Abstract: PB1977

Type: Publication Only

Background
Complex chromosome aberrations (≥3 cytogenetic aberrations, cA) are associated with an inferior outcome in patients (pts) with myelodysplastic syndromes (MDS). About half of those pts show mutations in TP53 which might be an important factor for development of cA and further worsen the prognosis (Haase et al., 2019). One of the proteins being part of the p53 pathway is Wip1, encoded by the gene PPM1D. Wip1 is a direct and indirect inhibitor of p53. In former single nucleotid polymorphism (SNP) analyses of pts with cA we found 17q where PPM1D is located being affected in a higher incidence as by chance. PPM1D is already known to be mutated in pts with clonal hematopoiesis of indetermined potential (CHIP) as one of the most frequently affected genes. In therapy-associated myeloid neoplasms it is even mutated in up to 15% (Lindsley et al., 2017)

Aims
The aim of our study was to find out the prevalence of PPM1D mutations in MDS pts with cA in comparison to pts without complex karyotypes. We also wanted to find out if variants of PPM1D impact on prognosis and whether PPM1D mutations might provide a pathway towards cA alternative to mutations of TP53.

Methods
We included 115 pts with cA (75 x MDS, 36 x sAML, 4 x CMML) and 39 pts without cA (34 x MDS, 4 x sAML, 1 x CMML) all characterized by conventional cytogenetic analyses. 26 (22.6%) of the cA pts and 16 (41%) of the non-cA pts had a therapy-associated neoplasia. The median number of aberrations in the group of cA was 8 (range 3-50). A deletion of 5q was present in 69.6% of the pts with cA and in 7.6% of the pts without cA. The TP53 status was known of all pts and has been analyzed by fluorescence in situ hybridization (FISH) and next generation sequencing (NGS). In the cohort of cA we found TP53 being altered in 79/115 pts (68.7%, 4 deletions, 38 mutations, 37 biallelic affection), in the control-cohort 2/39 pts showed alterations of TP53 (2 deletions). All pts were subjected to NGS of PPM1D. Exons 1-6 were amplified by a multiplex polymerase chain reaction and then sequenced on a MiSeq platform (Illumina, San Diego, CA, USA). We used our local bioinformatics pipeline to identify single-nucleotide variations (SNVs) and indels.

Results
In the cohort with cA we identifiedd 8 pts (7%) with single nucleotid polymorphisms of PPM1D. The median number of aberrations in those pts was x (range x-xx). Six of those variants have already been described as very rare SNPs, the other two are not listed in any common databases. One of those two pts had a TP53 mutation additionly, the other had a TP53 wildtype. Those two might be potentially pathogenic mutations. The variant allele frequency (VAF) was 33.6% and 6.9% resp.. None of these cases was therapy-associated.
In the cohort without cA there were 3 pts with variants of PPM1D (7.6%). None of them had any affection of TP53, but two out of three were therapy-associated.  The VAF were 8.1%, 45.3% and 48%.

Conclusion
We were able to show that variants of PPM1D occur in different frequencies in pts with or without cA. Interestingly, pathogenic mutations seem to be very rare in pts with cA. The two pts having potentially pathogenic variants had a completely different course of disease (1 vs. 22 months) so, as yet we cannot see a clear effect on the prognosis. In pts without cA PPM1D variants seem to be more common and, as already been reported, even more prevalent in therapy-associated neoplasia. We further want to characterize more pts and continue observing our pts to find out which effect those variants might have on the prognosis of pts with or without cA.

Session topic: 09. Myelodysplastic syndromes - Biology & Translational Research