Over the past decade, advances in the understanding of tumor biology and driving mutations have led to the discoveries of numerous biomarkers and targeted agents that have revolutionized the fight against cancer. With recent news being so heavily centered on these targeted agents, it seems as though people often forget about the value that chemotherapy has provided over the years and, more importantly, that there is still room for improvement in these older regimens to generate significantly better outcomes for patients. Celgene was looking to take advantage of this sentiment when the company acquired Abraxis BioScience and their leading drug candidate, Abraxane®, an injectable suspension of nanoparticle albumin bound (nab)-paclitaxel.
Pancreatic cancer is a notoriously difficult malignancy to treat, with historical median overall survival hovering around six months from first-line therapy. Pancreatic cancer has remained stubbornly resistant to the modern drug development paradigm, highlighted by the numerous failed studies of targeted agents in combination with gemcitabine (the standard of care). An EGFR monoclonal antibody (Erbitux®, Bristol-Myers Squibb/Eli Lilly), VEGF inhibitors (Avastin®, Genentech/Roche; Nexavar®, Onyx/Bayer; and Inlyta®, Pfizer), an IGF-1R inhibitor (ganitumab, Amgen), and a c-Kit inhibitor (masitinib, AB Science) have all failed to show any additive survival benefit when combined with gemcitabine. The only targeted agent to show a statistically significant benefit and gain regulatory approval was the EGFR tyrosine kinase inhibitor, Tarceva® (erlotinib, Astellas/Roche), and that benefit was only about 10 days of extended overall survival at the median. These failures prompted academics and clinical physicians to re-examine combination chemotherapy, and the quintessential result was presented two years ago at the American Society of Clinical Oncology (ASCO) conference, where the FOLFIRINOX regimen wowed the audience with an 11.1-month median overall survival as first-line therapy (Conroy, J Clin Oncol, 2011). However, the well-publicized toxicity issues associated with FOLFIRINOX do leave the door open for more tolerable regimens to gain utilization, even if they are somewhat less efficacious.
The stage is now set for the Abraxane plus gemcitabine combination to enter the fray. On Friday at the 2013 ASCO Gastrointestinal Cancers Symposium, results were presented from the Phase III MPACT study (NCT00844649) that showed adding Abraxane to gemcitabine increased median overall survival to 8.5 months compared to 6.7 months in the gemcitabine-alone arm (HR=0.72, p=0.000015; Von Hoff, Abstract LBA148). Compared to the historical standard, 8.5 months is an excellent result, but it is also a little bit of a letdown considering that the median overall survival in the Phase II study was 12.2 months. Also, these results will be contrasted with the FOLFIRINOX survival of around 11 months, considering that the trials both enrolled similar “good performance status” patients.
Despite the shorter survival, Abraxane still has the ability to contend with FOLFIRINOX through its other attributes. Comparing the toxicity profiles side by side, as the discussant of the MPACT study did, Abraxane plus gemcitabine looked to be generally more tolerable with lower rates of Grade 3/4 fatigue (17% vs. 24%), diarrhea (6% vs. 13%), neutropenia (38% vs. 46%) and febrile neutropenia (3% vs. 5%) compared to FOLFIRINOX. Although Abraxane plus gemcitabine showed higher rates of Grade 3/4 neuropathy (17% vs. 9%), the discussant pointed out that all neuropathy was not the same and that the neuropathy associated with Abraxane could be resolved relatively quickly through dose delays or reductions, which may not be the case for FOLFIRINOX. However, this side-by-side comparison of toxicities should come with another caveat: The fact that many centers are now favoring the modified FOLFIRINOX regimen, which has the key difference of administering 5-FU in a 48-hour infusion instead of the bolus injection that was used in the original French study. This change has attenuated some of the toxicity of the regimen, but it also highlights another competitive difference between the two regimens. FOLFIRINOX is a complex four-drug regimen requiring a 48-hour infusion, whereas Abraxane plus gemcitabine can be administered concurrently once a week on a three-weeks-on, one-week-off schedule. Could the convenience of administration make the Abraxane plus gemcitabine the regimen of choice outside of large cancer centers? Although not presented at ASCO GI, there’s also the possibility that biomarker data may identify subpopulations of patients who derive even greater benefit with Abraxane; patients with RAS mutations or whose tumors overexpress SPARC are potential biomarkers for patient selection. Stay tuned for this data later this year, possibly at ASCO 2013.
