Abstract
Objectives
Immunomodulatory drugs (IMDs) are crucial for treating autoimmune, inflammatory, and oncologic conditions. Data regarding the safety of IMDs in people living with HIV (PLWH) are limited. We describe outcomes in all PLWH prescribed these agents from 2000–2019 at two academic medical centers.
Design
Retrospective cohort study.
Methods
We systematically identified and reviewed charts of all PLWH receiving IMDs. We defined a treatment episode (TE) as an uninterrupted period on an IMD regimen. We quantified infections, blips (detectable viral load following an undetectable result), and virologic failure (progression from viral load <200 copies/mL to two consecutive values >200 copies/mL despite ART).
Results
77 patients contributed 110 TEs. Rheumatologic comorbidities were the most frequent indication. The most common IMD classes were TNF inhibitors, antimetabolites, and checkpoint inhibitors. 90% of TEs involved concomitant ART. Median pre-treatment CD4+ T cell count was 609 (IQR 375–861). Among 51 TEs on ART with undetectable pre-treatment plasma HIV RNA, HIV became detectable within one year in 21 cases (41.2%); there were no instances of virologic failure. Compared with other agents, TEs involving checkpoint inhibitors were more likely to involve a blip (77.8% vs 33.3%, p=0.015). Thirteen TEs (11.8%) were associated with concomitant infection; none was attributed to IMDs by the treating clinician.
Conclusions
PLWH treated with IMDs should be monitored carefully for virologic blips and incident infections. Checkpoint inhibitors may be associated with a higher rate of viral blips, although the clinical significance is unclear.
Keywords: HIV, biologic, TNF inhibitor, checkpoint inhibito
INTRODUCTION
Immunomodulatory drugs (IMDs) have altered the morbidity associated with debilitating autoimmune and inflammatory conditions such as rheumatoid arthritis, psoriasis and psoriatic arthritis, and inflammatory bowel disease [1–3]. IMDs work through diverse mechanisms to alter immunologic pathways and modify disease processes. This includes interfering with cellular metabolism and nucleic acid synthesis [4], antagonizing molecules in inflammatory pathways [5–9], interfering with intracellular [10, 11] or extracellular [12] proteins, and depleting certain cell types [13]. While most IMDs downregulate pathologic immune system processes, newer IMDs known as “checkpoint inhibitors” have been developed to enhance immune surveillance in oncologic conditions and are becoming increasingly important in cancer therapy [14–16].
IMDs can be used as monotherapy, in combination, or in sequence. Data on outcomes of IMD use in people living with HIV (PLWH) are limited [17]. In general, PLWH were excluded from studies that led to approval of these agents. While there are case reports on the use of these agents in PLWH, the literature is likely to be biased due to selective reporting of certain outcomes, including successful treatment or unanticipated complications. For this reason, a previous systematic review was unable to make firm conclusions about the use of IMDs in PLWH in real-world practice [17].
We performed an institution-level analysis of all PLWH prescribed these drugs at two academic medical centers. We determined the characteristics of PLWH prescribed IMDs and complications associated with IMD use. We hypothesized that IMDs would not result in virologic or immunologic failure and that there would be few infectious complications attributed to these agents. We further hypothesized that checkpoint inhibitors, which may reverse HIV-1 latency [18, 19], would be associated with an increased rate of clinically insignificant viral blips.
METHODS
Study population
We utilized the Clinical Data Research Consultations service at the University of California, San Francisco (UCSF) to use the electronic medical record (EMR) to systematically identify all PLWH prescribed non-steroidal IMDs from January 1, 2000 to July 31, 2019 at the UCSF Medical Center and Zuckerberg San Francisco General Hospital & Trauma Center.
We searched for all patients with an ICD-9 or ICD-10 coded diagnosis related to HIV infection who had documented prescription or administration of at least one of 35 IMDs (Supplemental Table 1). We manually reviewed all charts to ensure that the following criteria inclusion were met: age ≥ 18 years, positive HIV antibody/antigen test or detectable value on HIV plasma HIV RNA assay, and documentation of immunomodulator use. We included patients who acquired HIV infection while already on IMDs, but excluded individuals who acquired HIV infection after discontinuing IMDs. We excluded individuals who received concurrent systemic cytoreductive chemotherapy for malignancy or whose primary indication for therapy was a lymphoproliferative disorder. We also excluded those who received concomitant mycophenolate mofetil, cyclosporine, or tacrolimus; these medications are not typically used as first-line therapy for non-transplant conditions at our institution.
Definition of “treatment episode”
We defined a treatment episode (TE) as an uninterrupted period on a particular IMD regimen. A patient could contribute multiple TEs if interrupting or switching therapy. For example, a patient who started a TNF inhibitor, was lost to follow-up, and then resumed the agent contributed two TEs. Similarly, a patient who had a poor response to etanercept and switched to adalimumab contributed two TEs. We considered periods of combination therapy (i.e., an antimetabolite plus a TNF inhibitor) as one TE. These episodes were analyzed separately, with the exception of the combination of ipilimumab and nivolumab, which are often administered concurrently.
