Enhanced Prognostic Value of the SKY92 Classifier over High-Risk Cytogenetic Aberrations in Multiple Myeloma: Results from the Prommis Real-World Study

Introduction: Multiple myeloma (MM) is a hematological malignancy characterized by a highly complex molecular landscape. Despite advancements in therapeutic strategies, patients identified as high-risk (HR), experience limited benefits from modern treatments. Various risk factors have been established to assess the risk in MM, with cytogenetic aberrations (CA) and their co-occurrence being frequently used in clinical practice for prognostication. Additionally, gene expression profiling (GEP), particularly the SKY92 classifier, has been recognized as a prognostic marker to identify HR MM. PROMMIS (NCT02911571), a real-world prospective clinical trial conducted across 9 academic cancer centres in the US, provides an opportunity to compare the prognostic utility of SKY92 with CA and evaluate its added value in clinical practice.

Methods: The gene expression profiles of 251 newly diagnosed MM patients were assessed using the SKY92 assay (SkylineDx, San Diego). This involved analysing RNA extracted from plasma cells, which were isolated using immunomagnetic separation to achieve a purity of ≥80% CD138. CA status for t(4;14), t(14;16), del(17p), and gain(1q) were assessed at each center, and were available for 250 patients. Data on progression-free survival (PFS) and overall survival (OS) were collected for 242 patients, measured from the date of diagnosis. The median follow-up period at the time of analysis was 44 months. Survival analyses were conducted using the Cox proportional hazards model to evaluate the prognostic significance of SKY92 compared to CA.

Results: Despite sufficient numbers of patients in the analysis to detect relevant effects, no significant differences were observed in PFS between patients when stratifying on CA status (del(17p), p=0.8; t(4:14), p=0.14; t(14;16), p>0.9), with the exception of gain(1q) (p=0.005). The prognostic impact of gain(1q) appears to be age-related; this aberration only exhibits a significantly reduced PFS in elderly patients (<65yr-old: p=0.25, >65yr-old: p=0.01).

Consistent with previous studies, SKY92 has shown independent prognostic value from CA (p<0.05). Patients with a SKY92-HR profile have significantly worse PFS compared to those with a SKY92-SR profile (p<0.001).

A low agreement is seen between CA and SKY92 risk status. Among 152 patients classified as HR based on either CA (R2-ISS definition) or SKY92-HR, only 38% exhibited both markers, 9% presented with only SKY92-HR profile while the remaining 53% presents with HR CA only.

Patients with a SKY92-HR profile (both with and without CA) consistently show significantly worse PFS compared to those with a SKY92-SR profile, regardless of their CA status (p<0.001). Patients with SKY92-HR without 1q gain showed similar PFS to those with SKY92-SR. Although this trend was not statistically significant this may indicate that a normal 1q21 status might mitigate the poor prognostic SKY92 status, warranting further investigation in future studies.

Conclusions: Based on these real-world data, the SKY92 classifier demonstrates significant prognostic value in predicting shorter PFS in MM patients, regardless of CA. The findings indicate that patients classified as HR by SKY92 consistently exhibit worse PFS compared to those classified by CA alone. This underscores the potential of SKY92 to enhance risk stratification in clinical practice compared to traditional CA HR markers. SKY92 has the potential to replace FISH testing moving forward for risk stratification.