Use of mycophenolate mofetil for chronic, refractory immune cytopenias in children with autoimmune lymphoproliferative syndrome


Dr Stephen E. Straus, LCID/NIAID, 10 Center Drive, Room 11N228, NIH, Bethesda, MD 20892-1888, USA. E-mail:


Autoimmune lymphoproliferative syndrome (ALPS) is a disorder of apoptosis associated most often with heritable FAS mutations leading to lymphadenopathy, hypersplenism and chronic refractory autoimmune cytopenias. Mycophenolate mofetil (MMF) was used to treat cytopenias in 13 ALPS patients aged 9 months to 17 years from a cohort of 118 children (aged < 18 years) and 82 adults. Twelve responded for a median follow-up of 49 weeks (range 38–240 weeks), defined by maintenance of adequate blood counts and reduction in dosage or cessation of other immunosuppressive agents. This preliminary experience suggests that MMF may spare steroid usage in patients with ALPS-associated cytopenias.

Autoimmune lymphoproliferative syndrome (ALPS) is an inherited disorder of lymphocyte apoptosis that manifests in early childhood with chronic non-malignant lymphadenopathy, hepatosplenomegaly and recurring autoimmune complications, particularly autoimmune haemolytic anaemia (AIHA), immune-mediated thrombocytopenia (ITP) and autoimmune neutropenia (AIN) (Sneller et al, 1997). This lymphocyte dysregulation is shown by an elevated percentage of CD3+CD4CD8TCR αβ+ double-negative T cells (DNT) in the peripheral blood and lymphoid tissues, defective lymphocyte apoptosis in vitro, fairly characteristic histopathology and a significantly elevated lifetime risk for lymphoma (Straus et al, 2001).

Inherited or somatic heterozygous mutations in the tumour necrosis factor receptor superfamily member 6 (TNFRSF6) gene encoding Fas account for ∼ 75% of all cases of ALPS, being classified as Type Ia (Holzelova et al, 2004; Puck & Straus, 2004). Mutations in Fas ligand (FasL), caspase 10 or 8 underlie cases classified as types Ib, IIa or IIb respectively. The roughly one-quarter of cases who meet the clinical and laboratory criteria for ALPS but in whom no causative mutations have been identified are termed as type III.

Nearly half (87 of 200) of the patients we have studied at the National Institutes of Health (NIH) Clinical Centre have undergone splenectomy, typically before their diagnoses of ALPS had been established, to manage severe, refractory cytopenias, of whom five (two adults and three children) have died. Two of them, children aged 10 and 15 years, died of Streptococcus pneumoniae sepsis; while the other three died of opportunistic infections associated with multidrug immunosuppressive therapy for aggressive, infiltrative polyclonal lymphoproliferation. We have been exploring therapeutic options for the cytopenias in ALPS patients that may avoid the need for splenectomy and/or long-term corticosteroid use in growing children.

Mycophenolate mofetil (MMF) is a prodrug of mycophenolic acid, an inhibitor of inosine monophosphate dehydrogenase, and a key enzyme in de novo purine synthesis, especially in proliferating T and B lymphocytes. MMF ameliorates acute rejection of transplanted organs (Jacqz-Aigrain et al, 2000) and graft-versus-host disease in allogeneic bone marrow recipients (Bolwell et al, 2004). It is a second-line immunosuppressive agent for many autoimmune diseases. Recently, MMF has been used in immune-mediated cytopenias (Howard et al, 2002; Hou et al, 2003). Here we describe 13 consecutive ALPS patients with refractory chronic cytopenias in whom MMF enabled the discontinuation of all other immunosuppressive medications including corticosteroids and may have helped some of them avoid splenectomy.

Patients and methods

Study subjects and case definitions

Individuals were evaluated with informed consent under institutional review board-approved research protocols. Among them, 182 met all case criteria for ALPS: chronic lymphadenopathy and/or splenomegaly, defective lymphocyte apoptosis in vitro, and elevated (≥1% or ≥0·018 × 109/l) circulating DNT cells (Sneller et al, 1997). The total study cohort included 200 patients, of whom 125 had ALPS associated with Fas mutations and seven had mutations in other apoptosis pathway genes including FasL and caspases 8 or 10; 50 patients had ALPS type III; and 18 subjects were classified as having an ‘ALPS phenotype’ in manifesting typical historical and clinical features of ALPS and elevated DNT cell numbers, but no in vitro apoptosis defects have been demonstrated (Table I).

Table I.  Demographic and diagnostic features of 13 ALPS patients who received MMF therapy.
Patient number*ALPS diagnosis†Apoptotic mutation in TNFRSF6: cDNA change, amino acid codon changeDNT‡% Apoptosis§Age MMF started (years)AspleniaAssociated cytopeniasPrevious treatmentsDuration of MMF (weeks)Current MMF dose (mg/m2/b.i.d.)Current prednisone dose (mg OD)
  1. ALPS, autoimmune lymphoproliferative syndrome; DNT, double-negative T cells; MMF, mycophenolate mofetil; AIN, autoimmune neutropenia; AIHA, autoimmune haemolytic anaemia [direct antiglobulin (Coombs’) test was positive for both IgG and C3d in all patients]; ITP, idiopathic thrombocytopenic purpura; G-CSF, granulocyte-colony stimulating factor; 6-MP, 6-mercaptopurine; PRBC, packed red blood cells; IVIG, intravenous immunoglobulin; WinRho, anti-D immunoglobulin; PLT, platelet; i.v., intravenous administration; NA, not available.

