Monday, December 13, 2010

Letters to the Editor on the NEJM Palliative Care in Lung Cancer Trial

The trial published earlier this year in NEJM on early palliative care in metastatic non-small lung cancer remains highly publicized after the initial brisk response to the study.  As I review my Google Reader RSS feed that searches for the term "palliative," I still routinely come across references to this study, most which highlight the survival advantage seemingly conferred to the palliative care group.  (See our initial reactions to the study herehere, and here.)

NEJM recently published four letters to the editor regarding the study and yes, almost everyone is focused on the survival result and how to explain the survival advantage.  Points made about the survival result included the following:

  •  Could the palliative care wing have received more chemotherapy (not less as I hypothesized), thus resulting in improved survival?  Could the mechanism of improved survival have been that palliative care leads to improved performance status, thus leading to more chemo, thus improved survival? (Authors respond that a preliminary analysis suggests that the number of chemo regimens did not differ between the groups.  They highlight some logistical challenges in measuring PS independently in this study over time. )
  • Could the palliative care group have had fewer comorbidities? (Authors acknowledge this possibility but also point out that randomization which resulted in similar pt characteristics between groups should have negated this possibility.)
Only one published letter tried to poke holes in the HQROL conclusion, with the letter writer suggesting that the statistically significant results might not be clinically relevant or perhaps borderline.   In response, the study authors point out that results for the the FACT-L and TOI scores differed enough to be considered clinically significant but did agree that the Lung Cancer Subscale (of the FACT-L) difference was not clinically significant.  This is one of the most challenging aspects of HRQOL research, and those of us who are not researchers need to be aware of meaningful cutoffs.

The letters do nothing to change the evolving significance of this study but should be reviewed by those developing future research to examine similar interventions (perhaps especially the co-morbidity issue). 

Pallimed | Blogger Template adapted from Mash2 by Bloggermint