Secondary Logo

HLA Class I restricted CD8+ and Class II restricted CD4+ T cells are implicated in the pathogenesis of nevirapine hypersensitivity

Keane, Niamh M.a,*; Pavlos, Rebecca K.a,*; McKinnon, Elizabetha; Lucas, Andrewa; Rive, Craiga; Blyth, Christopher C.b,c; Dunn, Davida; Lucas, Michaelad; Mallal, Simona,e; Phillips, Elizabetha,e

doi: 10.1097/QAD.0000000000000345

Objectives: This study sought to examine nevirapine hypersensitivity (NVP HSR) phenotypes and their relationship with differing major histocompatibility complex (MHC) Class I and Class II alleles and the associated CD4+ and CD8+ T-cell NVP-specific responses and their durability over time.

Methods: A retrospective cohort study compared HIV-positive patients with NVP HSR, defined by fever and hepatitis and/or rash, with those tolerant of NVP for more than 3 months. Covariates included class I (HLA-A, B, C) and class II (HLA-DR) alleles. Cellular studies examined NVP-specific CD4+ and CD8+ T-cell responses by interferon-gamma (IFNγ) ELISpot assay and intracellular cytokine staining (ICS).

Results: NVP HSR occurred in 19 out of 451 (4%) NVP-exposed individuals between March 1993 and December 2011. HLA associations were phenotype dependent with HLA-DRB1*01 : 01 associated with hepatitis (P = 0.02); HLA-B*35 : 01 and HLA-Cw4 associated with cutaneous NVP HSR (P = 0.001, P = 0.01), and HLA-Cw*08 was associated with NVP HSR with eosinophilia (P = 0.04) and multisystemic NVP HSR (P = 0.02). NVP-specific INFγ responses waned significantly more than 3 months from the original reaction and were diminished or completely abrogated when either CD4+ or CD8+ T cells were depleted from the peripheral blood mononuclear cells culture.

Conclusion: The association of specific class I and II allele pairings with specific phenotypes of NVP HSR, and cellular studies showing both CD4+ and CD8+ T-cell NVP-specific responses suggest that specific combinations of NVP reactive class I restricted CD8+ and class II restricted CD4+ T cells contribute to the immunopathogenesis of NVP HSR.

Supplemental Digital Content is available in the text

aInstitute for Immunology and Infectious Diseases, Murdoch University

bDepartment of Paediatrics and Adolescent Medicine, Princess Margaret Hospital for Children

cSchool of Paediatrics & Child Health

dSchool of Medicine & Pharmacology and School of Pathology & Laboratory Medicine, The University of Western Australia, Perth, Western Australia, Australia

eVanderbilt University School of Medicine, Nashville, Tennessee, USA.

*Niamh M. Keane and Rebecca K. Parlos contributed equally to this study.

Correspondence to Elizabeth Phillips, MD, Division of Infectious Diseases, Vanderbilt University School of Medicine, 1161 21st Avenue South, A-2200 Medical Centre North, Nashville, TN 37232-2582, USA. E-mail:

Received 13 March, 2014

Revised 6 May, 2014

Accepted 6 May, 2014

Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal's Website (

Back to Top | Article Outline


Nevirapine (NVP) is a nonnucleoside reverse transcriptase inhibitor (NNRTI) used in combination HIV therapy with an excellent long-term safety profile. A treatment-limiting hypersensitivity syndrome (NVP HSR) characterized by fever, and/or rash and/or hepatitis and often accompanied by eosinophilia occurs in 5% of those starting NVP [1,2]. NVP HSR is suggested to be T-cell mediated by both human and rat models. In rats, the sensitivity to NVP-induced skin rash can be transferred with CD4+ T cells from NVP-rechallenged rats to naive recipients and partial depletion of CD4+ T cells delays and decreases the severity of the rash. This is consistent with the decreased incidence of rash in patients with a low CD4+ T-cell count [3–9].

A CD4+ dependent-major histocompatibility complex (MHC) Class II restricted immune response directed against NVP was first reported in Whites as the association of hepatic symptoms with a combination of CD4+ T cells at least 25% and HLA-DRB1*01 : 01 [10]. Multiple class I and/or II MHC associations have now been described with different phenotypes of NVP HSR across several ethnicities [11–19]. These differing Class I and Class II HLA associations with different NVP HSR phenotypes across distinct populations and the increased risk of cutaneous phenotypes of NVP HSR in African Americans with CYP2B6 516 G>T [20] suggest that genetic, immunological and metabolic pathways may be important. NVP may trigger class I restricted CD8+ or class II restricted CD4+ T-cell mediated immune responses in the presence of the relevant class I and II MHC alleles, respectively.

Back to Top | Article Outline

Materials and methods

As part of a retrospective cohort study, we identified the association of class I and II HLA alleles with clinical and immunological phenotypes of NVP HSR.

