January 16, 2007
To Our Shareholders:
This is an important time for Northfield Laboratories. We have received many calls and messages regarding our preliminary Phase III trial results. I want to take this opportunity to summarize recent events and provide a clear picture of where Northfield stands today and our plans for the near future.
Background
On December 19, 2006, we reported preliminary top-line data from our pivotal Phase III trial. The trial was designed to assess the safety and efficacy of PolyHeme®, our human hemoglobin based oxygen-carrying red blood cell substitute, in the treatment of severely injured and bleeding patients when blood is needed but not immediately available. The reporting of these preliminary results was confounded by the fact that we discovered discrepancies in the dates of death for two patients in the database upon which the analysis was conducted. Since mortality is the primary endpoint, this is a critical variable, and it is therefore necessary to unlock the study database to make the corrections.
Since then, Northfield and our Contract Research Organization have developed and implemented a rigorous quality assurance program both internally and externally to verify all data that contribute to the primary mortality analysis. This program is underway with the goal of absolute accuracy. The internal audit is proceeding quickly, while the external review will take somewhat longer. It is our goal to have the entire process completed sometime next month. Once the database has been completely verified, it will be relocked and the data reanalyzed. It is possible that the final results may differ from the preliminary results. We will provide an updated report following completion of the data reanalysis.
--------------------------------------------------------------------------------
The PolyHeme Trial
The trial was designed to seek an indication for the use of PolyHeme in patients with urgent, life-threatening hemorrhage, when transfusion is indicated but blood is not immediately available. The goal was not to obtain approval for the use of PolyHeme in place of blood when blood is available. This is a critical distinction. The indication of unavailability of blood addresses a critical, unmet medical need, with the potential to provide life-sustaining capability when no adequate alternative is available.
It was not possible to conduct a randomized, controlled trial in a setting where no blood was used. The study therefore compared the survival of trauma patients receiving PolyHeme, starting at the scene and continuing for up to 6 units or 12 hours following injury before receiving blood, to the survival of patients who received standard treatment that did include early blood transfusion upon arrival at the hospital. This trial design is the basis of the agreed upon trial endpoints and provides the framework to accurately interpret the preliminary results.
The Trial Endpoints
The primary efficacy endpoint was Day 30 mortality. Before the trial began, an agreement with FDA was reached regarding the acceptable statistical boundary for noninferiority. The term “noninferior”, meaning not different or not worse, is a relative term that represents the predefined boundary for the comparison of mortality between the PolyHeme and control groups. In our study, that boundary for Day 30 mortality for PolyHeme patients, selected based on historical data and the trial setting, was set at up to 7% more than those patients with early access to blood. This does not mean that the actual observed mortality rate in the PolyHeme group could be 7% greater, but rather that the upper limit of that boundary, which is a statistical calculation known as the confidence interval, could not exceed that level. The observed mortality difference had to be considerably less than 7%, as occurred in the trial. The agreed upon boundary was based on the potential to provide a life-sustaining benefit in situations where transfusion of blood, the standard of care, was indicated but not available. Once again, this is a critical point.
The primary safety endpoints in the study were Day 1 mortality, Day 30 mortality, and durable serious adverse events (SAEs). So day 30 mortality was both an efficacy and safety endpoint. The predefined statistical assessment for safety was inferiority to control, meaning that the difference between the PolyHeme and control groups could not be statistically significant. This is a different statistical assessment than for efficacy.
Efficacy
In the primary modified intent to treat (MITT) population of 712 patients (those who were randomized and received some treatment) the 7% boundary was exceeded by only 0.3%. Furthermore, in the pre-specified per protocol (PP) population of 586 patients, the outcome was well below the 7% boundary, at 5.8%. The per protocol group represents those patients appropriately enrolled, randomized, treated, and followed as specified in
--------------------------------------------------------------------------------
the study protocol. This population represents the clearest opportunity to assess the potential therapeutic benefit of PolyHeme in the setting of this study.
The difference between the MITT and per protocol populations is comprised of the 126 patients considered “protocol violations.” Protocol violations include such issues as enrolling subjects that are not eligible for the study, giving enrolled patients the wrong treatment either by accident or intentionally, or not treating the patient at all after enrollment. We are reviewing the data from these patients in detail in an effort to understand the nature and type of violations that occurred, and to identify any patterns that might emerge.
Safety
With respect to Day 1 and Day 30 mortality, as well as durable SAEs, the preliminary analysis reveals there was no statistically significant difference between the treatment and control groups in the MITT population. Therefore the safety analysis indicates that the outcome in patients receiving PolyHeme was not inferior to the outcome of patients receiving blood.
Going Forward
I know shareholders are interested in our future plans. We intend to submit a Biologics License Application, or BLA, to FDA later this year. To accomplish this there are some intermediate steps we need to take. First, we need to completely analyze and understand the final data, including additional safety data from the study. We will then prepare a summary for FDA, and submit this for their review and feedback as we refine our future actions.
In summary, this study is a comparison of trauma patients receiving PolyHeme without early blood to patients receiving early blood transfusion. This trial is not a comparison of PolyHeme to salt water or placebo treatment. We believe PolyHeme has the potential to provide life-sustaining capability when blood is not immediately available, and thus address an unmet medical need. In addition, PolyHeme offers many additional benefits: universal compatibility, immediate availability, viral reduction, and an extended shelf-life. Since we are seeking an indication for which there is no alternative, these are important attributes that will be part of the risk-benefit analysis for PolyHeme. That was the basis of this trial and will be the basis of our BLA.
We appreciate your continued support of our goal of bringing the life-sustaining potential of PolyHeme to realization. We remain encouraged with our status and confident about our future, and look forward to providing further updates as the year progresses.
Sincerely,
/s/ Steven A. Gould, M.D.
Steven A. Gould, M.D.
Chairman and Chief Executive Officer |
