Mixed Results for Immunotherapy Approaches in Myelodysplastic Syndromes

Findings from trials evaluating new treatment strategies for patients with myelodysplastic syndromes (MDS) produced mixed results: Combining an immune checkpoint inhibitor with azacitidine was associated with high response rates, although perhaps ones that don’t differ much from azacitidine monotherapy, while the PD-L1 inhibitor atezolizumab plus azacitidine was associated with a high early mortality rate.

Immune Checkpoint Inhibitors With or Without Azacitidine

In the first trial, researchers found that treatment with the immune checkpoint inhibitors ipilimumab or nivolumab, with or without the hypomethylating agent (HMA) azacitidine, showed clinical activity – but distinct toxicity profiles – in patients with previously treated MDS or disease that failed to respond to HMA treatment.

Lead author Guillermo Garcia-Manero, MD, from the University of Texas MD Anderson Cancer Center, presented the findings from this trial, which enrolled adults with untreated MDS or with MDS that did not respond to HMA treatment. In the HMA-failure cohort, patients had received HMA therapy within four months of enrollment and no other therapy after HMA exposure.

A total of 76 participants were divided into frontline (n=41) and HMA-failure groups (n=35). Patient characteristics were balanced; most patients had an abnormal or complex karyotype and high-risk genomic features, according to Dr. Garcia-Manero.

In the HMA-failure group, patients were assigned to receive either nivolumab (3 mg/kg every 2 weeks) or ipilimumab (3 mg/kg every 3 weeks). In the frontline group, patients received a backbone of azacitidine (75 mg/m2 for 5 days) plus nivolumab or ipilimumab.

In the HMA-failure cohort, patients could receive azacitidine again if they had no response to six cycles of immune checkpoint inhibitors.

At a median follow-up of 20.1 months (range = 1-33 months), the overall response rate (ORR) appeared to be highest in the nivolumab plus azacitidine combination for previously untreated patients and lowest for HMA-failure patients who received single-agent nivolumab (TABLE). Treatment of this cohort was stopped early for a lack of response.

In the frontline group, median overall survival (OS) was not reached in the ipilimumab plus azacitidine group but was 11.8 months in the nivolumab plus azacitidine group (ranges not reported). In the HMA-failure cohort, median OS was 8.5 months with ipilimumab and 8.0 months with nivolumab (ranges not reported).

“These agents can be associated with significant toxicities including colitis and other inflammatory processes that require steroid treatment,” Dr. Garcia-Manero stressed. The most commonly reported grade 3/4 adverse events (AEs) in both the frontline and previously treated cohorts included infection, rash, musculoskeletal pain, and transaminitis.

The safety and efficacy results observed in this trial suggest that “the incorporation of immune checkpoint inhibitor therapy is feasible in patients with MDS,” Dr. Garcia-Manero concluded. The findings suggested a potential relationship between PD1 expression and response, but he noted that this needs to be explored further. The results also need to be confirmed in larger randomized trials in the frontline and relapsed settings, he added.

Atezolizumab Plus Azacitidine

Next, Aaron Gerds, MD, from Cleveland Clinic’s Taussig Cancer Institute, shared findings from a trial evaluating atezolizumab, with or without azacitidine, in patients with HMA-naïve, higher-risk MDS (defined as intermediate, high, or very high-risk per the Revised International Prognostic Scoring System) or who had disease that was relapsed/refractory to prior HMA therapy. After researchers observed a high number of deaths within the first three treatment cycles in the HMA-naïve cohort, the study was terminated early.

This phase Ib trial enrolled 42 adult patients into three cohorts:

  • cohort A1: HMA-failure MDS and atezolizumab alone (n=10)
  • cohort B1: HMA-failure MDS and atezolizumab plus azacitidine for 6 cycles, followed by atezolizumab maintenance alone (n=11)
  • cohort C1/C2: HMA-naïve MDS and atezolizumab plus azacitidine until clinical benefit (n=21)

In the safety analysis, grade 3/4 AEs were common, occurring in 50 percent (n=5/10), 91 percent (n=10/11), and 57 percent (n=12/21) of patients in cohorts A, B, and C1/2, respectively. Immune-related AEs included infusion-related reactions and hypothyroidism in cohort A and rash and hemolysis in cohorts B and C.

There were 21 deaths across the three cohorts: seven in cohort A, eight in cohort B, and six in cohort C1/C2. Most deaths were related to progressive disease in cohorts A and B, but all deaths in cohort C were attributed to AEs.

Deaths occurred early in the HMA-naïve patients in cohort C1/C2, Dr. Gerds reported. All six occurred within the first three treatment cycles, three of which occurred within the first treatment cycle.

When asked about potential mechanisms for the increased toxicity with atezolizumab, compared with other immunotherapies, Dr. Gerds noted that there was “no clear biologic plausibility [to explain this association], although the fact that antibiotic prophylaxis was left to investigator discretion may have led to an increased infection risk.”

No patients with HMA-failure MDS responded to treatment with atezolizumab (cohort A), while only one patient with HMA-failure MDS responded to treatment with atezolizumab plus azacitidine (cohort B). In cohort C, however, 13 patients responded to treatment, for an ORR of 62 percent.

Median OS was not reached in cohort C but was 177 days in cohort A and 361 days in cohort B. Six patients with HMA-naïve MDS are continuing to receive the atezolizumab/azacitidine combination, and the investigators also are continuing to follow patients who discontinued the study early to identify potential markers of response.

“The number of deaths we saw early was certainly concerning,” Dr. Gerds said. However, given the responses to atezolizumab plus azacitidine seen in the HMA-naïve population, he added that “a better understanding of the different toxicity profiles observed between HMA-naive and HMA-failure patients with higher-risk MDS patients is crucial for future developments of this combination.”

The authors of the first trial report relationships with Bristol-Myers Squibb, which supported the trial. The authors of the second trial report relationships with Roche, which supported the trial.


Garcia-Manero G, Sasaki K, Montalban-Bravo G, et al. A phase II study of nivolumab or ipilimumab with or without azacitidine for patients with myelodysplastic syndrome (MDS). Abstract #465. Presented at the 2018 ASH Annual Meeting, December 2, 2018; San Diego, CA.

Gerds A, Scott B, Greenberg P, et al. PD-L1 blockade with atezolizumab in higher-risk myelodysplastic syndrome: an initial safety and efficacy analysis. Abstract #466. Presented at the 2018 ASH Annual Meeting, December 3, 2018; San Diego, CA.