Ultimately, physicians and patients will have to weigh a number of options when choosing their first-line pancreatic cancer therapy, from cost, to ease of administration, to performance status of the patient. It seems likely that FOLFIRINOX will remain the standard of care for the most “fit” patients who are willing to endure the complex regimen, but outside of those patients, Abraxane plus gemcitabine could prove to be a favorable option in the community clinic.
By: Stephanie Hawthorne, PhD, Director, Kantar Health and Josh Garcia, Associate Consultant, Kantar Health
When Avastin® (bevacizumab, Genentech/Roche) burst onto the oncology scene in the mid-2000s, the potential of VEGF inhibition became one of the hottest trends in cancer therapy. The ensuing race to find additional methods of disrupting VEGF signaling led Eli Lilly to throw their hat into the ring with ramucirumab, a fully human, immunoglobulin G1 (IgG1) monoclonal antibody that binds to the extracellular domain of VEGFR Receptor-2 (VEGFR-2). This mechanism of action is unique when compared with Avastin and Zaltrap® (ziv-aflibercept, Regeneron/Sanofi), which inhibit the VEGF ligands. Lilly advanced the development of ramucirumab quickly and aggressively, focusing the Phase III trials in indications where other anti-angiogenic agents are either already approved or in late-stage development: breast, colorectal, gastric, hepatocellular and non-small cell lung cancers. In gastric cancer, the company took ramucirumab directly from Phase I into Phase III, despite limited data suggesting substantial efficacy. Later, the failure of Avastin plus first-line chemotherapy in the AVAGAST study cast a large shadow over the hopes for VEGF inhibitors in gastric cancer.
In October 2012, however, Lilly announced that ramucirumab met its primary endpoint of improving overall survival in the Phase III REGARD study (NCT00917384), which compared best supportive care (BSC) with or without ramucirumab in patients with previously treated gastric or gastroesophageal junction cancer. On Thursday at the 2013 ASCO Gastrointestinal Cancers Symposium, full results from the REGARD study were presented as a late-breaking poster abstract (Fuchs, Abstract LBA5). The poster generated a large amount of attention at the conference, but ultimately the results seemed underwhelming. The study showed that adding ramucirumab to BSC improved progression-free survival by 0.8 months (2.1 months versus 1.3 months; HR = 0.483, p<0.0001) and overall survival by 1.4 months (5.2 months versus 3.8 months; HR = 0.776, p=0.0473), just meeting statistical significance.
With an interesting juxtaposition (as these posters were adjacent to each other), physicians at the conference could easily contrast the REGARD results with those of the Phase III COUGAR-02 study, which evaluated docetaxel plus active symptom control (ASC; essentially this was BSC) in relapsed esophago-gastric cancer (Ford, Abstract LBA4; this abstract was also presented in an oral session). The efficacy results in this study were strikingly similar to those of ramucirumab. The COUGAR-02 study showed that adding docetaxel to ASC provided an overall survival benefit of 1.6 months (5.2 months versus 3.6 months with ASC alone; HR = 0.67, p=0.01).
It looks as though docetaxel and ramucirumab provide similar levels of efficacy benefit in relapsed patients, and the choice of second-line agent may come down to how comfortable physicians feel with treating the myelosuppression of docetaxel or the hypertension and bleeding associated with ramucirumab. Rates of hypertension and bleeding in the REGARD study were unremarkable, 7.6% and 3.4%, respectively, very similar to what was observed with Avastin in the AVAGAST study. This hypothetical choice presupposes that Lilly will file for regulatory approval based upon the REGARD results, but this course of action may not be on the agenda.