Back to Top | Article Outline

Patient and control population

Definitive ascertainment of NVP-associated reaction status was achievable in 375 of 451 individuals in the West Australian HIV Cohort identified as NVP-exposed. Ethnicity and/or HLA genotyping was unavailable for 54 of these individuals and 41 individuals lacked clinical information required as part of the inclusion criteria. Associations between HLA alleles, severity of CD4+ T-lymphocyte-dependent immune deficiency and predisposition to NVP HSR were explored in analyses restricted to Asians and Whites without symptoms after 3 months NVP exposure (n = 262) or who had developed NVP-induced drug reactions (n = 19) (Table 1). NVP-associated reactions were identified prospectively in the database, and the case definition was retrospectively validated by a clinician blinded to HLA typing, who utilized standardized diagnostic criteria, including a minimum of fever in combination with rash and/or hepatitis (grade 2 toxicity or greater: alanine aminotransferase >2.5 times the upper limit of normal) and/or eosinophilia (eosinophils > 0.5 x 109/l). HBV/HCV serology was also examined for all patients displaying hepatotoxicity. Serology results taken together with clinical symptoms such as eosinophilia or rash that indicated drug HSR and not hepatitis B virus (HBV)/hepatitis C virus (HCV) infection excluded viral infection in all cases that exhibited hepatic symptoms. Selection of patient PBMCs for cellular experiments was based on availability. Experiments were conducted with the understanding and consent of each participant as approved by the Royal Perth Hospital and Murdoch University Human Research Ethics Committees.

Table 1

Table 1

Back to Top | Article Outline

Human leukocyte antigen typing

HLA-A, B, C and DR, DQ typing was performed using sequence based typing as previously described [21].

Back to Top | Article Outline

Isolation of nevirapine and 12-OH-NVP

NVP solution was prepared by dissolving 200 mg NVP tablets (Viramune; Boehringer-Ingelheim, North Ryde, Australia) in dimethyl sulfoxide and the absorbance at 260 nm was checked with the nanodrop spectrophotometer. The NVP stock solution was diluted to 1 mg/ml in PBS, filtered and resuspended 1 : 10 in culture medium (10%FCS-RPMI-1604) and added to cell cultures in the ELISpot assay at a final concentration of 40 μm. The 12-OH-NVP metabolite was prepared similarly and then added to cell cultures at 50 μm final concentration.

Back to Top | Article Outline

T-cell depletions and interferon gamma ELISpot assay

PBMCs were depleted of T cells using CD4, CD8 or CD25 Dynal beads (Invitrogen, Carlsbad, California, USA) according to the manufacturer's instructions. Cryopreserved peripheral blood mononuclear cells (PBMCs) were thawed and left to settle overnight or freshly isolated PBMCs were used in the T-cell depletion assays. The T-cell depleted fractions were counted prior to use in the ELISpot assay. Specific CD4+, CD8+ or CD25high T-cell depletions were confirmed by flow cytometry (Gallios Flow Cytometer; Beckman Coulter, Gladesvile, Australia).

The IFNγ ELISpot assay was performed in triplicate as described previously using NVP and 12-OH NVP at final concentrations of 40 and 50 μm, respectively [22]. A positive response was defined as greater than 50 spots/million cells after background removal [23].

Back to Top | Article Outline

Intracellular cytokine staining

Intracellular cytokine staining (ICS) was set up at 2 million cells/ml with either 40 μm NVP or 50 μm 12-OH-NVP, and positive (phytohemaglutinin) and negative controls (cells alone). Brefeldin A (20 μg/ml) was added after 2 h. The reaction was stopped after 6 h. Cells were then washed and 10 μl each of CD4 and CD8 fluorochrome-conjugated antibodies (CD4-PE and CD8-APC-H7) were added (20 min room temperature). Cells were fixed and permeablised using Intraprep reagents 1 and 2 (Beckman coulter); then, antihuman IFNγ antibody was added (ALEXA Fluor 488) 15 min room temperature. Cells were then washed and analysed on the Gallios flow cytometer. A nine-colour flow antibody panel (CD3-V450, CD4-PE, CD8-PE CF594, CD45RA-APC, CCR7-PECY7, IFNγ-ALEXA Fluor 488 IL-2- PerCP Cy5.5, TNFα AF 700 and MIP-1β-APC-H7) was used to assess the immune response after overnight stimulation with 40 μm NVP and/or unstimulated cells. Staining procedure was similar to that described with the exception of Brefeldin A (10 μg/ml), which was added at set up of the overnight cultures. All antibodies were supplied by BD Pharmingen, In addition, for membrane staining for T-regulatory cell markers (CD4+CD25high CD127low), antibodies (CD3 PerCP Cy5.5, CD4 PE-Cy7, CD8 FITC, CD25 APC and CD127 PE) were added to isolated PBMC and incubated for 20 min at room temperature. Cells were then washed, resuspended in flow buffer (PBS-1% FCS) and analysed on the Gallios flow cytometer (Beckman Coulter, Miami, Florida, USA).

Back to Top | Article Outline

Statistical analysis

Statistical analyses were restricted to individuals of known White or East/Southeast Asian race, with corresponding stratification, because of the known impact of race on genetic associations and the potential ascertainment bias arising from incomplete cohort information. Demographic and clinical/immunological differences between NVP-sensitive and tolerant groups within each racial group were assessed by either a Fisher's exact test (sex, HCV coinfection, undetectable viral load at baseline, treatment experience), or Wilcoxon test (age, CD4+ and CD8+ T cells). Initial genetic analyses that investigated specific HLA alleles by means of a race-stratified Mantel–Haenszel test were confirmed by logistic regression analyses incorporating adjustments for race, NRTI backbone and CD4+ T-cell count at initiation of NVP.