Definition of “blip” and “virologic failure”
We defined a viral “blip” as a scenario in which an undetectable (“no copies detected” or “0”) plasma HIV RNA result was followed by any detectable plasma HIV RNA result (including “detectable below assay limit”). This allowed us to characterize changes below the linear range of clinical assays, in addition to those in the quantifiable range. We recorded blips within one year of the beginning of each TE. We defined virologic failure as the inability to achieve or maintain suppression of the plasma HIV RNA level below 200 copies/mL despite documented ART adherence over the year following IMD initiation [20].
Definition of “infectious complication”
We defined an infectious complication as any documented infection or infectious syndrome (e.g., sepsis syndrome) following immunomodulator initiation. When an individual contributed multiple TEs, we associated each infectious complication with the most proximate preceding TE. We also recorded whether the treating clinician attributed the infection to HIV or IMDs.
Data abstraction
An HIV clinician-scientist (MJP) abstracted patient demographics (age, gender, race), HIV history (year of infection, risk factor, CD4+ T cell nadir, history of opportunistic infection, and antiretroviral therapy [ART] history), concomitant infections (hepatitis B, hepatitis C, tuberculosis), HIV parameters (plasma HIV RNA levels and CD4+ T lymphocyte counts), immunomodulator history (comorbid indication for therapy, prescribing specialty, and TE date of initiation and discontinuation), and infectious complications during immunomodulator therapy. If there was uncertainty regarding a clinical outcome, the case was discussed with the senior clinician-investigators (TJH and PVC). Three non-clinician reviewers (JC, AR, SM) independently abstracted longitudinal HIV laboratory parameters. No discrepancies were identified.
Statistical analysis
We summarized baseline characteristics using descriptive statistics. To compare the rate of blips to that reported in the literature, we used a one-sample binomial test. To test the hypothesis that checkpoint inhibitors would be associated with increased blips, we calculated a comparison of proportions using the chi-square test to compare TEs involving checkpoint inhibitors with TEs involving all other agents. We used STATA version 15.1SE for statistical analyses and GraphPad Prism version 8.1 to display results.
Study approval
The study was approved by the UCSF institutional review board.
RESULTS
Characteristics of study participants
The characteristics of 77 individuals contributing 110 TEs are displayed in Table 1. Median age at the first TE was 54.2 years. The majority of patients were white men and the primary HIV risk factor was sexual activity between men (54.5%), reflecting the HIV epidemic in our region.
Table 1.
Baseline characteristics of 77 patients with HIV prescribed immunomodulator therapy. Values listed are median (interquartile range) unless otherwise specified.
| Demographic Characteristics | |
| Age at IMD Initiation (years) | 54.2 (46.5–59.3) |
| Gender, n (%) | |
| Cisgender man | 67 (87.0) |
| Cisgender woman | 7 (9.1) |
| Transgender woman | 3 (3.9) |
| Race/Ethnicity, n (%) | |
| White | 47 (61.0) |
| Black | 8 (10.4) |
| Latinx | 14 (18.8) |
| Asian | 1 (1.3) |
| Other/Unknown | 7 (9.1) |
| HIV Disease History Characteristics | |
| HIV Risk Factor*, n (%) | |
| Men who have sex with men | 42 (54.5) |
| Injection drug use | 8 (10.4) |
| Heterosexual intercourse | 4 (5.2) |
| Blood transfusions for hemophilia | 1 (1.2) |
| Not specified in medical record | 24 (31.1) |
| Duration of HIV Infection at IMD Initiation (years) | 18.1 (8.9–25.5) |
| Nadir CD4+ T Lymphocyte Count (cells/uL) | 199 (100–333) |
| Co-infection History, n (%) | |
| History of opportunistic infection | 20 (26.0) |
| Positive hepatitis B core antibody | 36 (46.8) |
| Positive hepatitis C antibody | 17 (22.1) |
| Positive latent TB Testing | 5 (6.5) |
| Indication for initial treatment episode with immunomodulatory therapy | |
| Class of Disorder, n (%) | |
| Rheumatologic | 26 (33.8) |
| Oncologic | 19 (24.7) |
| Gastrointestinal | 13 (16.9) |
| Dermatologic | 7 (9.1) |
| Neurologic/neuromuscular | 4 (5.2) |
| Pulmonary | 4 (5.2) |
| Allergic | 2 (2.6) |
| Hematologic | 1 (1.3) |
Note: Some patients had multiple risk factors.