After the initiation of the REGARD trial, Lilly seems to have become aware of the potential difficulties in gaining an approval with a placebo-controlled trial in second-line gastric cancer. The results of the COUGAR-02 study highlight the benefit that patients get from an active second-line therapy, and in reality most fit patients are being actively treated in clinical practice today with either a fluoropyrimidine- or a taxane-based regimen (Kantar Health, CancerMPact, Treatment Architecture). Recognizing this weakness, one year after the initiation of the REGARD study, Lilly initiated the Phase III RAINBOW study (NCT01170663) to compare paclitaxel with or without ramucirumab in patients with previously treated gastric cancer. In the eyes of the oncology community, this is the make-or-break study for ramucirumab in gastric cancer. Will ramucirumab be able to show an additive benefit when combined with an active agent in a tumor type where Avastin failed to do so? This is a huge unanswered question for ramucirumab: is the mechanism of action different enough to surpass the activity of Avastin? Perhaps so, since the fact that ramucirumab demonstrates single-agent activity is something Avastin and Zaltrap have never demonstrated in a gastrointestinal tumor. If RAINBOW is positive, approval will be a foregone conclusion, and given the fact that it was initiated only one year after REGARD, it seems likely that top-line results should come later this year. Lilly may be waiting for these results to strengthen their case for approval, but one thing is for sure: in 2013, we will learn whether ramucirumab has what it takes to succeed where prior VEGF inhibitors have failed.
By Neesha Suvarna, PhD, Consultant, Kantar Health and Josh Garcia, Associate Consultant, Kantar Health
By Caroline Helwick
Prior to the 10th annual Gastrointestinal Cancers Symposium, being held January 24-26, 2013 in San Francisco, CA, a January 22 press cast featured studies related to
The symposium is jointly sponsored by American Society of Clinical Oncology, American Gastroenterological Association, American Society for Radiation Oncology, and Society of Surgical Oncology.
S-1 Treatment Out-Performed Gemcitabine in Pancreatic Cancer
Preliminary results from a Phase 3 trial conducted in Japan showed that pancreatic patients who received adjuvant chemotherapy with the oral drug S-1, an oral fluoropyrimide currently used in Asia (and to some extent Europe) but not approved in the United States, had a 44% lower risk of dying, compared with standard treatment with gemcitabine (Abstract 145).
“Our survival data were much stronger than expected,” said Katsuhiko Uesaka, MD, PhD, of Shizuoka Cancer Center Hospital. “Our interim analysis showed S-1 to be superior to gemcitabine in overall survival, and therefore it may be considered as a new standard treatment for resected pancreatic cancer patients in Japan.”
While the study was designed as a non-inferiority trial, the data found S-1 to be superior to gemcitabine. For this reason, Dr. Uesaka said the independent data monitoring committee recommended that the interim results be presented at this time.
The study included 385 patients with stage I-III pancreatic cancer who received post-operative treatment with gemcitabine or S-1. The interim analysis found an overall survival benefit for the experimental arm, with S-1 associated with a 44% reduction in mortality. The 2-year survival rates were 70% with S-1 versus 53% for gemcitabine (P<0.0001 for non-inferiority; P<0.0001 for superiority). The 2-year relapse-free survival rates were 40% and 29%, respectively.
S-1 was well tolerated in the Asian population, with more than 70% of this Asian population able to complete treatment.
Due to differences in toxicity according to ethnicity—probably a result of differences in how the drug is metabolized—the findings are not immediately applicable to non-Asian populations. Previous studies have shown that Caucasians have considerably more diarrhea as a side effect. However, Dr. Uesaka maintained that with dose adjustments, “I expect S-1 may be applicable to all patients with pancreatic cancer,” although at this time the use of the drug in the U.S. remains uncertain. A Phase 3 trial in the United States is currently exploring the drug for the treatment of stomach cancer.
Three Molecular Subtypes of CRC Identified
Similar to what has been done in breast cancer, tumor gene expression patterns can now classify colorectal patients into three distinct molecular subtypes, an international research team is reporting. The tumor subtypes are associated with different prognoses and responses to adjuvant chemotherapy, said Joseph Tabernero, MD, of Vall d’Hebron Institute of Oncology in Barcelona, Spain (Abstract 333).
“This study clearly shows that there are different subtypes in colorectal cancer with completely different biological and clinical characteristics,” he said. In the future, this classification system might be used to identify patients who require aggressive treatment and to predict which drugs will be most benefit in an individual patient.
