P-179: Immunoparesis remains a negative prognostic factor in newly diagnosed Multiple Myeloma by competing risk analysis

Abstract

<h3>Background</h3> Recent reports describe worst outcome in patients with multiple myeloma (MM) and immunoparesis (IP). However, infections can be considered a competing event when analyzing the impact of IP on overall survival (OS). Here, we analyzed the prognostic impact of IP in newly diagnosed MM (NDMM) with a different approach. <h3>Methods</h3> This is a single-center retrospective study of NDMM patients diagnosed between 2000 to 2018. We defined IP as suppression of both uninvolved polyclonal immunoglobulins in IgG and IgA MM or three uninvolved polyclonal immunoglobulins in light-chain MM. <h3>Results</h3> A total of 651 patients with NDMM were diagnosed and treated at our institution. Median age at diagnosis was 65 years old (56 – 74). The M-protein isotype was IgG in 355 (55%) patients, IgA in 184 (28%), light – chain in 98 (15%) and IgD in 14 (2%). At diagnosis, IP was present in 318 (47%) patients, and increased with each ISS stage (37% in ISS-1, 54% in ISS-2 and 63% in ISS-3; p<0.01). 347 (53%) patients were fit for autologous stem cell transplant (ASCT). With a median follow-up of 5 years, median PFS in patients with IP was shorter compared to those without IP (22 vs. 27 months; p<0.01). Cox univariate analysis for known risk factors of progression showed that high β2-microglobulin (≥3.5 mg/L, HR 1.4; p<0.01), low serum albumin (< 35 g/L, HR 1.2; p=0.1), high LDH (≥ 450 U/L, HR 1.6; p<0.01) and IP (HR 1.3; p<0.01) had similar negative impact on PFS. Multivariate analysis showed that high β2-microglobulin, high LDH and IP were the main risk factors with independent negative impact on PFS. Given this data, we aimed to create a prediction model on PFS by two different approaches. The first one consisted of evaluating all possible equations (Stata 16.0. College Station, TX) with AIC and C-Harrell as main criteria, and the second used backward stepwise elimination process. Both resulted in a parsimonious model that contained high β2-microglobulin, high LDH and IP. At the time of this analysis, 45% patients were still alive. Severe infection as a cause of death was seen in 7% patients. Cox univariate analysis showed that patients with IP at diagnosis had negative impact on OS (HR 1.2; p=0.04). To evaluate the impact of IP on OS, we included severe infections deaths as competing events. With this approach, risk of death among patients with or without IP did not change significantly, maintaining its prognostic value both in eligible and non-eligible patients for ASCT (Peppe and Mori test; p=0.12 and p=0.08 respectively). These results suggest that severe infections do not play a major role in OS of patients with NDMM and IP. <h3>Conclusion</h3> In patients with NDMM, IP is an important negative biomarker with shorter median PFS. Its prognostic impact was similar to other known potent risk factors of progression. Regarding OS, IP kept its negative impact even in the presence of severe infections deaths as competing events.