Sirolimus is effective in relapsed/refractory autoimmune cytopenias: results of a prospective multi-institutional trial

Abstract

Patients with autoimmune multilineage cytopenias are often refractory to standard therapies requiring chronic immunosuppression with medications with limited efficacy and high toxicity. We present data on 30 patients treated on a multicenter prospective clinical trial using sirolimus as monotherapy. All children (N = 12) with autoimmune lymphoproliferative syndrome (ALPS) achieved a durable complete response (CR), including rapid improvement in autoimmune disease, lymphadenopathy, and splenomegaly within 1 to 3 months of starting sirolimus. Double-negative T cells were no longer detectable in most, yet other lymphocyte populations were spared, suggesting a targeted effect of sirolimus. We also treated 12 patients with multilineage cytopenias secondary to common variable immunodeficiency (CVID), Evans syndrome (ES), or systemic lupus erythematosus (SLE), and most achieved a CR (N = 8), although the time to CR was often slower than was seen in ALPS. Six children with single-lineage autoimmune cytopenias were treated and only 2 responded. Sirolimus was well tolerated with very few side effects. All of the responding patients have remained on therapy for over 1 year (median, 2 years; range, 1 to 4.5 years). In summary, sirolimus led to CR and durable responses in a majority of children with refractory multilineage autoimmune cytopenias. The responses seen in ALPS patients were profound, suggesting that sirolimus should be considered as a first-line, steroid-sparing treatment of patients needing chronic therapy. The results in other multilineage autoimmune cytopenia cohorts were encouraging, and sirolimus should be considered in children with SLE, ES, and CVID. This trial was registered at www.clinicaltrials.gov as #NCT00392951.

© 2016 by The American Society of Hematology.

Figures

Figure 1

Sirolimus improves multilineage cytopenias in patients with ALPS. Panels depict changes in (A) hemoglobin, (B) reticulocyte count, (C) platelet count, and (D) ANC. All ALPS patients had multilineage cytopenias. Each line represents a different subject’s cell count at each time point measured. The dotted line depicts normal thresholds for hemoglobin and platelets. Although patients may have started below the normal threshold, all but one of the patients reached a CR across all cell lines measured. ANC, absolute neutrophil count.

Figure 2

Time to CR, to relapse, and on sirolimus, in ALPS and non-ALPS patients. (A) Overall time to achieve a CR on sirolimus in patients with ALPS compared with non-ALPS patients separated by multilineage vs single-lineage autoimmune cytopenias. For patients whose best response was PR or NR, the curve represents time on sirolimus plus 1 year of observation after discontinuation. The time to CR was not significantly different in ALPS (median, 4 weeks; range, 2 to 60) compared with non-ALPS patients with multilineage cytopenias (median, 9 weeks; range, 2 to 260) (P = .135). In contrast, the time to CR was significantly different in ALPS compared with non-ALPS patients with single-lineage cytopenias (P = .0065). Asterisk (*) represents 1 patient with non-ALPS single-lineage disease who achieved a PR and remained on drug for 260 weeks. (B) Relapse incidence in ALPS and non-ALPS patients with multilineage and single-lineage cytopenias, for all patients enrolled. Events include patients who achieved a CR and then relapsed, and patients who discontinued sirolimus for lack of efficacy. Most of the ALPS patients remained disease-free while on sirolimus, including 1 patient who stopped sirolimus after 3 years but maintained a durable CR. The median time to an event in non-ALPS patients was 13 months. Of note, 2 non-ALPS patients with multilineage cytopenias electively stopped sirolimus, had a transient drop in counts, and then were re-challenged with sirolimus, reverting back to a CR. These patients were not counted as having relapsed. (C) Histogram demonstrating the total time on sirolimus separated by ALPS, non-ALPS multilineage vs non-ALPS–single lineage. The different colors reflect whether they achieved a CR (black), PR (gray), or NR (white).

Figure 3

Sirolimus resolves the percentage of DNTs to a normal range in ALPS patients. (A) Patients with ALPS included in the clinical trial. (B) Patients with ALPS included in a biology study, not enrolled in the clinical trial. For both panels, the dotted line represents the normal threshold of circulating DNTs (<2.6%).

Figure 4

Sirolimus improves multilineage cytopenias in non-ALPS patients. Panels depict changes in (A) hemoglobin, (B) reticulocyte count, (C) platelet count, and (D) ANC. Each line represents a different patient’s cell count at each time point measured. The dotted line depicts normal thresholds for hemoglobin and platelets. ANC, absolute neutrophil count.

Figure 5

Immune function is not affected by long-term treatment with monotherapy sirolimus in ALPS patients. (A) ALC significantly improved pre- vs post-sirolimus treatment (P = .039). (B) Quantitative IgG was measured in 6 subjects. Prior to starting sirolimus, the IgG was elevated in 3 children, low in 2, and normal in 1. The IgG normalized in 5 of the 6 children with time (P = .5625). One subject had persistent hypogammaglobulinemia and was diagnosed with comorbid CVID prior to starting sirolimus. Interestingly, the other subject with hypogammaglobulinemia developed it while on chronic MMF, requiring chronic IVIgG replacement for many years. After starting sirolimus, the hypogammaglobulinemia improved. The child was subsequently challenged with vaccines and demonstrated normal antibody response, and was able to discontinue IVIgG supplementation. (C) CD4 and (D) CD8 counts were also studied in 5 ALPS patients pre- and post- long-term administration of sirolimus, with no decrease in number (P = .1875 and P = .3125, respectively). Although not depicted here, 5 ALPS patients had mitogen stimulation measured between 2 and 5 years after initiation of sirolimus, with robust responses to phytohemagglutinin A, concanavalin A, and poke weed mitogen. Measurements post-sirolimus treatment varied per subject, between 1 to 5 years of treatment.

Figure 6

Immune function measured following long-term treatment with monotherapy sirolimus in non-ALPS patients. (A) ALC. Although not statistically significant, 5 of the 12 subjects demonstrated an increase in the ALC, 2 with no change, 2 with decreases but remained normal, and 3 subjects whose ALC dropped to a more moderate lymphopenia range (post-sirolimus ALC range, 320 to 3456 μL; P > .999). (B) IgG quantification; it remained in a normal range for most, and were unchanged before and after the use of sirolimus (P = .8438). (C) CD4 and (D) CD8 counts were monitored pre- and post- long-term administration of sirolimus. CD4 and CD8 counts mostly either remained the same or increased (P = .8438 and P = .6250, respectively). Three subjects had a decrease in the CD4 count; however, only 1 fell to a level <200 μL. That same patient also demonstrated a drop in the CD8 count to a low level (from 492 to 96 μL), after 3 months of sirolimus treatment. Also not depicted here, the response to mitogens was available on 3 subjects: 1 remained normal at 1 year, whereas 2 were reduced at 1 and 4 years. Measurements post-sirolimus treatment varied per subject, between 3 months to 4 years of treatment.