Back to Top | Article Outline


Nineteen patients (13 Whites, six Asians) were identified with NVP HSR (Fig. 1). Reactions to NVP were more prevalent amongst the fully ascertained Asians (P = 0.009, Fisher's exact test), but comparisons of the NVP-sensitive and tolerant patient groups did not show any significant within-race differences in terms of age, sex or clinical and immunological factors (Table 1).

Fig. 1

Fig. 1

Back to Top | Article Outline

Genetic analysis

The association of the alleles HLA-Cw*04, HLA-Cw*08 and HLA-B*35:01 with NVP HSRs move in the same direction in both Asians and Whites (Table 2, Table S2, The allele HLA-DRB1*01:01 did show differences in its association with NVP-induced hepatotoxicity between Whites and Southeast Asians. The carriage of HLA DRB1*01 : 01 was associated with hepatic symptoms of NVP HSR (P = 0.02), particularly for Whites commencing therapy with CD4+ T-cell counts at least 25% [odds ratio OR) 10.2, P = 0.005, Fisher's exact test], but notably lacking amongst Whites with lower baseline CD4+ T-cell counts (P > 0.9), and amongst Asians (Table 2). The presence of HLA-B*35 : 01 was significantly associated with NVP HSR (P = 0.005) and specifically with NVP HSR with cutaneous symptoms (P = 0.001). The HLA-Cw*04 allele was also associated with NVP HSR with rash (P = 0.01). The association of HLA-B*35 with cutaneous phenotype NVP HSR is similar to observations in Thai populations with HLA-B*35 : 05, although our observation specifically associating HLA-B*35 : 01 with cutaneous phenotype HSR across White and Southeast Asian populations is new. HLA-Cw*08 was associated with multisystemic reactions including eosinophilia (P = 0.02).

Table 2

Table 2

Back to Top | Article Outline

Cellular responses in genetically susceptible individuals

Cellular experiments with available PBMC collected from 12 out of 19 NVP HSR patients (median of 14.4 months ranging from 2 days to 144 months after NVP stop) were conducted to evaluate the phenotype/HLA class I and II specific host immune responses.

Back to Top | Article Outline

Interferon-gamma response wanes with time since nevirapine hypersensitivity

NVP-specific responses were detected by IFNγ ELISpot assay in PBMC from four out of 12 patients evaluated. Mean responses for the four patients were 115, 400, 300 and 1600 SFU/million cells in the first PBMC samples evaluated post-NVP HSR for patients 1, 2, 3 and 4, respectively (mean days from reaction 26, range 2–92 days). Patient 1 and patient 2 expressed the B*35 and C*04 allele pairings and experienced the rash associated phenotype. Patient 1 (HLA-A*24 : 07, HLA-A24 : 10, HLA-B*35 : 30, HLA-B*38 : 02, HLA-C*04 : 01, HLA-C*07 : 02, HLA-DR*12 : 01, HLA-DR*15 : 02) experienced fever and cutaneous symptoms of NVP HSR within 7 days and was rechallenged with NVP 1 month later with recurrence of symptoms. NVP-induced IFNγ responses were detected in five longitudinal PBMC samples collected from this patient, after rechallenge, up to 144 days after NVP HSR (Fig. 2a, Table S1, A stored sample available from patient 1 collected 1 year prior to first NVP administration showed a negative IFNγ response to NVP in an ELISpot assay. Patient 2 (HLA-A*11 : 01, HLA-A*24 : 02, HLA-B*13 : 01, HLA-B*35 : 01, HLA-C*03 : 04, HLA-C*04 : 01, HLA-DR*11 : 01, HLA-DR*16 : 01) presented with fever, rash, altered liver function and eosinophilia within 14 days of NVP start and NVP-induced IFNγ responses were observed on day 97 (400 SFU) and had diminished by day 476 after NVP HSR (20 SFU, less than the positive cut off of 50 SFU/million cells) (Table S1, [23]. Patient 2 had experienced trimethoprim-sulfmethoxazole HSR 2–3 months prior to the NVP HSR and showed evidence of elevated and activated CD4+ and CD8+ T cells (32% CD4+ T cells/μl, 31% CD4+HLA-DR+, 37% CD8+HLA-DR+, 71% CD8+CD38+ T cells) in a sample tested 1 month after NVP HSR compared with a sample tested 4 months later (24% CD4+ T cells/μl, 5% CD4+HLA-DR+, 1% CD8+HLA-DR+, 24% CD8+CD38+ T cells).

Fig. 2

Fig. 2

Patient 3 was a 5-month-old infant with Asian ancestry diagnosed with NVP HSR (HLA-A*02 : 06, HLA-A*34 : 01, HLA-B*15 : 21, HLA-B*56 : 01, HLA-C*04 : 03, HLA-C*07 : 01, HLA-DR*04 : 06, HLA-DR*15 : 02), with fever, rash, eosinophilia and hepatitis on day 7 of NVP/abacavir/3TC treatment. NVP-specific IFNγ responses were detected in patient 3 on day 2 (300 SFU) and had diminished by day 62 after NVP HSR (45 SFU, <positive cut-off of 50 SFU/million cells).

