Cancer chemotherapy and obesity

In medicine, particularly oncology, it's often the little things that matter. Sometimes, however, the "little things" aren't actually little; they just seem that way. I was reminded of this by a story that was circulating a couple of weeks ago in the national media, often under titles like “Obese cancer patients often shorted on chemo doses”, ”Are obese people with cancer getting chemotherapy doses too small for them?”, and “Obese Cancer Patients Not Getting Full Doses of Chemotherapy Drugs”. It's also interesting to me because it stands in marked contrast to something I've written about a lot on this blog: The overtreatment of cancer. In this case, this story is about the undertreatment of cancer in patients who are obese, and it's a problem that has definite adverse effects on an obese person's odds of surviving cancer.

I've been aware of this issue for some time and had been thinking of blogging about it for at least three years. The reason is that the oncologist who is best known for sounding the alarm on this issue is Jennifer Griggs at the University of Michigan and I've seen her speak on the topic several times at local breast cancer conferences. I was actually rather puzzled why this issue bubbled up enough to be reported widely on the national news a while back when the Nature Clinical Oncology paper by Gary H. Lyman and Alex Sparreboom that drew attention to the issue was published in August, and the original American Society of Clinical Oncology (ASCO) guidelines were published last year. Whatever the reason this issue has been getting more attention, it's a good thing.

Although the issue is simple in concept, namely that obese patients are not always getting full weight-based doses of adjuvant chemotherapy after cancer therapy, it's not so simple to change attitudes. In chemotherapy dosing, most doses are calculated based on body weight or body surface area (BSA), which depends upon the weight. However, which "weight" to use in the BSA calculation has been a matter of controversy. When the actual body weight (ABW) is used, the BSA calculated can be huge, making it understandable why some oncologists might balk at the doses that they calculate, which can look staggering and provoke a reaction of "Holy crap, I can't give anyone that much chemotherapy!" (Of course, the word used is usually not "crap.") As a result, many oncologists "cap" chemotherapy doses, regardless of how large the patient is and how much chemotherapy the patient's weight mandates, or use various forms of "adjusted" body weights in the BSA calculation, such as the ideal body weight.

The issue is summed up nicely in an AP story by Marilynn Marchione:

Obese people are less likely to survive cancer, and one reason may be a surprising inequality: The overweight are undertreated.

Doctors often short them on chemotherapy by not basing the dose on size, as they should. They use ideal weight or cap the dose out of fear about how much treatment an obese patient can bear. Yet research shows that bigger people handle chemo better than smaller people do.

Even a little less chemo can mean worse odds of survival, and studies suggest that as many as 40 percent of obese cancer patients have been getting less than 85 percent of the right dose for their size.

Now, the largest organization of doctors who treat cancer, the American Society of Clinical Oncology, aims to change that. The group has adopted guidelines urging full, weight-based doses for the obese.

Being reluctant to give full doses in such patients is not entirely unreasonable on the surface. For instance, in breast cancer, one of the most commonly used drugs is Adriamycin (generic name: doxorubicin), and one of its most serious potential complications is cardiac toxicity that in extreme cases can lead to congestive heart failure. Also, given the association of diseases like type II diabetes with obesity, many oncologists will see multiple comorbidities in their obese patients and decide that the better part of valor is to dial back the doses a bit in order to avoid doing more harm than good. Recent evidence, however, is suggesting that these concerns are overblown. Worse, recent evidence is also suggesting that there is a very real potential for doing harm by not giving full, weight-based chemotherapy doses, hence the recent ASCO guidelines and the article by Dr. Lyman in Nature Clinical Oncology.

The derivation of these guidelines is actually interesting in that it shows how tricky and complex it is to derive science- and evidence-based guidelines for oncology. Because there has never been a randomized clinical trial comparing full, weight-based dosing of chemotherapy versus dose-capped regimens, what we are left with are observational and retrospective studies, which have a variety of shortcomings that can be hard to overcome, particularly confounding by other factors. However it is certainly plausible, from both a basic science and clinical perspective, that the common practice of underdosing the obese could potentially lead to poorer outcomes. Moreover, there is now good evidence that not only is obesity itself is a risk factor for various cancers (e.g., breast) but that it is a negative prognostic factor for patients who do develop cancer, meaning that such patients are more likely to die of their disease. Indeed, a recent meta-analysis concluded that obesity very likely contributes significantly to the risk of death due to breast, prostate, and colorectal cancer.

