) --Last week, I laid out in detail seven reasons why the U.S. Food and Drug Administration could decide to reject
lymphoma drug pixantrone.
On Monday, the FDA posted a substantially
in advance of Wednesday's FDA advisory panel meeting. The FDA review hits on almost all
Cell Therapeutics' CEO Jim Bianco
FDA review of pixantrone
is posted to the agency's web site.
Add it all up, and Cell Therapeutics faces a very difficult task convincing the FDA's panel of cancer experts to recommend pixantrone's approval as a treatment for patients with advanced, aggressive non-Hodgkin's lymphoma.
Let's break down the issues I raised about pixantrone last week and see how that compares to what the FDA's reviewers said about the drug Monday:
1. Missing patients?
Last week, I questioned whether the FDA would accept the "positive" results from the phase III "EXTEND" study of pixantrone given that Cell Therapeutics only enrolled 140 of a planned 320 patients.
Recall that Cell Therapeutics has long claimed the FDA was OK with the smaller-than-expected enrollment in the study and that the study's statistical plan was adjusted accordingly. That turns out not to be true.
Monday, the FDA wrote:
"The planned sample size was 320. However, the study stopped early at an unplanned time point, due to poor accrual. A higher level of evidence is usually required in trials which discontinue prior to the final analysis. Based on the Rho family error spending function (Rho parameter = 2) used in the sponsor's statistical analysis plan and with 44% of planned enrollment, the significance level allocated for the submitted analysis would be 0.0096 (0.0014 based on the O'Brien-Fleming-type error spending function). Therefore, the submitted primary analysis would not be significant."
Translation: The pixantrone study failed, even using Cell Therapeutics' own statistical plan due to the low patient enrollment. In order for the pixantrone study to have been positive and statistically significant, the "p value" generated by the final analysis needed to be lower than 0.0096. The actual "p value" was 0.021.
2. Does Cell Therapeutics really have a Special Protocol Assessment?
A Special Protocol Assessment (SPA) is essentially a formal agreement reached between a drug company and the FDA that the design and endpoints of a phase III clinical trial are sufficient for a drug's approval.
Cell Therapeutics executives, including CEO Jim Bianco, have stated repeatedly, in public, that the EXTEND study of pixantrone was conducted under an SPA from the FDA. Many of the company's recent press releases, including one from
, also make the claim.
Here's what the FDA said Monday:
"On March 28, 2008, Cell Therapeutics notified the FDA of an early halt to enrollment for PIX301
the EXTEND study.. The study was not stopped at a planned interim analysis and early study stopping
invalidated the applicant's Special Protocol Assessment.
The applicant subsequently analyzed their data and began submission of a rolling NDA on April 13, 2009 with the last module submitted on June 22, 2009."
Let's repeat that. The FDA states, "The study was not stopped at a planned interim analysis and early study stopping
invalidated the applicant's Special Protocol Assessment.
How is Cell Therapeutics going to explain why it lied about having an SPA for the pixantrone study? An SPA was in place at one point in time, but Cell Therapeutics neglected to tell investors that the agreement was yanked.
Perhaps we now know why the company never mentioned having an SPA for the pixantrone study in its filings with the Securities and Exchange Commission, as I outlined last week.
3. How reliable are the response rates in the EXTEND study?
Recall that Cell Therapeutics says the "EXTEND" met its primary endpoint, with 25.7% of pixantrone patients achieving a complete response (CR) or an unconfirmed complete response (CRu) compared to 7.1% of patients in the control arm achieving a CR/CRu.
Last week, I said to watch how the FDA adjudicates the patients deemed to be unconfirmed complete responders (CRu) because in recent years, at least one influential academic group of lymphoma researchers no longer uses CRu as a valid measurement of disease response. Patients are now characterized as complete responders (CR) or partial responders (PR.)
The FDA's conclusions were even worse than I had anticipated:
"A non-aggressive histology was found, by central review, in 12/64 (18.8%) patients on the pixantrone and 16/66 (24.2%) patients on the comparator arm. This level of disagreement is consistent with the literature (Pathol Res Pract 1989 184:242, Cancer 1977 39:1071). Among patients with a non-aggressive histology by central review, 5 patients on the pixantrone and 1 patient on the comparator arm achieved a CR/Cru."
Translation: Cell Therapeutics enrolled a large number of patients who didn't suffer from aggressive NHL, per the eligibility rules of the phase III study. Therefore, five pixantrone patients deemed to have a complete response by Cell Therapeutics were thrown out in the FDA's analysis. Likewise, one patient in the comparator arm was similar excluded.
4. Were the patients in the comparator arm of the study treated with the best drugs available?
This was the only issue in my column last week not raised by FDA in Monday's pixantrone review.
5. How sick really were the NHL patients in the EXTEND study?
The FDA tackles this issue above when it found that not all the patients enrolled had aggressive NHL. Moreover, FDA found that the vast majority of patients enrolled outside the U.S. were less heavily pre-treated than the small handful of U.S. patients enrolled in the study.
6. Where in the world did pixantrone actually work?
Last week, I said to be wary of responses to pixantrone by geography, especially if no U.S. patients respond to the drug. That turned out to be completely true:
The FDA, in it review, notes that, "Most patients were accrued outside of the U.S., with 8 patients accrued from 6 sites in the US."Wow! Only eight patients out of 140 total were enrolled in the U.S.
The FDA also states, "Of concern, no patients enrolled in the US attained a CR or CRu."
7. Is pixantrone less cardio-toxic?
Pixantrone belongs to the anthracycline class of chemotherapy drugs, which are well known to cause heart failure at high doses and/or prolonged exposure. Pixantrone is designed to be less toxic to the heart, allowing it to be used in patients who have been treated previously with other anthracyclines, according to Cell Therapeutics.
But Monday, the FDA raised serious questions about pixantrone's safety, noting that, "deaths, SAEs (serious adverse events), and grade 3-4 events were all more common on the pixantrone arm.
Moreover, three pixatrone pateints died to due to heart failure compared to a single patient in the study's comparator arm, according to the FDA's analysis.
The FDA goes on to state, "All of this suggests that pixantrone is indeed cardiotoxic, but no conclusions can be drawn concerning its toxicity relative to other anthracyclines/anthracenediones.
When the FDA's advisory panel convenes Wednesday to review pixantrone, the cancer experts are being asked to vote on two big questions:
- The randomized study was stopped at less than 50% of its planned accrual because of poor accrual. Do the efficacy data support accelerated approval of pixantrone for the proposed indication?
- Is the risk:benefit ratio favorable for the proposed indication?
The advisory panel is going to vote "no" on both questions unless Cell Therapeutics pulls off a miracle.
-- Reported by Adam Feuerstein in Boston.
Follow Adam Feuerstein on
Adam Feuerstein writes regularly for TheStreet.com. In keeping with TSC's editorial policy, he doesn't own or short individual stocks, although he owns stock in TheStreet.com. He also doesn't invest in hedge funds or other private investment partnerships. Feuerstein appreciates your feedback;
to send him an email.