HIV predisposes patients to drug hypersensitivity reactions, with an estimated 100-fold increase in the risk of drug rashes compared with the general population.

ADRs are the 4th to 6th leading causes of death in the developed world. In a meta-analysis of inpatient ADR prospective studies, 15.1 % of patients sustained ADRs during their hospitalizations, 6.7 % experienced serious ADRs, and 0.32 % fatal ADRs. ADRs result in death in 0.1 % of medical and 0.01 % surgical inpatients, adversely affect surviving patients’ quality-of-life (QOL), and cause patients to lose confidence in their providers. They also often mimic other diseases, resulting in unnecessary investigations and delays in treatment.

Cutaneous adverse drug reactions (cADRs) are the most frequent ADRs comprising 10–30 % of all ADRs. cADRs account for 1 % of outpatient antibiotic prescription and 1–3 % of inpatient admissions. These range from mildly discomforting to life-threatening. Prior to the recognition of HIV infection and AIDS in the early 1980s and continuing into the HIV era cADRs were/are most commonly associated with anti-infective and anticonvulsant drugs, especially the aromatic anticonvulsants (phenytoin, phenobarbital, carbemazepam) but in the AIDS era have been joined by several ARV agents (Tables 37.1 and 37.2).

Adverse drug reactions can be classified by mechanism of the reactions (or by their clinical manifestations (Tables 37.3 and 37.4).

The most common cADRs are mild to moderate maculopapular (morbilliform) eruptions without systemic/other organ system involvement. These can frequently be managed without drug discontinuation or with later drug reintroduction (graded challenge/test dosing/direct rechallenge) while tolerance induction procedures such as dose escalation and true desensitization are sometimes indicated and performed. Patients with more severe, potentially life-threatening, systemic drug hypersensitivity reactions (with multi-system/multi-organ involvement) as in DIHS/DRESS and those with SJS/TEN should never undergo graded challenge or direct rechallenge and rarely, if ever, should tolerance induction procedures be attempted.

Risk factors for cADRs include being on multiple medications (polypharmacy), HIV infection, other viral infections (Epstein Barr virus [EBV]/cytomegalovirus [CMV] mono, human herpesvirus 6 [HHV-6], parvovirus B19, viral hepatitis viruses [in some but not all studies]), female sex, and age (especially age <3 years), immune status, and the presence of various genetic polymorphisms.

Most cADRs occurring in those living with HIV are Type B (idiosyncratic) reactions and of these, the vast majority are immunologically mediated Type IV (delayed, T-cell-mediated) hypersensitivity (allergic) reactions, although basically all categories of the immunologic reactions within the Gell and Coombs classification have been described. Clinically, reactions involving T-lymphocytes are characterized by prominent skin findings because the skin is a repository for an enormous number of T-cells.

It is unclear why HIV patients are so prone to adverse drug reactions, but the underlying reasons are likely multifactorial. Mechanisms include altered metabolism affecting the rate of inactivation and behavior of toxic metabolites, enhanced sensitivity of HIV-infected cells to cytotoxicity, immune system dysregulation, and activation including activation of T-cells via increased antigen presentation by major histocompatibility complex molecules or direct activation by parent drugs or metabolites. The “danger signal” hypothesis endorses the idea that HIV infection itself induces cytokine release and inflammatory signals, alerting the immune system to danger and promoting hyperactivity.

Heterogeneity in the reactions seen with individual ARV and OI agents and between agents can be explained by preferential recruitment and activation of various effector cells, distinct and varied T-cell recruitment and cytokine profiles, as well as host genetic predisposition influences.

Further research into the mechanism of hypersensitivity to commonly used medications in HIV infected patients is vital to improving tolerability and adherence to these therapies. Further identification and characterization of specific genetic risk profiles and additional specific risk factors/scores for more common as well as the rare severe cADRs should help to mitigate these burdensome and potentially dangerous reactions.

Adverse cutaneous drug reactions in HIV-infected patients include severe cutaneous adverse drug (SCAR) reactions including DiHS/ DRESS, SJS/TEN, angioedema/anaphylaxis, and acute generalized exanthematous pustulosis (AGEP) (Fig. 37.5) which are potentially life-threatening, and bullous fixed drug eruption [bFDE] which is not; non-SCAR reactions including the very common exanthematous (morbilliform, maculopapular) eruptions (which can frequently be managed without drug discontinuation or with later drug reintroduction), other less common reactions and manifestations, and non-immune ADRs such as lipodystrophy (Fig. 37.6). Prototypical drug reactions involving the skin and/or skin structures in patients with HIV infection are discussed further below.

The diagnosis of cADRs is mostly clinical. The medical history regarding the reaction is extremely important as it helps determine if the patient has had or is having a cADR versus another clinical condition in the differential diagnosis (see Table 37.5), guide choice of diagnostic tests, determine which medicine is the likely culprit drug and determine whether it might be safe to continue or reintroduce the medication. If possible, the original medical record that describes the reaction should be reviewed. Expertise is essential to optimizing the outcomes for our patients with cADRs. Consultation should be sought early and urgently in order to safely and effectively manage both the reaction and the patient’s ARV and other HIV-specific needs.

