Pomalidomide in chronic GvHD – efficacy and safety

Chronic graft-versus-host disease (cGvHD) following allogeneic hematopoietic stem cell transplantation (allo-HSCT) for hematological malignancy is a significant source of non-relapse mortality and morbidity.¹ Skin manifestations and joints/fascia involvement are particularly difficult to treat, especially when the disease is refractory to steroids and subsequent treatments, such as extra-corporeal phototherapy, topical agents, and other immunomodulators.^1,2 Pomalidomide is an immunomodulating drug (IMiD) licensed by the U.S. Food and Drug Administration (FDA) for the treatment of multiple myeloma¹ and has shown efficacy in the treatment of chemotherapy-induced skin fibrosis in murine models.³ The immunomodulatory effects of pomalidomide include increasing CD4⁺ T-cell counts, stimulation of anti-inflammatory cytokines, suppression of proinflammatory cytokines and Th2 cells, and putatively suppression of B-cell proliferation.²  

In this single-center, randomized phase II trial (NCT01688466), Lauren Curtis and colleagues investigate the safety and efficacy of pomalidomide in treating moderate to severe cGvHD refractory to prior therapies and determine the best recommended dose. Results have been published recently in Blood.²

Study design

Dosing:

• Subjects were randomized to receive oral pomalidomide either 0.5 mg/day (low dose) or 2 mg/day (high dose). The high dose group started on 0.5 mg/day for 2 weeks, increasing by 0.5 mg/day every 2 weeks. Pomalidomide was administered continually on a 28-day cycle until progression, consent withdrawal, or unacceptable toxicity. No crossover between groups was allowed, and patients were to be followed up for 24 months following initial dose of pomalidomide.

Primary outcome measure:

• cGvHD organ response at Month 6 as measured by the 2005 National Institutes of Health (NIH) cGvHD Response Criteria (skin, mouth, eyes, gut, and liver) and healthcare provider global rating (0–10), plus Photographic Range of Motion (P-ROM) for lung, fascia, and joints.

Secondary outcome measures:

• Safety (assessed by National Cancer Institute [NCI] Common Terminology Criteria for Adverse Events version 4.0), compliance with pomalidomide dosing, pharmacokinetics of pomalidomide, changes in T-cell compartment as assessed by flow cytometry, and health-related quality of life, using the Short Form-36 Health Survey, Functional Assessment of Cancer Therapy-Bone Marrow Transplant tool, Human Activity Profile survey, and Lee Symptom Scale.

 Additional treatments

• All patients were allowed concurrent topical GvHD treatments and oral systemic steroids.
• Patients were also allowed a maximum of two steroid pulses (prednisolone 0.5-2 mg/kg/day with a taper to baseline within 3 weeks) for the treatment of cGvHD flare.
• Patients received aspirin 325 mg/day or enoxaparin 40 mg/day for the prophylaxis of coagulation-related adverse events (AEs).

Inclusion criteria

• Patients aged 18–75 years in complete remission of their primary cancer with moderate to severe cGvHD.
• Adequate organ function and Karnofsky Performance Score (KPS) ≥ 60%.
• cGVHD unresponsive or progressive in at least one organ on high dose corticosteroids (≥ 0.5 mg/kg/day for ≥ 8 weeks) and/or subsequent therapies.
• On a stable or reducing dose of corticosteroids and/or subsequent systemic immunosuppression in the ≥4 weeks before study-entry
• Agreement to methods of contraception and birth control measured as per the PREMS® risk management program.

Results

A total of 34 subjects were recruited and randomized to the low dose (n = 17) or high dose (n = 17) treatment cohorts. All patients had received a transplant due to either a myeloid, lymphoid, or mixed-lineage hematological malignancy. The median KPS was 80%, with the majority of patients receiving peripheral blood as the stem cell source (82%), from a matched unrelated donor (53%), and undergoing myeloablative conditioning (71%). There was no statistical difference between the two cohorts for any described characteristics (p > 0.05).

Clinical cGvHD and transplant-related characteristics of patients included in the final analysis are summarized in Table 1.

Table 1. Baseline clinical and GvHD characteristics of patients included in final analysis²

Additional treatments   

At baseline, 32 of 34 patients were on concurrent, stable steroid therapy. At 3 months, nine patients received a lower dose, two had a dose increase, and one started new maintenance. At 6 months, 12 of the 24 remaining evaluable patients had a decrease in systemic steroid dose, and 11 had a stable dose.

Three subjects in each cohort received corticosteroid pulses due to cGvHD flare, possibly related to pomalidomide treatment.

Primary outcomes

With a median follow-up of 24 months, partial remission as best response was seen in 16 subjects at 6 months in an intention-to-treat analysis (47%; 95% CI, 30–65%) while responses were 67% (95% CI, 45–84) in the 24 patients evaluated at 6 months. There was no difference in the overall response rate (ORR) between the two dose groups.

