarchical clustering. Putative neoantigens were then further identified by extracting altered peptides with ANNOVAR predicting MHC binding against patient-specific HLA type using NetMHCpan-4.1 and NetMHCIIpan-4.1. Bioinformatics analyses are implemented in standardized WDL/Cromwell workflows, ensuring reproducibility.
This set of putative neoantigens was further filtered to a conservative list of variant peptides with the highest probability to elicit tumor-specific immunity using criteria determined by the TESLA consortium, including: variant IC50 <34nM, RNA gene expression transcript per million (TPM)>33, and variant peptide affinity/reference peptide affinity (agretopicity) <0.1.(13) Overlapping variant peptides per somatic mutation were possible.
Statistical Analyses.
Comparisons were limited to tumor sites with at least 5 cases recorded among PWH in the ORIEN dataset. Output for TMB and putative neoantigen counts was log-transformed, and a Wilcoxon rank sum test was used to compare average tumor-specific metrics between PWH and cancer and their matched, HIV-uninfected counterparts. This was done overall and repeated for each cancer site separately. The proportion of total tumor-specific mutations classified as putative MHC class I neoantigens was also calculated overall and by tumor site and compared according to HIV status using a Wilcoxon rank sum test. We then examined the association between tumor-specific metrics and cancer patient HIV status using logistic regression models adjusted for age at specimen collection, sex [except for breast and prostate cancer]), race, smoking status, and cancer site.
Prior to conducting survival analyses, we classified cancers as ‘high’ versus ‘low’ TMB by separating them according to whether the average TMB for that respective cancer fell above or below the overall median TMB (‘high’ TMB cancers: lung, bladder, colorectal, gastric, head and neck, liver, and other GU; ‘low’ TMB cancers: prostate, kidney, breast, sarcoma, pancreatic, thyroid, other GI). For comparisons of survival following a cancer diagnosis between patients with versus without HIV, follow-up time began at the date of cancer diagnosis and continued until death or end of follow-up, with the current dataset inclusive of ORIEN follow-up through 2022. Unadjusted comparisons were conducted using a log-rank test. Adjusted survival comparisons were conducted using a Cox proportional hazards model that included age, sex, race, smoking status, and cancer site.
Data Availability Statement.
The data analyzed in this study were generated by Avatar. Restrictions apply to the availability of these data, which were used as part of the ORIEN network of institutions (https://www.oriencancer.org/contact). Data are available from the authors upon reasonable request with the permission of Avatar, collaboration with an ORIEN institution, and after appropriate approvals are in place.
RESULTS.
A total of 140 eligible PWH and cancer were identified in the ORIEN dataset (2005-2022), along with 140 matched patients with cancer without HIV. After further review, the HIV status of one PWH could no longer be verified, leaving 279 participants with clinical data. Forty-nine patients with cancer (30 PWH and 21 without HIV) did not pass pre-specified WES QC criteria, leaving 229 patients with cancer (110 PWH and 119 without HIV) in the analysis reported here. Matching on sex, age, race, cancer stage, and smoking status was efficient (Table 1). This patient population was majority White (90%); more than half (57%) indicated either current or former smoking history; and the average age was 58 years. The three most prevalent cancers were breast (N=40), head & neck (N=31), and lung (N=28).
Table 1.
