I learned of a new study implicating stress in reduced breast cancer survival by Twitter. A line in my feed alerted me that CNN’s health blog, “Paging Dr. Gupta,” broke embargo on the soon-to-be-published paper in the journal Clinical Cancer Research. The story – that women who undergo a stress relief program live longer after breast cancer recurrence – couldn’t wait.
“Less stress helps breast cancer patients” is the title at CNN. Researchers, based at Ohio State’s Comprehensive Cancer Center, reported that psychological intervention helps to increase the quality of life and survival among women with recurrent breast cancer.
The intervention at issue is this: weekly, small-group meetings of BC patients for 4 months after their initial surgery and diagnosis. Led by clinical psychologists, the women met 18 times and discussed strategies to reduce stress, improve mood, strengthen social networks, eat better, exercise and adhere to medical treatments.
The current report is an extension of previous findings among an initially larger group of women at the time of BC diagnosis. Then, the researchers randomized 227 patients after surgery, all with stage II or III disease, either to receive the intervention, or not, after an initial psychological assessment and blood tests to check their immune function. The women were overwhelmingly Caucasian, mostly with stage II tumors, well-educated and, for the most part, had ER/PR+ tumors. What the researchers noted was that:
…As predicted, patients receiving the intervention showed significantlowering of anxiety, improvements in perceived social support,improved dietary habits, and reduction in smoking…
In 2008, the same Stress and Immunity Cancer Project investigators reported in the journal Cancer that the psychological intervention reduces BC recurrence and prolongs survival. Here’s where the results become both exciting and suspicious. In 2007, with a median follow-up of 11 years, 62 women (29%) in the initial study had recurrent disease. The proportion among those who’d received the intervention was indeed lower, confirming the authors’ hypothesis that the intervention would help prevent BC from coming back. But the p-value for this difference was 0.034, barely meeting the threshold for statistical significance. Similar results were observed for overall survival among the women who’d attended the group sessions: they had a reduced risk from death with a minimally-significant p-value (0.028) for the difference.
Now, the Ohio group reports on the 62 patients who relapsed. Before going further, I should say that it’s a bummer of a result from a medical perspective and from mine as a BC survivor out at seven years. Group support aside, 44 (71%, yikes!) of the patients with recurrent disease have died with a median time until death (after recurrence) of 2.8 years (range: 0.9 – 11.8 years).
What the authors conclude is that the psychological intervention improved survival after BC recurrence. How the data flowed is this: among the recurrences, there were 33 women who’d been randomized to the assessment only, “A” arm of the original trial and 29 who’d been randomized to the intervention, “I” arm. These numbers were whittled down to 18 patients who could be followed for continued study on the “A” arm and 23 women on the “I” arm. So the total number of women evaluated in this new report is small: just 41 women.
Among those, the women who received the psychological intervention were more likely to survive, with what’s called a hazards ratio of 0.41. Here again the p-value is valid but marginal (p= 0.014). The authors show a very limited amount of data regarding test-tube based studies of natural killer (NK) and T lymphocytes in the different patient groups, and suggest in the paper’s abstract that “immune indices were significantly higher for the intervention arm.”
What would have killed this paper (pardon the verb) had I been a reviewer is this: among the 41 women with recurrent disease, there was a major difference in the treatments they received. According to the “Patients and Methods” section of the paper, in the section on “adherence, chemotherapy dose intensity,” the authors indicate that just 6 of the “A” patients received chemo in the 12 months following the recurrence, while 13 of the “I” patients got chemo in the same period.
My math: only 6 of the 18, or 33 percent of the “A” group (assessment-only) patients received chemo, while 13 of the 23 (56 percent) of the “I” (intervention) patients got chemo in the year after recurrence. That’s a huge difference in medical treatment among a very small number of patients.
My point: the small difference in survival after recurrence among those women who received structured psychological support, years earlier, may be attributable to the prompt chemotherapy they received upon relapse of the cancer. This seems a more plausible explanation than that group therapy-type sessions make a difference in tumor biology or treatment resistance. (Neither outcome is proved by this study.) The difference may also derive from better overall health in the women who initially received the intervention that included advice and support regarding diet, exercise, smoking cessation and medication compliance.
I think there can be tremendous value in psychological support for people with illness of any kind, whether that’s provided casually by supportive families and friends, one-on-one psychotherapy, medication and/or group meetings. And it’s easy to envision that meetings in which women with similar disease situations and concerns get together and discuss coping mechanisms, how to stay healthy, eat better and exercise could have positive effects on overall survival.
But the immune benefits of stress relief, like those presented in David Servan-Schreiber’s like-minded Anti-Cancer, are bogus. There’s no scientific evidence that the sort of NK or T cell changes tested in this study help or hurt breast cancer growth. It’s possible that a revved-up, stress and steroid-driven immune system might help kill cancer cells, or might favor their expansion. It could go either way. The effects of stress steroids on tumors vary and are complex. Prednisone and similar steroids, for example, which resemble the body’s natural stress steroid cortisol, are well-established and effective components of most lymphoma treatment regimens and once were a mainstay of breast cancer treatment. My point is not that the immune system doesn’t affect tumor growth. (I think it can and does.) But the effects are complicated and differ among individuals and according to the specific tumor type.
On linking stress and breast cancer: this argument, which is all that it is in the absence of better data, is patronizing and demeaning to women. It’s the kind of advice we offer children, that if they think and do the right things the outcome will be favorable. Oncology doesn’t work that way. If a woman’s breast cancer comes back, it’s not because she didn’t go to support groups or relax sufficiently. Tumors grow due to inherent, malignant properties of the cancerous cells and other biological factors in the body, such as other illnesses that may, indeed, weaken the immune system or limit a person’s capacity to receive effective treatment.
I’m all for stress reduction, as an aim in itself. But it’s not a cure for cancer.