Clinical implications of TP53 mutations in myelodysplastic syndromes treated with hypomethylating agents

Abstract

We screened TP53 mutations in 168 MDS patients who were treated with HMA and evaluated predictive and prognostic value of TP53 mutations. Overall response to HMA was not different based on TP53 mutation status (45% vs. 32% in TP53-mutated and wild type [WT], respectively, P = 0.13). However, response duration was significantly shorter in TP53-mutated patients compared to WT patients (5.7 months vs. 28.5 months, P = 0.003). Longitudinal analysis of TP53 mutations after HMA showed that TP53 mutations almost always persisted at times of disease progression. TP53-mutated patients showed significantly worse overall survival (OS) compared to WT patients (9.4 months vs. 20.7 months, P <0.001). Further, TP53 mutations distinguished prognosis in the subgroup of patients with complex karyotype and Revised International Prognostic Scoring System (IPSS-R) defined very high-risk disease. Multivariate analysis showed that TP53 mutation status is significantly prognostic for OS after adjusting prognostic effect from other factors. The current study provides evidence that TP53 mutations are independently prognostic in MDS patients treated with HMA. While TP53-mutated MDS patients initially respond well to HMA, their duration of response is significantly shorter than WT patients. Novel strategies to improve duration of response in TP53-mutated MDS are urgently needed.

Keywords: TP53; hypomethylating agents; myelodysplastic syndromes.

Figure 1. Lollipop figure of TP53 mutations detected in 168 patients with MDS and CMML

Green dots indicate missense mutations and red dots indicate nonsense mutations. The figure for was created using cBioPortal website (http://www.cbioportal.org/).

Figure 2. Landscape of well characterized myeloid driver mutations in 53 MDS/CMML patients whose bone marrow samples were sequenced by WES

TP53 mutated cases had less co-occuring mutations. TP53 mutation and splicing gene mutations had trend to be mutually exclusive (P = 0.07).

Figure 3

A. Kaplan-Meier curve comparing time to best response for TP53 mutated patients and TP53 WT patients who responded to HMA therapy. B. Kaplan-Meier curve comparing duration of response for TP53 mutated patients and TP53 WT patients who responded to HMA therapy.

Figure 4

A. Kaplan-Meier curve comparing OS of TP53 mutated patients and TP53 WT patients. Kaplan-Meier curve comparing overall OS of TP53 mutated patients and TP53 WT patients among. B. patients with IPSS-R high or very high risk (N = 52). C. patients with complex karyotypes (N = 47), D. patients with monosomal karyotypes (N = 43).

Figure 5. Representative cases of longitudinal TP53 follow up

Except for PT3 case, the same TP53 mutations persisted at times of disease progression or relapse. A. The patient PT2 had TP53 p.H178D mutation. Received 4 cycles of 5-azacitidine and vorinostat achieving complete response (CR). When disease progressed, the same TP53 p.H178D mutation was detected. B. The patient PT3 had TP53 p.H193R mutation Received 7 cycles of standard of care decitabine and after 3 cycles, TP53 mutation became negative on bone marrow. When disease transformed to AML, TP53 was still wild type (WT). C. The patient PT8 had TP53 p.R196* nonsense mutation and received 1 cycle of guadecitabine (SGI-110) followed by allogeneic hematopoietic stem cell transplant (HSCT). One year later, when disease relapsed, the same TP53 p.R196* mutation was detected in bone marrow. D. The patient PT12 had TP53 p. H179R mutation. Received 6 cycles of guadecitabine (SGI-110) and achieved CR. TP53 sequencing showed WT. However, when disease relapsed, the same TP53 p. H179R was detected in bone marrow.