We’ve received tons of emails and comments about Pixantrone (Pix) over the last several weeks. The tension regarding the Pix data at ASCO and an impending? deal over the next few months has been mounting. There has also been no shortage of battles between bulls and bears.

To provide further insight, we’ve attempted to dig a bit deeper and analyze the clinical data generated to-date that will be used to support (or is being used) an NDA filing of Pix. Pix has a long history and has been in a number of different clinical trials, most of which showed mixed results or were stopped for slow enrollment. Therefore, we will not address every study that’s ever been done with Pix, but will just address the studies that are being used to support registration.

To better understand the status of the drug and postulate what might happen, we have to look at Pix’s history and the events that led up to now. Let’s be clear on something: this post will not discuss commercialization potential, deal structures or potential scenarios with partners. This post is strictly to organize the history of events of Pix, analyze the publicly available clinical and regulatory information and make a hypothesis regarding approvability.

OK, let’s do it…

About Pixantrone

Pix is a novel anthracycline derivative that is presumed to be devoid of the putative cardiac toxicity generating 5,8-dihydroxy groups. It was designed to have a better activity and safety profile than other anthracyclines/anthracenediones.

Special Protocol Assessment (SPA)

In our opinion, the SPA is what is the key to approval. Remember, even if the primary endpoint is met, the FDA is not obligated to approve Pix. There could be a red flag in the data that was not disclosed to the public, or there could be something in the SPA that was not addressed or fully elucidated in the clinical trials. The actual SPA documentation/communication between the company and the agency is confidential, as it is with most companies, and the details have never been released to the public. It’s hard to fully appreciate the registration path without examining these materials and comparing them to the trial results. Also, CTIC met with the FDA several times, outside of the SPA, since it began developing Pix and without knowing what was discussed, we can only assume so much. Companies don’t disclose the documentation to the public and reviewing of an NDA, FDA meeting minutes, etc is reserved for investors and potential partners to review under CDA (there’s even some resistance to show these to investors at times).

Per the SPA, the FDA is allowing data from two trials and two trials only to support efficacy and safety. In our opinion, these two trials are all that matters for approval. The other dozen + trials that have been conducted and failed, stopped, etc. are not put into our equation.

The SPA – 2004

CTIC announced in March 2004 that the FDA had “provided guidance” through the SPA process for a randomized pivotal trial of Pix in patients with relapsed, aggressive NHL. What’s interesting is that based on our research, we didn’t find a press release from the company that confirmed it had actually received an SPA. The March 4, 2004 press release said that “the trial protocol and supporting data are in the final stages of review with the FDA,” but we couldn’t find confirmation after that. Many press releases from various sources (Reuters, BioSpace, etc) since then always refer to CTIC receiving an SPA in 2004, but we failed to find any verification of an SPA for PIX-301 (which we’ll get to). CTIC did, however, put out a press release in 2007 that it received an SPA for its PIX-303 trial (fludarabine, Pix and rituximab vs. fludarabine and rituximab) in relapsed or refractory indolent NHL. Note: PIX-303 was stopped in 2008 to conserve resources.

Let’s make this clear: we’re not saying CTIC doesn’t have an SPA for the EXTEND trial, we just couldn’t find any evidence/confirmation that one was received.

OK, let’s go over the EXTEND trial…

PIX301 – EXTEND Trial

The trial was initiated in 2004 (under an SPA?). Here’s the highlights:

• Phase 3 registration study

• Single-agent treatment as salvage regimen in patients with relapsed aggressive NHL who have failed standard induction chemotherapy and at least one other combination chemotherapy regimen

• Randomized patients get either Pix or another single-agent drug of the physician’s choice presently used for the treatment of this patient population (choose: vinorelbine, oxaliplatin, ifosfamide, etoposide, mitoxantrone, gemcitabine, rituximab or other options, depending on the country).

• Pix 85 mg/m2, (days 1,8,15) q 3 weeks.

