NeuVax is a peptide-based vaccine, under development by Galena Biopharma (GALE), which aims to prevent recurrence of breast cancer in women who have undergone standard of care therapy, including surgery, radiation and chemotherapy. It is the lead drug in Galena's pipeline of cancer immunotherapy and hematology agents. The Special Protocol Assessment or SPA agreement with the FDA for the current NeuVax study specifies a minimum 30% efficacy (i.e., reduction in relative risk of recurrence) for approval. The Phase 3 NeuVax PRESENT trial calls for an Early Interim Analysis after the 70th event - the possible outcomes include a halt for futility, proceed with modifications, or unconditional proceed. If Galena proceeds, the study will reach its primary efficacy endpoint in April 2018 after a three-year follow-up of the last enrolled patient. This represents the roadmap for approval of NeuVax.
Galena is a biopharmaceutical company with all the attendant issues of an extended timeframe without commercial revenue and high development costs. The current balance sheet cash position for Galena is $34.8 M as of September 30, 2015 with a quarterly burn rate of $11 M to $13 M (2016 Outlook). With the settlement of overhanging shareholder derivative and securities class action lawsuits, a major remaining downside risk is the possible failure of its drug pipeline, especially NeuVax. In the balance, a successful PRESENT trial and subsequent FDA approval of the first-in-class NeuVax vaccine point to annual sales in the U.S. market alone of greater than two billion dollars (Galena presentation January 4, 2016).
Recapping the first installment of this analysis, the NeuVax Phase 1/2 breast cancer combined disease-free survival or DFS results reveal several striking signal features, including (1) no recurrences for the first eight months in any woman who received at least one dose of NeuVax vaccine, and (2) a persistent difference in the long-term steady relapse curves which narrowly missed a p-value of 0.05 (p<0.08) for all women treated with the NeuVax vaccine (including suboptimally-dosed and unboosted patients). Pictured in the above graph is the Phase 2 response of women with node-positive breast cancer whose optimized treatment regimen most closely resembles that of the final Phase 3 PRESENT trial design, the Phase 3 Target Patient Population. Beginning with these results, we will develop a model to provide perspective to the upcoming interim analysis anticipated in Q1 2016.
The salient data points from the Target Patient Population chart are (1) the baseline breast cancer DFS rate at 60 months (20/27 = 74.1%), (2) the baseline breast cancer DFS rate at 36 months which is the Primary Endpoint of the PRESENT trial (21/27 = 77.8%), and (3) the elbow of the baseline breast cancer DFS curve at 20 months (22/27 = 81.5%). A historical hazard function and corresponding recurrence curve from a large study of women with Stage II breast cancer (i.e., invasive breast cancer which generally involves local lymph nodes) is superimposed on the graph as the starting point for a model of NeuVax efficacy (from Cheng et al., Cancer Epidemiol Biomarkers Prev 2012: 21(5), 800-9).
A fitted hazard function and corresponding recurrence curve for the NeuVax PRESENT trial model is shown in red. The peak hazard rate is about 2% higher than historical controls while the long-lasting tail is much lower; nevertheless, it replicates the best estimate of the baseline breast cancer DFS rate for the current PRESENT trial.
Two limitations of the original Phase 1/2 studies deserve mention. Firstly, these trials vaccinated all HLA-A2/A3+ patients and utilized HLA-A2/A3- patients as controls. HLA-A2 is an independent negative clinical prognostic factor in patients with certain cancers (Masucci et al., J Nucleic Acids Investig. 2010: 1(1)) - this would tend to lower the noise floor of baseline breast recurrences, but its significance in the PRESENT trial is unknown. Secondly, the trials did not use GM-CSF (an immunoadjuvant) in their control groups. Although this raises the question of whether the clinical responses were the result of GM-CSF alone, the investigators thought this unlikely based on the Phase 2 GP2 trial with a GM-CSF-only arm (Mittendorf et al., Cancer 2012 May 15: 118(10), 2594-2602). In fact, the DFS rate in the GP2 node-positive HLA-A2+ GM-CSF-only Intent-to-Treat control group was only 81% at a median 34 month follow-up period (Mittendorf et al., ASCO Breast Cancer Symposium 2014), very similar to the 79.5% DFS rate of the NeuVax SN-33 Phase 1/2 Intent-to-Treat control group at 34 months follow-up. The current Phase 3 PRESENT trial corrects both of these deficiencies.
