Thursday, May 28, 2009

A couple for the prognosis-completists

For the completists, because while intriguing these aren't particularly clinically helpful.

Cancer has an interesting paper about the role of number of metastatic sites in non-small cell lung cancer and prognosis. Interesting because it provides some general prognostic information, as well as looks at tumor burden/volume and how it relates to prognosis.

The data come from a single institution's cancer registry (MD Anderson) and looks only at patients with documented metastatic disease within 3 months of being registered (~1200 patients, median age 62 years). (This means it's impossible to know for sure how far out from diagnosis these patients were; certainly many were initially diagnosed with metastatic disease and registered basically at the time of diagnosis, however some were likely diagnosed substantially prior and were only registered at the time of referral to this tertiary cancer center - we don't know from what's presented here. That is, these data describe survival from time to presentation at MD Anderson as opposed to necessarily time from diagnosis of metastatic cancer, which makes practical interpretation more difficult.) They looked at number of organ sites of metastases, presence of absence of brain metastases, and tumor burden, and how this all related to survival.

Gross survival was terrible: median 8 months overall, 7 months with brain mets. For patients with solitary organ spread (other than the brain), having brain mets made things worse (6 months vs. 10 months without brain mets). For patients with multiple organs with metastases (3 or more) brain mets didn't make much difference (~7 months).

They then looked at patients with only either brain mets (n=32) or lung mets (e.g. the only site of metastasis was to other areas within the lung parenchyma, n=137) to see if the actual volume of tumor at the time of registration correlated with survival. They only included patients for whom they could get decent volumetric data. Basically: size mattered and the greater tumor volume/burden the worse the survival.

While it's nice to know that this is the case (size matters), it isn't much of a surprise and I'm not sure this is of anything other than academic interest at least for palliative clinicians. The data would be more helpful if they were presented as time to death from time of diagnosis with metastatic disease as opposed to presentation at MD Anderson, but such are the limits of their dataset. 7 months vs. 10 months, at least to me, is all basically terrible

Annals of Internal Medicine has a study about the prognostic importance of kyphosis in older women. The data come from prospective, long-running study of osteoporosis in older women (mean age ~72 at time of enrollment, for some frustrating reason only white women were studied, mean follow-up ~13 years) in which the degree of kyphosis was measured. Basically, more pronounced kyphosis was linearlly associated with worse prognosis (hazard ratio of death approaching 3 for those with the worst kyphosis) for these women, even after adjusting for age, smoking, self-reported health status, and other confounders. The authors speculate it is a combination of kyphosis being a measure of general frailty, increased risk for other osteoporotic fractures which can increase mortalilty, as well as respiratory compromise.

Pallimed | Blogger Template adapted from Mash2 by Bloggermint