The past several years have seen the fruition of a new era in cancer therapy-targeted approaches using biologic modifiers. However, as the clinical experience with novel inhibitors grows, some of the premises on which the treatments were designed are being challenged, and clinical findings are leading to new paradigms. Drs. Song and Raben provide a forward-thinking review of the status of epidermal growth factor receptor (EGFR)-targeted therapy in head and neck cancer, a paper that serves to both highlight progress and raise issues that continue to challenge the implementation of targeted therapy.
Predicting Response to EGFR-Targeted Therapies
One of the most important issues in the optimal application of EGFRtargeted therapies is that of appropriate patient selection. Unlike conventional cytotoxic therapies, which are typically directed against ubiquitous, vital, cellular processes, targeted therapies focus on a single pathway that presumably plays a pivotal role in the proliferation and survival of the tumor cells. Ideally, this same pathway is not necessary for the function of normal tissues, thus providing a favorable therapeutic ratio. A key question then is, How does one determine that the EGFR pathway plays a pivotal role in an individual tumor?
At first glance, the logic for targeting EGFR in squamous cell carcinoma of the head and neck (HNSCC) appears fairly straightforward. There is substantial evidence that EGFR plays an important role in disease pathogenesis and response to therapy in HNSCC. As reviewed by Drs. Song and Raben, EGFR overexpression is an important prognostic and predictive factor.
Functional knowledge of EGFR inhibitors derived from preclinical studies would lead one to predict that tumors that overexpress EGFR, indicating dysregulated EGFR function, would be those most likely to respond to EGFR inhibitors. However, the correlation between EGFR expression and response to inhibitor therapy is poor. Unlike HER2 in breast cancer, where amplification of the gene leads to massive overexpression, there is as yet no clearly identified means of selecting patients for EGFR-targeted therapies based on the degree of receptor expression. One possible explanation for this is that EGFR can be pathologically activated even when expressed at modestly low levels.
Therefore, perhaps the more relevant measure of potential response would be to evaluate tumors for inappropriate EGFR activation, by measuring constitutive expression of the phosphorylated (active) form of the receptor. While this may still prove to be the case, initial studies do not indicate a clear correlation even between receptor activation and response to EGFR inhibitors. Interestingly, inhibition of downstream signals activated by the EGFR may be a more sensitive correlate for efficacy than inhibition of the receptor.[1-3] This makes sense if the ultimate control of tumor cell proliferation and survival is due to activation of key downstream pathways. If key pathways are inhibited, the upstream inhibitor should work. If not, perhaps other alterations are causing the activation of the key pathways despite EGFR inhibition. In reality, of course, it is not likely to be this simple.
One intriguing explanation for the lack of logical correlation between tumors that respond and those our models predict should respond is that we don't yet understand enough about EGFR pathology. An intriguing example of this is presented by the lung cancer experience with gefitinib(Drug information on gefitinib) (Iressa).[ 4] In the delightfully ironic setting of going from patient to laboratory (as opposed to laboratory to patient) to determine how these mechanistically derived targeted therapeutics really work, investigators identified a novel aspect of EGFR pathogenesis. Tumors from the relatively few responding patients were evaluated, and the vast majority of them were found to have mutant EGFR receptors. The individual mutations varied, but the common result was a supersensitive receptor; the kinase domain had been altered to render it both more active (hence the pathology) and more susceptible to inhibition by gefitinib (hence the response). Does this mean that we now have the answer to how to appropriately select patients for EGFR inhibitor therapy? Not yet, but it does provide an explanation for some of the unanswered questions, while raising a host of others.
EGFR Inhibitors as Radiosensitizers
The vast majority of studies investigating EGFR-targeted approaches to date have focused on their antitumor potential. A perhaps distinct issue is the potential utility of these agents to enhance conventional cytotoxic therapy. It is conceivable that the mechanisms responsible for an antitumor effect are distinct from those involved in promoting tumor cell death in the face of DNA damage. Thus, EGFR inhibitors may be effective in combination with radiation therapy in settings where they are not effective as single agents, or even in combination with chemotherapy. Drs. Song and Raben review the growing body of preclinical evidence indicating that EGFR family members are important mediators of radiation response.
In many ways, locally advanced head and neck cancer is an ideal setting in which to confirm the promising preclinical radiosensitization studies. It is a disease for which radiotherapy is the primary local modality, but where there is significant need for improvement. Radiosensitization with conventional cytotoxic chemotherapy clearly improves local control, but at the expense of normal tissue toxicity. Finally, tumor and normal tissue are readily available for correlative investigation to confirm or deny proof of principle.
Indeed, the first phase III clinical trial of an EGFR inhibitor as radiosensitizer has been carried out in advanced head and neck cancer patients, and the efficacy results recently released.[ 5] A total of 425 patients with previously untreated, nonmetastatic, advanced head and neck cancer were randomized to receive radiotherapy plus cetuximab(Drug information on cetuximab) vs radiotherapy alone. A primary end point was locoregional control. Secondary end points included time to progression, overall survival, and toxicity. A clear improvement in local control in the radiotherapy-plus-cetuximab arm translated into a significant survival benefit, with near doubling of the survival rate at 3 years. The addition of cetuximab to radiotherapy was associated with only modest toxicity. This trial provides the first definitive evidence of radiosensitization by EGFR inhibitors, and will likely spur additional research with other EGFR inhibitors.
As clinical studies of EGFR inhibitors are teaching us more about basic EGFR pathology, radiosensitization studies provide an unparalleled opportunity to increase our understanding of radiation response and radioresistance. What are the mechanisms of radiosensitization by EGFR inhibition? From the initial cell culture work, we predicted that perturbation of cell-cycle regulatory mechanisms and survival would be key factors in EGFR-mediated radiosensitization. This indeed appears to be the case, but as studies advanced to tumor models, we gained an appreciation for the role that EGFR plays in angiogenesis.
EGFR activation upregulates vascular endothelial growth factor (VEGF) production, promoting angiogenesis. Conversely, EGFR inhibition impedes VEGF production, resulting in an antiangiogenic effect, and perhaps in increased sensitivity of endothelial cells to radiation, which has been postulated to be a primary determinant of radioresistance.[6] Is the key role of EGFR in radiation response ultimately its indirect effect on the endothelium? As Drs. Song and Raben suggest, such mechanistic hypotheses provide the rationale for targeting the vascular system of tumors by dual EGFR- and VEGF-targeted approaches. Is the effect of EGFR inhibitors on endothelial survival after radiation one reason why EGFR inhibitors do not appear to be as effective chemosensitizers as radiosensitizers? We could soon know more from results of correlative queries investigating effects of EGFR inhibitors on tumor vascularity in responding and nonresponding tumors of patients treated in clinical trials.
Conclusions
Drs. Song and Raben raise the challenge "to understand when the application of these EGFR inhibitors is relevant to an individual patient and how or when these drugs should be combined with radiation or chemotherapy." While the initial results from clinical use of EGFR inhibitors have perhaps raised more questions than they have answered, these trials have enriched our understanding of the role of EGFR in cancer pathogenesis and response to therapy, illustrating the importance of translating not only from bench to bedside, but also from bedside back to bench. Through carefully designed clinical and correlative trials, we can build on the realizing promise of targeted radiosensitization.
Financial Disclosure: Dr. Sartor receives clinical trial support from Genentech and GlaxoSmithKline.
