Over the past quarter of a century, we have learned how to manage the infectious complications of chemotherapy and the disabling nausea and emesis that was such a feature of the early “platinum” era. What we have been left with, however, is profound fatigue as one of the major complaints of patients on chemotherapy. Brown and colleagues from Scotland have given us an intriguing study that sought to sort out many of the issues related to fatigue, but, like any good report, proposes more questions than it answers.1

Those authors examined a population of 38 patients with advanced lung cancer who were not on therapy and 15 age-matched and gender-matched “normal” controls. They included measures of depression, anxiety, inflammation, physical function, anemia, and performance status. Physical function was operationalized as either grip strength or chair-rise time to make it somewhat less subjective. Not surprisingly, the cancer patients were more fatigued, more depressed, had poorer physical function and performance status, and had elevations of several markers of inflammation. In a multivariate analysis of the relation to fatigue, only performance status, a weakness score, and the anxiety and depression scores were correlated significantly. Brown et al. concluded that physical functioning was poorer with increasing fatigue and that fatigue was related to performance status and psychological distress and not to weight loss or anemia.1

The failure to use chemotherapy in lung cancer (therapeutic nihilism) and the failure of nonsmall cell lung carcinoma to cause significant anemia on its own removed the impact of anemia from the trial by Brown et al. and, no doubt, led to the lack of a correlation with fatigue. This helps us sort out the other contributory factors but makes the results less universal.

Indeed, because there is such an extraordinarily lucrative market in the therapy of anemia, our focus on fatigue has been skewed in this direction. What this trial confirms is that, even absent significant anemia, there is significant fatigue.

The trial is not without its problems, however, not the least of which is the relatively small number of patients studied. There is only a sparse description of the cancer patients, with only a series of average scores with wide standard deviations presented. The attempt to use chair-rise time as a surrogate for physical function is not unreasonable; however, lacking a description of the numbers of patients with painful bone metastases or other conditions that may limit mobility, we are left to wonder about the ability to generalize this finding.

What is most unclear is the sequence of events. Does diminished physical functioning predate and contribute to fatigue, or does it always follow fatigue? If fatigue is the first step, then focusing on the psychological distress, anemia, and other contributors to fatigue should be more productive. If diminished physical functioning antedates fatigue, then focusing on the behavioral and inflammatory factors that may lead to diminished physical activity would be more productive. The ability of aerobic exercise programs to overcome fatigue suggests that there are elements of physical functioning that are important in the later development of fatigue.2, 3 Conversely, both fatigue and diminished physical activity may be secondary to an earlier “event” or series of “events,” including, for example, depression.

An array of possible antecedent events, including abnormalities of energy metabolism,4 neurophysiologic changes in muscle,5 chronic stress responses,6–8 anxiety and depressive disorders,9–11 specific treatments,12–14 concomitant systemic diseases,15 and hypothalamic-pituitary axis dysfunction,6–8 have been proposed and discussed extensively.16 What is lacking to date is the ability to assess whether these changes cause the later fatigue or whether they merely are associated with its development.

I have long had a certain amount of skepticism about the “mind-body connection” and its relation to cancer.17 However, there is some intriguing and, at first glance, unrelated data that suggest there is a systemic affect that precedes the development of the clinical fatigue syndromes we are accustomed to seeing. In the middle 1980s, the Lung Cancer Study Group took a relatively unsophisticated look at quality of life in patients with early-stage, resected lung cancer.18 That group found that a simple quality-of-life questionnaire administered to patients with early-stage lung cancer prior to any intervention was strongly predictive of survival. Several different groups have corroborated this finding in a number of different malignancies.19–21 The Lung Cancer Study Group report was all the more significant, because this finding was noted in a group of patients with early-stage disease who had not yet had any of the confounding effects of therapy. What is it that these patients sense that causes them to report a lower “quality of life” after controlling for disease stage, age, gender, and performance status? Is it a product of tumor growth and metabolism, such as proteolysis-inducing factor,22–24 that leads to a catabolic state that can be sensed but cannot be articulated well?

Depression has long been associated with cancer, although the evidence is strongest that the depression is reactive or secondary to the cancer diagnosis rather than etiologic for the cancer itself. A broad meta-analysis of cancer and depression showed that, overall, 24% (range, 15–42%) of cancer patients were depressed without regard to tumor type, disease stage, or disease status.9 This may have been an underestimate, particularly for a disease like lung cancer, in which therapeutic nihilism is rampant, and patients often are given little or no hope of cure or lasting remission. The current study by Brown et al. is compatible with the concept that depression precedes fatigue, in that all of the patients studied were end-stage lung cancer patients who were receiving no therapy, and depression was one of the significant correlates.

What is needed is a study of these events with physical, behavioral, biologic, and hematologic measures taken serially over time. One problem will be which measures to choose; however, hopefully, a consensus could be reached. More important is the population to study. It should be a population-based study using individuals without cancer at the onset. It is possible that much of these data reside in existing populations and serum samples, such as the Women's Health Initiative or the Framingham Heart Study.

Although there is no question that agents that stimulate erythropoiesis are useful in combating some of the symptoms of fatigue,25 it equally is clear that anemia is not the mechanism for all cases of fatigue. We need to move our exploration of the early events in the development of fatigue far earlier in the pathway that leads to clinical fatigue. If we are fortunate and there is a common pathway by which multiple events related to the tumor or its treatment can cause fatigue, then addressing that pathway will be more productive than trying to treat empirically.


  1. Top of page