This may be especially helpful in stage II patients, since clinical and pathological factors do not reliably indicate which of those patients are at greatest risk of relapse. Recent genomic classifiers—in particular, Oncotype® and ColoPrint®—can help in this regard, but current guidelines do not outline which should receive additional treatment. Current tests are also not helpful after relapse.
In the current study, gene expression data from 188 colorectal cancer patients (stage I-IV) were used to develop a molecular subtype classification system. The system was subsequently validated in tumor samples from 543 stage II and III patients. Investigators determined that 21.5% of the samples could be classified as subtype A, 62% as subtype B, and 16.5% as subtype C.
The three intrinsic subtypes differed in three biological hallmarks of colorectal tumors:
These features are known to independently affect outcomes in patients with cancer; in fact, the study did show that response to treatment varied by subtype, noted Dr. Tabernero.
Follow-up of the patients at 10 years revealed that subtype C had the worst outcome of the three and did not benefit from adjuvant chemotherapy. Subtype A had a good prognosis regardless of whether chemotherapy was given. Subtype B had a significantly better outcome with chemotherapy, versus no chemotherapy.
“In subtype B the effect of chemotherapy is pronounced. To the contrary, type C patients who got adjuvant chemotherapy did not benefit from it,” reported Dr. Tabernero.
“We have developed a diagnostic single-sample predictor that allows the classification of colorectal cancer tumors of different intrinsic molecular subtypes. These subtypes are potentially clinically relevant, as they differ in their underlying biology and clinical outcomes and consequently require different treatment strategies,” he concluded.
Researchers are currently evaluating the molecular subtypes in stage IV colorectal cancer.
Surgery After Imatinib Improves Survival in GIST
Patients with GIST who had residual disease resected after treatment with imatinib lived longer than those who received only imatinib, in a retrospective study from South Korea (Abstract 62). While previous studies have suggested as much, design flaws prevented conclusions to be drawn.
Seong Joon Park, MD, of Asan Medical Center in Seoul, reported the results of a study involving 134 patients with metastatic or recurrent GIST who were treated with imatinib alone or imatinib plus surgery to remove residual tumor (followed by more imatinib). Patients were eligible for the study if they responded to prior imatinib for at least 6 months.
Dr. Park reported that the median progression-free survival in the non-surgical group was 42.8 months, increasing to 87.7 months in patients who underwent surgery (P=0.001). Median overall survival has not been reached in the surgery arm, and was 88.8 months in the imatinib-alone arm (P=0.001).
The addition of surgical management reduced the odds of dying five-fold, he added, noting that about one-third of patients can undergo a successful resection for residual disease, depending upon tumor and patient characteristics. In a multivariate analysis, low initial tumor burden was associated with longer overall survival and low initial tumor burden, female gender and a specific alteration in the KIT gene (KIT exon 11 mutation, which is associated with improved responses) were associated with delayed disease progression.
“Considering the retrospective design of this study, it is difficult to established treatment guidelines based on this, but it provides good evidence supporting a clinical practice that is already widely adopted,” said Dr. Park.
Gene Expression Profiling of CTCs in Pancreatic Cancer
Gene expression profiling (GEP) of circulating tumor cells (CTCs) may help to personalize chemotherapy for patients with pancreatic cancer, researchers suggested after showing that CTCs predict for response to standard chemotherapy regimens (Abstract 142).
CTCs, shed by the tumor and accessible in the blood, can determine the molecular makeup of a tumor less invasively than biopsy and can be used to monitor treatment response. Tests for CTCs are already applied in breast cancer and are well suited to pancreatic cancer, which is difficult to biopsy. The study employed pharmacogenomics modeling, which has been used in other cancers; this is the first study to use this method to examine genetic response to chemotherapy regimens.
“This research lays important groundwork for customizing treatments according to a patient’s genetic composition,” said Kenneth Yu, MD, of Memorial Sloan-Kettering Cancer Center, New York. “We’re seeing encouraging signs that this strategy can help us determine which patients will benefit most from chemotherapy and more quickly identify the development of treatment resistance, even before physical changes in the tumor appear.”
