Cancer treatment today is undergoing as great a sea change as when we first successfully used cytotoxic therapies in the 1970s and when we introduced adjuvant therapies in the 1980s. The impressive improvements in outcomes coupled with a marked diminution in toxicity have been well summarized by Franco Muggia and his colleagues in their two-part review of targeted therapy.
Targeted Therapies Raise New Issues
A few key points that have emerged from the development of these new therapies need to be underscored.
Early randomized trials are needed for drugs that do not induce tumor regression
The phase II trials evaluating these treatments often produced a very weak signal. Traditionally we’ve relied on evidence of objective response and the response rate to identify drugs worthy of phase III/IV evaluation. At one time it was thought that a drug with a response rate < 20% should be abandoned. That has not been true for many targeted therapies. For example, the phase II response rate for trastuzumab(Drug information on trastuzumab) (Herceptin) in breast cancer was 11%, for erlotinib (Tarceva) in lung cancer it was 9%, for sorafenib(Drug information on sorafenib) (Nexavar) in renal cancer it was 2% to 18%, and for temsirolimus (Torisel) in renal cancer it was 7%. There were many who argued against development of each of these drugs—which were eventually shown to improve time to progression, disease-free survival and overall survival substantially. This history shows the value of performing randomized trials early in the developmental process, since this is the only way to demonstrate disease control for drugs that do not induce tumor regression.
The shift from optimal dose to optimal duration of treatment
Because these treatments target diseased rather than normal tissues, they have less toxicity and can be given for longer periods of time. This introduces the possibility of long-term suppression of tumor growth as opposed to our more traditional goal of eradicating tumor, which we have attempted to do by escalating dose. With targeted therapies, the key issue appears to be duration of treatment, not dose, since prolonged treatment may be necessary to optimize tumor control.
A new problem: how to get the target right early on
The story of drugs targeting the epidermal growth factor receptor (EGFR) in lung cancer illustrates the importance of getting the target right early on. There are two approaches to the identification of the correct target. One is to delay large randomized trials in selected patient populations until the target is fully characterized in preclinical or correlative studies. The other is to avoid prematurely excluding patients from clinical trials based on a presumed target. Of course, sometimes we’re just plain lucky. This was the case in human epidermal growth factor receptor 2 (HER2) studies. Prior to the initiation of phase III trials, there was considerable discussion of whether patient enrollment in the pivotal trial should be limited to those who had 3+ HER2 overexpression based on immunohistochemistry. In retrospect, it seems unlikely that we would have demonstrated the benefit of trastuzumab in the pivotal phase III study if we’d “taken all comers.”
Additional improvements in treatment may slow as the number of therapeutic options grows
New therapies—including targeted therapies—tend to be tested first in tumor types (or in tumor subsets) in which it is recognized that available therapies have very limited benefit, such as HER2-positive breast cancer or renal cancer. It is plausible that as the number of therapies available to treat a particular tumor grows, overall improvements in survival or quality of life will become marginal. Overestimation of the benefits from established treatments will make it more difficult to perform necessary phase II trials and to introduce therapies that may eventually prove to be much better.
Why the Delay in Extrapolating Targeted Therapies to Other Tumor Types With the Target?
Each of the targeted therapies has been developed for one tumor type and then, in some cases, been evaluated in other tumors only after demonstration of benefit in the first tumor tested. This is illustrated by the HER2 story. Overexpression or amplification of HER2 occurs in many tumor types (Table). However, it has been fully evaluated and shown to be useful in only two of these: breast, for which the phase II trial was published in 1996 and the drug approved by the US Food and Drug Administration (FDA) in 1998; and gastric cancer, for which the phase II trial was published in 2010 and the drug’s new indication approved by the FDA in 2010. Why the long interval? The answers given by investigators have included: “the number of patients with HER2 overexpression are too few to anticipate an anti-HER2 treatment will have much impact,” “the correlative studies fail to demonstrate consistent prognostic value for HER2,” “the response rate to trastuzumab was too low to predict success,” and “the responses achieved using anti-HER2 therapy in combination are uninterpretable.” All of these were true for the development of HER2 for both breast and gastric cancer. The remarkable success of trastuzumab in these two tumor types probably represents the continuing efforts of one or a few investigators who were so convinced by preclinical evidence and well developed hypotheses that they persisted in the face of challenges.
The two-part Muggia review has also been organized by tumor type rather than by target across tumor types. When is this going to change? When will we initiate trials in which the participants are defined by the molecular marker without regard for the tissue of origin? Isn’t it time to initiate phase II and phase III trials, maybe adaptive trials, in which patient selection is determined primarily by target?
I see a time in the not too distant future when we’ll define tumors this way. What will our subspecialties be? Rather than a breast clinic or a lung clinic, will we perhaps be attending a “HER2 clinic” or an “mTOR clinic” instead?
Financial Disclosure: The author has no significant financial interest or other relationship with the manufacturers of any products or providers of any service mentioned in this article.