Study participant characteristics
| Total (N=229) | PWH (N=110) | No HIV (N=119) | |
|---|---|---|---|
| Age (Mean [SD]) | 57.8 (11.9) | 57.4 (11.9) | 58.2 (11.9) |
| Sex | |||
| Female | 119 (52.0%) | 56 (50.9%) | 63 (52.9%) |
| Male | 110 (48.0%) | 54 (49.1%) | 56 (47.1%) |
| Race | |||
| White | 205 (89.5%) | 96 (87.3%) | 109 (91.6%) |
| Black | 15 (6.6%) | 11 (10.0%) | 4 (3.4%) |
| Other | 9 (3.9%) | 3 (2.7%) | 6 (5.0%) |
| Smoking status | |||
| Current | 43 (18.8%) | 21 (19.1%) | 22 (18.5%) |
| Former | 88 (38.4%) | 43 (39.1%) | 45 (37.8%) |
| Never | 95 (41.5%) | 45 (40.9%) | 50 (42.0%) |
| Missing | 3 (1.3%) | 1 (0.9%) | 2 (1.7%) |
| Cancer stage | |||
| Stage 1 | 49 (21.4%) | 24 (21.8%) | 25 (21.0%) |
| Stage 2 | 65 (28.4%) | 29 (26.4%) | 36 (30.3%) |
| Stage 3 | 21 (9.2%) | 11 (10.0%) | 10 (8.4%) |
| Stage 4 | 38 (16.6%) | 19 (17.3%) | 19 (16.0%) |
| Missing | 56 (24.5%) | 27 (24.5%) | 29 (24.4%) |
| Cancer type | |||
| Breast | 40 (17.5%) | 18 (16.4%) | 22 (18.5%) |
| Head and Neck | 31 (13.5%) | 15 (13.6%) | 16 (13.4%) |
| Lung | 28 (12.2%) | 14 (12.7%) | 14 (11.8%) |
| Thyroid | 19 (8.3%) | 9 (8.2%) | 10 (8.4%) |
| Colorectal | 22 (9.6%) | 12 (10.9%) | 10 (8.4%) |
| Kidney | 21 (9.2%) | 11 (10.0%) | 10 (8.4%) |
| Sarcoma | 16 (7.0%) | 7 (6.4%) | 9 (7.6%) |
| Prostate | 14 (6.1%) | 6 (5.5%) | 8 (6.7%) |
| Other GI a | 9 (3.9%) | 3 (2.7%) | 6 (5.0%) |
| Bladder | 8 (3.5%) | 4 (3.6%) | 4 (3.4%) |
| Pancreatic | 10 (4.4%) | 5 (4.5%) | 5 (4.2%) |
| Liver | 4 (1.7%) | 2 (1.8%) | 2 (1.7%) |
| Gastric | 4 (1.7%) | 2 (1.8%) | 2 (1.7%) |
| Other GU b | 3 (1.3%) | 2 (1.8%) | 1 (0.8%) |
Abbreviations: PWH (People living with HIV); SD (standard deviation); GI (gastrointestinal); GU (genitourinary)
Other GI: Jejunum (1), Ileum (3), Small Intestine (2), Anus (3);
Other GU: Penile (1), Urethra (1), not specified (1)
Tumor Mutation Burden (TMB).
The average TMB for PWH and cancer was 249 mutations, compared to 172 for the matched patients with cancer without HIV (Wilcoxon p<0.01; Figure 1). The overall association between HIV and TMB remained statistically significant after adjustment for age, sex, race, smoking status, and cancer site (OR=1.72; 95% CI: 1.26-2.43). When considering cancer sites separately, average TMB was elevated in PWH for thyroid cancer (HIV: 218; no HIV: 30; p<0.01), bladder cancer (HIV: 480; no HIV: 155; p=0.03), and sarcoma (HIV: 143; no HIV: 48; p=0.04). Marked differences in average TMB by HIV status were also observed for gastric, pancreatic, prostate, and head & neck cancers, although differences did not reach statistical significance (Figure 2a–b; Supplemental Table 1).
Figure 1. Average tumor mutation burden and class I neoantigen count in patients with cancer with (PWH) and patients with cancer without HIV.
Each tumor-specific genomic metric is represented by three aspects in Figure 1: (1) each dot represents the value for a specific patient; (2) histograms represent the overall distribution by group; and (3) box plots represent the mean and 95% confidence interval.
Figure 2. Cancer site-specific tumor genomic metrics by HIV status.
(A) Number of cases analyzed to generate (B) average tumor mutation burden in patients with cancer with (PWH) and without HIV. (C) Number of cases analyzed to generate (D) average class I neoantigen count in patients with cancer with (PWH) versus without HIV. Results are presented for cancer sites with statistically significant differences by HIV status (unadjusted, Wilcoxon p-value<0.05). Box plots represent the mean with 95% confidence intervals.
Putative Neoantigen Count.