The study enrolled 50 patients with unresectable stage II-IV pancreatic adenocarcinoma, who received 12 different drug combinations per physician’s choice. Circulating tumor progenitor cells were collected prior to chemotherapy and upon disease progression with the Vita-Cap™ device (Vitatex, Inc); total RNA was extracted and gene expression analysis performed. A pharmacogenomics model for the 12 chemotherapy regimens was created by CellPath Therapeutics and applied to the gene expression profiles to predict sensitivity to chemotherapy. Investigators compared gene expression differences between patients who continued to respond to a drug regimen and those who progressed while on treatment, looking at the individual gene level and also the genetic pathway level.
“The clinical questions were whether the pharmacogenomics model would predict treatment response and resistance, and whether the pharmacogenomics profile would change when the cancer progressed,” said Dr. Yu.
Differences in outcomes were apparent according to the model, among the 20 patients who had progressed at the time of the current analysis. Six patients received treatment that the model predicted they would be sensitive to; 6 received a regimen predicted to elicit an “intermediate” response; and for 8 patients, the model predicted complete resistance.
“Patients receiving treatment predicted by our model to be more effective did better,” he reported. Dr. Yu emphasized that all patients received standard treatment. The preliminary analysis showed that patients for whom chemotherapy was predicted to be effective had a median time to progression of 7.3 months, compared with 5.3 months for the intermediate group and 3.7 months for the resistant group. Interestingly, changes in chemotherapy sensitivity patterns were evident at disease progression, reflecting the development of treatment resistance, he added.
“We also did a sophisticated analysis of the gene pathways and showed that disruption in pathway E2F1 predicted for longer treatment response and disruption in the NFkB pathway predicted for shorter duration of response,” he said. Upon disease progression, the phospholipase C and retinoblastoma 1 pathways were disrupted.“
Gene pathway analysis provides insights into pancreatic cancer prognosis and the development of drug resistance, he concluded. “Ultimately, we hope this strategy can be used to determine which drug cocktail would be most suitable for the patient and to monitor patients during the course of therapy, so treatments can be modified at the earliest molecular sign of disease worsening,” said Dr. Yu.
Neal Meropol, MD, Chief of Hematology/Oncology at University Hospitals Case Medical Center at Case Western Reserve University in Cleveland, OH who moderated the session, said the studies by Dr. Yu and Dr. Tabernero highlight the direction that the field of oncology is going, both for research and clinical practice, that is, the future categorization of tumors molecularly. “We are starting to use genetic makeup to make treatment decisions regarding which drugs work best against a patient’s tumor and which may not work at all,” he said.
Second-Line Docetaxel Improves Esophageal and Gastric Cancer Survival
A Phase 3 study from the United Kingdom showed that second-line treatment with docetaxel improved overall survival in advanced esophago-gastric cancer. The strategy has already been adopted, but the study is the first to show definitive evidence of a survival benefit with docetaxel given in the second-line setting, said Hugo Ford, MD, of Addenbrooke’s Hospital in Cambridge, UK (Abstract LBA4).
Patients with advanced disease who relapse after first-line chemotherapy have a median overall survival of only 3 months, without second-line treatment, noted Dr. Ford. “The current practice in the United States and much of Europe is to give second-line chemotherapy, even though the evidence isn’t as strong as we would like,” he said. This is the first trial to show second-line chemotherapy extends survival, without causing deterioration in quality of life.
The COUGAR-02 study enrolled 168 patients with locally advanced or metastatic esophago-gastric adenocarcinoma whose disease progressed within 6 months of first-line treatment. Patients were randomly assigned to receive docetaxel (75 mg/m2 every 3 weeks for up to 6 cycles) or active symptom control, which included radiotherapy, steroids and/or supportive care. The median overall survival with docetaxel was 5.2 months, versus 3.6 months for patients receiving only symptom control, a 33% reduction in mortality (P=0.01).
Partial responses to chemotherapy were observed in 7% and disease stabilization in 46%. Docetaxel was associated with grade 4 toxicity in 21% of patients.
The survival benefit was evident across all subgroups; the most benefit was seen in “very fit” patients with performance status of 0. Preliminary data on quality-of-life suggest docetaxel led to improvements in pain and no loss in global quality-of-life or function. “However, benefits were seen across all groups, with no cohort of patients seeming to fare worse with chemotherapy,” said Dr. Ford.
