Originally published June 1.
CHICAGO (GenomeWeb) – Next-generation EGFR inhibitors for treating metastatic non-small cell lung cancer patients who have acquired resistance to first-generation drugs in this class accurately hit mutant EGFR tumor cells and caused fewer serious side effects, early data presented at a major cancer conference showed.
Researchers at the American Society of Clinical Oncology's annual meeting here this week, presented preliminary data from human studies on three next-generation EGFR inhibitors: AstraZeneca's AZD9291, Clovis Oncology's CO-1686, and Hanmi Pharmaceutical's HM61713. All three agents showed promising activity against patients who had EGFR mutations, had received prior treatment with a first-generation tyrosine kinase inhibitor – such as Roche's Tarceva (erlotinib) and AstraZeneca's Iressa (gefitinib) – and had T790M mutations.
Studies estimate that 10 percent of NSCLC patients in the US and 35 percent of East Asians have EGFR-mutated tumors. It's well established that patients with these EGFR mutations experience a robust benefit from therapies that block this eponymous receptor. But these first-generation drugs are associated with significant toxicities such as rash and diarrhea.
"First generation drugs began long before we knew about EGFR receptor mutations," Andrew Allen, Clovis' chief medical officer, told PGx Reporter at the meeting. "That was not in people's minds as [these drugs] were developed. Therefore they were developed in cell lines that carried wild-type EGFR and obviously are very potent wild-type EGFR inhibitors."
Then, a decade ago researchers discovered that patients with EGFR mutations responded particularly well to tyrosine kinase inhibitors. "We … found that tumors with these mutations … actually were exquisitely sensitive to the first-generation drugs, actually more sensitive than the wild-type [tumors] are," Allen said. "But the doses that were used were enough to hit the wild-type [cells] and hence [it caused] rash."
Now, with advancing knowledge of tumor biology, researchers have identified a number of markers that spur resistance in metastatic NSCLC patients previously treated by an EGFR tyrosine kinase-inhibiting drug. Approximately 60 percent of pretreated patients have T790M mutations which researchers believe is the most important resistant mutation in the context of EGFR tyrosine kinase treatment. "Those drugs were carefully designed based on what we knew at the time," Allen said of first-generation tyrosine kinase inhibitors. "We just know more now."
Thus, in battling resistance to these first-line therapies, drug makers are focused on developing agents that are effective against fighting the tumor, but without the toxicities of older drugs. "All three of the new drugs from the recent studies show specificity against T790M and EGFR mutations while relatively sparing EGFR wild-type," Yale Cancer Center's Thomas Lynch said during the symposium, adding "That explains the differential toxicity we see" from older drugs.
For example, CO-1686 was designed to have reduced binding affinities for EGFR proteins without one of the resistance mutations. "There is a warhead to the molecule that binds to the cystine residue and that's the piece responsible for the covalent binding, and the remainder of the drug is all about optimizing the fit so it fits well in mutant [EGFR] but not well in wild-type," Allen explained. "That's obviously what all of the companies have tried to accomplish."
For AstraZeneca's AZD9291, researchers led by Pasi Janne of Dana-Farber Cancer Institute evaluated nearly 200 patients in various dose escalation and expansion cohorts, and reported a 51 percent response rate in all evaluable patients. Out of 89 patients with T790M mutations, 64 percent responded, while 23 percent of those without T790M mutations responded to AZD9291. Disease control was achieved in 85 out of 89 T790M-positive patients. In this study, still being conducted, the longest enduring response is more than nine months.
A minority of patients (6 percent), experienced Grade 3 or higher adverse events, and patients most commonly experienced nausea, rash, and diarrhea. As researchers increased the dose of AZD9291 from 20 mg to 240 mg, more patients experienced Grade 3 and higher serious adverse events. “This suggests that these higher concentrations for AZD9291 are starting to inhibit wild-type EGFR,” Janne acknowledged during a symposium at the meeting. For Phase II studies, AstraZeneca will use an 80 mg dose of AZD9291.
During these early phases of study, researchers are still working out the mechanism of the drugs and why certain patients responded and why others didn't. Noting that 23 percent of T790M-negative patients responded to AZD9291, Janne posited one reason for the activity seen in this subset was due to a "re-treatment effect" to tyrosine kinase inhibitors – when patients previously treated and grown resistant to EGFR inhibitors become sensitive again after a break from therapy.
Janne concluded that the data so far suggests that AZD9291 will likely impart a progression-free survival advantage for patients in larger, late-stage studies. There was also a strong signal toward a significant progression-free survival advantage with Clovis' CO-1686.
