Exploring a Promising Option for Low-Risk MDS

Nearly 40 percent of patients with low intermediate-1 risk myelodysplastic syndrome (MDS) who were refractory to erythropoietin and granulocyte-colony stimulating factor (EPO/G-CSF) had hematologic improvements with erythroid response (HI-E) after being treated with Revlimid (lenalidomide) monotherapy, according to the results of the randomized phase 2 HOVON89 trial reported at the 2016 American Society of Hematology Annual Meeting in San Diego.

The overall HI-E rate included a third of patients without a 5q deletion (non-del[5q]). The addition of EPO/G-CSF did not significantly improve the HI-E rate overall or in the non-del5q subgroup. Improvement in HI-E had a significant association with better 12-month survival rates.

“Lenalidomide yielded sustained HI-E in 34 percent of patients with non-5q deletion low or intermediate-1 risk MDS that was refractory or unlikely to respond to EPO with or without G-CSF,” said Arjan A. van de Loosdrecht, M.D., Ph.D., a professor of Hematology at VU University Medical Center in Amsterdam, The Netherlands. “Use of erythropoietin, or WHO [World Health Organization] or IPSS [International Prognostic Scoring System] criteria, did not predict HI-E in lenalidomide-treated patients.”

Moreover, the researchers identified patients who may not benefit from Revlimid in EPO-refractory and transfusion-dependent low/intermediate-1 risk MDS with the use of mutational profiling and flow cytometry, said van de Loosdrecht.

Several lines of evidence provided a rationale for investigating Revlimid in low-risk MDS. Laboratory studies have suggested that Revlimid promotes erythroid lineage expansion of primitive erythroid precursors and may restore erythropoietin responsiveness in MDS progenitors buy interfering with JAK2/EPO signal transduction, van de Loosdrecht said. As an immunodulatory agent, Revlimid may interfere with the altered inflammatory microenvironment in low-risk MDS.

Treatment with Revlimid has been shown to improve erythropoiesis, including restoration of red blood cell transfusion independence to patients with non-del(5q) MDS, he continued. Some clinical evidence has suggested that the addition of EPO to Revlimid might improve HI-E and transfusion independence rates, compared with Revlimid alone.

Investigators performed a phase 2 trial to examine the safety and efficacy of Revlimid in low/intermediate-1 risk MDS that was refractory or unlikely to respond to EPO/G-CSF. All patients started treatment with Revlimid, and those who did not attain HI-E after four cycles received EPO, with or without G-CSF, in addition to Revlimid.

Patients who attained HI-E with Revlimid continued treatment for six months, and those who received EPO/G-CSF continued for 12 months, or until loss of response or disease progression. Data analysis comprised 184 patients, 84 percent of whom had non-del(5q) MDS.

The two groups had no substantive differences in demographic or clinical characteristics. The patients had received a median of 13 units of red blood cells, including four units in the eight weeks prior to enrollment.

Overall, 41 percent of the patients achieved HI-E: 38 percent of patients treated with Revlimid alone and 42 percent of those who received Revlimid plus EPO/G-CSF. HI-E rates were 34 percent among patients with non-del(5q) MDS versus 79 percent for those with a 5q deletion and did not differ between treatment groups. The median time to HI-E was 3.1 months in both arms, and the median duration of HI-E was 10.6 months.

Grade 3/4 adverse events occurred in 65 percent of patients treated with Revlimid and 72 percent for the group that received Revlimid and EPO/G-CSF. Rates of grade 3/4 toxicity during the first four cycles of therapy was 51 percent and 56 percent with Revlimid versus Revlimid /Epo/G-CSF.

Other comparisons of the two treatment groups showed that a similar proportion received the planned full dose of treatment (57 percent vs 59 percent), required dose reductions more than 10 percent (30 percent) and had dose reduction/delay (7 percent vs 8 percent). Hematologic toxicity was the most common reason for dose reduction/delay in both arms (52 percent vs 59 percent).

The median progression-free survival (PFS) was virtually identical in the two groups (about 15 months). Overall survival (OS) exhibited a trend in favor of the patients who received only Revlimid (45.1 vs 37.7 months).

The incidence of leukemia progression at two years was 16 percent and did not differ significantly between treatment groups.

Analysis of survival by del(5q) status showed no significant differences by treatment arm for PFS. However, patients with non-del(5q) MDS had significantly better overall survival with Revlimid alone. OS did not differ by treatment for patients with a 5q deletion.

An evaluation of potential predictors of response to treatment showed that endogenous EPO level, pretreatment with EPO/G-CSF, and MDS classification by WHO criteria did not predict HI-E, PFS, or OS. IPSS score did not predict HI-E or PFS, but was predictive for OS.

A landmark analysis at 12 months showed that attainment of HI-E had a significant association with prolonged OS (more than 67 months vs 28 months for nonresponders).

An analysis of results of next-generation sequencing and response to Revlimid showed that the presence of 2 or more mutations was inversely related to HI-E, as was the presence of one or more splicing-factor mutations. Analysis of the seven most frequently mutated genes showed that only SRSF2 and SF3B1 had significant associations with lack of response to Revlimid.