Individuals had complex HIV histories characterized by prolonged HIV infection and CD4+ T cell nadir values below 200 cells/uL. Twenty individuals (26%) had documentation of a prior opportunistic infection, which included thrush, Pneumocystis jirovecii pneumonia, disseminated or invasive herpesvirus infections (e.g., cytomegalovirus, varicella-zoster virus, herpes simplex virus), mycobacterial disease (pulmonary tuberculosis, mycobacterium avium complex adenitis), and AIDS-associated Kaposi sarcoma. None had an active opportunistic infection at IMD initiation. Most individuals had screening for concomitant infections including viral hepatitis (65/77, 84.4%) and tuberculosis (64/77, 83.1%). Nearly half had a positive hepatitis B core antibody, although none had detectable HBV DNA in blood. Five (6.5%) had positive latent tuberculosis testing; treatment was documented in 4/5 cases.
Comorbidities and immunomodulatory therapy
Tables 1 and S2 indicate the initial comorbid indications and IMD-prescribing specialties. Rheumatologic conditions were most frequent (33.8%) and the most common indication was psoriatic arthritis. A substantial proportion also had oncologic (24.7%) or gastrointestinal (16.9%) conditions. Of note, one subsequent TE involved a TNF-inhibitor for autoimmune complications precipitated by preceding checkpoint inhibitor therapy. The median duration of IMD therapy was 257 days (IQR 100–639), and the total duration of therapy was 159.82 patient-years.
Table 2 indicates the immunomodulatory classes utilized in this patient population. The most common class was TNF-inhibitors and the most common agent was etanercept. Antimetabolites and checkpoint inhibitors were also common. Ten TEs involved combination therapy, most commonly with concomitant use of an antimetabolite and biologic for management of rheumatoid arthritis (n=5).
Table 2.
Immunomodulatory drugs prescribed in 110 episodes of immunomodulatory therapy and pre-treatment HIV disease parameters preceding these treatment episodes.
| Therapeutic Class, n(%) | |
| TNF Inhibitor | 33 (30.0) |
| Antimetabolite | 21 (19.1) |
| Checkpoint inhibitor | 15 (13.6) |
| Interleukin inhibitor | 12 (10.9) |
| Protein kinase inhibitor | 6 (5.5) |
| Phosphodiesterase inhibitor | 5 (4.5) |
| Rituximab | 4 (3.6) |
| Integrin inhibitor | 2 (1.8) |
| Epidermal growth factor receptor inhibitor | 1 (0.9) |
| Immunoglobulin E inhibitor | 1 (0.9) |
| Combination therapy | 10 (9.1) |
| HIV Disease Parameters at Beginning of Treatment Episode | |
| On ART | 99 (90) |
| Pre-Treatment Plasma HIV RNA (n=92) | |
| No target detected | 64 (69.6) |
| Detected, <50 copies/mL | 13 (14.1) |
| 51–200 copies/mL | 5 (5.4) |
| >200 copies/mL | 10 (10.9) |
| Pre-Treatment CD4+ T Lymphocyte Count (cells/uL, n=88) | 609 (375–861) |
| <200 cells/uL | 9 (10.2) |
| 200–500 cells/uL | 30 (34.1) |
| >500 cells/uL | 49 (55.7) |
| Pre-Treatment CD4+ T Lymphocyte Percentage (n=88) | 29.5 (20.6–37.9) |
| Hematologic Parameters | |
| Absolute neutrophil count (x103 cells/uL, n=83) | 3.5 (2.8–4.5) |
| Absolute lymphocyte count (x103 cells/uL n=83) | 1.8 (1.3–2.4) |
| Hemoglobin (g/dL, n=88) | 13.5 (12.3–15.2) |
| Platelets (x103 cells/uL n=88) | 230 (189–289) |
Baseline virologic and immunologic parameters
In 99/110 (90%) TEs, patients were on ART at the time of immunomodulator initiation (Table 2); the remaining 11 episodes occurred in untreated “elite” controllers (n=4), those with incident HIV while on IMDs (n=4), or those beginning ART concurrently (n=3).
Ninety-two TEs (83.6%) had preceding plasma HIV RNA data documented (Table 2, Figure 1a). Of these, 64 TEs (69.6%) were initiated when the patient had no detectable plasma HIV RNA. Of the remaining 28 TEs initiated in the setting of detectable plasma HIV RNA, the patient’s viral load was greater than 200 copies/mL in 10.
Figure 1.
(a) Pre-treatment plasma HIV RNA. There were 10 episodes where the pre-treatment viral load was greater than 200 copies/mL. (b) Pre-treatment CD4+ T cell counts. (c) Plasma HIV RNA trends in individuals with undetectable starting levels, indicating detectable plasma HIV RNA levels within one year in a subset of individuals. Points are scaled to proportion of sample. (d) Longitudinal plasma HIV RNA trends in individuals experiencing an increase in plasma HIV RNA to greater than 200 copies/mL within a year of IMD initiation. Symbols represent unique participants. (e) Longitudinal plasma HIV RNA trends in individuals initiating IMD therapy with plasma HIV RNA levels greater than 200 copies/mL showing improvement in most cases. Symbols represent unique participants. (f) Comparison of proportion of TEs associated with a plasma HIV RNA blip within one year. Dotted lines represent upper and lower limit of blip rates reported in the literature (30% and 10%, respectively). (g) Pre- and post-treatment CD4+ T cell count among individuals with starting CD4+ T cell levels greater than 500 cells/uL. Thick line represents change in median between the two time points. (h) Pre- and post-treatment CD4+ T cell count among individuals with starting CD4+ T cell levels less than 500 cells/uL. Thick line represents change in median between the two time points. (i) Pre- and post-treatment CD4+ T cell percentages. Thick line represents change in median between the two time points.
