Assessment of Tumor Growth (g) and Regression (d) Rate Constants in Patients with Multiple Myeloma (MM) Shows That the Superiority of Bortezomib with Liposomal Doxorubicin (PLD+B) Over Bortezomib Alone (B) Is a Result of Reduced Growth of Refractory Tumor Cells and Not of Higher Regression Rates, and Provides An Earlier Efficacy Endpoint That Allows for Comparison Across Trials

Discerning amongst the many new therapy options in MM would be helped by novel assessment strategies that yield answers from smaller trials and allow for comparisons across trials. M-protein profiles during therapy reflect simultaneous tumor regression and growth. In breast, prostate and renal cell cancer we have validated a novel assessment method that quantifies tumor regression (d) and growth (g) rate constants using tumor quantities obtained while patients are treated in a trial. Unlike incremental measures of efficacy such as PFS/TTP or ORR, g is a continuous variable that accurately assesses differences between treatments. We have shown: (1) the principal effect of active therapies is a slowing of g, (2) g correlates with survival and (3) g can be accurately compared across trials because it is indifferent to assessment intervals and methodologies (Stein et al, Clin Can Res 2010). We utilized this method to evaluate a phase III trial of PLD+B vs B alone in relapsed/refractory MM (Orlowski et al, JCO 2007).

Using M-protein values and a two-phase mathematical equation we determined d and g.

The median g value in pts receiving PLD+B (0.000738) was significantly less (p=0.0012) than that observed in patients receiving B alone (0.00143); however both therapies had similar effects on d (PLD+B=0.01403; B alone=0.01400; p=0.456) indicating the superiority of PLD+B resulted from a greater effect on the growth of residual tumor and not greater cell kill. Consistent with this, the fraction of tumor sensitive to therapy was similar for PLD+B and B. A sustained statistical difference [p<0.05] in median g values between treatment arms was observed after only 370 pts enrolled, compared to 485 for a difference in PFS, suggesting using g, the difference between arms could have been determined in a smaller trial. Furthermore, in an individual patient, reliable, statistically valid g and d values could be estimated early in their evaluation, while M-protein values were still declining, long before M-protein levels increased. Thus, our approach discerns the concomitant growth of resistant clones as sensitive ones are eliminated, providing an earlier indicator of treatment failure.

The superiority of PLD+B over B alone is shown to be due to an effect on growth of resistant cells; not an acceleration of tumor cell kill. Statistically valid differences between PLD+B and B could be discerned early and the rate of tumor growth in individual patients could be quantitated before increases in M-protein were measured. Further collaborative studies could establish g as a valuable measure of efficacy in MM that allows for cross trial comparisons, the conduct of smaller trials and an earlier assessment of treatment failure.

Xiu:Johnson and Johnson: Employment, Equity Ownership. Zhuang:Johnson and Johnson: Employment, Equity Ownership.