Diet and Cancer: Here's What I Tell Patients

One of the most common questions my patients ask is, “What diet can help me beat this cancer?” It is a profoundly important question that is worthy of our efforts to answer. In this brief essay, I will take a deep dive into this question in depth and explore the broader clinical and scientific themes it brings into play.

Low-Hanging Fruit: Nutrition Science

A cancer diagnosis can be a deeply disempowering experience. Although I have not lived with cancer myself, I have seen this play out repeatedly over the past 5 years in my role as an oncologist treating patients with hematologic malignancies.

Our diet is an important part of our personal identity, culturally and spiritually. If lifestyle changes, such as a modified diet or more exercise, can contribute to cancer treatment, it may help us regain a sense of control over our lives, one that cancer so often cruelly strips away. I hypothesize that, among other factors, this is why diet is so important to our patients.

Another factor is exposure to a compelling diet-cancer narrative. Nearly every day, a media headline appears claiming that eating a particular food, or drinking coffee, can either increase or decrease your risk for a certain disease.

These claims, however, are often based on studies of large observational datasets where individuals fill out surveys about their dietary habits and are subsequently assessed for disease outcomes. In these studies, people aren’t asked to eat a particular diet; instead, their dietary habits are analyzed by researchers who have endless permutations to explore. This, in a nutshell, is the field of nutritional epidemiology.

In my opinion, nutritional epidemiology represents the collision of the well-intentioned effort to answer clinically meaningful questions with the ease — and near-infinite permutations — of dietary questions that can be asked from an increasingly larger number of different datasets.

Now, factor in the never-ending appetite (pun intended) of journalism and the public’s desire for dietary studies, and you create the perfect storm of incentives that drives a flood of low-quality nutritional science. These studies are highly malleable to analytical choices and can essentially produce results consistent with your prior beliefs, regardless of the philosophical inclination you have (pro keto-diet, pro-vegan, etc.). I love quoting this study to my trainees that, depending on what variables are included and how the analysis is conducted, the same dataset could be used to show that red meat either increases, decreases, or has no effect on all-cause mortality. Unfortunately, much of the evidence base for diet in cancer comes from similarly confounded, low-quality studies.

Diet and Cancer

So, what do randomized trials show for diet and cancer?

The highest-quality evidence is generated from randomized controlled trials. One of their key advantages is the ability to control both measured and unmeasured confounders.

Unfortunately, the evidence supporting diet as an anticancer modality in randomized trials in patients with cancer is bleak. We did a systematic review of all randomized trials of dietary intervention ever done in patients with cancer. Most of the trials measured outcomes such as feasibility (often small pilot studies that measure variables such as weight changes or lab values). The trials that measure clinical endpoints, such as survival, were largely negative and demonstrated no meaningful effect of diet on outcomes. Take trials exploring whether a Mediterranean diet helps prevent breast cancer recurrence, or whether a diet rich in fruits and vegetables improves prostate cancer outcomes. Although these diets may offer benefits, these studies found that specific diets did not change the natural history of cancer.

Myeloma and Diet

In my specialty, multiple myeloma, I am thankful that some trials are beginning to shed light on whether diet influences cancer outcomes.

One study, which was recently published in Cancer Discovery, explored whether a high-fiber, plant-based diet could potentially slow or delay progression from myeloma precursor conditions toward full-blown multiple myeloma. The trial enrolled 23 participants, with the primary endpoints of dietary adherence and changes to BMI. Measures of progression to multiple myeloma were exploratory at best. Yet, the media coverage, as well as the majority of the discussion and results sections of this study manuscript, claimed that the diet changes can prevent progression to myeloma.

However, the study design and conclusions were flawed. The paper focused on two patients who had some improvement in disease trajectory, while descriptions of patients who had an increase in their bone marrow plasma cell percentage were relegated to the supplemental section.

As a primary investigator of a trial in smoldering myeloma where we use advanced imaging as an alternative to pharmacologic treatment, I frequently see myeloma markers fluctuate and often decrease. I attribute these changes to random variation, or possibly regression to the mean, rather than the effect of any intervention.

Future randomized studies by this group used primary endpoints of stool butyrate level and implement dietary interventions for a limited period— 2 weeks in one study and 12 weeks in another — to again assess the impact of a high-fiber, plant-based diet on progression to myeloma. Although there are no data yet, the limited timeframes in these studies severely limits generalizability for outcomes that would truly matter, such as cancer control and longevity. There is also no evidence that changes in stool butyrate levels influence patient outcomes.

High-quality science — whether it is evaluating diet or other interventions—requires high-quality data, effort, funding, and time. It is not impossible.

We can draw inspiration from the CHALLENGE trial. This large, randomized trial, which took over a decade to complete, assessed the benefit of a structured exercise program in the adjuvant setting for colon cancer. The endpoint of this study was disease-free survival, and the intervention was deployed over a much longer period: 3 years, as opposed to a 2-week intervention. This trial took years from inception to completion, but it yielded a conclusive result and will probably lead to more dedicated efforts to facilitate exercise programs for patients with cancer.

Our patients deserve the same effort as the CHALLENGE trial to answer their important dietary questions. Until such trials are completed, we must acknowledge, with humility, that despite the common sense and feel-good factor that many diets offer us, their impact on cancer remains uncertain.

Conversely, we must recognize that even if diet does not cure or alter the course of a certain cancer, it can still impact quality of life, treatment tolerance, and other supportive care outcomes, making it an important factor in patient care.

This is what I tell my patients that it is unlikely any one diet will change the trajectory of your cancer. Focus on eating healthy, and remember that most things in moderation are fine. Your diet remains an important risk factor and determinant for health outcomes beyond cancer. Eat what makes you happy. You are going through a tough time, and this is not the moment to impose stringent restrictions on yourself.

A version of this article first appeared on Medscape.com.