Eighty-eight episodes (80%) had preceding CD4+ T lymphocyte count data (Table 2, Figure 1b). The median CD4+ T cell count at the time of treatment initiation was 609 (375–861); 9 TEs (10.5%) were initiated in the setting of a CD4+ T cell count < 200 cells/uL.
Changes in virologic parameters
Of 110 TEs, 78 (70.9%) had plasma HIV RNA levels documented before and after treatment. Among 51 TEs beginning with undetectable plasma HIV RNA in ART-treated individuals, HIV became detectable within one year in 21/51 cases (41.2%). This proportion was elevated in comparison to a rate of 10 to 30% in the published literature (p<0.001 to p=0.087) [21–23]. In an analysis restricted to only first TEs (n=33), 16 were associated with a blip (48.4%; p<0.001 to p=0.024).
Of 21 episodes with blips, 19/21 were to levels less than 200 copies/mL (Figure 1c). These were generally below the quantification threshold on assays with limits of 20 (n=2) and 40 copies/mL (n=16); in one case level was 78 copies/mL. Among two episodes with blips greater than 200 copies/mL, one was to 208 and one to 2557 copies/mL (Figure 1c–d). The latter was in an individual on sorafenib for hepatocellular carcinoma who discontinued therapy due to severe hand-foot syndrome and had documented non-adherence to ART; he re-suppressed with improved adherence.
Twenty-four TEs with follow-up data had detectable HIV RNA at IMD initiation. In all 24, the individual was on ART (n=22) or initiated ART concurrently (n=2). Plasma HIV RNA was subsequently suppressed below the limit of detection in 10 cases (41.7%) and below the limit of quantification in 5 cases (20%). There was only one case in which plasma HIV RNA rose substantially following therapy, from 65 to 23,045 copies/mL, in a patient being treated with apremilast for psoriatic arthritis who had ART non-adherence. No patients on ART experienced virologic failure. Suppression below the limit of detection occurred even in TEs in which the initial viral load exceeded 200 copies/mL (Figure 1e).
An analysis examining 15 TEs (n=14 patients) utilizing checkpoint inhibitors demonstrated a higher proportion of TEs associated with a viral blip in individuals receiving this therapy compared with the rest of the cohort (77.8% vs 33.3%, p=0.015; Figure 1f). Additional information regarding the subset of TEs involving checkpoint inhibitors is displayed in Supplemental Figure 1.
Changes in immunologic parameters
Figure 1g–i shows CD4+ T lymphocyte counts and percentages before and after immunomodulatory therapy. Sixty-seven TEs (60.1%) had documented counts in individuals on ART before and after therapy. Of these, 31 were associated with a decrease in CD4+ T cell count and 36 with an increase in CD4+ T cell count. The median change was +29 cells/uL (IQR −127 to +119 cells/uL).
Three TEs were associated with a decline to < 200 cells/uL. One individual on treatment with methotrexate for an undifferentiated autoimmune process experienced CD4+ T cell decline from 657 to 171 cells/uL (CD4%: 29% to 17%) in the setting of non-infectious liver function test abnormalities; viral load remained undetectable. A second experienced decline from 375 to 90 cells/uL (CD4%: 15% to 11%; Figure S1b) in the setting of severe non-infectious diarrhea on nivolumab for squamous cell carcinoma; viral load blipped from undetectable to detectable (<40 copies/mL). A third demonstrated CD4 decline from 339 to 96 cells/uL (CD4%: 17% to 33%) during a sepsis syndrome on azathioprine for interstitial lung disease; viral load remained undetectable.
Infectious complications
Thirteen of 110 TEs (11.8%) were associated with an infectious complication (Table 3). The majority involved skin and soft tissue infections (n=6). Others included secondary syphilis, recurrent sinusitis, and orchitis, although these were not attributed to the immunotherapy by the provider. There were several severe complications including Escherichia coli sepsis, disseminated aspergillosis and Clostridium difficile colitis, bacterial pneumonia, and pneumonia-associated sepsis syndrome. In each case, the complication was attributed to the underlying comorbidity rather than IMD therapy. There were no episodes of recurrent opportunistic infections, hepatitis B reactivation, or tuberculosis reactivation. Individuals with CD4 T+ cell nadirs < 200 cells/uL or prior opportunistic infections were not more likely to have an infectious complication (p=0.16 and p=0.23, respectively). Reports of infectious complications did not differ between checkpoint inhibitors and other therapies (20% vs 10.5%, p=0.3).
Table 3.
