One of the most frustrating aspects of taking care of cancer patients is that in general, with a handful of specific exceptions, we do not have good curative therapies for patients with stage IV cancer, particularly solid tumors. Consequently, we are forced to view patients with stage IV cancer as “incurable” because, the vast majority of the time, they are incurable. Over the years, we have thrown everything but the kitchen sink at patients with stage IV disease, largely with dissapointing results. That’s not to say that the few specific exceptions to which I alluded are not a reason for hope. After all, patients with colorectal cancer and liver metastases used to have a median survival of around 6 months, but these days, with newer chemotherapeutic regimens like FOLFOX plus Avastin, median survival has more than tripled. While expecting to live less than two years is cold comfort to cancer patients with this particular clinical situation, the prognosis is far better than it was.
Of course, I specifically mentioned Avastin because it’s been in the news a lot recently with respect to my area of clinical specialty, breast cancer. Specifically, beginning in July there started appearing a spate of stories about the FDA considering revoking the approval of Avastin for advanced breast cancer based on recent studies that demonstrate that it does not prolong survival in these patients. Many lay people and patients find this reconsideration of Avastin to be quite puzzling, given that the drug was granted accelerated approval in 2008 and has since gone on to be used fairly widely. Given that the case of Avastin in breast cancer is rapidly becoming a classic case study of how messy science-based medicine can be when practiced in the public eye and debated among pharmaceutical companies, the government, and patient advocacy groups.
The promise of angiogenesis inhibitors
Avastin (generic name: bevacizumab) is a class of drug that attacks cancer by attacking its blood supply. Basically, Avastin targets a protein called vascular endothelial growth factor (VEGF), which is growth factor that stimulates the cells lining the blood vessels to undergo a process of angiogenesis, sprouting out new blood vessels in the direction of the cells secreting VEGF. Angiogenesis is a normal physiological function used during embryogenesis and pregnancy, wound healing, during the proliferative phase of the female menstrual cycle, and a variety of other processes. Diffusion of oxygen and nutrients through an aqueous medium is limited to around 1 mm, and malignant tumors get around this limitation by hijacking the normal process of angiogenesis by secreting large amounts of pro-angiogenic factors like VEGF in order to provide themselves with the oxygen and nutrients that they need to grow. During the 1990s, largely inspired by the work of one of my scientific heroes, Judah Folkman, scientists discovered strategies to target the process of angiogenesis to treat cancer. After spectacular results in mouse models of various cancers, however, as is all too often the case the use of angiogenesis inhibitors in humans produced far less dramatic results and, in some cases, didn’t work at all. However, for some cancers, such as colorectal cancer, combining Avastin with conventional chemotherapy has been effective. To those of us “in the biz” since the mid-1990s, this is not a revelation or surprise. In fact, I was involved with some of the first research demonstrating that combining anti-angiogenic therapy with radiation resulted in synergistic anti-tumor effects.
Over the last decade, Avastin has tended to be a magnet for controversy. The major reason that Avastin has attracted a lot of attention is that it’s so expensive (up to $100,000 a year per patient), and part of the reason it’s so expensive is that it is a protein rather than a small molecule. Basically, Avastin is an antibody directed against human VEGF. More specifically, it’s a humanized monoclonal antibody against VEGF. What that means is that the antibody has been modified to get rid of most of the mouse sequences that can provoke an immune reaction against the protein in humans. In any case, developing and producing such a molecule, which is a large protein, are not cheap. Even so, there has been considerable criticism of Genentech, the manufacturer of Avastin, for allegedly overpricing it. In addition, Genentech has pulled some highly dubious stunts to protect its profits with regard to some off-label uses of Avastin, for example, for macular degeneration.
Be that as it may, the number of uses of Avastin have expanded to a number of cancers, now including colorectal, lung, breast, and renal cell cancers. Perhaps the most controversial expansion occurred in 2008, when the FDA approved Avastin on what is known as the “fast track” for use in metastatic breast cancer. This allows certain promising drugs for life-threatening conditions to be granted in essence provisional approval quickly, so that patients can have access to them before the usual large phase III randomized, double-blind trials are completed. To understand why this decision was controversial and why it is now being reconsidered, a brief review of terminology is required.
