In this issue of ONCOLOGY, Saba et al provide a comprehensive review of epidermal growth factor receptor (EGFR) biology, the scientific rationale behind inhibition of the receptor, and clinical trials using EGFR inhibitors in squamous cell carcinoma of the head and neck (SCCHN) and non-small-cell lung cancer (NSCLC). They conclude that EGFR is a valid target of anticancer therapy and that patient selection in clinical trials is important. We agree that further studies are required to fully understand the role of EGFR inhibitors in the treatment of SCCHN and NSCLC. In this commentary, we will elaborate further on the issue of patient selection and future directions in the development of EGFR inhibitors.

Patient Selection

SCCHN and NSCLC are biologically heterogeneous diseases with variable clinical behavior and outcome, even within comparable disease stages and treatment. On the other hand, targeted agents such as EGFR inhibitors, by definition, provide selective benefit depending on the presence of the target. Therefore, a uniform clinical response is unlikely with the expected heterogeneity of tumors. For example, tamoxifen(Drug information on tamoxifen) is not effective in breast cancer patients with estrogen receptor-negative tumors. Without patient selection based on the presence of HER2/neu gene amplification, the efficacy of trastuzamab (Herceptin) for the treatment of breast cancer would not have been detected because of dilution of the net-clinical activity.[1]

Proper patient selection for a given therapy can be achieved based on clinical criteria, or through biomarkers of response or resistance. Since we have more information in NSCLC compared to SCCHN regarding selection criteria, we will discuss lung cancer first.

Response Markers of NSCLC

As mentioned by Saba et al, several large trials have used the EGFR tyrosine kinase inhibitors (TKIs) erlotinib (Tarceva) and gefitinib(Drug information on gefitinib) (Iressa), with modest objective response rates (8.9%-19%) as single agents in recurrent/metastatic NSCLC.[2-4] However, a subset analysis of the multicenter phase II trial of gefitinib has shown that Japanese patients had higher response rates compared to non-Japanese patients (27.5% vs 10.4%, P = .0023). In addition, receiving prior immuno/hormonal treatment, being female, and having adenocarcinoma were associated with higher response rates.[3] Subsequent analyses suggest "never-smokers" respond at a higher rate as well.

In addition to these clinical characteristics of TKI response, somatic gain-of-function mutations in the tyrosine kinase (TK) domain of the EGFR gene were identified by three independent groups and shown to be associated with clinical response to gefitinib.[5-7] These mutations are thought to enhance TK activity; therefore, the tumors with the mutation are more dependent on the EGFR pathway and more sensitive to gefitinib. When the EGFR-TK domain was sequenced from genomic DNA collected from patients in Japan, Taiwan, the United States, and Australia, the mutations were detected in 21% of specimens with NSCLC, whereas none were detected in nonmalignant lung tissue from the same patients and other carcinomas.[8] The mutations were more frequent in patients of East Asian ethnicity, females, patients with adenocarcinoma, and in never-smokers.[8]

These findings were consistent with the clinical findings in the gefitinib trials.[3] Furthermore, Cappuzzo et al showed that the presence of EGFR gene amplification or high degree of polysomy and high levels of protein expression were associated with a higher rate of response to gefitinib (36% vs 3%) and longer survival (18.7 vs 7 months, P = .03).[9] A Southwest Oncology Group study of patients with bronchoalveolar carcinoma or adenocarcinoma with features of bronchoalveolar carcinoma also showed the association of increased EGFR gene copy number with prolonged survival (18 vs 8 months, P = .042) after gefitinib treatment.[10]

These mutation studies provided insight into resistance to TKI treatment as well, which may suggest a rationale for new agents or combination regimens to overcome the resistance. Pao et al found this acquired resistance to be associated with a second mutation in addition to the primary drug-sensitizing mutation in patients whose disease progressed after the initial response to gefitinib or erlotinib.[11] These data suggest that the acquired resistance is due to the emergence of a resistant cell clone developing the additional mutation during treatment. One other important finding is that EGFR mutations and K-ras oncogene mutations (which represent an important step in the carcinogenesis of NSCLC) were mutually exclusive.[8,12] While EGFR mutations are common in never-smokers, K-ras mutation is common in smokers.[7,8]