Originally published June 3.
NEW YORK (GenomeWeb) – Data from two large studies that tested the efficacy of widely used targeted drugs in molecularly defined breast cancer and colorectal cancer populations failed to show superiority of one therapeutic strategy over another.
During a symposium at the American Society of Clinical Oncology's annual meeting this week, researchers presented data from these two studies. Clinicians launched these studies in the mid-2000s with the hopes of gaining greater insights into the management of HER2-positive breast cancer patients in the adjuvant setting and the first-line treatment of metastatic KRAS wild-type colorectal cancer patients. While the failure of the large breast cancer study left one expert lamenting the incremental nature of advances in the life sciences field, the unclear findings of the colorectal cancer study raised hopes in another researcher that with greater molecular understanding of the disease, straightforward treatment guidance will come in the near future.
In the trial, called ALTTO, researchers from various institutions led by Martine Piccart-Gebhart of the Breast International Group in Belgium studied whether adjuvant treatment with two HER2 therapies – Genentech's Herceptin (trastuzumab) and GlaxoSmithKline's Tykerb (lapatinib) – improved HER2-positive breast cancer patients' outcomes compared to receiving Herceptin monotherapy. They found that regardless of whether patients received the combination regimen of Tykerb, Herceptin, or Herceptin then Tykerb sequentially, receiving two targeted HER2 treatments didn't significantly extend patients' time without relapse compared to those receiving just Herceptin.
In reflecting on the data at the symposium, past ASCO President and breast cancer expert George Sledge stated that given the findings from ALTTO, Herceptin plus chemo remains the standard of care for the treatment of HER2-positive breast cancer patients in the adjuvant setting, as it has been for a decade. Although outcomes for breast cancer patients have greatly improved with this strategy, there is still room for improvement, and researchers have put forth a number of strategies, such as dual HER2-blockade, prevention of HER2 dimerization, using HER2 receptor to deliver more lethal drugs to the tumor, or upstream and downstream blockade of the HER2 pathway.
The aim of the ALTTO study, Sledge said, was to determine if dual HER2 blockade in early-stage breast cancer patients improved disease-free survival and move the field closer to a cure for HER2-positive disease. "HER2 has reverted to the incrementalism that we see so commonly in the adjuvant field," Sledge said. "This is not a great leap forward. This is a serious disappointment, not just for investigators but for the entire field."
Meanwhile, in metastatic colorectal cancer patients with wild-type KRAS tumors, a Phase III study comparing two targeted agents with two types of chemotherapy – Avastin plus irinotecan-based chemo (FOLFIRI) against Erbitux plus FOLFIRI; and Avastin plus oxaliplatin-based chemo (mFOLFOX6) versus Erbitux plus mFOLFOX6 – failed to identify any one regimen as the clear winner in the first-line setting. The trial "didn't meet the endpoint of superiority," acknowledged Josep Tabernero of Vall d'Hebron University Hospital in reviewing the data at the symposium. Still, he didn't feel that this was a negative study "because it is very informative, and eventually with more sub-group analysis the interpretation may ... change."
In the 8,300-patient ALTTO trial, researchers followed HER2-positive breast cancer patients for a median of four and a half years. The hazard ratio for disease-free survival events (or relapses) in the Tykerb/Herceptin versus Herceptin arm was .84, while in the sequential treatment the hazard ratio was .93 — neither finding was statistically significant. During the study, 360 patients died. The four-year survival rate on the combination arm was 95 percent in both the combination and sequential treatment arms, and 94 percent on the Herceptin only arm.
"The small signal seen in favor of the combination of lapatinib and trastuzumab appears to be driven at this point in time by the hormone receptor-negative cohort and by the sequential design," Piccart-Gebhart said at the meeting. Approximately 20 percent of women with breast cancer have HER2-positive disease. Within this group, 50 percent are hormone receptor-positive and the other half are hormone receptor-negative.
The study ended up being underpowered because there were 555 patients who relapsed, far fewer than the around 800 disease-free survival events researchers had targeted initially. "It's difficult to mount any enthusiasm for dual-blockade of HER2 in the adjuvant setting, at least for the combination showed here," Sledge said. "Not only doesn't it work, but the combination … is significantly more toxic than trastuzumab alone."
When giving patients two targeted agents, researchers must adjust the doses carefully to manage toxicities. A maximum dose of Tykerb could only be given to around two-thirds of study subjects, while 90 percent of patients received an optimal Herceptin dose. Diarrhea, rash , and adverse events implicating the liver, gall bladder, and bile ducts were more frequent in the combination arm than with just Herceptin therapy. The diarrhea events seen in patients treated with Tykerb required aggressive management.
Although the ALTTO trial did not meet its primary endpoint, "patients did extremely well," Piccart-Gebhart reflected. The disease-free survival rates exceeded 80 percent in the three arms.
The hypothesis for treating HER2-positive breast cancer patients with two targeted agents emerged from past studies showing that more complete blockade of the HER2 pathway caused tumors to shrink and malignant cells to die. However, earlier studies such as the Neo-ALTTO, also led by Piccart-Gebhart, tested this hypothesis in the neoadjuvant setting (treatment before surgery), not in adjuvant setting (after surgery). In Neo-ALTTO, researchers reported that breast cancer patients had a doubling of pathologic complete response from dual HER2 blockade with Tykerb and Herceptin compared to treatment with Herceptin, and that this effect was particularly pronounced in women with hormone receptor-negative disease.
Researchers had hoped to see similar results in the adjuvant treatment of HER2-positive breast cancer patients. However, "the doubling of pathologic complete response observed with lapatnib and trastuzumab in the Neo-ALTTO trial did not translate into improved survival outcomes in ALTTO at 4.5 years median follow up," Piccart-Gebhart said.
