In the last several years we have seen an explosion in the knowledge and understanding of the molecular basis of cancer behavior. The identification of EGFR mutations, and their utilization in predicting response and survival in patients with untreated advanced lung cancer, provides lung cancer oncologists with the optimism that we may truly be able to offer effective “personalized therapy” for patients with lung cancer in the near future.

The majority of knowledge regarding EGFR mutations comes from studies of patients of East Asian ethnicity or of Caucasian backgrounds from either the US or Europe. It is evident and widely discussed from the data from these populations that lung cancer patients of East Asian ethnicity have a higher rate of EGFR mutation than Caucasian patients do.

What about patients however of other backgrounds? Do African American, Latino, or Native American patients have the same rates of EGFR mutations as Caucasians or East Asians, or are they different? How great is the disparity of mutation rate between never-smokers and those patients with a smoking history in these populations?

The unfortunate truth is that very little data exists for non-Asian, non-Caucasian populations. This is why a study by Dr. Leidner and colleagues recently published in the Journal of Clinical Oncology is a welcome addition. The investigators in this study tested tumor samples and normal tissue from areas without cancer samples from 53 African American patients, evaluating for EGFR activating mutations, EGFR overexpression as measured by FISH testing, and KRAS mutations. They compared results with a series of 102 Caucasian patients from Italy from previously-published studies.

In the African American patients, roughly equal numbers were men and women (compared with about 2/3 men and 1/3 women in the Italian comparisons). The distribution of different tumor types (squamous cell carcinomas, adenocarcinomas, and other) were essentially the same. Most of the African American patients being evaluated had early stage disease (primarily stage I cancers), as opposed to the Italian cohort in which all patients had advanced disease. This is primarily due to the requirement in this study of having normal tissue also to evaluate, which then placed the focus on patients who had their tumors surgically resected. Only seven (13%) of the African American patients were never smokers (16% in the Caucasian comparison arm).

Interestingly in the entire group of 53 African American patients, no patient had the commonly-recognized activating EGFR mutations (exon 19 deletion and L858R) that are linked with high response to EGFR inhibitor therapy with Tarceva or Iressa. One patient in the entire group had an EGFR mutation, which was a previously-unidentified mutation, termed S768N, and was counted as a positive finding to yield a 2% EGFR mutation rate for the group studied. More patients in this group had increased EGFR gene copy number by FISH (51% vs 32%), although no differences were seen in KRAS mutations (about 22% each).

EGFR mutations were identified in 17% of the Caucasian comparison patients, which was a statistically significant difference. Seven of the African American patients were treated with either Iressa or Tarceva. Although four of these patients had rapid disease progression, two were on treatment for a prolonged period of time (11 months and >28 months, respectively). The remaining patient was treated for a brief time on a clinical trial.

Obviously we cannot make broad generalizations about an entire patient population based on 53 patients, and we cannot make assumptions about the relative benefits of therapy based on information from seven patients. In a group of 14 African American patients studied in a cohort from Memorial Sloan Kettering (8 who were never-smokers), six had activating EGFR mutations. This study however provides some very important and intriguing information, highlighting a critical knowledge gap. It will be important to gather information on larger groups of African American patients as well as patients of all different ethnicities in order to understand the disparities in response to treatment and survival, and to provide educated therapy, not educated guesses.

If it is truly the case that EGFR mutations are uncommon or nonexistent in even the nonsmoking African American lung cancer patients, does that mean that other genetic changes, such as the EML4-ALK translocation, are more common? And the question, as always, is how much tissue we will need to fully evaluate and screen for different genetic markers that might predict for response and survival.

I find myself struggling with how to extrapolate data from studies in East Asian or Caucasian populations when I am discussing the chances of having an EGFR mutation, or possible responses to different therapies, with my never-smoking Latino or African American patients. With the results of this study, I will still be sending tumor samples for EGFR mutation testing, but my expectation for finding an EGFR mutation will be lower. I do not think it is useless to consider treating an African American patient without an EGFR mutation in the salvage setting (when cancer grows after chemotherapy) with Tarceva. I always prefer to offer patients clinical trial participation whenever possible, and results such as these hammer home how extremely important it is to improve enrollment on clinical studies for patients of all ethnic backgrounds.