Patient 4 (HLA- A*29 : 02, HLA-A*31, HLA-B*14 : 01, -B*44 : 03, HLA-C*08 : 02, HLA-C*16 : 01, HLA-DR*07 : 01) presented with rash, fever and eosinophilia 2 weeks after commencing NVP/abacavir/3TC. IFNγ responses to NVP were examined for patient 4 on days 5, 8, 12, 18 and 49 after NVP HSR (Table S1,, Fig. 2a,, with the highest frequency of responses detected on day 5 and no response by day 18 after HSR. Plasma cytokine profiles showed detectable INFγ release at days 1 and 5 post NVP HSR, which then sharply decreased (Figure S1, Alanine aminotransferase levels rose from 68 U/l 8 days before NVP start to a peak of 2670 U/l by day 3 after NVP stop coinciding near the highest NVP-specific IFNγ responses and then in parallel slowly declined to 29 U/l by day 49 after NVP HSR. Eosinophils peaked when the NVP-specific IFNγ responses had declined to undetectable levels (Fig. 2b).

When IFNγ responses were examined from all four patients, it was clear that NVP-induced T-cell responses decrease over time. The highest frequency of NVP-induced IFNγ responses were observed in PBMC collected from patients 1 and 2 within approximately 90 days of stopping NVP. In contrast, the peak response for patients 3 and 4 was within 1 week of stopping NVP and complete abrogation of NVP-stimulated responses was observed in PBMC collected from patient 4, 18 days after NVP stop (Fig. 2a). The abrogation of the IFNγ responses over time was not observed in samples from patient 1, possibly because he was re-challenged with NVP 1 month after first being administered the drug and this may have provided a potent stimulus to the memory T-cell pool and ensured their prolonged survival. Indeed, an increased NVP-stimulated IFNγ response of 870 SFU/million cells was detected in freshly isolated PBMC obtained from patient 1, 144 days after his initial reaction (4 months after rechallenge) compared with 115 SFU/million cells detected in a sample collected from this patient 6 days after first commencing NVP. Similarly, a prolonged response was seen in patient 2, 97 days after NVP HSR, and this patient had experienced sulfmethoxazole HSR only 2 months before being exposed to NVP. T-cell activation markers were elevated in patient 2 one month post-NVP HSR. Blocking of inhibitory signals during immune activation may have contributed to the persistence of NVP-specific T cells at day 97 post-HSR. T-cell activation markers decreased by 5 months post NVP HSR and NVP-specific T cells were not detected in this patient when tested at a later time point 476 day post-NVP HSR. For the eight patients with no detectable NVP stimulated IFNγ responses, PBMCs had been collected at least 11 months from the NVP HSR reaction (median 2 years, 11 months, range 11 months–12 years), consistent with the abrogation of NVP-induced IFNγ responses over time. The 12-OH-NVP metabolite did not illicit significant IFNγ responses in three of the four patients tested at time points when NVP responses were detected (patient 1, 50 SFU, day 144 after re-challenge; patient 2, 0 SFU day 97; patient 3, 0 SFU days 2 postNVP HSR).

Back to Top | Article Outline

Regulatory T-cell numbers increase in the recovery phase of nevirapine hypersensitivity

High frequencies of regulatory T cells have been reported following drug reaction with eosinophilia with systemic symptoms (DRESS) [24]; thus, we hypothesized that the abrogation of NVP-induced T-cell responses was due to expansion of CD4+ regulatory T-cell (Tregs) populations. The percentage of Tregs present in the day 5 sample (3.9%) collected from patient 4 was similar to levels of Tregs detected in three samples from two healthy controls (3.6, 4.0. and 5.1% Tregs, respectively) and was also similar to Treg levels detected in an HIV-infected patient controlling their viral load in the absence of medication (4.9%). However, Treg percentages in patient 4 increased to 7.3, 8.2 and 8.0% in samples collected on days 8, 18 and 49 post-NVP HSR, respectively (Fig. 2c) and this corresponded with the decline in the IFNγ response; however, depletion of CD25high cells did not restore the NVP-induced IFNγ responses (data not shown).

Elevated Tregs and relapse of DRESS symptoms may accompany viral reactivation [25,26]; however, viral serology for all NVP HSR patients failed to show any evidence of reactivation (i.e. IgM+) of human herpesvirus (HHV)-1, HHV-2, varicella zoster virus (VZV), cytomegalovirus (CMV), Epstein-Barr virus (EBV) or HHV-6 and whole blood qualitative PCR for HHV-6 was negative in patients 2, 3 and 4 (data not shown). The infant, patient 3, only remained positive for IgG to HHV-6 at age 14 months consistent with early infection with HHV-6 and waning of passive immunity and lack of infection with EBV, CMV, HSV 1/2 and VZV.