In terms of the association of obesity to various cancers, there is now robust evidence linking obesity to an increased risk of cancer. Indeed, in 2002 the International Agency for Research on Cancer (IARC), using obesity prevalence data from Europe and relative risks from a meta-analysis of published studies, concluded that obesity was a cause of 11% of colon cancer cases, 9% of postmenopausal breast cancer cases, 39% of endometrial cancer cases, 25% of kidney cancer cases, and 37% of esophageal cancer cases. Meanwhile, data from the American Cancer Society suggested that obesity was related to mortality from liver cancer, pancreatic cancer, non-Hodgkin's lymphoma, and myeloma. This effect on mortality reflects both the excess incidence and excess mortality among those with cancer. Overall, it has been estimated that obesity causes approximately 20% of all cancer cases. Over 32 years ago, Sir Richard Doll estimated that "overnutrition" (i.e., being overweight or obese) caused a combined 35% of all cancer cases. For the cancer to which I've devoted my professional life, breast cancer, as pointed out by Lyman and Sparreboom it's been known for over 40 years that in animal models a reduction in chemotherapy dose by as little as 20% can reduce complete remission and cure rates by up to 50%. It has also been known for at least 20 years that breast cancer recurrence and survival are negatively impacted by obesity. The dose-response curve for chemotherapy can be steep in many tumors.

Why it is the case that obesity is both a risk factor for the development of cancer, particularly more aggressive forms of the cancers for which it is a risk factor, as well as a risk factor for mortality from cancer is not fully understood yet. Various molecular mechanisms have been proposed, including insulin resistance common in the very obese and in people with type II diabetes, altered levels of insulin-like growth factors, oxidative stresses, adipokines (growth factors secreted by adipose—fat—tissue), and obesity-related inflammatory markers. Whatever the biological mechanisms behind the increased risk related to obesity for developing cancer and for any cancers that develop to be more aggressive subtypes, it is certainly plausible from a basic and clinical scientific standpoint that undertreating such cancers could potentially lead to higher rates of recurrence.

In obese patients, this appears to happen a lot. Back in 2005, Dr. Griggs published a paper in JAMA Internal Medicine looking at how often such dose reductions occurred. She found that dose reductions on the first dose (usually called "cycle") of chemotherapy, which would eliminate dose reductions due to toxicity, were instituted in 9% of the healthy weight, 11% of the overweight, 20% of the obese, and 37% of the severely obese women. She also found that first cycle dose reductions were independently associated with being overweight (P = .03), obese (P< .001), severely obese (P<.001), older than 60 years (P<.001), and having a serious comorbid condition (P = .03). Interestingly, she also found that severe obesity was actually associated with a lower likelihood of being admitted for febrile neutropenia (fever with a low white blood cell count, indicating potential infection in an immunosuppressed patient), was actually less common (odds ratio, 0.61; 95% confidence interval, 0.38-0.97) in the severely obese even in patients who received full dose, and there was a trend towards less admissions for febrile neutropenia associated with increasing body mass index that only became statistically significant at a BMI ≥ 35.

More importantly, it's been known since the 1990s that decreased chemotherapy dose delivered is associated with significant increases in recurrence and decreases in survival. For instance, in 1996, in a retrospective analysis of a randomized trial of women with lymph-node-positive early stage breast cancer, Rosner et al. found that obese women receiving adjuvant chemotherapy experienced no greater toxicity or poorer outcomes than similarly treated healthy weight patients, and overweight women receiving full dose chemotherapy experienced better failure-free survival than those receiving reduced doses. More recent studies, such as this meta-analysis by Hourdequin et al., show that obese patients receiving chemotherapy based on the use of their actual body weight experienced the same or fewer chemotherapy-related complications as compared to the non-obese when they received the full ABW-based dose of chemotherapy, and survival outcomes did not differ between the obese and non-obese who received their full dosage of chemotherapy. Couple that with studies linking decreased dose to poorer outcomes, and the likelihood that obese patients not receiving their full doses are at significantly increased risk of dying of their cancers. As ASCO put it in their guidelines:

Recommendation 1.1: The Panel recommends that actual body weight be used when selecting cytotoxic chemotherapy doses regardless of obesity status. There is no evidence that short- or long-term toxicity is increased among obese patients receiving full weight–based chemotherapy doses. Most data indicate that myelosuppression is the same or less pronounced among the obese than the non-obese administered full weight–based doses.

However:

Recommendation 1.2: The Panel recommends full weight–based chemotherapy dosing for morbidly obese patients with cancer, subject to appropriate consideration of other comorbid conditions. Data are extremely limited regarding optimal dose selection among the morbidly obese and other special subgroups. More studies are needed to evaluate optimal agents and agent combinations for obese and morbidly obese patients with cancer; however, based on available information, it seems likely that the same principles regarding dose selection for obese patients apply to the morbidly obese.

