ImmunoGen at ASH 2007 (Part II)

Regardless of IMGN901’s specific case, the impression I am getting from all the scientific material I come across that deals with Immunogen’s (IMGN) technology is that IMGN901 will probably be the last ADC (antibody-drug conjugate) powered by the cleavable DM1 linker. There are currently no ADC programs, except for IMGN901, that utilize this specific linker. As I mentioned in one of my SGEN’s (SGEN) pieces, Genentech (DNA) seems to prefer a noncleavable linker for the majority of its ADCs. Another example may be Centocor, who licensed Immunogen’s technology for arming a antibodies targeted against alpha integrin and evaluated both DM1 and DM4 cleavable linkers with the same antibody. Results from several animal experiments showed that the cleavable DM4 version was much more stable in the bloodstream and active in inhibiting tumor growth than the cleavable DM1 version.

However, linker stability is only one of numerous factors that affect the performance of an ADC and is certainly not a guarantee for success. Let’s take AVE9633 for example, which actually had good stability in the bloodstream, partly thanks to the DM4 linker.

Data from a phase I clinical trial that evaluated AVE9633 in AML patients was presented. AVE9633 was given on Day 1 and Day 8 of a 28-day cycle to a total of 17 patients. Doses ranged from 30 to 150 mg/m² per week and maximal tolerated dose was 150 mg/m². Of the 17 patients, one patient who received 105 mg/m² achieved a complete response [CR] which is still ongoing for 8 months, and another patient at the 130 mg/m² had a partial response [PR].

AVE9633 certainly didn’t fail, as it demonstrated clear clinical activity, but the efficacy it demonstrated is far from being impressive, especially as an ADC or when comparedto other anti-CD33 antibody based platforms such as Mylotarg and Seattle Genetics’ SGN-33. Both SGN-33 and AVE9633 were evaluated in small phase I trials, so results from more advanced trials could differ materially from the presented results. One factor that might explain the lower performance of AVE9633 is the challenging patient population, since all 17 patients were relapsed AML patients. In the SGN-33 trial, only 8 out of the 17 patients had received chemotherapy prior to the trial. Nevertheless, it still doesn’t change the big picture as we know that at least two of these eight relapsed patients had a CR after treatment with SGN-33. Furthermore, there were 5 patients in the SGN-33 trial who were considered secondary AML patients, one of which achieved a complete response as well. Secondary AML is a term used to describe AML that has developed from prior hematologic disorders, such as MDS. These patients have very poor prognosis, both in terms of survival and clinical response in comparison to patients first diagnosed with AML. Hence, either way we look at it, 2 out of 8 in relapsed AML and 1 out of 5 in secondary AML in the SGN-33 trial still looks better than 1 out of 17 in the AVE9633 trial.

Overall, results presented by IMGN this year at ASH were not phenomenal, however, that should not discourage the company’s investors. The field of ADCs is just beginning and it is unrealistic to expect such a disruptive field to have better success rates than the rest of the industry. Investors should always remind themselves that in drug development, most drug candidates fail. The companies who consistently innovate and learn from these failures are the companies to invest in, and Immunogen is definitely such a company. Putting Mylotarg aside, very little is known about ADCs as the only available data is from small uncontrolled trials, but either way we look at it, Immunogen is still involved in the most promising ADC in the clinic - Herceptin-DM1 (T-DM1). It does not mean that T-DM1 will be approved, it just means that for anyone who wants to invest in the ADC market and understands the tremendous risks involved, Immunogen is an obvious choice.

Disclosure: Author is long IMGN & SGEN