Back to Top | Article Outline

CD4+ and CD8+ T cells play dual roles in nevirapine hypersensitivity

T-cell depletion studies showed that both CD4+ and CD8+ T cells contributed to the NVP-specific immune response (Fig. 3 a). In patient 4, CD4+ T-cell depletion marginally reduced the responses detected (21% decrease) when compared with whole PBMC, whereas CD8+ T-cell depletion resulted in a greater reduction (83% decrease compared with whole PBMC) in the frequency of detectable NVP-specific T-cell responses (1600, 1265 and 270 SFU/million cells for whole PBMC, CD4+ depleted and CD8+ T-cell-depleted PBMC, respectively), data suggesting that CD8+ T cells may play a larger role in the NVP-induced immune response in patient 4. CD4+ and CD8+ T-cell depletion in cells from patient 1 showed a complete abrogation of the NVP-specific responses when either CD4+ or CD8+ cells were depleted (Fig. 3 a). The presence of cytokine-producing CD8+ and CD4+ T cells was confirmed and investigated further by ICS (patients 3 and 4). The frequency of IFNγ-producing cells was low but detectable by CD4+ and CD8+ T cells in NVP-stimulated PBMC (Fig. 3 b).

Fig. 3

Fig. 3

Fig. 3

Fig. 3

The phenotype of the NVP-stimulated CD4+ and CD8+ T-cell responses was further investigated by examining expression of CCR7 and CD45Ra in PBMC from patient 4 confirming that the active T-cell populations were central memory cells (Fig. 3c). In addition, NVP-induced tumour necrosis factor-alpha (TNFα)and IL-2 production was detected. TNFα was detected in CD4+ and CD8+ T cells; however, IL-2 was detected predominantly in central memory CD4+ T cells. IFNγ-producing cells were primarily single cytokine producing cells, although TNFα-producing cells were also IL-2 producing (Fig. 3d). We postulate that NVP-stimulated CD4+ T-cell production of IL-2 and TNFα induced the expansion of effector CD4+ and CD8+ T cells to produce IFNγ and TNFα.

Back to Top | Article Outline


Contrary to early animal and human studies supporting an HLA class II CD4+ T-cell-mediated model for NVP HSR, our work supports the role of both CD8+ and CD4+ T-cell mediated responses in the immunopathogenesis of NVP HSR. A role for MHC Class I specific CD8+ T cells is supported by the association of HLA-B*35 : 01 and HLA-Cw*04, in particular with cutaneous phenotype NVP HSR, the association of HLA-Cw*08 with multisystemic NVP HSR and eosinophilia, also the detection of IFNγ production by CD8+ T cells in an NVP HSR infant (patient 3) carrying HLA-C*04 and in an adult male patient with the B*14 : 01/Cw8 haplotype (patient 4). The significant reduction of NVP-specific responses observed with CD8+ T-cell depletion in patient 1 and patient 4 (B*35/C*04 and B*14/Cw*08) supports the contribution of the class I CD8+ T-cell restriction. A dual role for MHC Class II specific CD4+ T cells in NVP HSR has also been demonstrated through confirmation of the association of HLA-DRB1*01 : 01 with NVP HSR featuring hepatitis and the abrogation of, or decreased IFNγ production by the PBMC pool in response to NVP in CD4+ T-cell depletion studies.

It is still unclear whether the HLA-B*35 and HLA-Cw*04 associations with NVP HSR featuring rash are independent allele or haplotype effects and whether associations between HLA alleles and NVP HSR may be confounded due to linkage disequilibrium with other genes on chromosome 6. In our cohort, the numbers are too small to perform linkage analysis between the two markers; however, the six out of seven individuals who carried the HLA-Cw*04 allele also carried HLA-B*35 : 01. Some studies have identified HLA-Cw*04 as significant but not examined HLA-B*35 [15,19], although others have implicated both alleles in Asian populations [20]. Individual allele analysis and a genome-wide association study (GWAS) in relation to NVP-induced rash in a Thai cohort showed that HLA-B*35 : 05 and HLA-Cw*04 were higher in the HSR group; however, only HLA-B*35 : 05 remained significant with correction [18,19]. Another group more recently reported an association between NVP Stevens-Johnson Syndrome/toxic epidermal necrolysis (SJS/TEN) and HLA-C*04 : 01 in a Malawian population; however, SJS/TEN is a rare reaction with a differing immunopathogenesis that was not seen in our NVP HSR study [14].

Alleles outside of the MHC may contribute to phenotype-specific adverse reactions observed for NVP. A GWAS and independent replication studies revealed five single nucleotide polymorphisms (SNPs) on chromosome 6 that were significantly associated with NVP-induced rash within a 30 kb region containing the CCHCR1 gene, and two SNPs within the gene significantly associated with rash [27]. Complete screening of the CCHCR1 locus found one nonsynonymous SNP, rs1576, significantly associated with NVP-induced rash. The CCHCR1 locus has previously been associated with psoriasis susceptibility and may be involved in keratinocyte proliferation [28]. Thus, it is likely that this association is specific to the rash phenotype of NVP HSR. Although the CCHCR1 gene is located 110 kb telomeric from the HLA-C locus and 210 kb from the HLA-B locus, logistic regression analysis indicated that the association of SNPs in the CCHCR1 gene with NVP-induced rash was independent from that of the HLA-B*3505 allele [27]. This study raises the issue that other genetic associations may also contribute to phenotype-specific NVP HSR.