The sample size for assessing the association between HIV and putative class I and II neoantigen count was more limited given the need for patients to have both WES and RNAseq data available (74 PWH and 113 without HIV). The average count of conservatively filtered, putative class I neoantigens (i.e., neoantigens most likely to elicit an MHC class I tumor-specific immune response) for PWH and cancer was 3, compared to 2 for matched patients with cancer without HIV (Wilcoxon p<0.01); Figure 1). This association remained statistically significant after adjustment for age, sex, race, smoking status, and cancer site (OR=1.62; 95% CI: 1.10-2.41). When considering cancer sites separately, the putative class I neoantigen count was elevated in PWH for head & neck (HIV: 3; no HIV: 1; p<0.01) and thyroid (HIV: 1; no HIV: <1; p=0.01) cancers, and sarcoma (HIV: 2; no HIV: 0; p=0.04; Figure 2c–d). Of note, the average total class I neoantigen count, defined using the single criterion of having an IC50 <500nM rather than the full set of conservative filters, was 464 in PWH versus 436 in patients with cancer without HIV (Wilcoxon p-value=0.06).
Overall, we observed that 1.3% of all tumor-specific mutations were classified as conservatively filtered, putative class I neoantigens in PWH and cancer. This number was 0.9% in patients with cancer without HIV, and this difference was statistically significant (p=0.02). The proportion of the TMB resulting in putative class I neoantigens was observed to be elevated for PWH diagnosed with thyroid (HIV: 1.1%; no HIV: 0%; p=0.01) and head and neck (HIV: 1.3%; no HIV: 0.5%; p=0.02) cancers (Supplemental Figure 1).
We observed no difference in putative class II neoantigen count by cancer patient HIV status (Wilcoxon p=0.27), although the association after adjustment for age, sex, race, smoking status, and cancer site was elevated but not statistically significant (OR=1.28; 95% CI: 0.97-1.72; p=0.08). Prior to conservative filtering, the average total class II neoantigen count exceeded 4,000 in each study group (PWH: 4388; no HIV: 4690; p-value = 0.29).
Patient Survival.
Overall, the median survival in patients with cancer without HIV was 13.8 years, compared to 9.9 years for PWH (log-rank p-value=0.25). A survival analysis adjusted for age, sex, race, smoking status, and cancer site suggested a non-significant increase in mortality for PWH (HR=1.62; 95% CI: 0.90-2.91; p-value=0.11). The corresponding HRs for high-TMB versus low-TMB cancer sites were (HR=1.49; 95% CI 0.70-3.20; p-value=0.30) and (HR=1.28; 95% CI 0.44-3.72; p-value=0.65), respectively (Figure 3).
Figure 3. Overall survival in PWH and cancer, by tumor mutation burden.
Overall survival in PWH diagnosed with cancers classified as either (A) ‘high’ tumor mutation burden (TMB) or (B) ‘low’ TMB. Dashed lines represent the median survival in each group. Univariate comparisons (p-values) by HIV status were conducted using the log-rank test.
We computed 5-year survival rates by HIV for cancer sites with >10 patients with cancer in each study group defined by HIV status (breast, lung, and head & neck cancers). Five-year survival did not differ by HIV status for breast cancer (HIV: 89%; no HIV: 95%; log-rank p-value=0.95) or head & neck cancer (HIV: 60%; no HIV: 62%; p-value=0.80). Although limited by sample size and not statistically significant, 5-year survival was lower for lung patients with cancer with HIV (79%) compared to those without HIV (100%; p-value=0.07).
DISCUSSION.
We performed the first direct comparison of tumor mutation burden (TMB) between patients with cancer with versus without HIV by utilizing the multi-site Oncology Research Information Exchange Network (ORIEN) database. We observed that tumors diagnosed in PWH had a statistically significantly elevated number of total mutations and higher count of predicted class I neoantigens compared to tumors from patients without HIV.
Our findings were consistent with a study from the ACSR that examined TMB in twenty-one women with HIV diagnosed with either breast (N=13) or lung (N=8) cancer.(11) That study reported mutation rates of >10 mutations per mega-base in the majority of women with HIV. Our study examined 110 PWH and cancer, including 18 women with breast cancer and 14 PWH with lung cancer. We also designed our study to include a comparison set of patients with cancer without HIV matched to PWH on key patient features, including age and race, cancer site and stage, and smoking history. Perhaps the most crucial of these matching features was cancer site, which helped control for expected TMB differences across different cancer types. Even after adjustment for age, sex, race, smoking status, and cancer site, we observed a statistically significant association (OR=1.72) between HIV and TMB. These findings are compatible with our hypothesis that immune dysregulation in PWH results in tumors that are less immune edited (i.e., tumors with more mutations).
