Biomarkers signal true progress in war against lung cancer

March 18, 2010

Testing for EGFR and ALK mutations reveals tumor behavior and helps tailor treatment.

ABSTRACT: Testing for EGFR and ALK mutations reveals tumor behavior and helps tailor treatment.

For about half a century, outcomes for patients with solid tumors such as those of the lung and breast have improved by only 5%, despite a yearly investment of $5 billion in public sector funds for research. But targeted therapies are extending survival in patients with biomarkers such as EGFR mutations, and other biomarkers can predict chemotherapy sensitivity and toxicity.

Advances in tumor biomarkers signal true progress in the war against lung cancer, and many of these advances have already transitioned from the research lab to the clinic, according to an expert panel at the 2009 World Conference on Lung Cancer (WCLC) in San Francisco.

The panelists were:

Thomas Lynch, MD
Vincent Miller, MD
Tony Mok, MD
Jean-Charles Soria, MD, PhD
Joan Schiller, MD


EGFR mutation testing

The success of erlotinib (Tarceva) and gefitinib (Iressa) for patients with EGFR mutations has ushered in an era of molecular medicine, according to Dr. Lynch. These drugs provide dramatic responses in a subset of patients, and EGFR mutations with exon 19 or exon 21 deletions best predict this response. In such patients, the response rate for these TKIs is about 70%, he said.

Yet patients treated with these drugs can develop resistance and researchers are only now beginning to understand the pathways that make the tumors resistant. Biomarkers that signal resistance to TKIs include the T790M mutation, mesenchymal epithelial transition (c-MET) amplification, KRAS mutations, and PTEN loss, Dr. Lynch said. All these pathways are now under study as potential targets for new therapies.


©ASCO/Todd Buchanan 2005

But in the meantime, the ability to test for EGFR mutation status is an important tool for clinicians. "I would argue that EGFR testing is ready for prime time," Dr. Lynch said. "Almost always, you can get an EGFR mutation test from a larger resection specimen, and as the technology gets better, we will get to the point of testing from fineneedle aspiration."

Dr. Lynch noted that because of the dramatic response seen in patients with EGFR mutations, TKIs should be considered as a first-line therapy in these patients. "It should be a care standard for certain patients who harbor these mutations," he said. For certain subsets of patients, including light and never smokers with adenocarcinoma-who are more likely to harbor the EGFR mutation- testing for the biomarker should be done as early in the course of therapy as possible, he added.

One study that has added to scientists' understanding of the role of EGFR mutation status in predicting response to TKIs is the Iressa Pan-Asia Study (IPASS), Dr. Lynch said. In this phase III, open-label study, first-line gefitinib was compared with carboplatin/paclitaxel in patients from East Asia who had advanced pulmonary adenocarcinoma and were former light smokers or had never smoked (New Engl J Med 361:947-957, 2009).

According to the results, the 12-month rate of progression-free survival (PFS) was 24.9% with gefitinib and 6.7% with standard chemotherapy. Those who had EGFR mutations fared significantly better on gefitinib than those who were negative for the mutation. In the mutation-positive group, PFS was longer among patients who received gefitinib than among those who received chemotherapy (P < .0001).

"The response rate of 70% in the mutation-positive patients [treated with gefitinib] was striking, compared with 1.1% in the mutation-negative patients," Dr. Mok said.

The IPASS researchers also analyzed whether the EGFR gene copy number by fluorescence in situ hybridization (FISH) was predictive of outcomes. FISH-positive status did seem to predict a better outcome on gefitinib, but 70% of the FISH-positive patients in the study population also had an EGFR mutation, Dr. Mok said.

They concluded that the treatment effect that seemed to derive from the EGFR gene copy number was actually motivated by the underlying EGFR mutation. "We've learned the importance of the EGFR mutations, and future studies on TKIs should be driven by patients with these mutations," Dr. Mok said.

The ALK biomarker

Another biomarker that may soon prove important is anaplastic lymphoma kinase (ALK). It was first described in 2007, and Dr. Miller said that in that short time "we've got enough momentum in the field to get this diagnostic marker in a [clinical] trial that's working." The ALK translocation protein appears to be present in about 4% to 5% of adenocarcinoma cases, he said.

Last year, scientists presented results of a clinical trial using PF-02341066, which targets the ALK translocation and c-MET genes. Of 13 patients treated with the drug, eight were evaluable for response and the disease control rate in these patients was calculated at 87.5%. Most of the patients were heavily pretreated (ASCO 2009 abstract 3509; WCLC 2009 abstract A6.4).