  2. *Refers to family proband.

  3. †ALPS type I is associated with TNFRSF6 mutation; in ALPS type III, the associated apoptotic gene defect is unidentified; ALPS phenotype refers to patients with the clinical profile and DNT αβ% consistent with ALPS, but with normal Fas-mediated killing in vitro.

  4. ‡CD3+CD4CD8TCRα/β+ percentage and absolute numbers of cells in the lymphocyte gate. The normal range by percentage is 0·1–0·9%, and by absolute numbers is 0·002–0·017 × 109 cells/l.

  5. §By anti-Fas in vitro assay.

  6. ¶MMF stopped at week 240 to assess unexplained headaches.

  7. **Sibling of family proband.

  8. ††MMF stopped after 12 weeks because of lack of response.

30Type Ia1074delT, L278X30·5203567214·0+AIN, AIHA, autoimmune thrombocytopeniaG-CSF, 6-MP, decadron (oral)49555None
38PhenotypeNo mutation detected1·46558928·9+AIHA, pancytopeniaPRBC, IVIG, azathioprine, G-CSF, prednisone240¶NoneNone
43PhenotypeNo mutation detected5·318677NA12·1+AIHAPRBC, prednisone475006 mg
63Type Ia669del15insA, L143 frame shift8·715725113·7AIHA, hypersplenismPRBC, solu-medrol prednisone231338None
80**Type Ia942C → T, R234X13·137922845·2Anaemia, hypersplenismPrednisone75576None
84PhenotypeNo mutation detected1·29343701·4Anaemia, hypersplenismPRBC, prednisone199623None
113Type IIINo mutation detected1·588116117+AIHA, AIN, autoimmune thrombocytopeniaPrednisone, IVIG, G-CSF38600None
118Type IIINo mutation detected1·556153311·0+AIN, autoimmune thrombocytopeniaIVIG, G-CSF, WinRho, azathioprine, prednisone46316None
129PhenotypeNo mutation detected1·745867911·0AIHAPRBC, IVIG, prednisone, decadron (i.v.), PLT transfusions58333None
137Type Ia973A → G, D244G2·41330962·5Pancytopenia, hypersplenismPRBC, prednisone, IVIG decadron (i.v.)4962910 mg
159Type Ia942C → T, R234X4·82299600·8AIHA, hypersplenismPRBC, solu-medrol (i.v.), IVIG12††NoneNone
160Type Ia214T → C, L(–10)P3·52224910·9AIHA, neutropenia, hypersplenismPRBC, prednisone, G-CSF49576None
161Type Ia535 A → T, E98V3·71760492·1AIHA, ITP, hypersplenismPRBC, prednisone47600None

Among these patients, 13 boys (selection of boys was coincidental, as the first 13 patients with ALPS for whom MMF therapy was considered happened to be boys) with refractory cytopenias, and a median age of 10·9 years (range: 9 months to 17 years), were treated with MMF; seven patients had ALPS type Ia, two had type III, and four had the ALPS phenotype. Nine of these MMF recipients had AIHA associated with positive Coombs’ direct antiglobulin tests and median nadir haemoglobin (Hb) of 6·1 g/dl (range 3·3–10·7 g/dl); nine had ITP with a median nadir platelet count of 17 × 109/l (range1–74 × 109/l). AIN was found in three asplenic patients, and sequestration in grossly enlarged spleens was seen in eight patients with a median nadir ANC of 0·3 × 109/l (range 0–1·14 × 109/l). Many interventions for cytopenias had been attempted prior to instituting MMF therapy for these patients (Table I), including splenectomy in five cases. All patients received both parenteral and oral corticosteroids and/or granulocyte colony-stimulating factor (G-CSF) before adding MMF to their treatment regimen. Hence, the laboratory values immediately prior to commencing MMF, when compared with the nadir counts, reflect the benefits of ongoing, aggressive immunosuppression (Table II).

Table II.  Salient laboratory findings in ALPS patients before and during MMF therapy.
Patient number*Hb (g/dl)‡Platelet count (×109/l)§ANC (×109/l)¶Lymphocyte phenotype (cells × 109/l)**Serum IgG (g/l)††
CD3+ T cellsB cells (CD20+)
  1. ALPS, autoimmune lymphoproliferative syndrome; ANC, absolute neutrophil count; MMF, mycophenolate mofetil.

  2. *Refers to family proband.

  3. †Sibling of family proband.

  4. ‡Hb normal range for males 0–6 years = 11·0–15·0 g/dl, 7–14 years = 11·5–15·5 g/dl, >15 years = 12·7–16·7 g/dl. Nadir, lowest documented lab value ≤1 year prior to MMF initiation.

  5. §Platelet count normal range for males = 154–345 × 109/l.