Meanwhile, Lecia Sequist of Massachusetts General Hospital presented data on 72 patients pretreated with tyrosine kinase inhibitors. At the time of evaluation, the objective response rate was 58 percent in those with T790M mutaitons. "In patients with longer follow up, the responses do seem to deepen over time and can be quite durable," Sequist said. There are patients currently in this trial that have been on CO-1686 and responsive for almost one year. "While these data are not mature ...the median progression-free survival hasn't been reached and the current Kaplan-Meier [probability] estimate is that it will reach 12 months," she said.
Specifically designed to target mutant EGFR and T790M tumor cells and spare wild-type EGFR, CO-1686's mechanism has so far yielded a positive toxicity profile, as only three patients had any form of rash. In his review of the data, Lynch noted that the rate of rash with this drug was similar to that of patients receiving placebo. The US Food and Drug Administration has granted breakthrough therapy designation to CO-1686.
Meanwhile, more than 20 percent of patients in the study experienced hyperglycemia. "As the study has evolved and we have realized that hyperglycemia is an observed adverse event, we have gotten better at detecting it and preventing severe episodes," Sequist said. "Most patients have been well controlled with dose reduction and an oral agent, typically metformin."
Finally, Dong-Wan Kim at Seoul National University Hospital presented data on HM61713 showing that 18 out of 83 patients in the study had a partial response, all of whom were positive for T790M. The response rate in the overall population was 21.7 percent. In T790M-positive patients, the response rate was 29 percent, while in T790M-negative patients the response rate was under 12 percent. Approximately 60 percent of the T790M-positive patients are still continuing on the trial. Researchers are still studying increasing doses of HM61713.
More than 10 percent of patients in this study had common drug-related adverse events, such as skin exfoliation, nausea, diarrhea, and rash. However, most adverse events were Grade 1 or 2, and easily manageable, the study authors reported in the abstract.
Sparing wild-type EGFR cells
Considering it has been 10 years since EGFR mutations were first discovered to impact patients' response to tyrosine kinase inhibitors, "how can we improve therapy for these patients?" posited Lynch. "Unfortunately, not all of our patients are cured from first-line tyrosine kinase inhibitors." He articulated that the goal of next-generation tyrosine kinase inhibitors is to remain potent against the tumor while sparing patients the toxicities of earlier agents.
While he wouldn't outright back a single drug, noting that the data are evolving and preliminary, Lynch noted HM61713 had "impressive" wild-type EGFR sparing, with milder rash and diarrhea than what's been observed in older tyrosine kinase inhibitors. According to Kim, HM61713 is 1,000 times more selective for EGFR mutant cells over wild-type cells.
Lynch also said that CO-1686 had "outstanding" clinical activity, pointing out that the treatment of hyperglycemic patients with metformin – a drug used to control blood sugar in type 2 diabetics – might actually be bolstering the efficacy seen in the study. "All you have to do is open up … the ASCO abstracts and search for metformin and cancer and you see spectacular interest in this agent as an anticancer drug," he reflected. "I wonder if what Dr. Sequist is showing us is really a combination trial of metformin and CO-1686 in NSCLC.”
In the course of developing CO-1686, Clovis changed the formulation from a free base form to a hydrobromide salt formulation. Changing the formulation has made a difference in being able to dose the drug, he Lynch added.
Similarly, he noted that AZD9291 also had outstanding clinical activity and that the potential of the drug to impact progression-free survival was "very promising." Without picking a favorite, Lynch seemed to linger on the duration of responses seen with the Clovis drug.
If the PFS holds, "it won't matter that [CO-1686] is a [twice-a-day] drug. It won't matter if you have to use metformin," he said, adding that clinicians are fairly good at managing hyperglycemia. “What will determine the winner, is which is the better drug in terms of progression-free survival and response rate.”
Meanwhile, drugmakers continue to improve EGFR targeting agents in the first-line setting. The FDA a year ago approved a new agent, Boehringer Ingelheim's Gilotrif (afatinib) for first-line metastatic NSCLC patients whose tumors harbor EGFR exon 19 deletions or exon 21 L858R substitutions. Researchers presented pooled overall survival analysis involving two Gilotrif studies at this meeting.
The sponsors of these three next-generation drugs – AstraZeneca, Clovis, and Hanmi – are continuing to study these agents in later phases of study and in different patient populations. Although these initial studies are focused on investigating these newer agents in heavily pretreated patients, the drugmakers are on the lookout for signals that suggest they can also be effective as first-line NSCLC agents.
For example, Sequist noted that data from xenograft models suggest the CO-1686 might be a viable front-line attack against metastatic NSCLC. These studies do "offer the opportunity for us to think about first-line use for these new agents," Lynch agreed.