Overweight, Obesity, and Pancreatic Cancer: Title and subTitle BreakBeyond Risk Alone

Overweight and obesity have well-established public health implications by increasing risk for myriad diseases, most notably cardiovascular disease and diabetes, and also play a role in increasing the risk for multiple types of cancer,¹ including pancreatic cancer.² Obesity has also been reported to be associated with poorer prognosis in multiple cancers,^{3 - 4} perhaps most notably breast cancer.^{5 - 8}

In this issue of JAMA, Li et al⁹ have confirmed a well-established association between pancreatic cancer risk and obesity, emphasizing the potential role of early adulthood obesity. Their report includes 2 further observations. First, overweight and obese patients are diagnosed with pancreatic cancer at a younger age than patients with normal weight, and second, overweight and obese patients have lower rates and duration of survival once pancreatic cancer is diagnosed.

The authors used information obtained from a large case-control study based in an academic cancer center, conducted from 2004 to 2008. They conducted in-person patient interviews, with body mass index (BMI) computed from self-reported recall of weight in decade intervals throughout life. The study included 841 patients with pancreatic cancer and 754 healthy control participants, who were recruited from convenience and were frequency matched to cases by age, sex, and race. The association between increased BMI and age at diagnosis of pancreatic cancer was adjusted for clinical variables that are potentially related to diagnosis. Similarly, the survival analyses included clinical stage, with the adverse prognostic association with overweight and obesity consistently seen among patients with resectable, locally advanced, and metastatic cancer.

In each of these analyses, there was a dose-dependent association between increasing BMI and each of these end points. Indeed, the results of the survival analyses are similar to a study from the Mayo Clinic using a comparable clinic-based case-control study,¹⁰ suggesting these findings are indeed reproducible.

Observational studies have limitations and several potential explanations may contribute to these findings. The increasing level of obesity in the general population, especially at a younger age, could affect the findings of younger age at diagnosis because all of the patients were enrolled within a short period (ie, younger patients at diagnosis have higher weights at a particular age because their entire population birth cohort does). Additionally, the association of overweight and obesity with adverse survival for patients with pancreatic cancer may simply affect comorbidities, and hence result in increased adverse effects of treatment, as well as decreased ability of the patient to bear the burden of disease.

These analyses have important implications both for public health and for increasing the biological understanding of pancreatic cancer. Due to the increased risk of many illnesses and morbidities associated with increased weight, it may be impossible to expect widespread behavior changes by the public based simply on future risk of developing pancreatic cancer. However, the additional evidence of knowing that obesity could contribute to earlier onset of disease and worsened survival adds further to the call for interventions at the public health level to stem the increasing rates of obesity at all ages in the US population.¹¹ Perhaps even more important in this particular disease process, the association that increased weight may accelerate outcomes throughout the disease process from risk of developing cancer to its ultimate outcome may provide biological insight into why pancreatic cancer portends such a poor prognosis.

As Li et al correctly state, hyperinsulinemia and insulin resistance can be causes or consequences of pancreatic cancer,¹² so it has been difficult to adequately study these factors in this disease setting. Nonetheless, the authors' hypothesis that perhaps hyperinsulinemia and activation of the insulin–like growth-factor pathway contributes to tumorigenesis and, subsequently, more rapid progression of cancer has some plausibility. Other mechanisms are likely at work, potentially including differing immune function or variation in circulating estrogen levels in overweight or obese patients. A recent meta-analysis also did not show any association between diabetes and survival in patients with pancreatic cancer.¹³

Pancreatic cancer causes significant weight loss, often prior to diagnosis. The authors did not examine whether prognosis differed depending on a patient's weight loss. Rather, any effect of increased weight on cancer development and progression appears to be fixed prior to the onset of weight loss from cancer. If the finding of obesity and survival is confirmed in other populations, it may have implications for the design and conduct of future clinical trials of pancreatic cancer.

The study by Li et al⁹ represents an incremental advance in the understanding of clinical factors contributing to pancreatic cancer development and progression. In particular, the survival analyses illustrate how epidemiological studies that include retrospective information gathering, combined with prospective follow-up, are helpful in establishing survival factors outside of the clinical trial setting. The biological bases for how overweight and obesity contribute to younger age of diagnosis, increased risk for pancreatic cancer, and poorer survival in pancreatic cancer require further investigation. Understanding these associations will provide much needed clues for targeting potential preventive and therapeutic strategies for this extremely aggressive and resistant type of cancer.