A series of breast cancer DFS curves, representing varying levels of NeuVax vaccine efficacy (i.e., reduction in relative risk of recurrence), was generated using a Monte Carlo simulation with 10,000 trials per data point. Of note is (1) the 78% effective DFS curve shown in dark blue - this is the expected efficacy demonstrated in the Phase 2 Target Patient Population group of women; and (2) the 30% effective DFS curve shown in yellow - this is the threshold efficacy required to satisfy the Special Protocol Assessment agreement for approval of NeuVax by the FDA. Also shown are the 95% confidence intervals for each level of NeuVax vaccine efficacy in the PRESENT trial model.
While survival curves display the progression of breast cancer in these women as if they all commenced their NeuVax treatments at the same time, we now need to account for their differing enrollment dates over the course of the ongoing NeuVax PRESENT trial. Six key data points have been publicly disclosed: (1) the 1st patient enrolled January 19, 2012, (2) the 70th patient enrolled May 24, 2013, (3) the 533rd patient enrolled September 22, 2014, (4) the 599th patient enrolled October 31, 2014, (5) the 700th patient enrolled February 9, 2015, and (6) the 758th patient enrolled April 15, 2015.
The 16-month undocumented stretch from May 2013 to September 2014 presents a challenge to reproducing the enrollment curve. To make a better estimate, a conservative smooth enrollment curve (in blue) with late entry for the bulk of volunteer breast cancer patients was first developed. The simulation model assumes that recruitment of breast cancer volunteers for the treatment and control groups occurs in parallel in a 1:1 ratio.
An aggressive smooth enrollment curve (in yellow) with early entry for a larger portion of volunteer breast cancer patients was then developed to bracket the probable true enrollment curve. The resulting simulated breast cancer recurrence curves show the projected 70th breast cancer recurrence date to be not extremely sensitive to these variations. Focusing on the 70th breast cancer recurrence date is key because this event triggers the Early Interim Analysis in the Phase 3 PRESENT trial protocol. The final model enrollment curve represents a rough average of the two curves.
This chart shows the final enrollment curve and resulting projected breast cancer recurrence curves. As you can see, increasing NeuVax efficacy shifts the recurrence curve down and to the right (i.e., later in time) because women in the treatment half of the study group have a reduced rate of recurrence for their breast cancer. Two important features are revealed: (1) the 78% efficacy recurrence curve, equivalent to the Phase 3 Target Patient Population efficacy, crosses the 70th event threshold in October 2015, while (2) the 100% efficacy recurrence curve crosses that threshold in March 2016. These two time points delimit the projected range of expected results based on the Phase 1/2 findings and coincide with the original Q4 2015 / Q1 2016 guidance provided by Galena Biopharma. Dr. Mark Schwartz, President and CEO of Galena, has publicly stated that the 70th breast cancer recurrence event is expected in Q1 2016 (third quarter 2015 earnings conference call, November 9, 2015) - this corresponds to a mean efficacy of greater than 92% in the current simulation model. By comparison, Herceptin had a range in efficacy (i.e., reduction in relative risk of recurrence) of 33% to 52% in four pivotal trials (source).
To avoid the Flaw of Averages (à la Dr. Sam Savage), the NeuVax PRESENT trial model was used to generate the distribution of all possible dates for the 70th breast cancer occurrence event. While the output of the program is a series of cumulative distribution functions, orthogonally summing the monthly 70th events from that data and normalizing over the range of efficacy categories produces the probability density functions shown above. Finally, computing the area under each probability density function yields the desired cumulative probability distributions for the 70th breast cancer recurrence by month.
A 70th breast cancer recurrence event in early January 2016 (the most conservative projected date) thus predicts (1) no possibility of NeuVax efficacy equaling 30% or less (the minimum Special Protocol Assessment threshold required for approval by the FDA), (2) less than a 0.5% chance of NeuVax efficacy equaling 50% or less, (3) a 35% chance of NeuVax efficacy between 51% and 78%, and (4) a 65% chance of NeuVax efficacy greater than 78%. This is indeed a strong and highly encouraging signal!
In conclusion, I hope I have increased the signal-to-noise ratio so you too can clearly appreciate the signs of a successful Phase 3 NeuVax vaccine PRESENT trial. This is a singular achievement which promises to help women who live each day with the real threat of their breast cancer recurring and stealing them away from us all.
David M. Fujii, M.D.
Pediatric Cardiac Anesthesiologist and Electrical / Mechanical / Aeronautics & Astronautics Engineer
Disclosure: I am/we are long GALE.
I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.
Editor's Note: This article covers one or more stocks trading at less than $1 per share and/or with less than a $100 million market cap. Please be aware of the risks associated with these stocks.