Physician University Hospital Würzburg, United States
Introduction: SKY92 gene expression profiling (GEP) has been developed to identify high-risk (HR) multiple myeloma (MM). This study aimed to evaluate the combination of SKY92 and cytogenetics according to R2-ISS (del(17p), t(4;14), 1q CNA) for HR detection in newly diagnosed (ND) and relapsed/refractory (RR) MM.
Methods: We performed a single-center prospective study. Cytogenetics were analyzed on purified CD138 positive cells by FISH. SKY92 risk status was determined with MMprofiler gene expression assay. Whole genome sequencing (WGS) was performed to elucidate the discrepancy between the both risk stratification systems SKY92 and FISH.
Results: Overall, 258 patients were included (NDMM: n=109; RRMM: n=149). SKY92 classification was available for 216 (83.7%) patients. Samples of 26 (17.7%) patients, who showed significantly lower bone marrow infiltration than the remaining patients (median: 20% vs 50%, P=0.006), did not meet the SKY92 quality control criteria. HR SKY92 was significantly enriched in RRMM (57/121, 47.1%) compared to NDMM (17/95, 17.9%) (P < 0.001). In RRMM, HR SKY92 was significantly more frequent in patients with ≥4 prior lines of therapies (32/52, 61.5%) than those with < 4 therapy lines (25/65, 36.2%) (P=0.009). Moreover, HR SKY92 was more common in patients who received high-dose melphalan and autologous stem cell transplant (48/89, 53.9%) than the remaining patients (9/32, 28.1%) (P=0.01). RRMM with HR SKY92 showed significantly shorter progression free survival (PFS) (P < 0.001) and overall survival (OS) (P < 0.001) than standard-risk (SR). In NDMM, HR SKY92 also indicated a significantly inferior PFS (P < 0.001) in comparison to SR. We then combined SKY92 with FISH according to R2-ISS in 181 patients (NDMM: n=79; RRMM: n=102). We found a discrepancy between the both risk stratification systems, with 67 (37.0%) and 99 (54.7%) patients being defined as HR by SKY92 and FISH, respectively. Overall, 13 (16.4%) NDMM and 36 (35.3%) RRMM patients showed HR in both SKY92 and FISH (“double-HR”). Double-HR presented a negative prognostic factor for PFS in both NDMM (P < 0.001) and RRMM (P < 0.0001). Furthermore, “double-HR” patients showed the worst OS (P < 0.001) in RRMM. To elucidate the discrepancy between FISH and SKY92, we performed WGS in 16 patients who exhibited either only HR SKY92 (n=7) or only HR FISH (n=9). Interestingly, 1 patient with bi-allelic TP53 inactivation (del + mut) and 6 patients with 1q CNA were determined as SR by SKY92 but as HR by FISH. Moreover, 4 out of 7 patients with only HR SKY92 but SR FISH displayed 1q CNA, which was detected only by WGS, and del1p32 was found in 1 patients. Of note, we found CRBN mutation in 3 out of 7 patients with only HR SKY92 but SR FISH. The remaining 2 patients did not show any known HR genomic alterations.
Conclusions: We provide the first prospective evidence that “double-HR” (SKY92 + FISH according to R2-ISS) indicates the highest-risk MM.