Infectious complications in 110 treatment episodes of immunomodulator therapy.
| Clinical Scenario | Immunomodulator | Infectious Complication |
|---|---|---|
| 62M with melanoma | pembrolizumab | Skin abscess |
| 43M with ulcerative colitis | methotrexate + vedolizumab | Skin abscess |
| 62F transgender with interstitial lung disease | azathioprine | Pneumonia Sepsis syndrome Death |
| 54F with autoimmune hemolytic anemia | rituximab | Pasteurella cellulitis following cat bite |
| 53F with undifferentiated autoimmune process | azathioprine + rituximab | Recurrent sinusitis |
| 59M with melanoma | ipilimumab + nivolumab | Orchitis |
| 55M with Kaposi sarcoma | nivolumab | Cellulitis |
| 55M with melanoma | pembrolizumab | Skin abscess |
| 61M with psoriasis | ustekinumab | Bacterial pneumonia Dental abscess |
| 61M with autoimmune hepatitis | azathioprine | Disseminated aspergillus* Clostridium difficile colitis |
| 44M with rheumatoid arthritis | adalimumab + leflunomide | Secondary syphilis |
| 53M with autoimmune hepatitis | azathioprine | Escherichia coli sepsis |
| 57M with autoimmune hepatitis | azathioprine | Skin abscess |
Note: attributed to prior steroid use rather than IMD therapy.
Other notable cases
There were four elite controllers in our sample. While one individual was maintained on ART during two TEs, the others were not initiated on ART. Of these, one experienced a blip from undetectable plasma HIV RNA to detectable levels below the linear range of the assay after sequential therapy with adalimumab, infliximab, and methotrexate for ulcerative colitis; one experienced blipped from <20 to 51 copies/mL after initiating ustekinumab for psoriasis, and one transitioned to hospice care after therapy with ipilimumab and nivolumab and did not have further HIV care.
We identified four individuals who acquired HIV infection while on IMDs. All initiated ART. Three continued immunomodulator treatment without issue. In a fourth case, a 30-year-old man being treated for psoriasis acquired HIV and discontinued etanercept due to safety concerns.
We also identified two cases in which a patient received IMDs in the absence of HIV testing, and was later found to have had undiagnosed HIV infection once testing was performed. In one case, refractory arthritis symptoms uncontrolled with IMDs resolved completely with ART initiation. In the other, effective IMD therapy was discontinued due to concern for elevated risk of infection; symptoms related to hidradenitis suppurativa progressed.
DISCUSSION
We analyzed all PLWH receiving immunomodulatory agents at two medical centers that care for a large population of PLWH. While further work is necessary to systematically compare outcomes of interest in controlled studies, our results suggest that, in general, these agents can be used in PLWH without problematic virologic or immunologic failure, although a substantial proportion of individuals experienced viral blips and several experienced infectious complications. This analysis contributes to the existing literature on this topic by providing an unbiased review of a sizeable population of PLWH receiving these agents [17].
Most individuals were on ART when IMDs were initiated, although a small fraction (mainly “elite” controllers) was not. Modern ART is extremely effective at achieving virologic suppression. However, individuals occasionally experience viral “blips,” characterized by transient low-level viremia. While the clinical significance of such “blips” is unclear [22], detectable plasma HIV RNA on ART could be associated with adverse outcomes [21–23]. Historically, blips have been documented to occur at a rate of 10%−30%, although the exact incidence in the era of modern ART and newer-generation PCR testing is less certain and depends upon both sampling frequency and assay sensitivity [24]. In our sample, nearly half of TEs initiated when the patient had undetectable plasma HIV RNA were associated with a blip. This could be explained by comorbidity-associated inflammation, adherence or pharmacodynamic issues, or a direct effect of the medications on HIV latency. It could also be related to the sensitivity of the assays used or increased monitoring in these individuals. Because it was not feasible to construct an adequately matched control group, we could not assess this proportion in the context of our population of PLWH. Future prospective studies should aim to do so.
We found an increased proportion of blips associated with checkpoint inhibitors. While it is not possible to draw firm conclusions related to this observation, these agents have been hypothesized to cause HIV latency reversal [18, 19]. In a recent systematic review of checkpoint inhibitors in PLWH with advanced cancer [25], a small proportion (7%) exhibited detectable plasma HIV RNA after therapy. The higher blip rate here may be due to a stricter definition of “undetectable” plasma HIV RNA or a population with more severe underlying disease. There has been recent interest in identifying increased HIV transcriptional activity in the setting of checkpoint blockade [18, 19, 26]. Although the mechanisms are not entirely clear, our results could support these preliminary findings. Further investigation is warranted.
A substantial proportion did not have undetectable or suppressed plasma HIV RNA at the time of IMD initiation. These individuals generally demonstrated downtrending plasma HIV RNA after starting ART; those who did not had documented non-adherence. While the majority of individuals had CD4+ T cell counts >500 cells/uL at the time of therapy initiation, a substantial proportion were treated with lower counts, including counts <250 cells/uL. These individuals still appeared to do well, with stable or improving counts.