According to him, the study confirms docetaxel as a standard second-line therapy in the United Kingdom, and informs treatment choices for physicians in the United States.
Dr. Meropol commented that the study is “a model for providing an evidence base to help guide our treatment decisions at points in time when the goals are palliative and not curative, and a model for the type of study we would like to see more of…Treatment toward the end of life is an area where we have fallen a little short (regarding evidence) and we need to pay additional attention.”
On November 29, 2012, CMS released draft guidance updating the status of Coverage with Evidence Development or CED at the agency. Clearly, there continues to be an effort within CMS to justify and ease in the more regular use of CED in policy setting. The following will address a brief history of CED and important statements by CMS in its latest draft guidance.
CED, in CMS’s own words, is a mechanism “through which we provide conditional payment for items and services while generating clinical data to demonstrate their impact on health outcomes”. In theory, it is a reasonable attempt by Medicare to make a promising technology available earlier in its developmental process to patients in need while, at the same time, requiring the ongoing collection of data on the clinical outcomes associated with the technology’s use. It is clearly important for clinicians and patients that there be the continual analysis of data to strictly define the risks and benefits attending the use of any drug, device or procedure. It is also important that there be predictability, timeliness and certainty in the coverage process for patients, clinicians and for the innovator companies that have invested the monetary and intellectual capital to bring an innovation to patients who need it.
CED was formally introduced by private payers in the literature (McGivney, JNCI, May 20, 1992) as a concept twenty years ago. The impetus for the private payer postulate was the tremendous diffusion and utilization of a drug based procedure called high dose chemotherapy with bone marrow transplant (HDC/BMT). The use of this therapeutic strategy in the early 90s, with its serious toxicity and cost of at least $100,000, targeted patients with hematologic cancers and solid tumors such as breast cancer. Thus, it is important to note that the intended focus of the initial application of CED was to be on procedures for which no regulatory agency had responsibility, unlike the FDA with its oversight of drugs and devices. The initial CMS application of CED in 1996 was also on a procedure, lung volume reduction surgery. However, the resurgence in CED in 2005 and 2006 focused on devices and, also, drugs for colorectal cancers.
The most recent CMS pronouncement continues to highlight important issues to watch.
Firstly, the guidance is issued so that CED “can be used more widely and effectively as a driver for innovation”. While the term “driver” is encouraging, it is viewed skeptically by some who suggest that the phrase “manager of innovation” might be substituted more accurately. Notwithstanding the semantics, the quoted phrase is yet another signal that CED is indeed going to be applied more often at CMS.
Secondly, a basic argument and focus of CMS is that often drugs, biologics, devices and procedures are evaluated scientifically in studies where the study populations are not related or relevant to the Medicare beneficiary population. CMS posits the need for CED based upon a principal CED function of generating new evidence “on the benefit or harm of an item or service among the beneficiary population based upon rigorous scientific inquiry”. For drugs and biologics this could eventually mean that FDA approval does not guarantee complete coverage if CMS were to adjudge that the studies for FDA clearance were not conducted in a population close enough to the Medicare beneficiary population. Also, off-label use is a likely target for further evaluation through CED. The “prospective” CED studies could take years given that CMS indicates that upon completion of the study the collected data must be communicated within 2 years. One only needs to do the math in terms of total time for study design, design approval, the accrual of patients, final accrual of the last patient, collection of the final data points, analysis, and publication.
An interesting twist is the verbiage spent on justifying the authorization for the AHRQ to conduct and support needed studies. Adding AHRQ more directly in the mix can serve only to extend a process in juxtaposition to an era of rapid advancement in science and technology in cancer care. Remembrances of the Office of Health Technology Assessment (OHTA) in the predecessor agencies, AHCPR and NCHCTR, and the attendant delays in reaching coverage policy determinations raise further concern.
While there are many other noteworthy points, it is important to highlight that CMS points out the “complementary roles of randomized controlled clinical trials (RCTs) and practical observational studies to close outstanding evidence gaps and allow coverage after an RCT ends where appropriate”. The era of observational data based upon sound methodology being accepted as a basis for important policy decisions is approaching.
The sixty day comment period ends on January 28, 2013. Informal comments are accepted here.
By William T. McGivney, Ph.D., Principal, McGivney Global Advisors