Introduction

Higher-risk myelodysplastic syndromes (MDS) progress to secondary acute myeloid leukemia (sAML) within months. Treatment with azacitidine (AZA) has been shown to prolong survival and delay progression to sAML. TP53 mutations are found in MDS with abnormal chromosome 5 or complex karyotype and are associated with poor prognosis. In particular, TP53 mutations confer resistance to lenalidomide in low-risk MDS with del(5q). However, whether TP53 mutations also influence response to AZA is unclear. Therefore, we analyzed the prevalence of TP53 mutations in a cohort of 100 patients with higher-risk MDS or sAML treated with AZA and correlated this to outcome.

Patients and Methods

100 cases of higher risk MDS (IPSS INT-2 or High, n= 53), MDS/MPN (n= 8) or sAML (≥ 20% marrow blasts, n= 39) were included who had received at least one complete cycle of AZA (75 mg/m2 d1-7 q28) and for whom a bone marrow biopsy was available before the start of treatment. Presence of TP53 mutations was determined by immunohistochemistry [Iwasaki et al, Pathol Int 2008; Jädersten et al, JCO 2011; Kulasekararaj et al, Br J Haematol 2013]. Staining was defined as positive if ≥5% of CD34+ cells showed strong nuclear staining. In addition, 37 randomly selected cases were confirmed by sensitive next-generation amplicon deep-sequencing, demonstrating good correlation between both methods. Response to AZA was determined using IWG2006 criteria. Kaplan-Meier curves were estimated from the first day of AZA treatment with the log-rank test to determine significance. Patients who subsequently underwent allogeneic transplantation were censored at that date.

Results

Thirty-five patients (35%) had TP53 mutations. Both TP53 positive and negative cohorts were balanced in age (median 71 years) and number of administered AZA cycles (median 5/4, TP53 positive/negative, respectively). Of the TP53 positive patients, 68% had higher-risk MDS, 6% had MDS/MPN and 26% had sAML. Significantly more patients diagnosed with sAML had unmutated TP53 (46%, p=0.046). With regard to cytogenetics, only 23% of patients with TP53 mutations showed a good risk karyotype (KT) according to IPSS cytogenetic risk classification whereas 74% were diagnosed with poor risk KT. A closer look at the TP53 positive patients with poor risk cytogenetics revealed involvement of chromosome 5 aberrations in 77% compared to 33% in wild-type TP53 patients (p < 0.001). Of all 100 patients a total of 18 fulfilled the criteria for monosomal KT and 14 of these harbored TP53 mutations (77%, p < 0.001). There was no difference in pretreatment absolute neutrophil count or hemoglobin value in patients with or without TP53 mutations. The pretreatment median platelet count in TP53 mutated patients was significantly higher than in patients with wild-type TP53 (90 G/l vs. 43 G/l, respectively, p=0.021). The overall response rate (ORR) for patients with and without TP53 mutations treated with AZA was 74% (26/35) and 68% (44/65), respectively (n.s.). Quality of response did not show any differences between cohorts. However, hematologic improvement was significantly more frequent in patients with MDS and TP53 mutations than in MDS with wild-type TP53 (46% vs. 14%, p=0.015). Patients with complex KT and TP53 mutations responded as well as patients with wild-type TP53 to treatment with AZA (ORR 65% versus 53%, respectively, n.s.). Regarding the subgroup of patients with any kind of chromosome 5 abnormality within the complex KT group, there was a better ORR to AZA but no statistical significance between those with TP53 mutations compared to those with wild-type TP53 (75% vs. 43%, respectively). A similar result with respect to ORR was seen in the subgroup of patients with monosomal KT and the presence or absence TP53 mutations (ORR: 64% vs. 25%, respectively, n.s.). The median OS of the total cohort was 288 days, with a median OS of 212 days for TP53 positive patients and 353 days for TP53 negative patients (p = 0.731).

Conclusion

TP53 mutations in higher-risk MDS and sAML are frequent and predominantly associated with complex karyotype involving chromosome 5 abnormalities. In our cohort of 100 patients treated with AZA, the presence of TP53 mutations did not negatively impact on response to epigenetic therapy. However, although this analysis was not created for survival assessment it confirms the very poor survival for higher-risk MDS and sAML patients with TP53 mutations, in spite of response to AZA.

Disclosures:

Haferlach:MLL Munich Leukemia Laboratory: Employment, Equity Ownership. Germing:Celgene: Honoraria, Research Funding. Platzbecker:Celgene Corp: Honoraria. Götze:Celgene Corp: Honoraria.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution