Abstract

Identifying factors that independently correlate with important clinical endpoints is critical to evaluating patients (pts) with CLL, helping with treatment planning and providing insights into the fundamental biology of disease. CLL is typically a disease of multiple relapses with expected response to salvage treatment lower with subsequent salvage treatment and progressively shorter remission durations, usually despite escalating treatment intensity. The expectation for outcome in previously treated pts is lower than for untreated pts. Certain prognostic factors may emerge to be dominant over the clinical course of the disease. We evaluated 484 previously treated pts who had an NCI-WG ‘96 indication for treatment and received salvage therapy to identify significant independent predictors for achieving complete remission (CR), time to progression (TTP), and overall survival (OS). We evaluated both traditional prognostic factors such as age, Rai stage, absolute lymphocyte count (ALC), beta-2 microglobulin (β2M), number of prior treatments, refractoriness to alkylating agents or fludarabine, etc., as well as newer prognostic factors including IgVH mutation status (MS), chromosome abnormalities by FISH, and ZAP70 expression by flow cytometry. Pts included in this analysis received diverse investigational salvage therapy including chemoimmunotherapy-based (n=364), monoclonal antibody-based (n=83) or other (n=37) (eg. lenalidomide). The overall estimated median follow-up time was 60mo, TTP was 25mo and OS was 39.5mo. Pre-treatment characteristics were evaluated in univariable analyses for associations with CR, TTP, and OS; many factors were significant for each endpoint. Next, significant factors were evaluated in multivariable (MV) analyses for each endpoint. Table 1 shows the significant (p<.05) traditional characteristics identified in MV models as independent factors by rank order.

Table 1 Multivariable Model

 CR TTP OS 
# Pts in Model 462 306 455 
# Events 106 213 260 
Characteristic Rank Order Significance in MV Model 
# Prior Tx 
b2M 
Salvage regimen 
Hemoglobin 
Albumin 
Platelet 
Alk Phos 
 CR TTP OS 
# Pts in Model 462 306 455 
# Events 106 213 260 
Characteristic Rank Order Significance in MV Model 
# Prior Tx 
b2M 
Salvage regimen 
Hemoglobin 
Albumin 
Platelet 
Alk Phos 

We evaluated the newer prognostic factors (IgVH MS, del 17p, ZAP70, and complex abnormalities by metaphase karyotype) separately in each final MV model to assess independent significant associations (Table 2).

Table 2 Significance of Factor in MV Model

Characteristic CR TTP OS 
Del 17p Not Sig. p<.05 p<.05 
IgVH MS Not Sig. p<.05 Not Sig. 
ZAP70 Not Sig. p<.05 p<.05 
Complex karyotype Not Sig. p<.05 p<.05 
Characteristic CR TTP OS 
Del 17p Not Sig. p<.05 p<.05 
IgVH MS Not Sig. p<.05 Not Sig. 
ZAP70 Not Sig. p<.05 p<.05 
Complex karyotype Not Sig. p<.05 p<.05 

When evaluating new factors in the TTP model, # of prior treatments remained significant. In the OS model, there was more variability regarding what remained significant in the model when each new factor was analyzed. Continued follow-up may identify IgVH MS as a significant factor in time to event models. The ideal method to evaluate these factors will be to include them all in the same model. Continued pre-treatment characterization of the newer as well as traditional factors in pts receiving clinical trial will enable such analysis. These analyses are important to develop expectations for pts going on salvage treatment as well as in understanding the biology of CLL. They indicate that traditional factors should not be abandoned. The full analysis with nomograms will be presented.

Disclosures: No relevant conflicts of interest to declare.

Author notes

Corresponding author