Dynamics of PD-1 expression are associated with treatment efficacy and prognosis in patients with intermediate/high-risk myelodysplastic syndromes under hypomethylating treatment.

Hypomethylating agents (HMAs) are widely used in patients with higher-risk MDS not eligible for stem cell transplantation. However, the general response rate by HMAs is lesser than 50% in MDS patients, while the relapse rate is high. Emerging evidence indicates that demethylating effects committed by HMAs may facilitate the up-regulation of a range of immune checkpoints or cancer suppressor genes in patients with MDS, among which the programmed death protein 1 (PD-1) and its ligands are demonstrated to be prominent and may contribute to treatment failure and early relapse. Although results from preliminary studies with a limited number of enrolled patients indicate that combined administration of PD-1 inhibitor may yield extra therapeutic benefit in some MDS patients, identifications of this subgroup of patients and optimal timing for the anti-PD-1 intervention remain significant challenges. Dynamics of immune checkpoints and associated predictive values during HMA-treatment cycles remained poorly investigated. In this present study, expression levels of immune checkpoints PD-1 and its ligands PD-L1 and PD-L2 were retrospectively analyzed by quantitative PCR (Q-PCR) in a total of 135 myelodysplastic syndromes (MDS) cohort with higher-risk stratification. The prognostic value of dynamics of these immune checkpoints during HMA cycles was validated in two independent prospective cohorts in our center (NCT01599325 and NCT01751867). Our data revealed that PD-1 expression was significantly higher than that in younger MDS patients (age ≤ 60) and MDS with lower IPSS risk stratification (intermediate risk-1). A significantly up-regulated expression of PD-1 was seen during the first four HMA treatment cycles in MDS patients, while similar observation was not seen concerning the expression of PD-L1 or PD-L2. By utilizing binary logistic regression and receiver operating characteristic (ROC) models, we further identified that higher or equal to 75.9 PD-1 expressions after 2 cycles of HMA treatment is an independent negative prognostic factor in predicting acute myeloid leukemia (AML) transformation and survival. Collectively, our data provide rationales for monitoring the expression of PD-1 during HMA treatment cycles, a higher than 75.9 PD-1 expression may identify patients who will potentially benefit from the combined therapy of HMA and PD-1 inhibitors.