Chemotherapeutic Drugs in The Clinic – Competitors or Potential Partners?
Obviously, Seattle Genetics' (SGEN) SGN-33 was not directly compared to any other agent, so insight gained from comparing SGN-33 to other agents from different clinical trials is far from being conclusive. In addition, a comparison of a naked antibody (that will likely be given in combination with other drugs), to other chemo and combination regimens is not a fair one. Nevertheless, these comparisons are the only means researchers and investors alike have when evaluating the prospects of SGN-33.
The efficacy/safety ratio of SGN-33 is very impressive when compared to available treatments as well as other treatments currently evaluated in clinical trials. The cornerstone treatment for older AML patients is low-dose araC which has less than 20% complete response rate as a single agent (compared with 29% for SGN-33 in the current trial). araC is typically administered with other agents and is currently evaluated in combination with some novel drug candidates. These combinations result in a much better response rate, in the range of 30-60% among a variety of patient populations.
The most promising combination in the clinic is araC + Genzyme’s (GENZ) Clofarabine, which demonstrated a CR rate of 60%, but this was achieved in previously untreated patients with a median age of 61. In contrast, 8 out of the 17 AML patients in the SGN-33 trial had been previously treated and the median age of the total patient population was 75. Clofarabine alone demonstrated impressive results in older AML patients as well (CR rate of 38%-48%).
Vion Pharmaceuticals’ (VION) Cloretazine is another promising agent that achieved a CR rate of up to 49% in older AML patients. With regard to safety issues, SGN-33 seems to be much safer than all the above treatments, which often result in severe side effects and even deaths in some cases. In addition, all the above treatments should not necessarily be viewed as direct competitors to SGN-33, which, thanks to its excellent safety profile, may be administered in combination with any of them.
Antibody Based Drugs Targeted Against CD33 - Direct Competitors
There are two direct competitors to SGN-33. The first one is Wyeth’s Mylotarg, which is an approved anti-CD33 ADC (Antibody-drug conjugate). The second is Sanofi-Aventis’ (SNY) AVE9633, another anti-CD33 ADC based on Immunogen’s (IMGN) technology. The reason why both should be regarded as competitors is clear - They are both based on antibodies that bind CD33. Therefore in a theoretical scenario where each one of them is co-administered with SGN-33, it will literally compete with SGN-33 for the binding of CD33. Since all three drugs are dosed in a way that they saturate the majority of available CD33 molecules, this idea sounds even more unrealistic. Allegedly, Mylotarg and AVE9633 should be much more effective than a naked antibody since they comprise of an antibody conjugated to a toxic agent. In reality things are quite different.
A comparison to the clinical data for AVE9633 implies that SGN-33 is superior, both in term of activity and safety. Although I have to admit I was never a fan of AVE9633, the comparison with SGN-33 serves as another validation for me. The doses of AVE9633 were lower (0.8-4 mg/kg) compared to the evaluated doses of SGN-33 (1-8 mg/kg), and the two studies had different dosage timing. AVE9633 was administered twice every 4 weeks, while SGN-33 was dosed every week for 5 weeks in the first stage of the trial. This means that patients in the SGN-33 trial received 5-fold more antibody molecules than patients in the AVE9633 trial. Both trials had 17 evaluable older AML patients, which makes the comparison very easy. There were 5 complete responses in the SGN-33 trial whereas only one patient in the AVE9633 study achieved a complete response. In addition, AVE9633’s MTD was set at 4 mg/kg, in contrast to no MTD, even at weekly 8mg/kg injections of SGN-33. In my opinion, this comparison leads to the inevitable conclusion that, regardless of the dosage regimen, SGN-33 will always outdo AVE9633 in AML. The difference is even more striking considering the fact that AVE9633 is an ADC.
Mylotarg is the only approved ADC in the market consisting of an anti-CD33 antibody linked to the potent drug, Calicheamicin. Mylotarg is administered twice at very low doses (0.24 mg/kg), with a gap of two weeks between each administration. It was evaluated as a single agent in several phase II trials and demonstrated CR rate in the range of 20%-40%, which is comparable to that of SGN-33 in the phase I trial. This time, it would be fair to say that Mylotarg is much more potent than SGN-33, as substantially lower amounts of Mylotarg achieve similar response rate to that of SGN-33. However, Mylotarg cannot be given at higher doses due to side effects, in particular liver toxicities. In fact, despite being approved for older AML patients, Mylotarg is not widely used for the treatment of these patients, especially not patients who are older than 75. A tragic example for that might be a clinical trial that evaluated Mylotarg among 40 AML patients, and led to the death of 7 patients, all of whom aged above 75 years. Mylotarg is also given in combination with araC, mainly as first-line treatment to younger AML patients, leading to better response rates but to increased rate of side effects as well. Although there is absolutely no data regarding the co-administration of SGN-33 and araC, it seems likely that it will be better tolerated by patients than the Mylotarg-araC combination. There is, of course, the issue of how effective the combination will be, but it seems likely that SGN-33 will have an additive effect over araC as well.
A naked antibody with a superior efficacy/safety ratio than of ADCs targeted against the same target is not something you see everyday. We can speculate as to the reasons for this phenomenon but at the end of the day, the best explanation is that drugs and biological systems are very unpredictable. There are so many variables which eventually decide which drugs work and which don’t, that it sometimes seems like a miracle some drugs do work at all. One possible approach is to compare all three antibody-based platforms against CD33 and their possible mechanisms of action. We know that SGN-33 can recruit the immune system to attack the cancer cells it binds, in addition to a possible signal transduction it activates by binding the CD33 receptor.
Perhaps both Mylotarg and AVE9633 bind the CD33 receptor in a way that does not lead to its activation and the signal transduction. In addition, Mylotarg is an IgG4 type of antibody, which means its ability to mediate an immune response against cancer cells is quite limited. Another factor in favor of SGN-33 is the fact that naked antibodies stay much longer in the bloodstream as oppose to ADCs, so perhaps SGN-33’s extended presence in patients is the most crucial factor in this case. On the other hand, even if both ADCs lack these effects, they are supposed to have a much more deadly effect due to the toxic drugs that are attached to them. While Mylotarg does have potent activity (and toxicity), AVE9633 doesn’t manage to even achieve that task.
Disclosure: Author is long SGEN