Azathioprine is a purine analogue prodrug that requires metabolic conversion into active compounds. After oral uptake about 88 % of azathioprine is ingested and 12 % excreted via the gut [10]. Nearly all of the incorporated azathioprine is metabolised, since only 2 % is being excreted, unchanged, in the urine. The highest serum levels can be found approximately 2 h after oral application, and the half-life is approximately 5 h. The first step of azathioprine activation is the nonenzymatic removal of the imidazole ring in erythrocytes [34]. The active metabolite 6-mercaptopurine can be released from erythrocytes and taken up by other cells for further metabolisation. Although most of the therapeutic effects of azathioprine are dependent on the enzymatical modification of 6-mercaptopurine, there is experimental evidence that the imidazole derivatives of azathioprine might also be effective [69].

Three enzymes have been reported to compete for the cleavage of 6-mercaptopurine in the cell (Fig. 47.2) [34]. Thiopurine S-methyltransferase (TPMT) is able to catabolise 6-mercaptopurine into the non-toxic 6-methyl mercaptopurine. The enzyme is active in the duodenum, the liver and in erythrocytes. Xanthine oxidase (XO) produces the non-toxic, inactive metabolite 6-thiouric acid. A lack in one of these enzymes, XO and TPMT, leads to an increased production of metabolites via the hypoxanthine phosphoribosyltransferase (HPRT) pathway. A lack of TPMT activity is caused by genetic mutations, while XO might be blocked by XO inhibitory drugs such as allopurinol, which is one of the most commonly prescribed drugs in Europe and North America. The enzyme HPRT catalyses the addition of ribose 5-phosphate to 6-mercaptopurine and generates 6-thioinosine 5-monophosphate (6-TIMP). This product is processed by TPMT to active methylated metabolites or is phosphorylated to 6-thioinosine triphosphate (6-TITP). 6-TITP is converted to 6-TIMP by inosine triphosphate pyrophosphohydrolase (ITPA). A lack of ITPA, as it is often seen in Asian populations [34, 79], leads to an increase of the toxic 6-TITP and induces corresponding side effects, such as leukopenia, gastrointestinal disturbances or elevated liver function tests (for details concerning side effects see below). 6-TIMP and its active methylated metabolites can inhibit the glutamine phosphoribosylpyrophosphate transferase, the key enzyme of the purine de-novo synthesis. Thereby azathioprine contributes especially to the inhibition of lymphocyte proliferation, which depends on the purine synthesis for proliferation [1].

6-TIMP is converted by the inosine monophosphate dehydrogenase (IMPD) and guanosine monophosphate synthetase (GMPS) into thio-guanosine-monophosphate (TGMP). TGMP can be inactivated by TPMT into methylated metabolites. TGMP escaping methylation is processed by kinases and other enzymes into the active 6-thioguanine nucleotides, which are substrates for incorporation into RNA and DNA. DNA polymerases incorporate 6-thio-GTP with similar efficacy into DNA like unmodified dGTP. 6-thio-GTP is chemically more reactive than canonical nucleotides and undergoes methylation in DNA. Targeting of modified 6-thio-GTP by DNA repair pathways can lead to apoptosis and is responsible for the cytotoxic effects of the compound [15, 23, 34].

In lymphocytes 6-thio-GTP is able to bind to a small GTPase Rac1, thereby modulating costimulation by CD28 in T cells into apoptosis. This seems to be a major pathway for selective T-cell inhibition [76]. In addition, inhibition of Rac1 in endothelial cells exerts an anti-inflammatory effect by modulation of cytokine secretion [48].

It has further been shown that addition of azathioprine to T-cell cultures led to an inhibition of the mRNA expression of TRAIL, TNFRSF7 and alpha-4 integrin [74]. TRAIL is expressed on the surface of cytotoxic T cells and NK cells and can mediate a proliferative signal for these cells [21]. TNFRSF7 is also expressed by lymphocytes and enhances activation [85]. Inhibition of this effect could contribute to the immunosuppressive potential of azathioprine. A reduced expression of alpha-4 integrin could inhibit T cell and monocyte migration. Azathioprine not only affects T cells but also dendritic cells by reducing their proinflammatory cytokine-producing potential and maturation [4].