"Unlike EGFR mutations, early data suggest that this marker may not be prognostic," Dr. Miller said. The mutation tends to be more common in patients who do not have EGFR or KRAS mutations, have never smoked, and have adenocarcinoma. The incidence can be as high as 25% in this population, he added.

"So EGFR mutations are not the end of the story but the beginning of the story," Dr. Miller said. As with EGFR mutations, ALK translocation genes are a biomarker that may soon prove to be sensitive to targeted therapies.

However, Dr. Miller noted that the discovery of EGFR mutations sensitive to targeted therapies has led to a fundamental paradigm change in lung cancer research: Scientists are now searching different types of lung cancers for molecular markers that could be targeted by new medicines.

"Other kinase mutations will continue to be identified that can be addressed by new drugs, but that is not the final solution to our problems," he said. Identifiable kinase mutations are present in only about a third of advanced non-small-cell lung cancer, he noted.

"Lung cancer has really moved from being one disease to being four or five diseases, and we may think in even broader terms than that in the coming decades," Dr. Miller added.

Angiogenesis pathway

Another area that has attracted interest is biomarkers for the angiogenesis pathway. Recent studies have shown that the VEGF pathway is an active target for investigation. The Eastern Cooperative Oncology Group conducted a phase III randomized trial (ECOG E4599) of carboplatin/paclitaxel (Taxol) with or without bevacizumab (Avastin) in patients with non-squamouscell lung carcinoma. The E4599 trial revealed an improvement in response rate, PFS, and overall survival (OS) in those who received bevacizumab, said Dr. Schiller, who was an investigator on the trial (Clin Cancer Res 14:1407-1412, 2008).

But questions remain about appropriate biomarkers, she added. "We need to ask what biomarkers, if any, do we have that can guide us in terms of which patients are most likely to derive benefit from these types of therapies," she said.


©ASCO/Todd Buchanan 2009

A 2006 analysis of biomarkers present in the blood of patients in the E4599 trial found that only one was a prognostic indicator. Those patients who had a low baseline soluble intercellular adhesion molecule-1 (sICAM-1) fared better, whether they received bevacizumab or not, than those with a high baseline sICAM-1. So although this marker was useful for its prognostic value, it was not predictive of benefit from bevacizumab, Dr. Schiller said (ASCO 2006 abstract 7027).

Other studies have analyzed whether patients with certain single nucleotide polymorphisms (SNPs) might have an improved response. Although some studies have found good SNPs that seem to predict an improved survival with bevacizumab, these studies have been limited by multiple comparisons, Dr. Schiller said. "When you do multiple comparisons, the chance of finding one by chance goes up," she noted.

Don't rule out clinical biomarkers

While genetic biomarkers are making inroads, clinical biomarkers may also be useful in predicting response to drugs that target angiogenesis. In a study presented last year, researchers showed data from the E4599 trial suggesting that onset of hypertension during therapy with bevacizumab may be associated with improved PFS and OS (ASCO 2009 abstract 8032).

"But if you were on chemotherapy alone, whether you developed hypertension was not a marker of survival at all," Dr. Schiller said. Why is hypertension a clinical marker of outcome in drugs that target angiogenesis? Because drugs that antagonize the VEGF pathway decrease nitric oxide production, causing constriction of the vasculature, she explained.

As well as biomarkers that predict for sensitivity to targeted therapies, biomarkers that can show sensitivity or resistance to certain chemotherapies have also come into clinical practice recently, according to Dr. Soria. These markers include the excision repair cross-complementation group 1 (ERCC1) gene for cisplatin efficacy and the ribonucleotide reductase M1 (RRM1) gene for gemcitabine (Gemzar) efficacy.

"We need to take into account the tumor's molecular heterogeneity," Dr. Soria said. "Even with the best pathology, clinicians cannot always gain a complete understanding of how a tumor will respond to certain therapies. We need to understand the different pathways that lead to the tumor's behavior, whether it is RAS-driven, EGFR-driven, or [driven] by other genetic abnormalities."

In an era of increasingly expensive pharmaceuticals, biomarker-driven therapy will ultimately reduce costs, Dr. Soria predicted. "When we can better select the therapy for the patient, costs will be reduced. And that's the ultimate expectation we should have for biologically driven therapy," he said.