  6. ¶ Normal range for males = 1·32–7·50 × 109/l

  7. **Peripheral blood lymphocyte phenotype by surface markers (absolute number × 109/l). Normal range at the NIH Clinical Centre for CD3+T cells = 0·65–2·108 × 109/l; for CD20+ cells = 0·049–0·424 × 109/l.

  8. ††Serum quantitative IgG immunoglobulin normal range at the NIH Clinical Centre = 6·39–13·49 g/l.


Treatment schedule

The MMF was administered orally twice daily at ∼ 600 mg/m2 per dose (range 290–774 mg/m2) once the cytopenias had been stabilized with corticosteroids (2 mg/kg) or G-CSF.


Responses to MMF

Adverse events were recorded, and treatment was monitored with periodic laboratory profiles (Table II). Twelve MMF recipients maintained adequate Hb (>10 g/dl), ANC (>1 × 109/l), and platelets (>75 × 109/l) without need for splenectomy (in eight of 13), intravenous immunoglobulin (IVIG), G-CSF, or additional immunosuppressive drugs. They were considered to have responded to MMF for a median of 49 weeks (range 38–240 weeks). Eight of 10 patients discontinued prednisone, while two remain on slowly tapering dosages (Table I). Only one of eight patients with persisting splenomegaly required splenectomy since starting MMF (patient 159).

With well-maintained blood counts on full-dose MMF, cytopenias exacerbated in two boys when their MMF doses were reduced. The ANC and platelet counts in patient 38 fell from 3·185 to 0·288 × 109/l and from 426 to 15 × 109/l, respectively, within a month of stopping his MMF. His counts rose and stabilised once full-dose MMF was reinstituted. Similarly, for patient 63 Hb fell from 11·6 g/dl to 8·7 g/dl over 5 months when his MMF dose was halved; it recovered and stabilised once full-dose MMF was resumed.

However, patient 159 failed to respond to MMF. When his Hb fell to 5·0 g/dl and his ANC to 1·0 × 109/l, the MMF was discontinued. Partial splenectomy and rituximab afforded no benefit. Still red cell transfusion-dependent, he underwent a total splenectomy at age 2 years with resolution of his anaemia, neutropenia and thrombocytopenia.

Adverse effects of MMF

Patients 113 and 118 complained of epigastric pain while receiving MMF. Pain resolved in patient 113 after stopping concomitant administration of oral iron. Patient 118 continues on MMF, along with H2 receptor blockers, with symptom relief. Patient 38 stopped MMF therapy after 240 weeks to determine whether it might be contributing to unexplained headaches that developed after a non-specific viral-like illness. His thrombocytopenia recurred, but responded to 4-weekly infusions of rituximab. His headache has resolved.

While on MMF, peripheral blood immunophenotyping showed modest reductions in DNT cell numbers or percentages (P-values 0·06 and 0·16, respectively, by Student's two-tailed paired t-test), as well as in CD3 and CD20 cell counts (P-values 0·18 and 0·09 respectively). However, declines in serum immunoglobulin G (IgG) levels during therapy were significant (P = 0·001). Nonetheless, because these patients – like most with ALPS (Sneller et al, 1997) – had elevated serum IgG levels (mean 17·29 g/l) before MMF, the changes discerned during therapy resulted in mean serum IgG levels of 9·79 g/l (range 3·3–18·6 g/l), still largely normal.

For two reasons, monitoring of all patients treated with MMF also included vigilance for changes in lymph node size and systemic symptoms suggestive of lymphoma. First, ALPS patients with intracellular FAS mutations are at a significantly increased risk for lymphoma than the general population (Straus et al, 2001). Secondly, despite recent reports to the contrary (Birkeland & Hamilton-Dutoit, 2003); there have been concerns that MMF is associated with an increased risk of lymphoma and infections when used in conjunction with other immunosuppressive agents (Mathew, 1998). During follow-up clinical evaluations, we noted no significant increase in lymphoid mass or infections among our MMF recipients.


Refractory cytopenias associated with splenomegaly and/or autoimmunity have been managed by splenectomy for almost 100 years and with corticosteroids for the last 50 years, while biological agents such as IVIG, antiD-IgG (WinRho) and rituximab have been used with some success and side effects (Cooper et al, 2004). Hence, we are exploring MMF for long-term maintenance immunomodulatory therapy in children with ALPS-associated cytopenias. Our initial experience, in which 12 of 13 consecutive ALPS patients with previously refractory cytopenias showed sustained responses to full-dose MMF, suggests it is an effective, convenient, and well-tolerated option for managing autoimmune cytopenias while avoiding splenectomy and/or growth-impairing doses of steroids in children. These preliminary findings are particularly encouraging, given that all five of our asplenic ALPS patients with refractory cytopenias responded to MMF therapy. Chronic ITP patients who relapse after splenectomy are among the most challenging to manage (McMillan & Durette, 2004). Further accrual and follow-up of ALPS patients is under way to define an optimal dose schedule, duration and long-term safety of chronic immunosuppressive therapy with MMF.


We are indebted to the patients for their participation in the NIH ALPS studies and to their referring physicians. We thank Brenda Rae Marshall for editorial assistance.