In vitro cytotoxicity as a marker of hypersensitivity to sulphamethoxazole in patients with HIV

In vivo and in vitro testing for adverse drug reactions

The goal of providing an accurate and reliable diagnostic test for the evaluation of adverse drug reactions remains elusive. Rechallenge remains a truly conclusive method for determining the contribution of drugs to adverse effects, but there are significant disadvantages in exposing patients to drugs that may have caused serious adverse effects. Skin tests are useful for the assessment and predictive risk of serious adverse reactions for a small number of drugs, but skin tests are limited in their utility to adverse drug reactions mediated by Type 1 hypersensitivity. Skin tests must be performed with care and rigor as well. There are a number of in vitro assays that have been used for the assessment of adverse drug reactions, but many of these assays are primarily research tools and have little utility in the clinical setting. Although in vitro tests are not yet readily available for the diagnosis of adverse drug reactions, our increasing understanding of the biology of adverse drug reactions and interactions between the immune system and xenobiotics offers promise that more clinically relevant assays can be developed. At the present time, the primary diagnostic tools for the assessment of adverse drug reactions remain a careful history and physical examination and an astute, insightful physician.

Antibiotic-associated hepatitis: update from 1990

OBJECTIVE

To review the literature on the recent available evidence of antibiotic-associated acute liver injury.

DATA SOURCES

All published articles from January 1990 to July 1995 were extracted from the monthly updated HEPATOX database. Additional articles were found using MEDLINE, EMBASE, and PASCAL searches. Hepatic injuries associated with antituberculous, antimycotic, antiviral, antiprotozoal, and antiseptic compounds were excluded form this review.

STUDY SELECTION

As the amount of literature was large, only case reports, series and epidemiologic data were used. Results from clinical trials were reviewed only when other information was available.

DATA EXTRACTION

Original articles were reviewed to select relevant material. Information regarding the clinical description, histologic features, severity, outcome, and possible risk factors was extracted. Data on incidence were provided by epidemiologic studies or spontaneous reporting to regulatory agencies.

DATA SYNTHESIS

Antibiotic-associated acute injury is rare, with an incidence not exceeding 1 case 10,000 users for most drugs. Among beta-lactams, amoxicillin/clavulanic acid and penicillinase-resistant penicillins are associated with predominant and sometimes protracted cholestasis. The hepatotoxic potential of all available erythromycin salts is confirmed, and recent evidence suggests that roxithromycin could be added to the list of antibiotic-induced liver injury. Among fluoroquinolones, only ciprofloxacin has been associated with serious hepatitis. Trimethoprim/sulfamethoxazole-induced hepatitis often reported, but trimethoprim alone also appears as a possible cause of acute liver injury. Finally, acute bile duct injuries and ductopenia have been described with several antibiotics.

CONCLUSIONS

The most important recent information is the possibility of protracted liver cholestasis with bile duct injuries induced by several antibiotics, particularly penicillinase-resistant penicillins, and the identification of new potentially hepatotoxic antibiotics, namely, roxithromycin, ciprofloxacin, and trimethoprim.

Role of viral infections in the induction of adverse drug reactions

A spectrum of adverse drug reactions that are caused by the combined action of drugs and viruses has been described: ampicillin rash in acute infectious mononucleosis; Reye's syndrome; hypersensitivity reactions to sulphonamides in patients with HIV infection; drug-induced agranulocytosis; paracetamol (acetaminophen) hepatotoxicity; aspirin (acetylsalicyclic acid)-induced asthma; Epstein-Barr virus-associated lymphoma and methotrexate; and AIDS-related Kaposi's sarcoma and nitrite use. Changes in pharmacokinetics have been reported for: caffeine, sulfamethoxazole and fluconazole in patients with HIV infection; theophylline, following influenza and influenza vaccination; and recently, dipyrone metabolites in carriers of the hepatitis B virus. In addition increased drug- and drug metabolite-related toxicity has been observed in virally infected cells. Pathogenetic mechanisms for the interaction between drugs and viruses are varied, and include biological mechanisms (often immunological) and changes in drug metabolism. The combined effects of chemical and biological exposure provide a unique model for the study of disease induction.

Mechanisms of drug reactions: the metabolic track

Hypersensitivity syndrome (HSR) describes a drug-induced symptom complex consisting of fever, rash, and internal organ involvement. Although these reactions are rare, they are very important because of their severity and unpredictability. The metabolic conversion of drugs to chemically-reactive products is now established as a prerequisite for many idiosyncratic drug reactions. In the setting of HSR, an imbalance in the rates of formation of reactive metabolites and of enzymatic detoxification can lead to accumulation of these byproducts. Reactive metabolites could act as haptens eliciting an immune response, covalently bind target proteins causing cell death, or interact with nucleic acids leading to mutations. The lymphocyte toxicity assay (LTA) provides an in vitro assessment of host susceptibility to reactive metabolites of a given drug. It has validated the clinical finding of increased risk of HSR in first degree relatives of patients. It is hoped that the LTA will be used to predict host susceptibility before drug exposure. Ultimately it is hoped that the genetic defects that lead to drug reactions will be identified. This would improve drug development safety and allow primary prevention of serious reactions.