How to evaluate an anticancer therapy
Cancer therapies are generally evaluated using a number of endpoints. The most commonly used include overall survival (OS) and progression-free survival (PFS). There are, of course, other endpoints related to survival that are measured, but these two are the most relevant for this discussion. Overall survival is what it sounds like: How long do patients survive their cancer after diagnosis? Period. It’s hard for an endpoint to be more objective than that: Either the patient is alive or he is dead. This number is usually expressed in terms of median survival, which is the period of time after which half of the patients under study are still alive and half have died. This includes all causes, not just cancer. If a patient with cancer under study dies of a heart attack that is not related to his cancer or his cancer treatment, that counts. Traditionally, OS has been the “gold standard” endpoint in measuring the efficacy of a cancer therapy, because the primary goal has been to prolong survival, the ideal case being prolonging survival to the point where it is indistinguishable from life expectancy if the patient never had cancer in the first place. PFS is survival without progression; i.e., how long the patient with cancer survives before his or her tumor starts measurably growing again or metastasizes. While PFS is often measured as well as OS, it’s generally considered less useful because it is entirely possible for a treatment to prolong PFS without prolonging OS. This sort of result can happen when the treatment is effective at shrinking a tumor or slowing its growth but its toxicities can result in death. Thus, PFS improves with no improvement in OS.
Which brings us to Avastin.
In 2008, that’s exactly the sort of data upon which the FDA’s fast track approval of Avastin was based. Back in 2005, the National Cancer Institute (NCI) announced that in a randomized trial Avastin combined with paclitaxel had increased PFS approximately 4 months over the use of paclitaxel alone. Eventually, by the time an FDA panel voted to approve Avastin in 2007, the E2100 study published in the New England Journal of Medicine had reported encouraging results that Avastin improved PFS in women with metastastic breast cancer. Unfortunately, E2100 also failed to find any benefit in terms of OS. Specifically, Paclitaxel plus Avastin significantly prolonged progression-free survival compared to paclitaxel alone (median, 11.8 vs. 5.9 months; hazard ratio for progression, 0.60; P<0.001). Adding Avastin to paclitaxel also increased the objective response rate (36.9% vs. 21.2%, P<0.001). Unfortunately, however, the overall survival rate (median, 26.7 vs. 25.2 months; hazard ratio, 0.88; P=0.16). Complications were also more frequent in the group receiving Avastin. Grade 3 or 4 hypertension (14.8% vs. 0.0%, P<0.001), proteinuria (3.6% vs. 0.0%, P<0.001), headache (2.2% vs. 0.0%, P=0.008), and cerebrovascular ischemia (1.9% vs. 0.0%, P=0.02), all complications common from antiangiogenic therapies, were more frequent in patients receiving paclitaxel plus Avastin. Infection was more common in patients receiving paclitaxel plus Avastin (9.3% vs. 2.9%, P<0.001).
These data were thin gruel indeed to approve Avastin for the indication of metastatic breast cancer, which the FDA did in February 2008. Although the increase in PFS was impressive, the lack of effect on OS was very disturbing, strongly suggesting that drug toxicity was "taking back" all the benefit the addition of the drug was providing against the cancer. Also, there were serious concerns about the E2100 trial. For example, there was 34% loss to follow up, 10% missing data, and discordance between radiological and clinical progression findings. The discordance between disease progression findings was of particular concern, given that it was the primary endpoint upon which Genentech's claim of efficacy was based.
Nonetheless, despite these concerns and despite the recommendation of the FDA's Oncologic Drugs Advisory Committee (ODAC) that Avastin not be approved, the FDA gave provisional approval to Genentech to market Avastin for metastatic breast cancer using the "fast track" program designed in the 1990s to speed the approval of promising drugs for life-threatening conditions, with the reasonable condition that Genentech had to perform additional studies to verify the results upon which the original approval had been based. At the time, there was a great deal of argument over whether this was a wise decision, particularly given how expensive Avastin is and the lack of evidence that it improved quality of life or overall survival. Indeed, I recall attending a talk at ASCO in June 2008 in which the approval of Avastin was strongly defended and the speaker (who it was now escapes me) attacked the use of OS as the basis upon which to approve drugs and argued that PFS should be enough, given the potential improvements in quality of life that can result from a longer period of time without tumor progression.
