Three studies presented this week at the American Association for Cancer Research conference in Philadelphia winnow down further which molecular markers make the best indicators of response to the AstraZeneca lung-cancer drug Iressa.
"In my opinion, it is now very clear that we now seem to have reasonable markers at hand to select a group of patients who have a very clear clinical benefit from this drug," said Fred Hirsch, a professor of medicine and pathology at the University of Colorado Cancer Center, in a press conference this week at the AACR meeting. "At this stage, I think that gene copy number, detected by FISH, seems to be the method that — if we're looking at a single marker — seems to be the best predictor, with a significant difference in outcome."
The three studies feature three separate lines of research. A group led by Fred Hirsch investigated the association of survival with epidermal growth factor receptor gene-copy number, EGFR protein expression, and Akt protein activation status in advanced non-small cell lung cancer treated by gefitinib or placebo. A second group of researchers, led by Brian Holloway of AstraZeneca Pharmaceuticals in England, focused on the association of survival with the mutation status of three genes — EGFR, K-Ras, and B-Raf in advanced non-small cell lung cancer treated by gefitinib or placebo. And a third team of researchers was led by Vanesa Gregorc, a medical oncologist at the Scientific University Hospital San Rafaelle in Milan, Italy. This last study examined the effect of different somatic, inherited variations in EGFR on the survival of patients treated with Iressa.
"However, if you are looking into the FISH-positive group treated with gefitinib, we can see substantial survival benefit also for [patients with non-adenocarcinoma, males, smokers, and non-Asian patients], and I think that is kind of an interesting point to make."
Iressa, also known by the generic name gefitinib, was approved by the US Food and Drug Administration in 2003 for use in non-small cell lung cancer. But in the Iressa Survival Evaluation in Lung, or ISEL, trial that concluded last year, Iressa did not result in overall patient survival, and its use was limited by the FDA to only those patients who had already shown some benefit from treatment. Since then, several investigations have attempted to find molecular markers that signal tumor susceptibility to the drug — ISEL trial results suggested that such markers of efficacy existed, since Iressa showed apparent survival benefit in some subgroups of patients, particularly never-smokers and East Asians.
EGFR Copy Number
In the research effort led by Hirsch, investigators studied EGFR gene-copy number, EGFR protein expression, and Akt-protein activation in formalin-fixed paraffin-embedded tissue samples collected from 170 patients who took part in the ISEL trial. Akt is a protein downstream of EGFR in the same biochemical pathway. All patients had previously failed chemotherapy.
"We have kind of a detailed classification [system], but it comes down to us including some of the groups together into what are called 'FISH positive' and 'FISH negative' groups," Hirsch said. FISH-positive groups included samples showing high polysomy — more than four EGFR gene copies in more than 40 percent of cells — and gene amplification, which is a clustering of many copies of the EGFR gene within each cell.
In the primary study endpoint of patient survival, "there is a statistically significant difference between FISH-positive and FISH-negative groups," as determined by using an interaction statistical test having a P value of 0.04, said Hirsch.
In the FISH-positive group, gefitinib treatment "reduced the mortality by 39 percent" compared to the placebo group," Hirsch said. "However, in the FISH-negative group, we could not see any significant difference in treatment outcome between gefitinib and placebo — as a matter of fact, patients treated with gefitinib had a slightly, slightly adverse outcome, but [essentially] no difference" in survival, he said.
FISH-positive tumor response was 16.4 percent versus 3 percent in the placebo group, and in the FISH-negative group tumor response rate was 3 percent versus zero, Hirsch said.
Previous Phase III trials suggested that patients with non-adenocarcinoma, males, smokers, and those with non-Asian origins had non-favorable outcomes, "however, if you are looking into the FISH-positive group treated with gefitinib, we can see substantial survival benefit also for these groups, and I think that is kind of an interesting point to make," Hirsch said.
