Aastrom Bio: Final RESTORE Data Looks Solid

| About: Vericel Corporation (VCEL)

Full Data From RESTORE-CLI Presented at AHA

On November 14, 2011, Aastrom Bio reported the final analysis from the phase 2b RESTORE-CLI trial. The RESTORE-CLI clinical trial assessed the safety and ability of ixmyelocel-T, Aastrom's expanded, bone marrow-derived cellular therapy, to restore damaged tissue in a lower extremity affected by critical limb ischemia (CLI). The results include the final analysis from all 86 randomized patients.

Of the 86 patients enrolled, 77 underwent the bone marrow aspiration procedure and 72 received treatment injections, 48 ixmyelocel-T and 24 placebo. Patient characteristics where as follows:

- Age 18-90

- Diagnosed with CLI

- Ischemic rest pain > 2 weeks duration

- Ulcerations or gangrene of toe / foot

- Toe systolic pressure <50 mmHg

- Ankle systolic pressure < 70 mmHg

- No previous amputation or exposed tendon or bone in wound.

- No active infection of target extremity

- HbA1c < 10%

- No option for revascularization

The study was powered as a safety trial with the primary endpoint in adverse events. However, there were significant efficacy endpoints, which include the primary endpoint time to first occurrence of treatment failure ((NYSE:TTF)) at 12 months. TTF is a composite endpoint that includes four components:

1 - Major amputation of the treated leg

2 - All-cause mortality

3 - Doubling of wound size from baseline

4 - De Novo gangrene

The Data

The chart below depicts the product-limit survival function showing a 62% risk reduction (hazard ratio 0.38, 95% CI = 0.20-0.74) in the composite endpoint of time to first occurrence of treatment failure (TTF) for patients taking ixmyelocel-T vs. the placebo.

There were 19 events (40%) for the ixmyelocel-T group vs. 16 events (67%) for the placebo group, this equated to a p-value = 0.0032.

Events within the composite TTF endpoint include:

Below is a graphical representation of each composite endpoint comparing ixmyelocel-T to the placebo control. It shows the time to individual treatment failure on study day vs. the median number of days to first event.

(Click to enlarge)

The secondary efficacy endpoint was Amputation-Free Survival (AFS) at 12 months. This is important because AFS at 12 months will be the primary endpoint in the companies phase 3 REVIVE trial. Based on the data above, the AFS-12 for the ixmyelocel-T group was 75% vs. 67% for the placebo control group. This accounts to a 32% risk reduction (hazard ratio 0.68, 95% CI 0.28-1.65). Results were not statistically significant (p-value 0.3880) given the small sample size.

Subset Analysis Sheds Additional Light On Data

The phase 2b RESTORE-CLI program enrolled both Rutherford-4 and Rutherford-5 patients. We note that the phase 3 REVIVE program will focus solely on “more sick” Rutherford-5 patients. From RESTORE-CLI, 45 of the 72 patients qualified as Rutherford-5 (29 in the ixmyelocel-T group and 16 in the control).

For the primary composite endpoint of TTF, Aastrom saw greater separation between the ixmyelocel-T group and the control. Results show a TTF for ixmyelocel-T at 44.8% vs. 87.5% for the control, p-value < 0.0001. This equated to a 77% risk reduction (hazard ratio = 0.225). This was greater statistical significance than in the total population.

(Click to enlarge)

The subset analysis of the Rutherford-5 patients on the secondary endpoint yielded impressive separation between the ixmyelocel-T group and the control. The Rutherford-5 AFS rate for the ixmyelocel-T group was 79%, lower than the AFS rate for the overall ixmyelocel-T population at 75%. Results showed even further separation from the control, which showed an AFS rate of 56%, down from the overall control population at 67%. This is a 61% risk reduction (hazard ratio of 0.391). Results were not statistically significant (p=0.0802), but we note the small sample size of only 45 patients, and remind investors that the phase 3 REVIVE program will be nearly 600 patients.

(Click to enlarge)

All in all we are pleased with the final results from RESTORE-CLI. We think the trial demonstrates excellent proof-of-concept in patients with no option CLI, especially those with confirmed tissue loss. The design of REVIVE is solid. If Aastrom can reproduce similar results in REVIVE to what was presented at AHA on November 14, 2011, we believe the trial will be successful. Ixmyelocel-T could be under review by the end of 2014.

REVIVE Ready To Initiate

REVIVE-NO will seek to enroll up to 594 patients and approximately 80 clinical centers across the U.S. with severe CLI. These “no option” patients must have confirmed existing tissue loss (e.g. ulcerations and/or dry gangrene) and disease progression to qualify for enrollment. The trial is 90% powered to show superiority of ixmyelocel-T over the control (no cells). Management used an event rate for the control group of 34.5%, while expecting that ixmyelocel-T will come in around 22%, to calculate the power assumptions. The primary endpoint of the program will be amputation-free survival at 12-months (AFS-12). Other secondary endpoints will include de novo gangrene and wound healing.

Management has taken a number of steps to improve the trial design compared to the phase 2b RESTORE trial.

1 - Aastrom will employ an independent steering committee of outside physicians who have direct experience with CLI programs to help oversee and monitor the trial.

2 - Aastrom will utilize a centralized review committee comprised of experienced vascular surgeons to monitor patients and set a standardized disease progression and amputation criteria.

3 - Management has established rigorous entry criteria to help standardize the baseline characteristics of each “no option” patient. This establishes a second layer of qualification beyond the treating physician. The goal will be to enroll only Rutherford-5 patients.

4 - To limit geographical variability and standard of care variability, the trial will be conducted at approximately 80 sites in only the U.S.

We expect this trial to begin enrollment shortly with the majority of sites to come online shortly after the trial initiates. We expect that Aastrom may need to screen as many as 750 patients in order to enroll (treat) near the target 594 number. The trial should take 18 months to complete and cost between $20 and $25 million. If the results are positive, based on the agreed upon SPA with the FDA, management plans to file the Biologic License Agreement seeking approval for this “no option” patient population. We consider the ability to file after only one phase 3 program as positive news, as management can now focus on this trial and seek approval with positive data while negotiations with the FDA on the “poor option” trial separately.

We note that tissue loss and de novo gangrene are the two biggest predictors of future amputation. In fact, patients with increased tissue loss and gangrene are 5X more likely to experience a major amputation than patients with stable disease. It is clear to us that Aastrom (and the FDA) are looking at tissue loss as a key enrollment criteria for the phase 3 trial. Final results from RESTORE-CLI show that patients with confirmed tissue loss (Rutherford-5) actually responded better than “less sick” patients.

Remaining Bullish On Aastrom

We’ve first upgraded Aastrom Bio after the stock sold off hard following release of the second interim update from the phase 2b RESTORE trial. At the time we believed the data was not representative of what the final outcome would be, and as it turns out we were right. The final data (presented above) is impressive in our view.

In the meantime, there are a number of catalysts to continue to drive the shares higher, including initiating enrollment in REVIVE-NO, finalizing the SPA for REVIVE-POOR, final 12 month data from the DCM surgical study, interim 6 month data from the DCM catheter study, and initiating the REVIVE-POOR trial. News that management may seek to file for approval of ixmyelocel-T after REVIVE-NO is a significant positive in our view and greatly lowers the risk of investing at this stage.

We have performed a discounted cash flow (DCF) analysis based on our assumptions for ixmyelocel-T in both CLI and DCM. We arrive at a full value of $395 million, or $6.43 per share. Our published target is $5 to be conservative ahead of potential dilutive financing in the near-term.

Disclosure: I have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.