In 1971, Greaves et al. introduced azathioprine into the treatment of bullous pemphigoid [27] (Table 47.1). Before that, most patients had been treated with systemic corticosteroid monotherapy to prevent recurrent blister formation. Greaves et al. showed that in eight out of ten patients with bullous pemphigoid, no prednisone maintenance therapy was needed for preventing relapses while azathioprine was given at a dose of 2.5 mg/kg body weight/day [27] (Table 47.1). Ahmed et al. first documented a steroid-sparing effect in patients with bullous pemphigoid treated with azathioprine and prednisone in comparison to prednisone alone and found that the treatment time was significantly reduced in the patients who received azathioprine [3].

In a small prospective clinical trial in 1978, Burton et al. examined azathioprine (2.5 mg/kg/day, n = 12) plus prednisone (30–80 mg/day) versus prednisone alone (n = 13). No significant difference between both groups was found concerning the overall disease control [14], but the prednisone-sparing effect was statistically significant. In the azathioprine group, a cumulative average dose of 3,688 mg of prednisone was used over 3 years, while 6,732 mg was used in the prednisone monotherapy group over the same time period. In 1993, Guillaume et al. found no significant difference in disease control between prednisone monotherapy (1 mg/kg/day, n = 31) and azathioprine (100–150 mg/day, n = 36) plus prednisone therapy [28]. The prednisone-sparing effect was not investigated in this study.

In 2007 Beissert et al. compared treatment with methylprednisolone (0.5 mg/kg/day) plus either azathioprine (2 mg/kg/day, n = 38) or mycophenolate mofetil (MMF, 1 g/twice/day, n = 35) in a national randomised trial [8] (Table 47.1). The results showed no significant differences in the primary outcome (i.e. complete healing of skin lesions), but time to remission was more prompt in the azathioprine group (azathioprine 23.8 ± 18.9 days vs. mycophenolate mofetil 42.0 ± 55.9 days, p = ns). Similar corticosteroid doses were used in both groups to control disease. Interestingly, mycophenolate mofetil was significantly less liver toxic compared with azathioprine, which can be of advantage especially in elderly patients. A further retrospective study demonstrated a slightly faster induction of remission in patients treated with azathioprine and prednisone compared with patients treated with dapsone and prednisone [77] (Table 47.1). In summary, the studies could show a steroid-reducing effect of azathioprine and demonstrated effective disease control by azathioprine in most patients with bullous pemphigoid.

The use of azathioprine for the treatment of pemphigus was introduced in 1969, when Krakowski et al. presented the first case report of a woman with pemphigus vulgaris treated successfully with azathioprine [36] (Table 47.2). At the same time, Wolff and Schreiner published a case series of four patients, describing the use of azathioprine as “steroid saving and beneficial” in pemphigus patients [84]. In 1977, Lever et al. published a retrospective analysis of 63 patients with pemphigus [41]. Lever et al. treated patients (n = 12) with prednisone monotherapy and the other patients in this cohort (n = 51) with a combination of either azathioprine, cyclophosphamide or methotrexate and prednisone. In that report, azathioprine had a steroid-sparing effect and controlled disease. Lever’s therapeutic approach is still today known as “Lever’s regime”. Initially, patients receive a high dose of prednisone (up to 2 mg/kg body weight) in combination with azathioprine 2–2.5 mg/kg. After cessation of new blister formation and reepithelialisation of erosions, the prednisone dose is reduced to 40 mg/day, while the azathioprine dose remains unchanged. Further proceedings depend individually upon the clinical development. Normally, the prednisone dose is gradually reduced over a period of several months. In most studies patients with pemphigus foliaceus were not specifically mentioned and might have been included in the pemphigus vulgaris group.