Holloway's group investigated the mutation status of three genes: EGFR; K-Ras; and B-Raf. EGFR mutations had been associated with increased survival in previous studies, while the two other genes, which lie downstream from EGFR in the same pathway, have been implicated with resistance to Iressa when mutated. The 215 patient samples examined under-represented East Asians and never-smokers, the two groups associated in the ISEL trial with survival, in order to eliminate other factors those patients may have shared, which increased Iressa-associated survival.
The researchers examined mutation with two different methods: direct DNA sequencing of tumor-cell genes, and an amplification refractory mutation system — the ARMs technology developed at AstraZeneca and employed by the spin-off of that company, DxS.
"When we did this, we found an overall frequency in EGFR mutation of about 12 percent, so samples from 26 patients had mutations in EGFR," said Holloway. "The distribution of these mutations was very similar to that published in the literature, in that there were mostly exon 19 deletions and L85HR substitutions," he said. "So, two mutations that have been shown to be associated with activation of the EGFR and to response to gefitinib."
The samples studied tended to follow the pattern established in the ISEL trial, with East Asians, never-smokers, and women harboring more EGFR mutations, said Holloway. "In fact, we only had samples from 16 people of Asian ethnicity in this study, and more than 50 percent of them have mutations," he said.
Unfortunately, since only 26 mutation-harboring tumor samples were available for analysis, "that was too few to do a formal survival analysis," Holloway said. But fewer patients in the mutation-positive, gefitinib-treated group died than in the mutation-negative and placebo-treated groups, he said.
The response rate of EGFR mutation-positive patients was approximately 40 percent, showing a frequency of response that is much higher than in patients lacking EGFR mutations, who showed a response rate of 3 percent, Holloway said.
Perhaps not surprisingly, "most of the FISH-positive patients were also positive for EGFR protein expression," and "most of the mutations occurred within this FISH-positive, protein-positive group," said Holloway. "So if we were to pursue a selection strategy based on FISH, it would also include most of the patients with mutations," he said.
There was not enough data on the mutations in K-Ras and B-Raf to allow the researchers to evaluate clinical outcomes.
Inherited EGFR Sequence Variations
The research team led by Vanesa Gregorc examined the effect of natural variations in EGFR-gene sequence on patient survival after gefitinib treatment — that is, somatic variations, and not acquired genomic mutations. "We concentrated on the promoter area of the epidermal growth factor receptor, and four versions of this gene [were] identified, and the most common was a 'G-C' version [of the gene], which was present in 66 percent of the patients," said Gregorc.
Using genotyping, Gregorc's group analyzed 170 patient samples, who represented very closely patients in all other studies of non-small cell lung cancer, she said. Anti-tumor activity in the gefitinib group was 33 percent, compared to control, and 9 percent showed a partial drug response, while a remaining 1.2 percent experienced a complete response, according to the study's abstract.
"What we find is that those patients who have G-C variants have a shorter time to progression, so they progress more rapidly," said Gregorc. "They have a 55-percent probability to progress by treatment, with respect to those patients who were non-G-C, [who] were favored during treatment," she said.
"The same was found in overall survival — those patients with G-C haplotypes, when compared to those with non-G-C-containing haplotypes, had an 86-percent probability to have a shorter survival, and the difference was seven months," said Gregorc. Those patients having a G-C haplotype were more likely to progress under treatment with Iressa, she said.
The group's research demonstrates a genetic contribution to gefitinib treatment, Gregorc said.
"As a medical oncologist, I think it is very important for the FDA to [require subgroup identification for] patients who may benefit from gefitinib, but also for Tarceva, because we have four possibilities in second-line for lung cancer today, two chemo and two tyrosine-kinase inhibitors, Gregorc said. "And for those patients who may benefit from tyrosine-kinase inhibitors, we have to know who they are, and for others, we have to give chemo. I expect both of [the groups of patients] to be specified."
— Chris Womack ([email protected])