Ways forward for treating depressed patients with cancer

Abstract

Jane Walker and colleagues' findings that major depressive disorder was more prevalent in patients with cancer than in the general community, and that less than 30% of those patients with cancer and depression were receiving any form of treatment for their depression are of concern. This finding is exacerbated in realisation that the success rates for initial treatments for major depressive disorder are between 27·5% and 32·9%, depending on the outcome measure used, and that recurrence of depression after recovery is as high as 60% after 5 years in specialised mental health settings and 35% after 15 years in the general population. In response to these gloomy data regarding treatment efficacy, a rethink of major depressive disorder has been urged on the basis that it is a heterogeneous disease with nearly 1500 possible combinations of the diagnostic symptomatology that meet the criteria for that diagnosis. Further, some have suggested that the present dichotomous diagnostic system (ie, major depressive disorder being present or not) should be replaced by a model that develops symptom profiles for individual patients with major depressive disorder and then traces neurobiological pathways to targets for behavioural change, forming diagnostic subtypes of the disease in the process. This exploration of diagnostic subtypes of depression that are based on clinical content has received attention in the wider research literature, and several possible methods of grouping the symptoms of major depressive disorder into meaningful clusters have been reported, with validity, reliability, and prevalence of one set of these subtypes of major depressive disorder for one form of cancer. If treatment for depression in patients with cancer is to progress, then the "best practice" recommended by Walker and colleagues needs to also progress beyond reliance on simplistic dichotomous models of identifying depression as present or not, to models that examine the spectrum of symptomatology for major depressive disorder, identify homogeneous clusters of symptoms that are linked by their neurobiological underpinnings and external behaviour, and then target specific aspects of those symptom clusters and behaviours for change, based on neurobiology and behaviour rather than epidemiology alone. To remain with a single-variable diagnostic and classification system (ie, a total score on a screening inventory with or without a dichotomous major depressive disorder vs not-major depressive disorder decision) and then prescription of a universal antidepressant or psychotherapy is akin to treatment of all abdominal pain as due to gastric influenza.