There is concern that immunomodulatory therapy can be associated with increased risk of infectious complications [27]. In a recent study of > 4000 patients with psoriatic arthritis, incidence of serious infections ranged from 1.0–3.01 per 100 patient-years in individuals treated with TNF inhibitors or methotrexate [28]. In another study using hospital administration data from 5596 patients with rheumatologic diseases, 289 (4.2%) were hospitalized due to severe infections [29]. There were relatively few infectious complications in our population. Many were skin and soft tissue infections; most did not require hospitalization. The majority were attributed to comorbid disease rather than HIV or IMDs. Due to methodologic differences and the heterogeneity of comorbidities and IMDs within our study, it is difficult to compare the proportion of infections here with those previously described. However, given prior data linking severe infections, including skin and soft-tissue infections [30], to immunomodulator therapies, further studies are needed to fully understand the infectious risk of IMDs in the setting of HIV infection.
Elite controllers, who comprise <1% of PLWH [31], were over-represented in our sample. This could be related to enrichment among PLWH at our institution, from which several cohorts studying elite controllers are based, or to genetic characteristics that both promote autoimmunity and spontaneous HIV control (HLA*B27 positivity). While many elite controllers are now treated with ART, 3/4 in our study did not initiate ART with their IMDs and two experienced viral blips.
In several instances, previously effective immunomodulatory therapy was discontinued due to newly acquired or diagnosed HIV infection. The rationale for this was not clear and IMD discontinuation often resulted in worsening morbidity. The handful of episodes in which IMDs were initiated without prior HIV testing are particularly concerning and demonstrate the importance of excluding this diagnosis before initiating immune-modifying therapy.
This study has several limitations. First, because the population these centers serve is dynamic, it was not possible to reliably calculate a denominator of PLWH that would allow us to identify the total proportion of individuals receiving immunomodulatory therapy during this period. Second, due to limitations in our EMR system and the fact that few individuals with autoimmune disease or cancer are not on therapy for these conditions, we were similarly unable to construct a well-matched and clinically relevant comparator group, which would have further contextualized the findings. Third, because internal correlations between individuals contributing multiple TEs are also possible, we attempted to control for these by conducting sensitivity analyses limited to a single TE for each individual. This further emphasizes the value of future studies utilizing a control group. Fourth, we were unable to reliably identify whether individuals had received concomitant steroid therapy, and so could not conduct a planned sub-analysis evaluating whether this had an effect on outcomes with immunomodulators. In addition, study of emerging immunotherapies or those excluded from our analysis (e.g., tacrolimus, sirolimus, mycophenolate, cyclosporine, etc) might be of interest. Fifth, there was substantial variability in the rate of laboratory testing, the specific type of plasma HIV RNA testing performed, the timing at which these tests were performed, and the length of follow-up, each of which could serve as a source of bias. Sixth, CD8+ T cell counts and CD4/CD8 ratio data were not available in most cases, but analysis of these parameters may be informative in future studies. Seventh, only infections that came to medical attention and were included in the EMR were assessed. More minor infections such as upper respiratory tract infections would have been missed by our chart abstraction. While in most cases, instances in which patients presented for care outside of the hospital system were readily identifiable using specific features of the EMR systems or through documentation in the provider notes, it is possible that instances in which an individual sought care out-of-state or out-of-network and did not report this to their provider were missed. Finally, evaluating immune-related adverse effects of checkpoint inhibitors would also potentially be informative, but was beyond the scope of this study which focused on infectious complications.
Further studies, including inclusion of PLWH in controlled trials examining the safety and efficacy of these agents is warranted. Until these data are available, we propose the recommendations for clinicians caring for PLWH as outlined in Table 4.
Table 4.
Recommendations for use of immunomodulatory drugs in people living with HIV.
| All patients with HIV being prescribed immunomodulatory drugs for comorbid conditions should be counseled. There are limited data on the safety of these agents in people living with HIV, but there is growing evidence that they can be used safely with appropriate monitoring. |
| For previously HIV-negative individuals or those with unknown HIV status, an HIV test should be part of the workup prior to initiation of immunomodulator therapy. |
| For HIV-positive individuals, immunomodulatory drugs should be prescribed with careful monitoring and in consultation with an HIV provider. Before starting immunomodulatory drugs: |
|
| Patients with HIV initiating immunomodulator therapy should be treated with concomitant ART, regardless of their virologic or immunologic status. |
|
| There are no data to guide frequency of monitoring for HIV parameters in people living with HIV on immunomodulatory therapy. |
|
| Clinicians and patients should remain vigilant for infectious complications given the risk associated with these agents in general. |
|
| Acquisition of HIV infection should not be an automatic contraindication to continuation of immunomodulatory therapy. |
|
Supplementary Material
Figure S1. (a) Plasma HIV RNA trends in individuals receiving checkpoint inhibitors, indicating the highest subsequent plasma HIV RNA value recorded in the following 1 year. Note two TEs did not have follow-up plasma HIV RNA testing. (b) Pre- and post-treatment CD4+ T cell count in individuals receiving checkpoint inhibitors. Thick line represents change in median between the two time points. (c-e) Illustrative examples of individual participants receiving checkpoint inhibitors demonstrating virologic blips and CD4+ trajectories. Preceding data was included where available. IMD indicates time at which checkpoint inhibitor was initiated.