Components of a drug allergy history are contained within the “RASHES ITCH” mnemonic device:

Additional history details are described below:

With the need to simultaneously suppress HIV as well as prevent and/or treat opportunistic infections, and the multiple comorbidities often encountered in these patients, it is important to anticipate and manage drug–drug interactions and avoid, if possible, initiating at the same time more than one medication with higher chances of a cutaneous drug eruption. For example, if you anticipate the need for Pneumocystis jiroveci (formerly P. carinii) pneumonia (PCP) and/or toxoplasmosis prophylaxis and a neviripine-containing or an abacavir-containing HAART regimen (in an HLA-B*5701 non-screening setting), consider either avoiding TMP/SMX DS (i.e. use either atovaquone or dapsone for PJP prophylaxis with concomitant pyrimethamine if Toxoplasma gondii prophylaxis is also needed [i.e., Toxoplasma IgG antibody positive and CD4 <100]) or staggering these by starting the TMP/SMX DS first and then starting HAART 1–3 months later. Use of both abacavir (in an HLA-B*5701 non-screening setting) and neviripine in the same HAART regimen is not recommended and neither drug is recommended for PEP, sPEP, or PrEP.

The use of fixed-dose combination (FDC) agents, while decreasing pill burden, improving patient quality of life (QOL) and adherence to ARV therapy, can prove to be challenging and problematic in the setting an adverse drug reaction. First the culprit medication is not always clear. HIV suppression requires combination, multi-drug regimens. HIV replication is incredibly dynamic and this, along with differential intracellular half-lives of component drugs, the risk of resistance mutations occurring during “management” of drug reactions is quite real.

Discontinuing a fixed-dose combination ARV drug when the component medications within the FDC-ARV have vastly different half-lives entails a large concern for resistance development to the agent with the longer half-life due to “virtual monotherapy,” which occurs as the shorter half-life component drugs dissipate and are cleared. This is made even more difficult when in a severe reaction stopping the culprit medication early on can be critically important. Stopping one drug at a time or alternatively discontinuing all potential medication causes of a hypersensitivity reaction and reintroduction of one agent at a time is often not tenable, due to resistance concerns even when the current or proposed new regimen is comprised of agents which all have similar half-lives.

Both lipodystrophy and immune reconstitution inflammatory syndrome (IRIS) have cutaneous manifestations and affect the patient’s appearance, which is often of great concern as these appearance-related adverse effects (AEs) can be quite stigma-inducing, and therefore hold the potential to lead to drug discontinuation and/or poor adherence both of which causes loss of viral control, potential for HIV drug resistance, and concomitant clinical and immune system deterioration.

Even prior to the HAART era, nucleoside analog-induced lipodystophy (also called HIV- associated adipose redistribution syndrome [HAARS] and HIV-associated metabolic and morphological abnormality syndrome [HAMMAS]) consisting of increasing visceral adiposity with concomitant facial and extremity fat wasting/lipoatrophy with or without trunchal fat accumulation/lipohypertrophy were seen and described.

Lipoatrophy is considered a consequence of mitochondrial toxicity seen especially with earlier nucleoside analogue reverse transcriptase inhibitors, didanosine (ddI), zalcitabine (ddc), and stavudine (D4T), collectively referred to as “D drugs” and somewhat with zidovudine (ZDV, also initially know as “AZT” for azidothymidine).

Lipohypertrophy and cardiometabolic syndrome/insulin resistance subdivisions of lipodystrophy are more associated with protease inhibitor-based HAART. Carr et al. identified lipodystrophy in 83 % of 113 HIV-infected patients treated with protease inhibitors (PIs) compared with 4 % of controls (HIV-infected, not using PIs). The lipodystrophy occurred after a mean of 21 months of therapy and failed to resolve following cessation of treatment.

Treatment strategies for lipodystrophy are limited and until now have had only a modest impact on those already affected. The incidence of HIV lipodystrophy appears to be declining as a result of the use of newer antiretroviral drugs. Recent studies confirm that stavudine and zidovudine are the nucleoside analogues responsible for most lipoatrophy, but also suggest that different protease inhibitors have opposing effects on lipoatrophy. Antiretroviral regimens excluding stavudine and zidovudine offer substantial protection against lipoatrophy. Established lipoatrophy improves gradually with the cessation of these drugs.

Thiazolidinediones, uridine, and pravastatin may also improve lipoatrophy, although their effects have not been shown to be sustained. Metformin and growth hormone and its analogues are effective in reducing abdominal fat accumulation, although they aggravate lipoatrophy and generally have only transient effects. Tesamorelin, a synthetic growth hormone releasing hormone, which is FDA approved for lipodystrophy was shown after 6 months of treatment to decrease substantially both visceral adipose tissue and hepatic lipid measurements. No medical intervention has been shown to have a sustained and substantial benefit on either lipoatrophy or visceral fat accumulation, so results with this agent at longer follow up are eagerly awaited. Cosmetic surgery can modestly improve facial lipoatrophy.