Idiosyncratic drug reactions. Metabolic bioactivation as a pathogenic mechanism

The metabolism of drugs to chemically reactive metabolites may play a pivotal role in the pathogenesis of idiosyncratic drug toxicity. A large number of in vitro studies and a limited number of in vivo studies have demonstrated that many drugs are not toxic per se, but produce toxicity after undergoing enzyme-mediated bioactivation to chemically reactive species. Such reactive species may inflict a toxic insult on the cell either directly or indirectly by acting as a hapten and initiating an immune-mediated reaction. The enzymes responsible for bioactivation have been widely studied, both quantitatively and qualitatively, the most important being the enzymes of the cytochrome P450 (CYP) mixed function oxidase system. CYP enzymes are the most predominant drug metabolising enzymes in the liver and are also present in most other tissues of the body. The diversity of this enzyme system means that a wide range of xenobiotic substrates can be bioactivated by either a single CYP isoform or multiple isoforms of this enzyme superfamily. Other enzymes do, however, play an important role in drug bioactivation. In white blood cells, for example, myeloperoxidase has been shown to bioactivate a wide range of drugs. In other tissues low in CYP activity, prostaglandin H synthase may also be responsible for bioactivation; e.g. in the kidney paracetamol (acetaminophen) toxicity is though to result from activation via this enzyme. The phase II or conjugation enzymes may also be important in the ultimate bioactivation of drug molecules. Whilst activation by these enzymes is, to date, apparently confined to chemicals, most drugs are also substrates for these enzymes and bioactivation by them must remain a possibility.

Hypersensitivity anaphylactoid reaction to pefloxacin in a patient with AIDS

OBJECTIVE

To report a life-threatening anaphylactoid reaction to oral pefloxacin in a patient with AIDS and to review the pertinent literature.

CASE SUMMARY

A 32-year-old woman with AIDS developed an anaphylactoid reaction following a second exposure to oral pefloxacin. This reaction was characterized by severe hypotension, dizziness, itching, and fever.

DISCUSSION

Fluoroquinolones are broad-spectrum antimicrobial agents. They are used frequently in patients with AIDS for numerous indications, including treatment of Mycobacterium avium complex. Pefloxacin, a broad-spectrum fluoroquinolone, was introduced in France in 1985. Since then, many patients with AIDS have been treated with this drug. Several cases of anaphylactoid reactions to ciprofloxacin have been documented in patients with HIV infection. To our knowledge, this is the first reported case of an anaphylactoid reaction to pefloxacin in a patient with AIDS.

CONCLUSIONS

There is a need for continued vigilance in the reporting of adverse drug reactions in patients with AIDS, especially with new drug. Also, care must be taken in introducing drugs, including fluoroquinolones, to this patient population.

Immune response to sulfamethoxazole in patients with AIDS

Antibody- and cell-mediated responses to sulfamethoxazole (SMX) were analyzed in AIDS patients with or without a history of hypersensitivity and in negative controls. In 20 of 20 (P < 0.01) human immunodeficiency virus (HIV)-seropositive patients with skin reactions to cotrimoxazole, we found SMX-specific antibodies, while only 9 of 20 and 17 of 20 HIV-seropositive patients without a history of hypersensitivity to cotrimoxazole had SMX-specific immunoglobulin M (IgM) and IgG, respectively. The levels of specific IgM and IgG were higher in patients with skin reactions than in patients without reactions (IgM, 1.0 +/- 0.19 versus 0.47 +/- 0.23 [P < 0.001]; IgG, 0.68 +/- 0.15 versus 0.47 +/- 0.14 [P < 0.001] [mean optical density values +/- standard deviations]). Seronegative controls with no history of exposure to sulfa compounds did not have SMX-specific IgG or IgM antibodies, and controls with a history of intake of SMX with or without reactions had low levels of IgG and IgM. The SMX-specific IgG subclasses were exclusively IgG1 and IgG3. None of the patients had detectable SMX-specific IgE or IgA antibodies nor did they exhibit a cell-mediated response as measured by a lymphocyte proliferation assay. Antibodies to SMX recognized N-acetyl-sulfonamide, N-(2-thiazolyl)-sulfanilamide, sulfadiazine, and sulfisoxazole but did not recognize sulfanilamide or 3-amino-5-methyl isoxazole in an inhibition assay. It is not known whether the SMX-specific antibodies associated with hypersensitivity reactions to SMX in HIV-seropositive patients have a pathogenic role in these reactions. Sulfanilamide or 3-amino-5-methyl isoxazole, on the other hand, could be potential alternative therapies in HIV-seropositive patients with a history of skin reactions to SMX.