Acknowledgements
We acknowledge the patients whose data we reviewed and the providers who cared for these patients. We also thank Dr. Paul Volberding and IAS-USA for support and advice in pursuing this clinical question. MJP is supported through the National Institutes of Allergy and Infectious Diseases, National Institutes of Health on training grant T32 AI60530-12. This project was supported by UCSF Academic Research Systems, and by the National Center for Advancing Translational Sciences, National Institutes of Health, through UCSF-CTSI Grant Number UL1 TR991872 and UL1 RR024131; the National Institute on Minority Health and Health Disparities, National Institutes of Health, through Grant Number U54 GM118986; and the National Institute of Allergy and Infectious Diseases, National Institutes of Health, through Grant Number R01 AI141003, UM1 AI126622, and R01 AI122862. It was further supported by a grant from the UCSF Resource Allocation Program (MJP) and funds from the Division of Experimental Medicine (TJH). Its contents are solely the responsibility of the authors and do not necessarily represent the official views of UCSF or the NIH.
Footnotes
Publisher's Disclaimer: Disclaimer: The contents of this work are solely the responsibility of the authors and do not necessarily represent the official views of University of California, San Francisco or the National Institutes Health.
REFERENCES
- 1.Danese S, Vuitton L, Peyrin-Biroulet L. Biologic agents for IBD: practical insights. Nat Rev Gastroenterol Hepatol 2015; 12(9):537–545. [DOI] [PubMed] [Google Scholar]
- 2.Her M, Kavanaugh A. Advances in use of immunomodulatory agents--a rheumatology perspective. Nat Rev Gastroenterol Hepato l 2015; 12(6):363–368. [DOI] [PubMed] [Google Scholar]
- 3.Iskandar IYK, Ashcroft DM, Warren RB, Evans I, McElhone K, Owen CM, et al. Patterns of biologic therapy use in the management of psoriasis: cohort study from the British Association of Dermatologists Biologic Interventions Register (BADBIR). Br J Dermatol 2017; 176(5):1297–1307. [DOI] [PubMed] [Google Scholar]
- 4.Cronstein BN. Molecular therapeutics. Methotrexate and its mechanism of action. Arthritis Rheum 1996; 39(12):1951–1960. [DOI] [PubMed] [Google Scholar]
- 5.Li P, Zheng Y, Chen X. Drugs for Autoimmune Inflammatory Diseases: From Small Molecule Compounds to Anti-TNF Biologics. Front Pharmacol 2017; 8:460. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Mitoma H, Horiuchi T, Tsukamoto H, Ueda N. Molecular mechanisms of action of anti-TNF-alpha agents - Comparison among therapeutic TNF-alpha antagonists. Cytokine 2018; 101:56–63. [DOI] [PubMed] [Google Scholar]
- 7.Bilal J, Berlinberg A, Bhattacharjee S, Trost J, Riaz IB, Kurtzman DJB. A systematic review and meta-analysis of the efficacy and safety of the interleukin (IL)-12/23 and IL-17 inhibitors ustekinumab, secukinumab, ixekizumab, brodalumab, guselkumab and tildrakizumab for the treatment of moderate to severe plaque psoriasis. J Dermatolog Treat 2018; 29(6):569–578. [DOI] [PubMed] [Google Scholar]
- 8.Wyant T, Fedyk E, Abhyankar B. An overview of the mechanism of action of the monoclonal antibody vedolizumab. J Crohns Colitis 2016; 10(12):1437–1444. [DOI] [PubMed] [Google Scholar]
- 9.Lenz HJ. Cetuximab in the management of colorectal cancer. Biologics 2007; 1(2):77–91. [PMC free article] [PubMed] [Google Scholar]
- 10.Cervello M, Bachvarov D, Lampiasi N, Cusimano A, Azzolina A, McCubrey JA, et al. Molecular mechanisms of sorafenib action in liver cancer cells. Cell Cycle 2012; 11(15):2843–2855. [DOI] [PubMed] [Google Scholar]
- 11.Schafer P Apremilast mechanism of action and application to psoriasis and psoriatic arthritis. Biochem Pharmacol 2012; 83(12):1583–1590. [DOI] [PubMed] [Google Scholar]
- 12.Kaplan AP, Gimenez-Arnau AM, Saini SS. Mechanisms of action that contribute to efficacy of omalizumab in chronic spontaneous urticaria. Allergy 2017; 72(4):519–533. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Weiner GJ. Rituximab: mechanism of action. Semin Hematol 2010; 47(2):115–123. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Darvin P, Toor SM, Sasidharan Nair V, Elkord E. Immune checkpoint inhibitors: recent progress and potential biomarkers. Exp Mol Med 2018; 50(12):165. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Wei SC, Duffy CR, Allison JP. Fundamental mechanisms of immune checkpoint blockade therapy. Cancer Discov 2018; 8(9):1069–1086. [DOI] [PubMed] [Google Scholar]