There are only few prospective, randomised trials reported in pemphigus. Beissert et al. examined 38 patients with pemphigus vulgaris or pemphigus foliaceus and found no significant differences between treatment with azathioprine or mycophenolate mofetil—both in combination with methylprednisone—concerning remission of disease and corticosteroid-sparing effects [9]. The azathioprine-treated patients received a median methylprednisolone dose of 8.916 ± 29.844 mg. In the mycophenolate mofetil-treated group, patients received a median of 9.334 ± 13.280 mg methylprednisolone (n = ns). The mean duration of follow-up was 438 days in both groups. However, the time needed to achieve disease control in 50 % of the patients was about 30 days in the azathioprine group compared to 75 days in the mycophenolate mofetil group. After 200 days of treatment, the patients in the mycophenolate mofetil group had a remission rate of 90 %, while those patients that were treated with azathioprine had a remission rate of 75 %. After 600 days, this trend persisted, because 20 % of the pemphigus patients were still not achieving effective control with azathioprine compared with 10 % of the patients using mycophenolate. The recurrence rate was similar in both groups.

In another study with 120 patients analysed, Chams-Davatchi et al. found no significant difference concerning disease remission between azathioprine and mycophenolate mofetil, but the patients treated with azathioprine showed significantly less steroid consumption [19]. They later extended their findings by a double-blind randomised trial comparing prednisolone and azathioprine with prednisolone and placebo and found a 50 % increased remission rate in the azathioprine group. Interestingly, the steroid-sparing effect was less prominent compared with previous studies [20]. Rose et al. compared dexamethasone/cyclophosphamide pulse therapy with oral methylprednisolone/azathioprine therapy in pemphigus [65] and described a tendency in favour of methylprednisolone/azathioprine concerning complete remissions (Table 47.2).

In the subtype of immunoglobulin A (IgA) pemphigus, azathioprine is not recommended in favour of dapsone as first-line treatment [26, 86]. For pemphigus vegetans, azathioprine appears to work in individual cases [53].

In paraneoplastic pemphigus, the therapy of the underlying malignancy is essential. Concomitant treatment with azathioprine and other immunomodulatory drugs has been reported [25]. Lam et al. presented a case of a 77-year-old man with chronic conjunctivitis, acanthosis nigricans and pemphigus-like mucocutaneous lesions [39]. Further examinations revealed an underlying bronchogenic squamous cell carcinoma. While skin lesions resolved with oral prednisone and azathioprine (100 mg/day) therapy, the conjunctivitis and mucous membrane erosions persisted. Verrini et al. described another patient with paraneoplastic pemphigus who showed good response to azathioprine (100 mg/day) but died shortly after initiation of treatment [82].

In summary, the importance of systemic corticosteroids in the treatment of pemphigus is clearly documented. Since the reports show a tendency in favour of azathioprine concerning corticosteroid-sparing effects and no significant differences with regard to disease control, azathioprine can be suggested as first-line therapy in mild-to-moderate cases of pemphigus. Mycophenolate mofetil appears to be a very valid second-line choice. In severe and rapidly progressing cases, dexamethasone/cyclophosphamide pulse therapy could be considered. Other established treatment opportunities include immunoadsorption and rituximab [7].

Since pemphigus is a chronic disease, especially long-term follow-up studies (>3 years) are clearly needed. A Cochrane review concluded from the available study results that the optimal immunomodulatory agent in the treatment of pemphigus is not found. While azathioprine and cyclophosphamide did show advantages concerning the steroid-sparing effect, mycophenolate mofetil showed superior disease control [49].

There are several case reports showing that azathioprine can be effective in cicatricial pemphigoid. These reports demonstrate that azathioprine is able to control disease and prevents progression. However, azathioprine showed no beneficial effect on already existing scars and cicatrising vegetations (Table 47.3).