Abstract
Although autologous stem cell transplant (ASCT) remains an important therapy for myeloma, outcomes are highly variable. MM is characterized by the loss of central mediators of immune surveillance and the development of an immunologic milieu that promotes disease progression and resistance to treatment. We aimed to analyze the impact of various clinically available immune-related markers on progression-free survival (PFS) after ASCT and formulate a combined immune score that could be used to predict early progression.
One hundred and thirty patients who received melphalan 200 mg/m2 and ASCT between 2002 and 2016 were included in the final analysis. The median age was 59 (range: 34-77) years with the majority of patients being male (n=71, 55%), African-American (n=61, 47%), IgG subtype (n=65, 50%), and ISS stage 3 at diagnosis (n=55, 57%). Sixteen (12%) patients were diagnosed with high risk FISH/karyotype. Median number of prior treatments was 2 (range: 1-5), the majority of which were IMiD- (n=82, 63%) or PI- (n=72, 55%) based. The median time to transplant was 8 months (range: 3-144), and approximately half of patients received post-ASCT maintenance treatment (n=60, 46%). Pre-transplant response was VGPR (n=23, 23%) or PR (n=52, 51%) in most patients.
We collected the following values at baseline (Day -2) and at Day +90 post-ASCT: absolute lymphocyte count (ALC); absolute monocyte count (AMC); lymphocyte to monocyte ratio (LMR); and the number of immunoglobulin (Ig) levels suppressed. Values were separated into upper and lower quartiles for analysis. A low ALC, AMC, and LMR were defined as less than 1100 cells/μL, 300 cells/μL, and 2.4 respectively. A high ALC, AMC, and LMR were defined as more than 2400 cells/μL, 500 cells/μL, and 5.7 respectively. For the entire cohort, the median PFS was 25 months. Baseline ALC, AMC, LMR, or Ig levels did not predict PFS. However, at Day +90, we found that a low ALC (18 versus 23 months, p=0.04) or AMC (13 versus 25 months, p=0.02) predicted for worse PFS. When combined as the LMR, a low ratio strongly predicted for worse PFS when compared to a high ratio (16 versus 52 months, p=0.004). Finally, patients with 2 or 3 suppressed Ig levels were found to have significantly worse PFS compared to those with normal levels (17 versus 51 months, p=0.04).
When LMR and Ig levels were combined, we were able to divide patients into poor risk (low LMR and 2-3 suppressed Ig), good risk (high LMR and 0-1 suppressed Ig) and intermediate risk (all other patients) groups. Median PFS for poor, intermediate and high risk groups was 7.5 versus 27 versus 79 months respectively (see figure, p=0.0004). In a multivariate analysis adjusted for prognostic factors (protein subtype, ISS stage, number of prior treatments, novel agents induction, and use of maintenance treatment), low LMR (95% CI 0.01-0.15, HR 0.01; p=0.001) and suppressed Ig levels (95% CI 1.45-39.71, HR 7.58; p=0.017) were strong independent predictors of poor PFS.
Here we demonstrate that clinically available markers related to immune recovery post-ASCT can predict PFS. We observed that at Day +90 a low LMR and suppressed Ig levels strongly predicted for a worse PFS. We propose an immune score combining these readily available immune-related tests to stratify patients at risk for early progression and identify those who may benefit from intensified post-ASCT consolidation and maintenance. Future investigation will focus not only on validating this score in an independent cohort of patients but also on identifying the important cell subsets upon which to focus clinical interventions.
Patel: Celgene: Consultancy, Honoraria.