In vitro analysis of metabolic predisposition to drug hypersensitivity reactions

Idiosyncratic hypersensitivity reactions may account for up to 25% of all adverse reactions, and pose a constant problem to physicians because of their unpredictable nature, potentially fatal outcome and resemblance to other disease processes. Current understanding of how drug allergy arises is based largely on the hapten hypothesis: since most drugs are not chemically reactive per se, they must be activated metabolically to reactive species which may become immunogenic through interactions with cellular macromolecules. The role of drug metabolism is thus pivotal to the hapten hypothesis both in activation of the parent compound and detoxification of the reactive species. Although conjugation reactions may occasionally produce potential immunogens (for example, the generation of acylglucuronides from non-steroidal anti-inflammatory drugs such as diclofenac), bioactivation is catalysed most frequently by cytochrome P450 (P450) enzymes. The multifactorial nature of hypersensitivity reactions, particularly the role of often unidentified, reactive drug metabolites in antigen generation, has hampered the routine diagnosis of these disorders by classical immunological methods designed to detect circulating antibodies or sensitized T cells. Similarly, species differences in drug metabolism and immune system regulation have largely precluded the establishment of appropriate animal models with which to examine the immunopathological mechanisms of these toxicities. However, the combined use of in vitro toxicity assays incorporating human tissues and in vivo phenotyping (or, ultimately, in vitro genotyping) methods for drug detoxification pathways may provide the metabolic basis for hypersensitivity reactions to several drugs. This brief review highlights recent efforts to unravel the bases for hypersensitivity reactions to these therapeutic agents (which include anticonvulsants and sulphonamides) using drug metabolism and immunochemical approaches. In particular, examples are provided which illustrate breakthroughs in the identification of the chemical nature of the reactive metabolites which become bound to cellular macromolecules, the enzyme systems responsible for their generation and (possibly) detoxification, and the target proteins implicated in the subsequent immune response.

Mechanisms of unpredictable adverse drug reactions

Adverse drug reactions are common problems associated with therapy, and are major sources of morbidity and mortality. There are numerous types of drug reactions, including predictable drug reactions such as side effects, toxicity, drug interactions and secondary effects that can be anticipated when planning therapy. There are also a number of unpredicted adverse effects, which are unexpected consequences of therapy. The least severe unpredicted adverse drug reaction is intolerance, which appears to be an exaggeration of pharmacological or toxic effects of the drug among vulnerable subsets of patients. Some of the most severe and life-threatening adverse drug reactions are allergic. These adverse effects can be mediated by a number of mechanisms, including the development of drug-specific IgE, serum-sickness-like reactions in response to drug-antibody complexes, direct release of inflammatory mediators, or involvement of the immune system by mechanisms that are poorly understood. Idiosyncratic adverse drug reactions are a heterogeneous group of adverse effects that are not predictable from the pharmacological actions of the drug. Many of these reactions occur as a consequence of pharmacogenetic variations in drug bioactivation and drug or metabolite detoxification or clearance. The physician must be vigilant for the possibility of unpredictable adverse drug reactions during or after therapy. Research currently underway may afford the opportunity to predict, and hopefully prevent, some of these adverse reactions in the future.

Immunohistological assessment of cutaneous drug hypersensitivity in patients with HIV infection

The pathogenesis of drug hypersensitivity in patients with HIV infection is unknown. To study further the nature of hypersensitivity, the histopathological features of morbilliform drug hypersensitivity reactions were examined in a group of HIV-infected patients. Skin sections from 23 HIV-infected subjects with morbilliform drug hypersensitivity reactions were examined by light microscopy, direct immunofluorescence and immunohistochemistry, to determine the nature of the inflammatory infiltrate and the role of immunoglobulin, complement and cytokines. The principal light microscopic findings were spongiosis, hydropic generation of the basal layer, civatte bodies, an epidermal lymphocytic infiltrate (48%), and a perivascular dermal infiltrate of lymphocytes (87%) and macrophages (52%). Two patients had findings consistent with toxic epidermal necrolysis. Immunohistochemistry demonstrated that the lymphocytic infiltrate consisted of CD8+, HLA-DR+ T lymphocytes (some of which also stained for CD38), a marked depletion of epidermal Langerhans cells (90%), and strong cytoplasmic staining of keratinocytes for IL-6 (60%), IL-1 beta (50%), tumour necrosis factor-alpha (TNF-alpha) (45%) and to a lesser degree, interferon-gamma (IFN-gamma) (35%). Immunofluorescence did not demonstrate any significant deposition of immunoglobulin or complement. The histological findings were independent of the responsible drug, the duration of either therapy or the rash, and of peripheral blood CD4+ and CD8+ cell counts. These findings suggest that activated CD8+ lymphocytes and perhaps epidermal production of cytokines are involved in the pathogenesis of cutaneous drug hypersensitivity in HIV-infected patients. The common histological features, regardless of the causative drug, suggest a common pathogenesis.