• Per press releases, outcome measures include CR (complete response) or uCR rates, PFS (progression-free survival) and overall survival.

• Patient size was designed for N=320, enough to power the study CR rate assumption of < 5% for the control arm and a 10% improvement in CR rate for the Pix arm.

PIX301 – 2006 Interim Analysis

CTIC announces that results from 40 patients treated to-date are “encouraging”; supports continuing and expanding the trial.

PIX301 – 2007 Interim Analysis

CTIC says, after discussions with the FDA, that it will conduct a full analysis of the trial instead of an interim analysis as previously planned. Though no reason was given, slow enrollment is presumed to be the reason for this. The FDA says that using less than N=320 “could be acceptable” for NDA submission if it can show a statistically significant (p<=0.05) difference between treatment arms.

Interestingly, the FDA also says that randomized safety data from the RAPID (PIX203) study (CHOP-R vs. CPOP-R), which we mention below, can be used to support the PIX301 results.

PIX301 – November 3, 2008

CTIC closes the data set for preliminary analysis of “the primary endpoint for patients with relapsed diffuse large B cell non-Hodgkin’s lymphoma.” CTIC also confirms here that the primary endpoint is CR and uCR.

PIX301 – November 11, 2008

CTIC says Pix works, achieving its primary endpoint (CR/uCR) in patients with “advanced, relapsed aggressive non-Hodgkin’s lymphoma” based on a preliminary intent to treat (ITT) efficacy analysis. Here’s the results:

• Pix arm: 70 patients with 14 uCR/CRs (20%) — 8 patients with confirmed CR.

• Standard chemo arm: 70 patients with 4 uCR (5.7%) — no confirmed CRs.

• p = 0.02.

• Side effects grade 3 and 4 neutropenia and febrile neutropenia, consistent with all earlier trials.

We should also note that this is the first time we get details on total patients: 140 patients at 130 trial sites, so about 1 patient per site. We assume again, this slow enrollment was the reason for going to full analysis and completing the study. Note: the study here loses half its power because we are not at the predetermined N=320 as described in the SPA.

PIX301 – 2009

CTIC says that after FDA discussions, it will begin a rolling NDA and request priority review for Pix to treat “relapsed aggressive non-Hodgkin’s lymphoma” in 1Q09. CTIC also says in January that the secondary endpoint was met: PFS. Patients on Pix had a statistically significant improvement in median PFS, vs. other single-agent chemotherapeutic agents (4.7 months vs. 2.6 months, p < 0.01) based on an intent to treat analysis.

CTIC says in its SEC filing that Pix patients had…more severe cardiac events (5 vs. 2) than in the control arm.

PIX-203 – RAPID Trial

Let’s backtrack a bit. The RAPID (A Trial in Patients With Diffuse Large B-Cell Lymphoma Comparing Pixantrone Against Doxorubicin) trial is a Phase 2 study of Pix substituting pixantrone (the 1st P) for doxorubicin (the H), or CPOP plus rituximab (CPOP-R) regimen vs. CHOP plus rituximab (CHOP-R). The trial was conducted in treatment-naive stage II–IV diffuse large B-cell lymphoma patients. The trial (safety?) results will be used to support the ongoing NDA submission of Pix in the EXTEND trial.

An interim analysis in July 2007, showed that:

To date, a majority of patients on both arms of the study achieved a major objective anti-tumor response (complete response or partial response). Patients on the pixantrone arm of the study had clinically significant less left ventricular ejection fraction (LVEF) drops, infections, and thrombocytopenia (a reduction in platelets in the blood), as well as significant reduction in febrile neutropenia.

The RAPID trial was closed in early 2008. CTIC said it had “adequate sample size to demonstrate differences in cardiac events and other clinically relevant side effects between pixantrone and doxorubicin.”

The full results will be reported in 4Q09, far behind anticipated acceptance of NDA filing by the FDA and coinciding with anticipated Pix approval. Here’s the ASH poster from 2007.