At 6 months, the site showing the greatest decrease in GvHD score was the joints/fascia (9 of 24 patients, p = 0.0018 for overall change), with improvements also seen in the gastrointestinal tract (7 of 24), mouth (5 of 24), and skin (5 of 24). The improvements in the joints/fascia was reflected in the P-ROM which detected increased scores in 14 of 24 patients, with the largest improvements seen in the elbow (p = 0.0097) and wrist scores (p = 0.0083). The 11-point healthcare provider severity score showed the strongest improvement from baseline with a median change of –2 (95% CI, –1 to –7; p = 0.0003).    

Secondary outcomes

Compliance with pomalidomide dosing

The median duration of treatment for the low dose and high dose treatment groups was 253 days (range, 36–356) and 168 days (range, 17–343), respectively (p = 0.03). Dose escalation in the high dose cohort to 2 mg/day was planned to be over 8 weeks (56 days), but 15 patients reached this dose level after a median 42 days. Eight of these patients subsequently required dose reduction due to AEs.

Safety

At the 6 month primary efficacy point, ten subjects were unevaluable for safety due to uncontrollable cGvHD progression (n = 2), withdrawal of consent (n = 4), AEs (n = 3), or for receiving < 75% of the intended pomalidomide dosing (n = 1). Fifteen of 17 patients in the low dose cohort were evaluated; nine of 17 in the high dose cohort (Table 2). Of the four patients who withdrew consent, one experienced cGVHD resolution after first cycle (meaning < 75% of planned dose was received), with the remaining three withdrawals a result of AEs. Overall, six out of eight patients who withdrew from the high dose cohort did so due to unacceptable toxicity. One patient died in the low dose cohort due to bacterial pneumonia.

The most common AEs were lymphocytopenia, infection, hypophosphatemia, and fatigue (Table 2). Nine patients in the high dose cohort require pomalidomide dose reductions or discontinuation due to toxicity, versus two patients in the low dose cohort (p = 0.03). Eight patients in the low dose cohort had severe AEs, compared with 11 in the high dose cohort (p = 0.49).

Table 2. Most common adverse events in the low and high dose cohorts²

HRQoL

Lee Symptom Scale data was available for 19 patients at baseline and 6-month follow-up, and showed a clinically meaningful decrease in median score from 40 to 26 (p = 0.035) across the 11 responders. Subjects with improvement in cGvHD symptoms at the 6-month assessment had improvements in mean symptom severity (p < 0.05) and self-reported physical health (p < 0.01). Human Activity Profile scores were unchanged from baseline.

Pharmacokinetic Data

Linear pharmacokinetics were identified through analysis across all samples. No significant differences in peak or predose-trough levels between patients who experienced cGvHD symptom response and those who did not.  

Biomarker assessments  

Of the 24 patients evaluable at baseline, 3 months, and 6 months, low baseline cell counts were identified, particularly in CD4 and B cell populations. There was no significant change over 6 months in the absolute numbers of CD4⁺ T cells, B cells, CD8⁺ T cells, or NK cells between the low and high dose cohorts or responders, following Wilcoxon matched pair analysis.

No difference was seen in CD25⁺CD127^-CD4⁺ Treg cells (Treg/µl) between baseline and the 6-month assessment (p = 0.17). However, an increase was seen in the percentage of Treg cells within the CD4⁺ T-cell populations in patients treated with pomalidomide. An absolute > 50% increase in the percentage of CD25⁺CD127^- cells within the CD3⁺CD4⁺ cells was seen in 16 out of 21 patients overall and seen in 73% of responders formally evaluated (p = 0.0001).  

IL-2 levels were increased in all patients receiving pomalidomide at 6 months compared with baseline (p = 0.005), but no difference between low or high dose groups was detected.  

Conclusion

• Pomalidomide is an effective treatment in patients with severe GvHD failing prior treatment lines, with an ORR of 47% in an intention-to-treat analysis and 67% in the 24 evaluable subjects at 6 months.
• The main benefit identified is improvement in sclerotic skin lesion and joints/fascia cGvHD.
• Partial responses were identified in both cohorts with no difference in ORR between high dose and low dose pomalidomide. There is no treatment benefit with higher-dose pomalidomide given the higher associated toxicities and AEs.
• The most frequent AEs were lymphopenia, infection, and fatigue.
• A limitation of this study is the high dropout rate, with 10 of 34 patients not reaching the 6-month primary assessment point, largely due to unacceptable toxicity and AEs.
• Pomalidomide 0.5 mg, orally, daily, is the recommended dose for pomalidomide as salvage treatment.