Polymorphism in drug-metabolizing enzymes and in particular CYP2B6 516G → T and 983T → C have been shown to be associated with NVP exposure [29–31]. Some who have not considered HLA have not shown a direct relationship between oral clearance and development of NVP HSR [29], although others have shown that cutaneous adverse events have been associated with CYP2B6 516G → T HSR in African Americans carrying HLA-Cw4 [32]. This suggests that the development of class I mediated NVP HSR is dependent on HLA carriage as well as accumulation of the parent drug. In contrast, hepatic symptoms have not been linked to such metabolism and suggest an MHC Class II restricted CD4+ T-cell dependent mechanism [20]. Although NVP has also shown to be a substrate for efflux transporter ABCC10, and genetic variants influence plasma concentrations, the clinical implication is not yet known [33]. Prolonged exposure to any antiretroviral therapy, coinfection with HCV and abnormal baseline levels of alanine aminotransferase place patients at a higher risk of developing hepatotoxicity due to NVP explosure [34,35]; however, HCV and HBV coinfection were excluded in our cohort for individuals with hepatotoxicity.

Animal studies also highlight the duality of the immune response to NVP in hypersensitivity. In the NVP rash model, rats show no evidence of hepatotoxicity and animal T cells produce IFNγ in response to NVP rather than to 12-OH-NVP [36]. Taken together with the responses seen here in systemic NVP-HSR patient PBMCs, the evidence suggests another pathway for induction of the systemic immune response. Keratinocytes can upregulate the expression of MHC Class I/II molecules on antigen-presenting cells and stimulate antigen-experienced T cells; however, the consensus is that they are unlikely to be able to prime new T-cell responses [37]. Thus, it would appear that the mechanism in the NVP-induced skin rash may propagate systemic stimulatory signals for T cells activated by NVP via another pathway.

The rapid waning of ex-vivo responses to NVP supports a delayed-type hypersensitivity model in which drug-induced T-cell activation is mediated by the associated HLA allotypes. Naive T cells are primed on initial exposure to NVP and a memory pool is restimulated on repeat exposure. We suggest that a required threshold of NVP-specific memory cells is necessary in order to stimulate a response ex vivo, or perhaps that a second antigen-specific T-cell activating signal/pro-inflammatory stimulus is required to sustain the memory T-cell pool to NVP. Nondrug-specific T-regulatory responses may play a role in the rapid waning of immune responses. Some studies have suggested reactivation of EBV, HHV-6 and CMV, and corresponding virus specific T-regulatory cells in patients with DRESS [25,26]. We found no evidence of viral reactivation in our patients with NVP HSR. In contrast to the HLA-B*57 : 01 restricted CD8+-dependent abacavir hypersensitivity reaction wherein CD4+ help is not needed, and abacavir-specific IFNγ responses are durable over time, both CD4+ and CD8+ T cells are required for the initiation and maintenance of the immune response to NVP, which appears to wane quite rapidly in most patients.

Work examining T-cell populations in anticonvulsant drug-induced hypersensitivity have observed CD4+ T-cell proliferation and drug-specific CD8+ cytotoxic T-lymphocyte in the acute stages of disease and switching between the dominant cell populations [38]. In one case, drug-specific CD4+ CD25+ Foxp3+ Tregs proliferated during the recovery stage after withdrawal of the drug and a similar effect was observed here for patient 4. However, suppressive activity of Tregs did not explain the wane in response observed after the withdrawal of the drug in cellular assays in this case. Further evaluation in a larger patient group using cell sorting to remove the Tregs may provide further insight.

For the first time, we demonstrate a role for both CD4+ and CD8+ T cells in NVP HSR through both HLA class I and II associations and characterization of cellular responses. Although future studies incorporating larger samples sizes are needed to validate these findings, the examples presented show that precisely phenotyped populations with differing HSR clinical characteristics are important tools to further define the immunopathogenesis of HLA-restricted drug HSR. The complexity of these findings and the low positive predictive value of HLA typing for NVP HSR illustrate the restrictions of implementing such testing prior to NVP prescription [39]. However, this and other studies examining NVP HSR highlight that drug metabolism genotyping, viral serology, precise phenotyping of the HSR reaction and HLA typing can potentially aid in clear diagnosis of the various drug HSR phenotypes for NVP and identify at-risk patients.

Back to Top | Article Outline


The authors would like to thank Professor Jack Utrecht for donating the 12-OH-NVP metabolite used in this study. E.P., R.P. and N.K. drafted the manuscript. B.Mc.K. performed the statistical analysis. D.D. conducted HLA typing. E.P., M.L. and C.B. identified and classified clinical phenotypes of individual patients. N.K., A.L. and C.R. performed the cellular studies. All authors were involved in the conception of the article, interpretation of the results, critical revisions of the article and approved the final version.

E.P. and S.M.'s work was funded in part by National Institutes of Health//National Institute of Allergy and Infectious Diseases and the National Health and Medical Research Council of Australia. A.L., E.Mc.K. and N.K. were funded in part by National Institutes of Health, National Health and Medical Research Council of Australia and the Bill & Melinda Gates Foundation. C.R. is supported by an Australian Postgraduate Award scholarship.

Back to Top | Article Outline

Conflicts of interest

The authors have no conflicts of interest to declare.