Implications of HIV-associated tumor mutation differences for response to cancer therapy warrant further examination. Specifically, a higher TMB in PWH may portend differences in the response to cancer treatment modalities such as immune checkpoint inhibitor therapy, which has been demonstrated to be safe to administer in the setting of HIV.(14–16) The more frequent presence of class I neoantigens in PWH also raises the future possibility of personalized treatment approaches to engage the immune system to combat malignancy. This is particularly crucial given the consistent cancer survival deficits that we and others have reported for PWH and cancer.(17–19)
The cancer types included in our study represented those with available data from the ORIEN resource and were primarily comprised of age-related cancers not linked to infections (e.g., breast, lung, thyroid). Many cancers caused by oncogenic infections, such as anal cancer linked to human papillomavirus, were not represented. Virus-associated cancers not only occur more frequently among PWH but also harbor fewer mutations in general than cancers etiologically linked to smoking or environmental exposures. Future research into the tumor mutational profile specific to virus-associated cancers that occur more often in PWH represents an important avenue for future work.
It should be noted that whereas the average TMB was elevated in PWH, we observed a range of TMB counts within each study group, as illustrated in Figure 1. For example, PWH with low TMB existed, and certain patients with cancer without HIV were observed to have a high TMB. A preserved range in the distribution across study group was also observed for putative class I neoantigen counts, although the average count per patient was low after application of conservative filters (PWH: 3; no HIV: 2). Of note, when comparing total, unfiltered, class I neoantigen counts, we observed an average of 464 in PWH versus 436 in patients with cancer without HIV. This emphasizes the degree of careful filtering applied, which decreased average putative class I neoantigen counts from >400 to <5 in both study groups.
Strengths of our study include novelty of the scientific question, a 5-fold increase in sample size compared to existing TMB data among PWH, and careful selection of a comparison group of matched patients with cancer without HIV. Although we conducted site-specific comparisons, sample size precluded adjusting these comparisons for patient demographics (e.g., thyroid cancer-specific estimates were not adjusted for age or sex). Limitations include lack of data on the degree of HIV control and ART use, which need to be included in future work to directly link immunosuppression to molecular tumor features, as well as a lack of data on molecular testing for tumors (e.g., HPV status of head & neck cancers) and detailed cancer treatment. Finally, although the published TESLA filtering criteria were applied to identify putative neoantigens most capable of triggering an anti-tumor immune response, these published criteria were optimized for class I, not class II, neoantigens.
Our findings indicate that tumors diagnosed in the setting of HIV harbor a higher total mutation burden (TMB) and a higher number of potential class I neoantigens predicted to elicit an anti-tumor immune response, when compared to tumors diagnosed at the same anatomic site in patients with cancer without HIV. A higher TMB in PWH may portend a more favorable response to cancer treatment modalities such as immune checkpoint inhibitor therapy, which would be a welcome avenue for future translational research given the higher rates of cancer-specific mortality in PWH and cancer that we and others have reported.
Supplementary Material
ACKNOWLEDGMENTS.
A. E. Coghill received funding from the National Cancer Institute (R01 CA268973).
Footnotes
Disclosures. Authors report no relevant conflicts of interest.
REFERENCES
- 1.Collaboration H-C, Cain LE, Logan R, Robins JM, Sterne JA, Sabin C, et al. When to initiate combined antiretroviral therapy to reduce mortality and AIDS-defining illness in HIV-infected persons in developed countries: an observational study. Ann Intern Med. 2011;154(8):509–15. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Collaboration H-C, Ray M, Logan R, Sterne JA, Hernandez-Diaz S, Robins JM, et al. The effect of combined antiretroviral therapy on the overall mortality of HIV-infected individuals. AIDS. 2010;24(1):123–37. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.French MA, King MS, Tschampa JM, da Silva BA, Landay AL. Serum immune activation markers are persistently increased in patients with HIV infection after 6 years of antiretroviral therapy despite suppression of viral replication and reconstitution of CD4+ T cells. J Infect Dis. 2009;200(8):1212–5. [DOI] [PubMed] [Google Scholar]