BIO 2009 & ASCO 2009

CTIC will be discussing the EXTEND results this Tuesday at BIO and at ASCO in two weeks.

Assessments, Questions & Answers

We had two big concerns as we traced the archives of Pix:

1. Why was there no announcement to confirm Pix had an SPA and if one was received, what are the details?

2. Pix was being touted as a safer and more efficacious anthracycline. Is this still the case? Let’s discuss.

Starting with #2 above, we’re still unclear if Pix is an anthracycline that works in patients who have failed or relapsed on other anthracyclines, but with some cardiac issues OR if it has similar efficacy to other anthracyclines, but has a better cardiac safety profile.

There is a difference here, and based on what we’ve learned so far, it’s difficult to assess how Pix can be profiled. If we had to make an assumption, we would say Pix has a solid efficacy profile that can be used in a salvage setting without a relatively high incidence of cardiac problems compared to other anthracyclines. The higher number of cardiac events (5 vs. 2) is concerning, although CTIC has communicated to us that there was no change in median cardiac function. This data will be presented in detail on Tuesday.

Previous studies of Pix in both animals and humans have demonstrated that the drug is not devoid of cardiotoxic potential. Decreases of the LVEF have been demonstrated by MUGA scan, although very few of these changes have been clinically significant or symptomatic even in patients previously treated with other anthracyclines [El-Helw et. al].

As for the SPA, we would like better clarity on the SPA details. Like we said at the beginning of the post, we didn’t find an actual press release that clearly stated an SPA was received, although every press release by the company and other sources since 2004 state that there is an SPA. Again, the FDA is not obligated to approve Pix and something could pop-up in the data analysis despite having met its primary endpoint.

Regarding the ongoing rolling NDA, we think it’s a positive that the FDA has been supportive of the trial modifications and the allowance of the RAPID data to support safety concerns. On the topic of the RAPID study, there’s still not a lot of clarity regarding the details or results aside from what we mentioned above. The RAPID trial was also completed over one year ago, but final results will not be available until 4Q09 — all this despite the fact that the data is being submitted along with the EXTEND data.

Our final questions, therefore are:

1. Is a sample size of 140 patients enough for approval and will the additional patient data from the RAPID trial make up for the lack/loss of power? Treanda was approved in a single arm study with about N=100. Is the RAPID trial needed to support safety issues and if so, why? This could be concerning (see #2 below).

2. The RAPID trial looked at treatment-naive patients. The EXTEND trial is in experienced patients. In our opinion, there’s a definite disconnect here and we ask if whether safety in treatment-naive patients can be compared to safety in experienced patients. Is this comparing apples and oranges?

3. Is CTIC going for advanced, relapsed aggressive NHL (a broad indication) or relapsed diffuse large B cell NHL, the latter being a specific subtype of NHL? If diffuse large B-cell, what are the results for that specific patient population?

4. What were the other treatment regimens/drugs used by investigators? Can comparing Pix against one drug vs. another skew the results in either direction? Were positive Pix results better for one drug vs. another drug?

5. Finally, based on communication with CTIC, all responses in heavily pre-treated anthracyclines were uCRs or PRs. No CRs. How does this affect the final analysis?

Conclusions

The registration path has been modified over the last several years, but CTIC appears to have stayed in constant communication with the FDA. The supportive response by the FDA is a positive and the achievement of the primary endpoint is excellent news. The drug appears to have worked. Being able to show a response in patients where palliative treatment is the only option is a +. In a salvage setting, the FDA will likely be positive if a drug can extend lives. If this were a me-too drug, it would not get approved, but again, because it’s demonstrating activity, albeit with questionable cardiac issues, it has a good shot based on what’s been disclosed to-date. Concerns still remain regarding the details of the FDA communications over the years, the low power of the EXTEND trial, the final RAPID details, and what the FDA will say regarding the cardiac events.

Pixantrone, Hype or Truth: More likely truth than hype based on the setting.

No positions or disclosures in CTIC, partners or competitors

Related Posts