Azathioprine is one of the most frequently used adjuvants to treat pemphigus [15]. Azathioprine has been suggested to have a steroid-sparing role in multiple case series [24–28], and it is suggested that it is therefore likely to reduce steroid-related complications [29]. However, most of the studies evaluating azathioprine have been case studies, and its use has not been evaluated in randomized controlled trials [30].

The dose of azathioprine is adjusted according to the level of thiopurine methyltransferase (TPMT) [15]. With normal TPMT levels, dosage between 2.0 and 2.5 mg/kg/day is usually applied, whereas 1 mg/kg/day is recommended for patients with reduced TPMT activity [31]. Common adverse events associated with azathioprine include nausea, vomiting, myelosuppression, arthralgia, and pancreatitis [15].

While there are case reports of successful cyclophosphamide therapy, there are also series that suggest it has limited utility as a monotherapy [32] and the most substantive evidence for its benefit is when used in combination with corticosteroids. Cyclophosphamide appears to have a steroid-sparing effect compared with prednisolone alone based on one study with 54 participants [33]. Cyclophosphamide can be administered orally or intravenously (IV). Pulse IV cyclophosphamide has been tried as an alternative treatment but failed to show any benefit [34, 35].

Cyclophosphamide possesses a poor safety profile, with patients at increased risk of sepsis, malignancy (especially bladder), infertility, and myocardial infarction [36]. Early use of alternate adjuvants would benefit patients with pemphigus and prevent the adverse events associated with cyclophosphamide use [3]. Refractory disease can then be treated with cyclophosphamide. It is rarely employed in our centers because of the safety issues.

Mycophenolate mofetil is quite effective as an adjuvant and a corticosteroid-sparing agent for the treatment of pemphigus [3]. Mycophenolate mofetil appears to be equivalent to azathioprine in inducing disease control based on one study with 40 participants [30]. However, mycophenolate mofetil had an inferior steroid-sparing effect compared with azathioprine based on two studies with 92 participants [30, 33, 37]. Mycophenolate mofetil also had an inferior steroid-sparing effect compared with cyclophosphamide based on one study with 54 participants [30, 33].

Mycophenolate mofetil is available in 250 and 500 mg capsules in an oral form to minimize gastrointestinal side effects but can also be administered in an intravenous form. It is typically prescribed as 2–3 g/day in divided doses with reduced dosages in patients with renal impairment [38, 39]. Therapy is initiated at 500 mg daily and increased as tolerated by 500 mg to the desired final dose. Improvement is usually noticed within 2 months of initiation [38]. Cost and availability are limiting factors in its use compared to other immunosuppressive agents [38, 40].

Though mycophenolate mofetil is generally well tolerated [39, 41], adverse effects have been reported. These are predominantly gastrointestinal in nature and include diarrhea, nausea, vomiting, and abdominal discomfort. Such gastrointestinal symptoms are dose dependent and can be reduced by alteration in dose, gradual dosage increments, or taking the tablet with food [38, 40]. Other reported adverse effects include genitourinary symptoms (frequency, urgency, dysuria), hematological abnormalities (<5 % of patients) [40], and opportunistic infections secondary to immunosuppression [42].

Mycophenolate mofetil interacts with multiple medications. Concomitant use with some antibiotics, immunosuppressive agents, and antacids results in a decrease in mycophenolate levels and inadequate immunosuppressive effect. Salicylates and probenecid cause an increase in mycophenolate level and hence a higher immunosuppressive effect. These effects need to be corrected with dosage alterations of mycophenolate mofetil [38, 40].

There is no consensus or guideline for dosing or laboratory monitoring of mycophenolate in dermatological conditions. Complete blood count with differential liver and renal function tests should be monitored fortnightly initially, then monthly for 3 months, and then every 3 months once on a stable dose [38]. Anemia, leucopenia, neutropenia, and thrombocytopenia have been reported with mycophenolate mofetil [43, 44].

Methotrexate is another corticosteroid-sparing agent commonly used in pemphigus [7, 45–48], but no controlled trials have yet been conducted investigating the efficacy of methotrexate in pemphigus. Though there are concerns about the adverse effects of methotrexate, doses up to 25 mg/week for symptom control are relatively safe. Larger doses can be associated with adverse effects. Nausea and infection, including pneumonia and activation of tuberculosis, were the most common side effects reported on the use of methotrexate in the treatment of pemphigus [49]. There is insufficient evidence to draw conclusions regarding the role of adjuvant methotrexate in pemphigus [29].

Cyclosporine has been evaluated in two studies, including comparisons with glucocorticoids alone and cyclophosphamide, but these studies have failed to demonstrate any advantage of cyclosporine as an adjuvant therapy for pemphigus [50, 51]. Moreover, it is possible that adverse events are more frequent in the setting of combined glucocorticoid and cyclosporine therapy [29]. It is suggested to use cyclosporine only after other agents have been tried [3].

