Mechanisms of ADR-Causing Cytopenias

MECHANISMS OF ADR-CAUSING CYTOPENIAS

A reduction, below the recognised reference range, in the numbers of any cell type in the peripheral blood must be because of either a reduction in the production of that particular cell type by the marrow (myelosup-pression) or a shortened survival of the cell type in the peripheral blood.

MYELOSUPPRESSION

Reduction in marrow output as an ADR may be caused by a reduction in marrow cellularity (hypopla-sia or aplasia, depending on severity). This may glob-ally affect all cell lines [as in aplastic anaemia (AA)] or may selectively affect only one lineage [e.g. pure red cell aplasia (PRCA)]. It may also be caused by interference with normal maturation in a cellular marrow (dysplasia), as in megaloblastic or sideroblas-tic anaemia.

CYTOTOXIC DRUGS

Most cytotoxic drugs cause ‘type A’ myelosup-pressive ADR by interfering with DNA synthesis or producing chemical damage to DNA that inter-feres with its replication. Others attack the mitotic spindle, inhibit protein synthesis or induce cell differentiation (Chabner and Wilson, 1995). Normal cells recover, but it is not surprising that dose-limiting toxicity is seen in the marrow that contains the most mitotically active normal cells in the body.

A rare indirect cause of drug-induced myelosup-pression is the late development of myelodysplasia or leukaemia because of genetic damage from previous exposure to cytotoxic and other drugs (Le Beau et al., 1986), but this is not considered further here.

OTHER DRUGS

Non-cytotoxic drug effects causing acquired marrow failure are more difficult to establish. Theoretical mechanisms include the induction of defects in the haemopoietic stem cells, damage to the stromal microenvironment of the marrow, inhibition of the production or release of haemopoietic growth factors or induction of humoral or cellular immunosuppres-sion of marrow cells (Young and Maciejewski, 1997).

CONSTITUTIONAL RISK FACTORS

Susceptibility to type A reactions varies between individuals because of differences in absorption and metabolism of the drug (pharmacokinetic changes) or differences in target organ sensitivity (Rawlins and Thomas, 1998). Some apparently idiosyncratic type B reactions may actually become more appropriately classified as predictable type A reactions for partic-ular individuals with constitutional risk factors, once mechanisms are elucidated and tests to identify those at risk become available.

The antibiotic chloramphenicol was one of the first drugs for which epidemiological evidence indi-cated a causal association with apparently idiosyn-cratic AA. An early report of the coincidence of this very rare reaction in a pair of identical twins suggested the possibility of genetic susceptibility (Nagao and Mauer, 1969).

The antipsychotic agent clozapine has an epidemio-logically established association with agranulocytosis (Amsler et al., 1977), which is considered further later in this chapter. An apparently increased risk of this complication correlated with human leucocyte anti-gen (HLA) phenotype (Dettling et al., 2001). Anal-ysis of a cohort of patients from the Long Island Jewish Medical Centre in New York (Lieberman et al., 1990) found that the HLA-B38 phenotype had an incidence of 83% in patients with agranulocytosis and 20% in clozapine-treated patients who did not develop the complication. The B-38 phenotype was part of a haplotype more common in the Ashkenazi Jewish population, and the subsequent work identified two different haplotype associations with clozapine-induced agranulocytosis, one in Ashkenazi Jewish patients and one in non-Jewish patients (Corzo et al., 1994). The association of both haplotypes with vari-ants of the heat-shock protein-70 (HSP-70), encoded by loci within the major histocompatibility complex (MHC) region, suggests linkage rather than direct association of the HLA in genetic susceptibility (Corzo et al., 1995).

6-Mercaptopurine (6-MP) is a thiopurine used extensively in the treatment of childhood acute lymphoblastic leukaemia. Azathioprine is a pro-drug of 6-MP in widespread use as an immunosuppres-sive agent in a variety of autoimmune conditions. 6-MP is inactivated by the enzyme thiopurine methyl-transferase (TPMT). Genetically determined varia-tions in TPMT activity were found to be associated with occasional unexpectedly severe myelosuppres-sion associated with 6-MP (Evans et al., 1991) and azathioprine (Lennard, Van Loon and Weinshilboum, 1989). The determination of TPMT activity, either by the measurement of enzyme activity or by the molecular detection of the polymorphisms associated with reduced activity, is feasible and could allow avoidance of drug in deficient patients and logi-cal dose stratification in heterozygotes. A pharma-coeconomic case has been made for this approach before the use of azathioprine in dermatological prac-tice (Jackson, Hall and McLelland, 1997). Polymerase chain reaction-based (PCR–based) techniques for rele-vant genotypic analysis offer an attractive alternative to the performance of radiochemical activity assays in pharmacogenetic screening (Coulthard et al., 2000).

Methotrexate (MTX) is a dihydrofolate reductase inhibitor used extensively as a cytotoxic agent in lymphoid and other malignancies and as an immuno-suppressive agent particularly in inflammatory arthri-tis. Polymorphisms in the methylenetetrahydrofolate (MTHFR) gene have been associated with variation in efficacy and toxicity of MTX in rheumatoid arthritis patients (Urano et al., 2002).

These examples suggest that technologies for predicting the risk of previously apparently completely idiosyncratic reactions may become available for at least some drugs that may help to reduce the incidence of these dangerous complica-tions.

SHORTENED PERIPHERAL BLOOD CELL SURVIVAL

Shortened survival of cells in the peripheral blood by ADR is most commonly mediated by immune destruction. Antibodies to the drug itself, alone or as a hapten in association with cell surface anti-gens or in immune complexes, may initiate effec-tor mechanisms that damage cells. Alternatively autoantibodies may occur because of altered immune regulation. Peripherally destructive immune mech-anisms in ADRs more commonly only affect one cell type but may involve red cells, granulocytes or platelets. A shortened red cell survival (haemoly-sis) may also be mediated by oxidant stress, particu-larly in more susceptible individuals [e.g. those with inherited glucose-6-phosphate dehydrogenase (G6PD) deficiency]. Red cell and platelet survival may both be shortened by endothelial damage causing inap-propriate intravascular plasma coagulation or platelet aggregation in disseminated intravascular coagula-tion (DIC) and thrombotic thrombocytopenic purpura (TTP), respectively.

Table 34.1 lists mechanisms of cytopenias in ADR together with examples of implicated agents.