In the press at the time, Dr. Kathy Albain at Loyola University defended the approval of Avastin thusly:
Dr. Kathy Albain, a breast cancer specialist at Loyola University Medical Center in Maywood, Ill., polled colleagues and patients and found overwhelming support for approving drugs based on delaying tumor progression. It would be ideal to show that a drug also prolongs life, but that may not be realistic, she said. The reason is that when a woman’s cancer progresses, doctors change the drugs they use, hoping to slow the cancer. That dilutes any impact of the first drug — in this case Avastin.
Personally, although I can see where Dr. Albain is coming from, I’ve always been uneasy with this view. Yes, I can see the utility of a drug that slows tumor progression but does not prolong survival. However, I can see its usefulness in only one situation: There must also be good evidence that that drug also improves quality of life, evidence that was lacking for Avastin. In fact, the evidence that existed was suggestive of increased complications due to Avastin. Now, it may be a bit of a judgment call whether it is worth it to slow down tumor progression if there is no improvement in survival and there are drug-related complications, but I remained unenthusiastic about a drug that didn’t prolong OS but did appear to produce markedly more complications. How is that improving the quality of life of a cancer patient? I can also understand the argument, which was made in 2008, that perhaps Avastin benefits certain subsets of patients; we just don’t know who they are yet. My response to that would be that we should be doing the studies to identify those patients rather than exposing large numbers of women to the risk of the drug outside the auspices of clinical trials that could identify them.
Avastin does not live up to even its modest promise
In any case, the followup studies, known as the AVADO and RIBBON-1 trials were indeed performed, and the results published. The AVADO trial was a phase III randomized double-blind clinical trial (N=736 subjects) comparing docetaxel with docetaxel plus Avastin in women with stage IV HER2-negative breast cancer. Its results were less than impressive. While it did demonstrate a decrease in PFS due to the addition of Avastin, PFS did not improve anywhere near the five months it did in the E2100 trial; rather it improved from 8.2 months to 9.0 months, a mere 0.8 month or 24 days. As in E2100, there was no detectable effect on OS. There was, however, a troubling trend towards a shorter OS in women who received Avastin. Even though the difference was not statistically significant, women who did not receive Avastin survived a median of 31.9 months; those who received the lower dose survived a median of 30.8 months; and those who received the higher dose level of Avastin survived a median of 30.2 months. In contrast to E2100, in the AVADO study Avastin did not increase the toxicity of treatment measurably. In the discussion, the authors of the AVADO study concluded:
Taken together with the results of the E2100 study, these data suggest that the combination of bevacizumab with taxane chemotherapy should be considered as an option for the first-line treatment of HER2-negative MBC.
I’m nots sure why one would recommend this based on such results. There is a difference between statistically and clinically significant improvements in DFS. Just because a small measurement can be measured does not mean that a drug has meaningful effects.
In any case, the initial results of the RIBBON-1 trial were reported at the American Society of Clinical Oncology (ASCO) meeting last year and were similarly disappointing. In the trial, Avastin was tested in a randomized, double-blind study enrolling 1,237 patients. Oncologists could choose from capecitabine, taxane, or anthracycline-based chemotherapy, and patients would be randomized to that plus or minus Avastin. The results were that Avastin improved PFS by 2.9 months when added to capecitabine, and 1.2 months when added to anthracycline-based chemotherapy. Again, there was no improvement in OS in women receiving Avastin. In fact, it was these studies that led the FDA advisory panel to vote 12-1 to revoke FDA approval for Avastin for use in breast cancer:
An FDA advisory committee voted 12 to 1 on July 20 to withdraw Avastin’s authorization for advanced breast cancer based on two new studies that the advisers concluded had not shown that the drug extends life. Not only that, the committee concluded that the studies indicated the drug slowed tumor growth for even less time — perhaps as little as about a month. “The vast majority opinion of the committee was that the drug was not doing very much, and what it was doing was more than offset by the negative,” said Wyndham Wilson of the National Cancer Institute, who chaired the committee. Avastin can cause a variety of potentially serious side effects, including blood clots, bleeding and heart failure. “In our best judgment, we did not feel this drug was safe to give relative to its benefits,” Wilson said.