Dapsone has been evaluated in one randomized controlled trial with 19 participants [52]. In this study, the authors presented some trends favoring the use of dapsone, but statistical significance was never reached. In a meta-analysis study which summarized 55 identified pemphigus patients who responded to dapsone, hemolysis was the most common adverse reaction observed [53]. In all patients, glucose-6-phosphate dehydrogenase deficiency must be excluded [15]. Dapsone appears to benefit some pemphigus patients in the maintenance phase of their disease; however, as there is little evidence to recommend the use of dapsone in pemphigus, it is difficult to evaluate its utility [29].

The evidence for the use of tetracyclines and nicotinamide in pemphigus is derived from small case series [54–56]. However, these studies present inconsistent evidence regarding the use of antibiotics and nicotinamide in pemphigus. They could perhaps be considered as adjuvant treatments in mild cases, but further research including controlled trials is warranted to elicit their utility in this setting [29].

Rituximab is a chimeric monoclonal antibody directed against the antigen CD20, which is present on the surface of pre-B cells and mature B cells. Rituximab destroys the CD20-positive B cells through a combination of complement, antibody, and cytotoxic activities [57]. The CD20 antigen is present on the surface of B cells but is not expressed by plasma cells, stem cells, and B-cell progenitors [58]. The highly selective nature of this murine-human monoclonal antibody means there is less chance of adverse effects. However, because of the close interaction between B and T cells, selective B-cell depletion indirectly results in reduction of a subset of T lymphocytes [59, 60]. Autoantibody levels are expected to decrease with clinical improvement [60], but a clinical response to treatment is not always reflected by changes in autoantibody levels [61]. Some patients who have achieved remission of their mucosal lesions continue to have elevated titres of anti-desmoglein 3 antibodies. On the contrary, changes in anti-Dsg1 antibodies are well correlated with the evolution of skin lesions, both in PF patients and in PV patients with mucosal and skin involvement.

So far, no prospective controlled clinical trial on rituximab treatment for pemphigus has been published [15], but Schmidt et al. [62] looked at a total of 136 cases of pemphigus (pemphigus vulgaris 103, pemphigus foliaceus 20, paraneoplastic pemphigus 13) treated with rituximab reported in the literature. About 95 % of them had at least partial remission following rituximab therapy with complete remission in 40 %. Clinical remission, defined as healing of all lesions but requiring further immunosuppression, occurred in 37 and 45 % of pemphigus vulgaris and pemphigus foliaceus cases, respectively. Only 12 % of the total patients had a relapse, and 3 % had worsening disease after treatment with rituximab.

There is no generally accepted protocol for the treatment of pemphigus with rituximab [15], but the typical dose of rituximab is 375 mg/m² weekly for 4 weeks [63]. Though some reports suggest the need for multiple cycles of rituximab [64, 65], a recent study demonstrated remission after a single cycle [66]. Rituximab is contraindicated in patients who are pregnant or breastfeeding, hepatitis B/C, HIV, and sepsis [14, 67]. An increased risk of infections [68] and fatality has been reported [69]. Premedication with paracetamol and prednisone may reduce the likelihood of infusion-related adverse effects such as hypotension, coryzal symptoms, nausea, and headache.

Because of the increased risk of infection, rituximab is recommended in refractory pemphigus, patients in whom it is impossible to decrease systemic corticosteroids due to relapse, and those who are unable to tolerate or experience complications with immunosuppressants [3]. The role of early treatment with rituximab relative to conventional therapies (e.g., immunosuppressants, IVIg, or systemic corticosteroids) needs further study.

Plasma exchange is a procedure in which plasma (usually large amounts) is removed and replaced with another fluid which is often albumin or fresh frozen plasma. It is a complex procedure due to the volume of exchange, disturbance of homeostasis, and potential electrolyte, volume, and pressure imbalances [70]. Plasmapheresis is a procedure in which plasma is removed (and not replaced) and is therefore a less complex procedure.

The role of removing plasma is to also remove the circulating antibodies from the body. The rationale behind plasmapheresis is based on the observation that there is a correlation between the levels of circulating autoantibodies and disease activity [71]. By removing plasma from the body, circulating antibodies are also removed, hence reducing the number of circulating antibodies and improving disease activity. But removal of the antibodies does not mean that production of antibodies is diminished. In fact, pathogenic B lymphocytes increase the production of antibodies in the 2 weeks following plasmapheresis to compensate for the loss of circulating antibodies. As a result, circulating levels of autoantibodies can be as high as or even higher than before the procedure. This rebound increase in antibody level can be minimized by administering immunosuppressive agents for a period of time prior to plasmapheresis [71, 72]. Despite this rebound effect, compared with other therapies, plasmapheresis results in lower long-term pemphigus antibody levels [73].

Though small trials have demonstrated the efficacy of plasmapheresis in obtaining partial or complete remission [9, 71], a randomized controlled trial with 40 participants suggested plasma exchange with low-dose steroids is ineffective as a therapy for pemphigus [74]. Plasmapheresis may be considered for rapid control of severe or recalcitrant pemphigus as should the use of concomitant immunosuppression [29]. In clinical practice, this treatment is very rarely used because of increased risk of infections and having been superseded by rituximab and IVIg.