Back to Top | Article Outline


1. Gangar M, Arias G, O’Brien JG, Kemper CA. Frequency of cutaneous reactions on rechallenge with nevirapine and delavirdine. Ann Pharmacother 2000; 34:839–842.
2. Pollard RB, Robinson P, Dransfield K. Safety profile of nevirapine, a nonnucleoside reverse transcriptase inhibitor for the treatment of human immunodeficiency virus infection. Clin Ther 1998; 20:1071–1092.
3. Bersoff-Matcha SJ, Miller WC, Aberg JA, van Der Horst C, Hamrick HJ Jr, Powderly WG, et al. Sex differences in nevirapine rash. Clin Infect Dis 2001; 32:124–129.
4. Bonjoch A, Paredes R, Domingo P, Cervantes M, Pedrol E, Ribera E, et al. Long-term safety and efficacy of nevirapine-based approaches in HIV type 1-infected patients. AIDS Res Hum Retroviruses 2006; 22:321–329.
5. Gonzalez de Requena D, Nunez M, Jimenez-Nacher I, Soriano V. Liver toxicity caused by nevirapine. AIDS 2002; 16:290–291.
6. Manfredi R, Calza L. Safety issues about nevirapine administration in HIV-infected pregnant women. J Acquir Immune Defic Syndr 2007; 45:365–368.
7. Peters PJ, Stringer J, McConnell MS, Kiarie J, Ratanasuwan W, Intalapaporn P, et al. Nevirapine-associated hepatotoxicity was not predicted by CD4 count >/=250 cells/muL among women in Zambia, Thailand and Kenya. HIV Med 2010; 11:650–660.
8. Shenton JM, Popovic M, Chen J, Masson MJ, Uetrecht JP. Evidence of an immune-mediated mechanism for an idiosyncratic nevirapine-induced reaction in the female Brown Norway rat. Chem Res Toxicol 2005; 18:1799–1813.
9. Taiwo BO. Nevirapine toxicity. Int J STD AIDS 2006; 17:364–369.quiz 70.
10. Martin AM, Nolan D, James I, Cameron P, Keller J, Moore C, et al. Predisposition to nevirapine hypersensitivity associated with HLA-DRB1*0101 and abrogated by low CD4 T-cell counts. AIDS 2005; 19:97–99.
11. Phillips E, Bartlett JA, Sanne I, Lederman MM, Hinkle J, Rousseau F, et al. Associations between HLA-DRB1*0102, HLA-B*5801, and hepatotoxicity during initiation of nevirapine-containing regimens in South Africa. J Acquir Immune Defic Syndr 2013; 62:e55–e57.
12. Vitezica ZG, Milpied B, Lonjou C, Borot N, Ledger TN, Lefebvre A, et al. HLA-DRB1*01 associated with cutaneous hypersensitivity induced by nevirapine and efavirenz. AIDS 2008; 22:540–541.
13. Yang KL, Chen IH, Hsiao CK, Cherng JM, Yang KZ, Chang CC, et al. Polymorphism of HLA-B27 in Taiwanese Chinese. Tissue Antigens 2004; 63:476–479.
14. Carr DF, Chaponda M, Jorgensen AL, Castro EC, van Oosterhout JJ, Khoo SH, et al. Association of human leukocyte antigen alleles and nevirapine hypersensitivity in a Malawian HIV-infected population. Clin Infect Dis 2013; 56:1330–1339.
15. Gao S, Gui XE, Liang K, Liu Z, Hu J, Dong B. HLA-dependent hypersensitivity reaction to nevirapine in Chinese Han HIV-infected patients. AIDS Res Hum Retroviruses 2012; 28:540–543.
16. Gatanaga H, Yazaki H, Tanuma J, Honda M, Genka I, Teruya K, et al. HLA-Cw8 primarily associated with hypersensitivity to nevirapine. AIDS 2007; 21:264–265.
17. Littera R, Carcassi C, Masala A, Piano P, Serra P, Ortu F, et al. HLA-dependent hypersensitivity to nevirapine in Sardinian HIV patients. AIDS 2006; 20:1621–1626.
18. Chantarangsu S, Mushiroda T, Mahasirimongkol S, Kiertiburanakul S, Sungkanuparph S, Manosuthi W, et al. HLA-B*3505 allele is a strong predictor for nevirapine-induced skin adverse drug reactions in HIV-infected Thai patients. Pharmacogenet Genomics 2009; 19:139–146.
19. Likanonsakul S, Rattanatham T, Feangvad S, Uttayamakul S, Prasithsirikul W, Tunthanathip P, et al. HLA-Cw*04 allele associated with nevirapine-induced rash in HIV-infected Thai patients. AIDS Res Ther 2009; 6:22.
20. Yuan J, Guo S, Hall D, Cammett AM, Jayadev S, Distel M, et al. Toxicogenomics of nevirapine-associated cutaneous and hepatic adverse events among populations of African, Asian, and European descent. AIDS 2011; 25:1271–1280.
21. Mallal S, Nolan D, Witt C, Masel G, Martin AM, Moore C, et al. Association between presence of HLA-B*5701, HLA-DR7, and HLA-DQ3 and hypersensitivity to HIV-1 reverse-transcriptase inhibitor abacavir. Lancet 2002; 359:727–732.
