Diffuse Large B-Cell Lymphoma (DLBCL) is the most common blood cancer with 18,000 new diagnoses each year (lymphoma.org). It is typically treated with the five-drug combination R-CHOP, which includes the following drugs:
Rituximab (monoclonal antibody targeting CD20 on B-cells)
Cyclophosphamide (causes crosslinking of DNA)
Hydroxydaunorubicin (doxorubicin, topoisomerase poison)
Oncovin® (vincristine, antimitotic drug)
Prednisone (Corticosteroid)
This drug combination was established in the early 2000s (Coiffer et al. 2002), and all trials that have attempted to improve on this combination have failed (Ibrutinib, Lenalidomide, Bortezomib). However, recently, the pharmaceutical company Roche announced that their Phase III POLARIX trial had overcome this hurdle, stating in a press release that the “study shows Roche’s Polivy plus R-CHP is the first regimen in 20 years to significantly improve outcomes in previously untreated [DLBCL] compared to standard of care.”
“study shows Roche’s Polivy plus R-CHP is the first regimen in 20 years to significantly improve outcomes in previously untreated [DLBCL] compared to standard of care.”
Polivy is an antibody-drug conjugate with the antimitotic drug vedotin linked to polatuzumab, a monoclonal antibody targeting CD79B on the surface of B-cells (Figure 1). Polivy replaces vincristine in the five-drug combination and acts as a much stronger antimitotic drug that has the advantage to specifically targeting B-cells. Thus, I hypothesize that R-CHP + Polivy will behave like R-CHOP, where a higher effective dose of the antimitotic drug reaches the cancer cells.
In my current research project, I have developed a mechanistic model of curative combination chemotherapy that explains the results of R-CHOP and other DLBCL clinical trials (Figure 2). I used this model and existing data about the activity of Polivy in relapsed/refractory (r/r) DLBCL patients (Figure 3, top panel) to predict the results of the POLARIX trial before the results were officially released at a major medical meeting.
Using the existing model of R-CHOP treatment (Figure 2), I simulated the results of an existing clinical trial comparing Rituximab + Bendamustine (a drug that, like cyclophosphamide, causes DNA cross-linkages) to Rituximab + Bendamustine + Polivy in patients who had already been treated with R-CHOP and had a progression event (e.g., relapse) (Figure 3). This trial was chosen because it shows the effect of Polivy compared to a control arm. Fitting the simulation (bottom panel) to the clinically observed response (top panel) suggested that, assuming Polivy acts like a stronger version of vincristine, the effect of Polivy was equivalent to giving 3.5 times the dose of vincristine.
I then used this to simulate what would happen if patients were given R-CHP + Polivy instead of R-CHOP assuming that Polivy is equivalent to giving 3.5 times the dose of vincristine. The results of this trial show an increase in progression-free survival (Figure 4), with a simulated hazard ratio (HR) of 0.812 with a 95% confidence interval of 0.668-0.988 when 400 patients were enrolled in each arm. This matches the minimal information included in the press release: R-CHP + Polivy is a significant improvement over the current standard of care, R-CHOP. To determine how well the progression-free survival curves match the clinically observed data, we’ll have to wait for the complete trial results to be published.
One of the main limitations of this result is that we do not know how well the R-CHOP control arm will perform. PFS with R-CHOP treatment varies from trial to trial depending on the trial's enrollment criteria and any potential enrollment bias. Across various historical trials of R-CHOP, the 5-year PFS varies as much as 30% (compare Coiffer et al. 2002 and Pfreundschuh et al. 2006). This estimation includes my personal estimate of where the trial will fall, with ~40% of patients having a progression event when treated with R-CHOP, supported by the statistic included in the Roche press release: "newly diagnosed DLBCL patients as currently 40% of patients relapse." The magnitude of benefit of Polivy and the HR values could change significantly if the PFS of the control arm is, in fact, higher or lower. However, this is something we can also predict (follow-up blog post coming soon!).
This result demonstrates the potential of the mechanistic model of curative combination therapies that I have developed to predict clinical trial outcomes before the trial is completed. Since the average oncology clinical trial costs $60k/patient (source), a trial enrolling 400 patients costs at least $24 million to conduct. When conducting a trial with such a high financial cost and a large participant burden, we want to be as sure as possible that a trial will succeed and have a reliable estimate of the magnitude of benefit a new treatment will confer compared to the current standard of care. This model has the potential to provide such insights before a trial is even conducted, reducing the number of failed clinical trials and providing guidance as to how the trial should be designed for the best possible success.
Disclaimer: This work is a work in progress. The methods and details of the model have been intentionally excluded since this work has not yet been published in a peer-reviewed journal.
Acknowledgements: I would like to thank my advisor, Dr. Adam Palmer, for his continued support of my research and the rest of the Palmer lab, Deb Plana, Sarah Cooper, Hannah Hwangbo, and Jacob Pantazis, for their feedback and moral support.
Comments