And finally:

The Panel recommends that full weight–based chemotherapy doses (IV and oral) be used in the treatment of the obese patient with cancer, particularly when the goal of treatment is cure. Selecting reduced doses in this setting may result in poorer disease-free and overall survival rates. There are compelling data in patients with breast cancer that reduced dose-intensity chemotherapy is associated with increased disease recurrence and mortality. Although data in other malignancies are more limited, based on improved survival observed with chemotherapy compared with controls, a dose-response relationship exists for many responsive malignancies. Therefore, although data are not available to address this question for all cancer types, in the absence of data demonstrating sustained efficacy for reduced dose chemotherapy, the Panel believes that the prudent approach is to provide full weight–based chemotherapy dosing to obese patients with cancer, especially those receiving treatment with curative intent. Most of the data in support of full weight–based dosing come from the treatment of early-stage disease. Data supporting the use of full weight– based doses in the advanced disease setting are limited.

At this point, it is worth reiterating that medicine is rarely black and white (one notable exception is the case of something like homeopathy, which violates multiple laws of physics and chemistry and is, without a doubt, total quackery). It exists in shades of gray, which is where these recommendations come from. The panel examined evidence from multiple sources, including preclinical evidence from cell culture and animal models that dose reduction of chemotherapy can compromise "cure" rates, coupled it with evidence from humans that obese patients are more prone to cancer and more likely to die of cancer if they get it, noted that obese patients frequently don't receive the full recommended dose of chemotherapy and that decreased dose of chemotherapy is associated with increased risk of cancer recurrence and death. From these data, they made the connection that quite likely at least part of the increased mortality due to cancer in the obese is because many of them don't receive their full dose of chemotherapy based on a BSA calculated from their ABW. They also noted that obese patients are no more likely—and, in fact, may be less likely—to suffer chemotherapy-related complications than the non-obese when they receive full chemotherapy doses. Putting it all together, the panel was quite reasonable to recommend that chemotherapy dosing based on BSA be calculated using ABW, not the various other "capped" dosing methods or methods using adjusted body weights.

At the same time, weaknesses and uncertainties were acknowledged. For instance, as has been pointed out before, there has never been a randomized controlled clinical study comparing full, ABW-based dosing to capped doses and/or doses based on various adjusted weights. Given the confluence of evidence above, such a trial would likely fail the ethical test of clinical equipoise and therefore never be approved by an institutional review board. As I've said many times before, the reason we call it "science-based" medicine rather than science is because ethical and practical considerations limit how rigorous we can be. Moreover, we don't know enough about the pharmacokinetics of chemotherapy drugs in the morbidly obese to determine whether a different strategy for calculating their doses would result in decreased recurrences and decreased toxicity.

My good bud Mark Crislip likes to quote a certain not-so-popular former defense secretary even though he doesn't like this not-so-popular former defense secretary, "As you know, you go to war with the army you have, not the army you might want or wish to have at a later time." It was one of the rare bits of actual near-wisdom that ever came out of the mouth of Donald Rumsfeld. Patients need to be treated now and clinicians can only use the tools they have at the time. Right now weight-based dosing is the tool that we have. Maybe someday there will be "personalized" (or "precision") dosing based on expression levels of various proteins that metabolize the drug, weight, and who-knows-what other personal factors. That doesn't help patients today; so we go with what we have, especially since the evidence that we have strongly supports not reducing chemotherapy dose in the obese just because they are obese. As Gary Lyman put it, while more research is clearly needed in this important area, it is essential that the science that we do have is applied appropriately and effectively in the optimal treatment of patients with cancer.

Finally, it's a mistake to refer to this problem as "discrimination" against the obese, as Melissa Ford at BlogHer did:

We've seen a lot of discrimination towards obese individuals in various areas of health care not to mention a wide-range of businesses from airlines to restaurants. Yet in this case, we're talking about a person's survival. For 40% of obese cancer patients, before they've taken their first dosage of chemotherapy, the odds are stacked against them -- not by science or fate, but by the doctor treating them.

While there is certainly discrimination against obese people because they are obese, dose reduction of chemotherapy in the obese is not an example of it. Rather, this practice derives from a concern that did not seem unreasonable in the past but that we now know to be almost certainly misplaced; that dosing obese patients the same way we dose non-obese patients will harm them. As Dr. Lyman says:

Don't call it supersizing; it's right-sizing cancer care, said Dr. Gary Lyman, a Duke University oncologist who led the panel that wrote the advice.

"There's little doubt that some degree of undertreatment is contributing to the higher mortality and recurrence rates in obese patients," he said.

Finally, there's another lesson to be had here. Based on these data, there is a growing consensus to link the receiving of adequate chemotherapy doses to quality metrics in oncology care. Most people, including physicians, think that it's all about saving money when they see evidence-based guidelines. Sometimes it is, particularly with initiatives like Choosing Wisely, which is designed to discourage the use of unnecessary tests and procedures. In this case, though, it isn't. Treating with the full dose of chemotherapy is likely to cost more in the short term. Whether the amount by which it will decrease recurrences will be such that there will be a net savings because of fewer relapsed cancer patients is currently unknown. It might not be. In this case, higher quality care is likely to cost more. Such is likely to be the cost of "right-sizing cancer care."