The immune reconstitution inflammatory syndrome (IRIS) has also been variably referred to as immune recovery disease, immune reconstitution disease, immune reconstitution syndrome, immune restoration disease, immune rebound illness, steroid withdrawal disease, immunorestitution disease, and immune response reaction.

IRIS is usually self-limited, however it may cause significant morbidity and is rarely fatal. IRIS is important as it may lead to HAART interruption/failure, tests/procedures, and require medical or surgical management. Early examples of immune reconstitution inflammatory syndrome targeting disseminated extrapulmonary tuberculosis with angry, hot, swollen, tender, and pointing anterior cervical lymph nodes with dual NRTI ARV clinical therapy and “hepatic IRIS” versus hepatitis B and C from blinded zidovudine plus neviripine versus zidovudine plus placebo clinical trial protocol-based therapy (leading to a temporary international halting of the trial) were seen in the early 1990s by this author.

IRIS is not unique to AIDS and HAART. It is well described as “paradoxical rxns” during tuberculosis treatment in patients without HIV/AIDS. Localized and disseminated infections with mycobacteria other than TB (MOTT) worsening during steroid withdrawal is another example.

IRIS is characterized by clinical features which are not particularly typical of naturally occurring AIDS-related OIs yet the similarities are close enough to sometimes cause confusion especially if IRIS is not initially suspected or considered. IRIS must be differentiated from drug reaction/toxicity, especially DIHS (Table 37.6)—e.g., neviripine HSR, efavirenz HSR, abacavir HSR in HLA-B*5701 negative patients or non-screening settings—OI progression due to drug resistant pathogen, non-adherence with OI prophylactic treatment with a lack of immune system restoration due to nonadherence with HAART or HIV resistance, and other non-immune/non-inflammatory processes.

It is important to understand that IRIS will resolve in most individuals with continued HAART, continued OI coverage, with or without the institution of anti-inflammatory medications (with fairly rapid resolution). However, IRIS causing significant morbidity including that due to perforation, obstructive phenomena, and mass effect, has been described.

The most common IRIS entities described include: mycobacterium avium complex lymphadenitis, pulmonary infiltrates and nodules, paradoxical exacerbation of pulmonary and central nervous system CNS) M. tuberculosis disease (with potential for serious CNS sequelae), IRIS-related dermatomal zoster (varicella zoster infection), Pneumocystis jiroveci (formerly P.carinii) pneumonia/respiratory failure, paradoxical worsening of viral hepatitis (HBV, HCV), progressive multifocal leukoencephalopathy (PML) worsening (including contrast enhancing PML and fatal PML), paradoxical exacerbation of cryptococcal disease (meningitis; pulmonary disease, necrotizing mediastinal lymphadenopathy), cytomegalovirus (CMV) uveitis, optic nerve neovascularization, cutananeous ulcers, and pneumonitis.

The antigenic target in IRIS, when known, is often but not always a component of an opportunistic microorganism or a more common cutaneous viral condition such as genital herpes and warts (human papilloma virus) and includes abscess formation with or without fluctuance and sinus tract formation, nodules, pustules, vesicles, exanthema, plaques, ulcers, verucous lesions, inflamed tender elevation of tattoos, and alopecia universalis.

Other less common variants include: IRIS-induced pre-eclampsia, IRIS-related sarcoidosis (new-onset or recurrent/relapsing), relapsing Guillian Barre syndrome, Reiter’s syndrome, tuberculoid leprosy with “reversal reactions” and autoimmune thyroid disease (AITD).

When IRIS is considered and recognized it is important to continue HAART attempting to “treat through” the IRIS. When recognized one can often defer/avoid diagnostic evaluation procedures and their attendant risks. Exceptions to this strategy include serious, life-threatening manifestions or non-IRIS conditions which are not easily differentiated from IRIS which may be occurring and/or prove difficult to manage and the results of the diagnostic procedure would likely result in a major change in therapy.

Usual IRIS management measures include reassurance, NSAIDs, corticosteroids, and, less commonly, IVIG, thalidomide, and TNF alpha drugs. Surgery is occasionally needed for MAC/Mtb lymphadenopathy that is threatening spontaneous suppuration and drainage/sinus formation, and other situations where there is uncontrolled inflammation in a closed space causing mass effect or other serious sequelae. Continued or renewed treatment directed at the pathogen is indicated when a live antigenic stimulus for IRIS is suspected while specific pathogen directed treatment would not be indicated or helpful (and in fact might be detrimental) when the IRIS is directed against non-viable, dead microbial antigenic products. It is unknown whether treatment which leads to lysis of the organism compared with non-lytic therapies (as in cell wall active bactericidal bacterial meningitis treatment and Jarish-Herxheimer reactions with syphilis treatment) are of importance in OI treatment or prevention in the setting of IRIS.