- 4.Lederman MM, Calabrese L, Funderburg NT, Clagett B, Medvik K, Bonilla H, et al. Immunologic failure despite suppressive antiretroviral therapy is related to activation and turnover of memory CD4 cells. J Infect Dis. 2011;204(8):1217–26. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Valdez H, Connick E, Smith KY, Lederman MM, Bosch RJ, Kim RS, et al. Limited immune restoration after 3 years’ suppression of HIV-1 replication in patients with moderately advanced disease. AIDS. 2002;16(14):1859–66. [DOI] [PubMed] [Google Scholar]
- 6.Chehimi J, Campbell DE, Azzoni L, Bacheller D, Papasavvas E, Jerandi G, et al. Persistent decreases in blood plasmacytoid dendritic cell number and function despite effective highly active antiretroviral therapy and increased blood myeloid dendritic cells in HIV-infected individuals. J Immunol. 2002;168(9):4796–801. [DOI] [PubMed] [Google Scholar]
- 7.Sachdeva N, Asthana V, Brewer TH, Garcia D, Asthana D. Impaired restoration of plasmacytoid dendritic cells in HIV-1-infected patients with poor CD4 T cell reconstitution is associated with decrease in capacity to produce IFN-alpha but not proinflammatory cytokines. J Immunol. 2008;181(4):2887–97. [DOI] [PubMed] [Google Scholar]
- 8.Castilho JL, Bian A, Jenkins CA, Shepherd BE, Sigel K, Gill MJ, et al. CD4/CD8 Ratio and Cancer Risk Among Adults With HIV. J Natl Cancer Inst. 2022;114(6):854–62. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Dunn GP, Bruce AT, Ikeda H, Old LJ, Schreiber RD. Cancer immunoediting: from immunosurveillance to tumor escape. Nat Immunol. 2002;3(11):991–8. [DOI] [PubMed] [Google Scholar]
- 10.Kim R, Emi M, Tanabe K. Cancer immunoediting from immune surveillance to immune escape. Immunology. 2007;121(1):1–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Caro-Vegas C, Ramirez C, Landis J, Adimora AA, Strickler H, French AL, et al. Molecular profiling of breast and lung cancer in women with HIV reveals high tumor mutational burden. AIDS. 2022;36(4):567–71. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Orenbuch R, Filip I, Comito D, Shaman J, Pe’er I, Rabadan R. arcasHLA: high-resolution HLA typing from RNAseq. Bioinformatics. 2020;36(1):33–40. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Wells DK, van Buuren MM, Dang KK, Hubbard-Lucey VM, Sheehan KCF, Campbell KM, et al. Key Parameters of Tumor Epitope Immunogenicity Revealed Through a Consortium Approach Improve Neoantigen Prediction. Cell. 2020;183(3):818–34 e13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.El Zarif T, Nassar AH, Adib E, Fitzgerald BG, Huang J, Mouhieddine TH, et al. Safety and Activity of Immune Checkpoint Inhibitors in People Living With HIV and Cancer: A Real-World Report From the Cancer Therapy Using Checkpoint Inhibitors in People Living With HIV-International (CATCH-IT) Consortium. J Clin Oncol. 2023;41(21):3712–23. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Lurain K, Zarif TE, Ramaswami R, Nassar AH, Adib E, Abdel-Wahab N, et al. Real-World Multicenter Study of PD-1 Blockade in HIV-Associated Classical Hodgkin Lymphoma Across the United States. Clin Lymphoma Myeloma Leuk. 2024; 24(8): 523–30. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Uldrick TS, Goncalves PH, Abdul-Hay M, Claeys AJ, Emu B, Ernstoff MS, et al. Assessment of the Safety of Pembrolizumab in Patients With HIV and Advanced Cancer-A Phase 1 Study. JAMA Oncol. 2019;5(9):1332–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Coghill AE, Shiels MS, Suneja G, Engels EA. Elevated Cancer-Specific Mortality Among HIV-Infected Patients in the United States. J Clin Oncol. 2015;33(21):2376–83. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Coghill AE, Suneja G, Rositch AF, Shiels MS, Engels EA. HIV Infection, Cancer Treatment Regimens, and Cancer Outcomes Among Elderly Adults in the United States. JAMA Oncol. 2019;5(9):e191742. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Horner MJ, Shiels MS, Pfeiffer RM, Engels EA. Deaths Attributable to Cancer in the US Human Immunodeficiency Virus Population During 2001-2015. Clin Infect Dis. 2021;72(9):e224–e31. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The data analyzed in this study were generated by Avatar. Restrictions apply to the availability of these data, which were used as part of the ORIEN network of institutions (https://www.oriencancer.org/contact). Data are available from the authors upon reasonable request with the permission of Avatar, collaboration with an ORIEN institution, and after appropriate approvals are in place.