22. Keane NM, Nolan, Phillips EJ, Lucas A, Almeida CA, Martin A, et al. Use of the ELISpot assay to detect abacavir and nevirapine induced IFN-gamma responses in PBMC from HIV-infected patients with abacavir or nevirapine. 4th International AIDS Society Meeting, Sydney, Australia, July 23–25, 2007. Abstract MOPEB007.
23. Keane NM, Roberts SG, Almeida CA, Krishnan T, Chopra A, Demaine E, et al. High-avidity, high-IFNgamma-producing CD8 T-cell responses following immune selection during HIV-1 infection. Immunol Cell Biol 2012; 90:224–234.
24. Hanafusa T, Azukizawa H, Matsumura S, Katayama I. The predominant drug-specific T-cell population may switch from cytotoxic T cells to regulatory T cells during the course of anticonvulsant-induced hypersensitivity. J Dermatol Sci 2012; 65:213–219.
25. Picard D, Janela B, Descamps V, D’Incan M, Courville P, Jacquot S, et al. Drug reaction with eosinophilia and systemic symptoms (DRESS): a multiorgan antiviral T cell response. Sci Transl Med 2010; 2:46ra62.
26. Criado PR, Avancini J, Santi CG, Medrado AT, Rodrigues CE, de Carvalho JF. Drug reaction with eosinophilia and systemic symptoms (DRESS): a complex interaction of drugs, viruses and the immune system. Isr Med Assoc J 2012; 14:577–582.
27. Chantarangsu S, Mushiroda T, Mahasirimongkol S, Kiertiburanakul S, Sungkanuparph S, Manosuthi W, et al. Genome-wide association study identifies variations in 6p21.3 associated with nevirapine-induced rash. Clin Infect Dis 2011; 53:341–348.
28. Asumalahti K, Veal C, Laitinen T, Suomela S, Allen M, Elomaa O, et al. Coding haplotype analysis supports HCR as the putative susceptibility gene for psoriasis at the MHC PSORS1 locus. Hum Mol Genet 2002; 11:589–597.
29. Dickinson L, Chaponda M, Carr DF, van Oosterhout JJ, Kumwenda J, Lalloo DG, et al. Population pharmacokinetic and pharmacogenetic analysis of nevirapine in hypersensitive and tolerant HIV-infected patients from Malawi. Antimicrob Agents Chemother 2014; 58:706–712.
30. Penzak SR, Kabuye G, Mugyenyi P, Mbamanya F, Natarajan V, Alfaro RM, et al. Cytochrome P450 2B6 (CYP2B6) G516T influences nevirapine plasma concentrations in HIV-infected patients in Uganda. HIV Med 2007; 8:86–91.
31. Wyen C, Hendra H, Vogel M, Hoffmann C, Knechten H, Brockmeyer NH, et al. Impact of CYP2B6 983T>C polymorphism on nonnucleoside reverse transcriptase inhibitor plasma concentrations in HIV-infected patients. J Antimicrob Chemother 2008; 61:914–918.
32. Mahungu T, Smith C, Turner F, Egan D, Youle M, Johnson M, et al. Cytochrome P450 2B6 516G-->T is associated with plasma concentrations of nevirapine at both 200 mg twice daily and 400 mg once daily in an ethnically diverse population. HIV Med 2009; 10:310–317.
33. Liptrott NJ, Pushpakom S, Wyen C, Fatkenheuer G, Hoffmann C, Mauss S, et al. Association of ABCC10 polymorphisms with nevirapine plasma concentrations in the German Competence Network for HIV/AIDS. Pharmacogenet Genomics 2012; 22:10–19.
34. Martinez E, Blanco JL, Arnaiz JA, Perez-Cuevas JB, Mocroft A, Cruceta A, et al. Hepatotoxicity in HIV-1-infected patients receiving nevirapine-containing antiretroviral therapy. AIDS 2001; 15:1261–1268.
35. Torti C, Costarelli S, De Silvestri A, Quiros-Roldan E, Lapadula G, Cologni G, et al. Analysis of severe hepatic events associated with nevirapine-containing regimens: CD4+ T-cell count and gender in hepatitis C seropositive and seronegative patients. Drug Saf 2007; 30:1161–1169.
36. Chen X, Harmanathan T, Mannargudi B, Gou H, Uetrecht J. A study of the specificity of lymphocytes in nevirapine-induced skin rash. J Pharmacol Exp Ther 2009; 331:836–841.
37. Gutowska-Owsiak D, Ogg GS. The epidermis as an adjuvant. J Invest Dermatol 2012; 132:940–948.
38. Hung SI, Chung WH, Jee SH, Chen WC, Chang YT, Lee WR, et al. Genetic susceptibility to carbamazepine-induced cutaneous adverse drug reactions. Pharmacogenet Genomics 2006; 16:297–306.
39. Daly AK. Human leukocyte antigen (HLA) pharmacogenomic tests: potential and pitfalls. Curr Drug Metab 2014; 15:196–201.

CD4+; CD8+; drug reaction with eosinophilia with systemic symptoms/drug-induced hypersensitivity syndrome; drug hypersensitivity; HIV; human leukocyte antigen; nevirapine; regulatory T cells

Supplemental Digital Content

Back to Top | Article Outline
© 2014 Lippincott Williams & Wilkins, Inc.