- 16.Ribas A, Wolchok JD. Cancer immunotherapy using checkpoint blockade. Science 2018; 359(6382):1350–1355. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Fink DL, Hedley L, Miller RF. Systematic review of the efficacy and safety of biological therapy for inflammatory conditions in HIV-infected individuals. Int J STD AIDS 2017; 28(2):110–119. [DOI] [PubMed] [Google Scholar]
- 18.Fromentin R, DaFonseca S, Costiniuk CT, El-Far M, Procopio FA, Hecht FM, et al. PD-1 blockade potentiates HIV latency reversal ex vivo in CD4(+) T cells from ART-suppressed individuals. Nat Commun 2019; 10(1):814. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Evans VA, van der Sluis RM, Solomon A, Dantanarayana A, McNeil C, Garsia R, et al. Programmed cell death-1 contributes to the establishment and maintenance of HIV-1 latency. AIDS 2018; 32(11):1491–1497. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Panel on Antiretroviral Guidelines for Adults and Adolescents. Guidelines for the use of antiretroviral agents in HIV-1-infected adults and adolescents. Department of Health and Human Services. October 17, 2017. [Google Scholar]
- 21.Henrich TJ, Wood BR, Kuritzkes DR. Increased risk of virologic rebound in patients on antiviral therapy with a detectable HIV load <48 copies/mL. PLOS One 2012; 7(11):e50065. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Lee PK, Kieffer TL, Siliciano RF, Nettles RE. HIV-1 viral load blips are of limited clinical significance. J Antimicrob Chemother 2006; 57(5):803–805. [DOI] [PubMed] [Google Scholar]
- 23.Sorstedt E, Nilsson S, Blaxhult A, Gisslen M, Flamholc L, Sonnerborg A, et al. Viral blips during suppressive antiretroviral treatment are associated with high baseline HIV-1 RNA levels. BMC Infect Dis 2016; 16:305. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Nettles RE, Kieffer TL, Kwon P, Monie D, Han Y, Parsons T, et al. Intermittent HIV-1 viremia (Blips) and drug resistance in patients receiving HAART. JAMA 2005; 293(7):817–829. [DOI] [PubMed] [Google Scholar]
- 25.Cook MR, Kim C. Safety and efficacy of immune checkpoint inhibitor therapy in patients with HIV infection and advanced-stage cancer: a systematic review. JAMA Oncol 2019; 5(7):1049–1054. [DOI] [PubMed] [Google Scholar]
- 26.Galanina N, Goodman AM, Cohen PR, Frampton GM, Kurzrock R. successful treatment of HIV-associated kaposi sarcoma with immune checkpoint blockade. Cancer Immunol Res 2018; 6(10):1129–1135. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Martin-Mola E, Balsa A. Infectious complications of biologic agents. Rheum Dis Clin North Am 2009; 35(1):183–199. [DOI] [PubMed] [Google Scholar]
- 28.Ritchlin CT, Stahle M, Poulin Y, Bagel J, Chakravarty SD, Kafka S, et al. Serious infections in patients with self-reported psoriatic arthritis from the Psoriasis Longitudinal Assessment and Registry (PSOLAR) treated with bioloics. BMC Rheumatology 2019; 3:52. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Quartuccio L, Zabotti A, Del Zotto S, Zanier L, De Vita S, Valent F. Risk of serious infection among patients receiving biologics for chronic inflammatory diseases: usefulness of administrative data. Journal of Advanced Research 2019; 15:87–93. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Dixon WG, Watson K, Lunt M, Hyrich KL, Silman AJ, Symmons DP, et al. Rates of serious infection, including site-specific and bacterial intracellular infection, in rheumatoid arthritis patients receiving anti-tumor necrosis factor therapy: results from the British Society for Rheumatology Biologics Register. Arthritis Rheum 2006; 54(8):2368–2376. [DOI] [PubMed] [Google Scholar]
- 31.Crowell TA, Hatano H. Clinical outcomes and antiretroviral therapy in ‘elite’ controllers: a review of the literature. Journal of Virus Eradication 2015; 1(2):72–77. [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
Figure S1. (a) Plasma HIV RNA trends in individuals receiving checkpoint inhibitors, indicating the highest subsequent plasma HIV RNA value recorded in the following 1 year. Note two TEs did not have follow-up plasma HIV RNA testing. (b) Pre- and post-treatment CD4+ T cell count in individuals receiving checkpoint inhibitors. Thick line represents change in median between the two time points. (c-e) Illustrative examples of individual participants receiving checkpoint inhibitors demonstrating virologic blips and CD4+ trajectories. Preceding data was included where available. IMD indicates time at which checkpoint inhibitor was initiated.