The recommendation has been praised by many cancer experts and by advocates for breast cancer patients.
“The FDA should never have approved Avastin for breast cancer to begin with,” said Fran Visco of the National Breast Cancer Coalition. “We don’t see evidence of benefit, but we do see evidence of harm.”
I tend to agree, although I must admit that back in 2008 I was ambivalent about the FDA approval of Avastin rather than necessarily opposed to it. Now, I’m more of the opinion that the FDA would be justified, based on science alone, in revoking Avastin’s approval, although I can also understand, if not agree, with a lot of the multitude of opinions in the breast cancer oncology world regarding the FDA’s decision. However, there is more than science at work, which is why where the rubber hits the road practicing science-based medicine in intersection with public policy can be so messy and contentious. Just practicing SBM alone is contentious enough, but add a biotech company trying to protect its profits and expand its market, regulatory agencies trying to fulfil their mandate in the case of ambiguous data, and advocacy organizations trying, well, to advocate for their members, and the whole debate can turn into a huge kerfuffle in a hurry. In this case, the FDA is scheduled to make a final determination in September. In the meantime, patients receiving Avastin are afraid, because if the FDA revokes its approval using Avastin for breast cancer will become and off-label use, and insurance companies will no longer pay for it.
Wingnut politics and burning stupid intrude
The whole controversy, where science, economics, politics, and dying cancer patients intersect in what are sometimes very ugly ways, is taking an even uglier turn, thanks to the entire healt insurance reform debate and the intrusion of partisan politics and the unrelenting opposition of Republicans to what they often derisively term “Obamacare.” The reason is that right wing bloggers have jumped on the bandwagon of criticizing the FDA’s decision on Avastin as “evidence” of the impending arrival of socialized medicine-style rationing (“Oh, noes, teh socializm!”), thanks, of course, to “Obamacare.” For example, Senator Dave Vitter (R-LA) characterized the decision thusly:
Reviving allegations of government death panels, Sen. David Vitter of Louisiana said Wednesday that an FDA advisory panel’s negative recommendation on a contested breast cancer drug amounts to rationing health care. “I shudder at the thought of a government panel assigning a value to a day of a person’s life,” Vitter said in a press release about the drug Avastin. “It is sickening to think that care would be withheld from a patient simply because their life is not deemed valuable enough.”
Several right wing and rightward-leaning blogs, such as Red State, Hot Air, and Big Government characterized the decision as the FDA being made complicit in “rationing” health care. Perhaps the most despicable post I’ve seen about AVastin is this one by Moe Lane at Red State, entitled Obamacare worth 17.5K dead women a year? Burning stupid does not even begin to describe the drippy turd consisting of utter demagoguery of the most vile and despicable sort. Avastin, even under the most optimistic scenario imaginable, only prolongs disease-free survival; it does not save lives of stage IV breast cancer patients, at least not as far as we can tell. It’s possible that it may save lives for other indications in breast cancer, such as inflammatory breast cancer that is still curable, but such results will await clinical trials. Also at Red State, Ed Morrissey resorted to exaggerating the benefits and downplaying the results of clinical trials that failed to find much, if any, benefit and did find potential evidence of harm, even going so far as to accuse the FDA thusly:
With the new ObamaCare regime in place, the issue of cost has now become openly part of the FDA process. This is a perversion of their mission, which is supposed to only involve product safety and effectiveness, not bean-counting. If Medicare doesn’t want to cover Avastin, that should be a separate issue handled by CMS and HHS. This strongly suggests that the FDA has become politicized to a degree where their recommendations lose credibility — a dangerous situation for consumers and providers alike.
