WASHINGTON, DC—Scientists would always hope their research projects show positive results, with findings that benefit patients, but as has been seen with chemotherapy in general and biological agents in particular, benefits are mostly incremental.
With a blockbuster—imatinib in CML, platinum for testicular cancer—there's no need to justify the cost-benefit ratio. But what about the novel, targeted agents that improve outcomes by single-digit percentages, cost tens of thousands of dollars per treatment cycle, and carry significant risks of debilitating side effects? In a Special Session here at the American Association for Cancer Research Annual Meeting on that much-discussed topic one speaker showed how data from trials with barely statistically significant benefits nonetheless became acceptable regimens currently found in practice guidelines.
Chemotherapy & Marginal Benefits
Antonio Tito Fojo, MD, PhD, Senior Investigator and Head of the Experimental Therapeutics Section at the National Cancer Institute, began his talk by warning the audience that he would be a little more critical than they were used to.
Dr. Fojo's principal message was that “increasingly, therapies that demonstrate at best marginal benefit are being approved for the treatment of cancer.”
Cetuximab was his first example. Dr. Fojo described the European FLEX study treating non-small-cell lung cancer with a cisplatin-vinorelbine combination, with or without cetuximab. The key result he cited was a 1.2-month survival advantage with the addition of cetuximab, 11.3 months vs 10.2 months.
The trialists' conclusions he quoted had a familiar ring: “The addition of cetuximab to platinum-based chemotherapy represents a new treatment option for patients with advanced non-small-cell lung cancer.”
How could a benefit that minimal be considered grounds for adding that regimen to the medical oncologists' armamentarium, Dr. Fojo questioned.
“The bar for what we call ‘significant’ has fallen so low that we risk tripping over it,” Dr. Fojo said.
He added that the minimal survival benefit must also be balanced against Grade 3/4 toxicities such as acne-like rash (10%) and diarrhea (5%), and Grade 1/2 toxicities “that will be occurring every day of the patient's life and can actually be quite detrimental.
“I would characterize this trial as discouraging, disappointing, and marginal, but what is it about these words that we're afraid to use them?” he said.
His next set of slides relating to cost were eye openers, showing how the average cost of drugs to treat one patient with cisplatin-vinorelbine per protocol in this trial added up to $1,920.
“For an extra 1.2-month survival advantage with cetuximab, add $71,712” per patient, Dr. Fojo said. “That's not the cost of life, that's the cost of that drug, and I would argue that it's a bit too steep.”
He said another cetuximab NSCLC trial, the US BMS099 study, using a taxane-carboplatin platform, showed no statistically significant survival advantage to adding cetuximab, and the overall response rate improved only from 17.2% to 25.7%. But it did show toxicities similar to FLEX, including a 10.5% rate of acneiform rash and a 5.5% rate of infusion reaction.
“What is it about ‘ineffective’ that we do not understand?” Dr. Fojo said.
Another example of questionable benefit is the E2100 trial of bevacizumab in breast cancer, he said. Combining bevacizumab with paclitaxel increased overall survival by 1.5 months—25.2 months with paclitaxel-placebo vs 26.7 months with paclitaxel-bevacizumab—but at an additional cost of $90,816 per patient for bevacizumab over the $3,029 cost of paclitaxel.
These trial results wouldn't mean much unless practitioners started using the costly, marginally effective regimens routinely, he continued, noting that in fact the FLEX study did influence ASCO's 2009 clinical practice guidelines for treatment of Stage IV NSCLC, which advise that the addition of cetuximab to cisplatin/vinorelbine can be considered in patients with tumors testing positive for epidermal growth factor receptor.
That's “institutionalized mediocrity,” Dr. Fojo said.
The guideline's reference to EGFR positivity implies that a subset of patients might achieve a better than median survival advantage with cetuximab, but “the problem is, in [FLEX], 85% of the patients qualify [as EGFR positive] because you can find a little bit of EGFR staining in just about anybody with lung cancer. And this way you can treat more patients [with cetuximab], if that's what you want to do.”
Getting back to the BMS099 cetuximab study, he said no efficacy parameter appeared to correlate with any EGFR-related biomarkers.
“In other words, now we recommend that we give it to somebody who has a positive test for which there is no correlation with response,” he said. “It just doesn't make sense.”
Dr. Fojo said the scientific community has yet to deliver on the promise of personalized medicine.
The pharmaceutical companies lack an incentive to conduct or fund this research, so it's up to the scientific community to do that.
“If [bevacizumab] and cetuximab are an important part of our armamentarium seven years from now—especially in breast and lung cancer—we will have failed miserably in our quest to bring better therapies to patients,” he said. “More so than personalized medicine, what we really need are better and less expensive drugs.”
Bring Patients into Decision-Making Process
Another speaker at the AACR session, a palliative care specialist, tackled the cost of cancer care issue from a different perspective.
“Costs don't just happen, they happen because of what we [physicians] do and don't do,” said Thomas J. Smith, MD, Professor in the Division of Hematology Oncology and Palliative Care at Massey Cancer Center of Virginia Commonwealth University. “But it will not be easy to get doctors and patients to discuss costs without some societal shifts.”
Bringing patients into the decision-making process when chemotherapy is being considered could reduce the cost of treatment, but Dr. Smith said this isn't happening enough.
“Physicians find it difficult to discuss end-of-life issues with patients,” he said, citing studies from his institution showing that physicians discuss prognosis with patients only 39% of the time, and discuss impending death only 37% of the time.
And in a study of 75 sick patients in the inpatient service there, “the oncologist had brought up advanced directives only two times,” Dr. Smith said.
To reverse the skyrocketing trend in cancer care costs, Dr. Smith said oncologists must begin by acknowledging that “we drive the costs of care by what we do and don't do.”
His pointed to oncologists choosing chemotherapies that make the most profit for the practice, and giving too much chemotherapy at the end of life with little benefit.
On the other hand, he said, “without chemotherapy and supportive drug profits, oncology is a poorly paid, difficult specialty, so [changing this trend] will require real-time monitoring of physician performance and new payment models.”
Another example he gave of overspending is ordering tests such as CA125, CEA, and bone scans for which there is no evidence of efficacy.
“People can be convinced they don't need tests…but it requires saying [to the patient that] ‘these don't work,’” Dr. Smith said.
Other recommendations from Dr. Smith:
* Limit active therapy for patients with high performance status.
* Sequential monotherapy can be used for metastatic disease in many cancer types.
* Provide tools to “reset” patients' unrealistic expectations of benefit.
* Offer better and less costly end-of-life care.
* Accept independent “yes/no” guidance from cost-effectiveness analyses.
“Get individuals and practices out of the rationing decisions,” Dr. Smith said.