Sledge reflected during the symposium that the negative results from ALTTO necessitate the field to rethink its approach to the development of new drugs for early-stage breast cancer. Although the rationale for ALTTO was based on Neo-ALTTO results showing an impact of dual blockade on pathologic complete response, Sledge put forth that the endpoint may not truly reflect whether a therapy will improve survival outcomes for breast cancer patients in early stages of disease.
The US Food and Drug Administration in September approved Roche/Genentech's HER2-targeted agent Perjeta (pertuzumab) in combination with Herceptin and chemo, making it the first neoadjuvant treatment for breast cancer applying a dual-HER2 blockade strategy. Roche gained accelerated approval for Perjeta by submitting data from a study with pathologic complete response as an endpoint. Ahead of that approval, the agency issued a draft guidance in 2012 advising drug developers on how to use pathologic complete response – when patients have no signs of invasive cancer in their breast and lymph nodes at surgery – as a surrogate endpoint to garner accelerated approval for a neoadjuvant treatment for high-risk, early-stage breast cancer.
Given the available data at the time, "ALTTO represented a reasonable test of a hypothesis that improvements in pathological complete response rates are associated with improved disease-free survival," Sledge said. If this hypothesis had proved true in ALTTO, then it could have afforded pharma companies and researchers more confidence in using pathologic complete response as a surrogate to survival endpoints in drug development trials, facilitating a shorter pathway to market. "These hopes have now been dashed," Sledge said. "The null hypothesis won, as so often happens in clinical trials."
Now, given the findings of ALTTO, Sledge expressed doubt that accelerated approval should be granted to early-stage breast cancer treatments based on studies with pathologic complete response as an endpoint. The results of the study for Perjeta's approval were all too reminiscent of the Neo-ALTTO study for Sledge. "The ALTTO trial is certainly not a confidence booster for this approach," he noted. "We currently lack convincing evidence that differences in pathologic complete response rates in smaller neoadjuvant trials predict the outcome of larger adjuvant trials."
It is well established that metastatic colorectal cancer patients with mutations in KRAS exon 2 codons 12 and 13 don't respond to EGFR-inhibiting monoclonal antibodies, such as Bristol-Myers Squibb/Lilly's Erbitux (cetuximab) and Amgen's Vectibix (panitumumab). Although Genentech's Avastin (bevacizumab) and Erbitux in combination with either FOLFIRI or mFOLFOX6 are both used by doctors for first-line treatment of metastatic colorectal cancer patients with wild-type KRAS tumors, there is no data on which strategy is better.
In an effort to identify the best strategy for this subset of patients, researchers led by Alan Venook of the University of California, San Francisco analyzed the responses of more than 1,100 patients with KRAS wild-type tumors. They reported median progression-free survival of 10.84 months versus 10.45 months in the Avastin plus chemo versus Erbitux plus chemo cohorts; median overall survival was 29 months and 29.9 months, respectively. "These are essentially the same, and in fact, one is no better than the other, based on this data," Venook said during the symposium.
Then, researchers drilled down to gauge if the particular chemotherapy combined with the targeted agent made a difference in survival. With mFOLFOX6, median overall survival was 30.1 months in the Erbitux arm and 26.9 months in the Avastin arm, but the finding wasn't statistically significant. "It is interesting to note that these [survival] curves diverge at two years," Venook said. "I would argue that [Erbitux] plus FOLFOX is a reasonable option for patients in first-line metastatic colorectal cancer."
In a smaller subset of those receiving FOLFIRI, median overall survival was 33.4 months on Avastin and 28.9 months on Erbitux. Although this finding also wasn't statistically significant, Venook highlighted that the survival curve of the Avastin arm separated from Erbitux at two years.
"There was a subset of patients who were rendered disease-free with surgery and chemotherapy," Venook added, noting that these approximately 120 patients had median overall survival of over five years.
Toxicities were in line with researchers' expectations. Half of the study subjects had Grade 3 adverse events and there were 10 deaths due to toxicities. Common reactions, such as rash and diarrhea, were associated with Erbitux, and gastrointestinal issues and hypertension were due to Avastin treatment.
Venook and his colleagues are conducting expanded RAS analysis with the hopes of identifying a subset of patients who might respond differentially to these various regimens. Researchers will analyze 44,000 specimens in the Southwest Oncology Group's biorepository.
Recently, investigators from Amgen and elsewhere reported data form the PEAK trial, in which they analyzed close to 300 metastatic colorectal cancer patients with wild-type KRAS tumors randomized to Avastin plus mFOLFOX6 versus Vectibix (panitumumab) and mFOLFOX6. Median progression-free survival was similar between the two cohorts, 10.9 months in the Vectibix arm versus 10.1 months in the Avastin arm, while median overall survival was 34.2 months versus 24.3 months, respectively. However, when researchers in PEAK conducted an expanded analysis on banked samples to identify and exclude patients with additional RAS mutations in KRAS and NRAS in exons 2, 3, and 4, median overall survival was 41.3 months in the Vectibix arm versus 28.9 months in the Avastin arm.
A growing body of literature suggests that not just KRAS mutated colorectal cancer patients, but also those with other mutations in RAS pathway, have limited response to Vectibix and Erbitux. "By applying this important molecular knowledge, we have been able to narrow down the target population for EGFR monoclonal antibodies," Tabernero said, noting that the expanded RAS analysis by Venook and colleagues might change the survival estimates in their study.
This trial "marks an important milestone in the roadmap of metastatic colorectal cancer treatment," Tabernero added during his review of the data. He noted that since neither Avastin, nor Erbitux plus chemo emerged as the clear choice in KRAS wild-type metastatic colorectal cancer, doctors should factor in safety, efficacy, cost, and individual patients' characteristics in deciding which strategy to go with.