The Wall Street Journal chimed in with a fallacy- and demagoguery-filled editorial whose deconstruction would be worthy of a separate post of its own if it weren’t already over two weeks old, entitled The Avastin Mugging, in which the anonymous editorialist harped on the earlier E2100 trial and cherry picked the very best numbers out of the AVADO and RIBBON-1 trial, completely ignoring the fact that OS was no different with Avastin and that it might even be worse. He or she made liberal use of straw men arguments and even misquoted findings, at one point even claiming, “At any rate, even the 31% reduction in the risk of disease progression or death is better than the status quo.” The problem is that none of the studies showed a 31% decrease in the risk of death–or any decrease in the risk of death for that matter. The editorialist also referred to to chemotherapy as “savagery” and asserted:
The Avastin mugging is really an attempt to undermine regulatory modernization like accelerated approval that offends the FDA’s institutional culture of control and delay. It is also meant to discourage innovations like Avastin that the political and medical left has decided are too costly, with damaging implications for the next generation of cancer drugs.
No, the reconsideration of decisions like the decision to approve Avastin is an absolutely necessary part of the fast track approval of drugs. If we are going to approve drugs on the basis of relatively scarce preliminary evidence with the quite reasonable requirement that more studies be done and the issue of FDA approval be revisited after those studies are completed, we have to expect that we will find in at least some cases where follow up studies will suggest that FDA approval wasn’t justified, that the drug isn’t particularly useful, and that FDA approval should be revoked. If that doesn’t happen from time to time, then why bother requiring additional studies? Just use fast track approval and forget it, which is, of course, what the pharmaceutical companies would very much like to see.
Although it may not be unreasonable to be concerned that the law may evolve to consider cost more strongly, as the law exists now and as has been reported in the Wall Street Journal and the AP, FDA advisory panels do not consider the costs of the drugs they are evaluating, only the evidence. Whether they should or not is certainly something we as a nation will have to decide. I don’t consider it at all unreasonable to ask whether it’s worth over $8,000 a month to produce a 1.2 month increase in DFS, no increase in OS, and an increased risk of complications such as bleeding, stroke, hypertension, and bowel perforation.
FDA approval, science-based medicine, and politics
So how will this all play out on September 17, which is when the FDA is expected to make its final determination. The FDA could decide to side with the panel recommendation to revoke approval for Avastin, ignore the recommendation, or chart a middle course, continuing approval and asking for still more studies. It wouldn’t surprise me if the FDA chose that last option. It is also important to remember that Avastin may well have uses in breast cancer other than in stage IV disease. One area where it shows promise is in the neoadjuvant therapy (therapy before surgery) of inflammatory breast cancer.
Whatever the FDA decides in this particular case, determining which drugs to approve for metastatic breast cancer (or nearly all metastatic solid malignancies) is a particular challenge to SBM because SBM cannot cure the disease. Consequently, the goals in treatment must be palliative; i.e., to relieve symptoms and prolong survival as much as possible. Sometimes these goals are in conflict. In the case of Avastin, this may well be the case. The drug improves PFS very modestly, but there must be a cost in terms of complications that decrease quality of life. The Avastin saga is a saga the likes of which we are probably going to see with increasing regularity.
Another lesson of the ongoing Avastin saga is about the very nature of science-based medicine itself. We have stated that we believe that medical care should be science-based. However, although medicine should be based on science, medicine itself can never be a pure science because so many non-science-based considerations impact on it. On the patient level, there is patient choice and how the doctor and patient weigh the patient’s personal situation and personal considerations in choosing from among science-based therapies. At the national level, considerations of cost, politics, and values cannot be separated from medical policy considerations. Science can tell us that Avastin does not prolong overall survival in breast cancer patients and that it only very modestly prolongs progression-free survival. It can tell us that even that modest increase in PFS comes at a cost of complications that prevent improved PFS from translating to improved OS. What science can’t tell us is whether that modest benefit is worth the cost. That’s a value judgment that must be made both at the level of society as a whole and at the level of each patient and physician.
With more research, I can only hope that science will soon be able to tell us which women with breast cancer are most likely to benefit the most from Avastin.