August 05, 2019
However intriguing the results, this study is unlikely to change practice until the reason for the link is better understood, said psychiatrist Gary Rodin, MD, who heads up the Department of Supportive Care at the Princess Margaret Cancer Centre in Toronto, Canada.
"The implication of the paper is that the treatment of depression improves mortality, but I don't think the paper necessarily proves that," said Rodin. "Assuming this relationship is valid, the question is, what accounts for it? Correlations do not prove causation."
For the study, Gal Shoval, MD, of Tel Aviv University, and colleagues analyzed data from 42,075 patients served by Clalit Health Services, the largest health provider in Israel. The researchers included all patients diagnosed with malignant neoplasm from the time each patient was first prescribed an antidepressant during his or her cancer. Adherence was defined as the ratio between the number of months the patient filled an antidepressent prescription and the number of months he or she was in the study.
The researchers reported that any adherence above 20% was linked to increased survival over the 4 years of the study. The apparent benefit maxed out at 50%-80% adherence, with a 23% reduction in the patient's risk of death over 4 years. Adherence above 80% did not appear to confer any additional benefit.
The methodology of the study was reasonable, said Rodin. "It's not a bad measure [of compliance]. If they filled the prescription, they're more likely to take it than people who didn't fill it. You still don't know they took it but it's a not an unreasonable measure, especially in a large study."
Rodin said that, in general, the sample size — with over 1 million person-years of follow up — was the study's major strength.
However, Rodin said that it is hard to apply the findings to clinical practice since the study shows an association but has failed to deliver on causality. He rejects any suggestion of a direct biological effect of antidepressants on cancer survival.
"The people who are compliant with [antidepressant] medications may be different to people who are noncompliant," Rodin said. "They may have different lifestyle behaviors, they may be more compliant with cancer treatment."
Rodin cites a classic breast cancer study published in The Lancet in 2000 that showed 92% of nondepressed women accepted adjuvant chemotherapy. By contrast, only half of depressed women in the study agreed to chemotherapy.
"The present study tracked compliance with antidepressant treatment, but they didn't track compliance with anti-cancer treatment." Rodin said. "It may be that people who were compliant with depression treatment [in this study] were more compliant with all treatment."
In addition, there are no details of what the antidepressants were prescribed for, or if the patient needed antidepressants at all, said Rodin — or how they responded to treatment.
"I would like to know more about the severity of the depression, which they didn't measure. That's a limitation of the study," he said. "If they showed that people who lived longer not only took the antidepressant medication but their depression remitted that would be a lot more convincing."
In the article, Shoval and coauthors acknowledged that one of the study's limitations was lack of data on the mental health diagnosis. They also conceded that unknown factors that stacked the odds against survival could have been more common in the noncompliant patients, noting, "there might be still residual confounding by unmeasured variables, such as grading and staging of the neoplasm."
Rodin concluded: "So I think's it's premature to say that more adherence causes reduced mortality, but it's an interesting finding and bears further investigation."Shoval and coauthors have disclosed no relevant financial relationships.