Consequent on the improvements in early identification  and intervention strategies for breast cancer there has been a marked increase in the number of women who undergo adjuvant therapy. The data reported here indicate that more than one third of these women experience prolonged morbidity characterized by fatigue after successful surgical and adjuvant treatments – termed here PCF. Previous studies have also highlighted the prevalence of PCF in women treated for breast cancer. For instance, Bower et al  in a study of 763 women who had completed treatment five to ten years previously, found that 34% of women reported fatigue up to 10 years post treatment. Andrykowski et al  used a case definition approach in a prospective study of 288 women who had received adjuvant treatment for breast cancer and identified a prevalence of fatigue of 26% at the completion of treatment. A prospective study  found that of 430 patients with a 3-year follow-up, 20% reported fatigue. There was no relationship with either chemotherapy regimen (high versus low dose) or mean hemoglobin level. Poor mental health (identified from Short-Form 36 – SF-36) was the best predictor of fatigue.
Fatigue is a common symptom accompanying both depressive and anxiety disorders . In our data set the only significant predictor of fatigue was mood disorder (Table 2). However, studies in primary care as well as in oncology  and have identified prolonged fatigue states unaccompanied by psychiatric disorder [7, 47]. A review of several studies of depression following adjuvant treatment in patients with breast cancer reported an overall prevalence of 26% . The International Classification of Diseases (ICD-10) criteria for Cancer-related Fatigue specifically designates exclusion of patients with depression, as this disorder may provide an alternative explanation for the symptom set [49, 50]. If applied to the questionnaire-based dataset reported here, 18% of women experienced PCF in the absence of psychological distress at 10 months post-completion of adjuvant treatment.
This is one of very few studies to examine the longitudinal evolution of PCF. This analysis indicated that the fatigue state was largely stable – that is the initial designation of fatigue was not likely to evolve into overt mood disorder on follow-up. This finding argues that PCF has independent determinants from mood disorder, and is consistent with similar findings in relation to CFS . A similar pattern of gradual resolution over time was evident for both the fatigue and mood disturbance illness dimensions. However, the kinetics of resolution of the fatigue state was generally slower than that of psychological disorder. During extended follow-up, women with a combination of fatigue and psychological disorder ran a more persistent course than either those with psychological disorder alone or those with fatigue alone. Emergence of new cases of fatigue, either alone (15/116) or in combination with psychological disorder (7/116), was relatively rare. This finding argues that the genesis of both aspects of the illness complex was at the time of cancer treatment, rather than being attributable to subsequent events.
These data suggest that CRF may be reasonably separated into two temporally-defined groups. Firstly, a treatment-associated fatigue state with at least one clear etiologic factor – anemia . The data for the negative influence of anemia on treatment-associated quality of life in general, and fatigue in particular, and improvement in these parameters following treatment with erythropoetin is well documented . Secondly, cancer-related fatigue encompasses a group who has ongoing fatigue post-treatment – termed here PCF. In this latter group several studies have suggested that that anemia is unlikely to account for a significant proportion of cases of PCF [35, 52]. Analysis of laboratory data from a subset of 47 subjects in the present report, revealed no association between lowered hemoglobin and the development of PCF (data not shown). Studies reported by Broeckel et al  and Bower et al  indicated that fatigue was more severe among patients with menopausal symptoms. Again, the data presented here showed no association between the fatigue state and induction of menopause. Similar findings indicating that fatigue states and menopausal symptoms are independent have recently been reported . The relationship to the intensity of the original treatment is unclear, with some studies suggesting chemotherapy [23, 30] and combination therapy , are more likely to be associated with protracted fatigue, and others finding no such association . Taken overall it would appear that anemia and menopause do not provide a sufficient explanation for PCF.
These findings suggest that alternative hypotheses of the genesis of PCF will need to be explored. For instance, excessive production of pro-inflammatory cytokines that are known to be associated with induction of fatigue in the context of acute infection or immunotherapy, has been proposed . Greenberg  correlated elevations in IL-1 levels with increasing fatigue scores in patients treated for prostate cancer with radiotherapy. Similarly, a cross-sectional study comparing 40 women with PCF five years post therapy for early stage breast cancer with a matched control group showed distinctive changes in cytokine profiles and T cell subsets . An alternative endocrine hypothesis has been derived from recent data examining changes in the cortisol response to stressors in women with PCF suggesting that changes in the hypothalamic-pituitary-adrenal (HPA) axis may be relevant . Such disturbances have also been implicated in relation to chronic fatigue syndrome .