C-reactive protein as a marker for prostate cancer survival
Posted Apr 28 2010 12:00am
Evidence continues to accumulate that serum levels of C-reactive protein (CRP) may have value as a prognostic marker for survival in men with very early and very late-stage forms of prostate cancer.
Knowledge about potential relationships between acute and chronic states of tissue inflammation and prostate cancer is evolving. CRP is a protein that is highly expressed in patients with acute-phase inflammation. Questions that have been posed include whether CRP can be used to classify prostate cancer patients according to their risk for lethal disease at the time of diagnosis and their risk for rapid progression of their disease after failure of hormone therapy.
In a study published in 2008, using data from the ASCENT trial, Beer et al. reported that, “Elevated plasma CRP concentrations appear to be a strong predictor of poor survival and lower probability of PSA response to treatment in patients with [castration-resistant prostate cancer (CRPC)] who are receiving docetaxel-based therapy.” Recent research by Prins et al. sought to confirm this finding in an independent data set.
Prins et al. accumulated data from 119 CRPC patients enrolled in six different phase II clinical trials of drugs and drug combinations. They were able to use these data to investigate the relationship of CRP level, alkaline phosphatase level, hemoglobin level, patient age, ECOG performance status, and PSA level with survival. The results of their analysis showed the following:
Median patient follow-up was 19.7 months (range, 0.9 to 98.5 months).
The majority of the patients (89 percent) have died.
Three different markers were independently associated with overall patient survival:
Specifically, a higher CRP concentration predicted a shorter overall survival with a hazard ratio (HR) of 1.106, and this result was statistically significant. (Stated another way, a doubling of the CRP concentration was associated with a 10 percent reduction in probability of survival compared to the median overall survival.)
In a very different study, McArdle et al. have looked at the relevance of CRP levels in patients with organ-confined disease at the time of diagnosis to their risk for prostate cancer-specific mortality. Their study was based on a series of 98 men who met the relevant diagnostic criteria and for whom clinical stage, tumor grade, serum PSA levels, and C-reactive protein concentrations at diagnosis were all recorded. Most of these patients were < 70 years of age at diagnosis and about half received radical local therapy.
Long-term follow-up of these 98 patients (median follow-up was 10 years) generated the following data:
38 patients died, of whom 18/98 (18 percent) died of prostate cancer; 6 died of other cancers, and 14 died of non-cancer causes.
On univariate survival analysis, the patients’ ages, Gleason scores, C-reactive protein levels, and treatment were all significant predictors of overall survival.
Also on univariate survival analysis, the patients’ ages, Gleason scores, and C-reactive protein levels were all significant predictors of prostate cancer-specific survival.
On multivariate analysis, the patients’ ages, Gleason scores, and C-reactive protein levels were still significant independent predictors of prostate cancer-specific survival.
In fact, this study showed that elevated CRP levels at the time of diagnosis nearly doubled a patient’s risk for prostate cancer-specific mortality (with a hazard ratio [HR] of 1.88 in multivariate analysis). The authors of this paper conclude that “the presence of a systemic inflammatory response, at diagnosis, independently predicts poor long-term cancer outcome in patients with localised prostate cancer.”
So exactly how important are all of these data?
In truth, we still don’t know. They do tell us that inflammation may play an important role in the natural history of prostate cancer … but this is hardly a revolutionary piece of information. What is going to be critical are answers to two questions: (a) Can CRP levels be used to accurately predict patient survival in prospective, randomized clinical trials? (b) If CRP levels can be so used, does this additional data provide us with prognostic information that is not already available (based on things like PSA levels, Gleason scores, alkaline phosphatase levels, hemoglobin levels, and other data).
It should be noted that we shouldn’t need to start special trials to investigate the potential of CRP as a prognostic marker. It should be possible to do this in association with other trials of investigational therapies in the treatment of CRPC or through long-term registry data in the case of early stage disease. But until we have prospective CRP data from such trials, the actual clinical value of measuring CRP levels in patients with localized or advanced disease remains unknown.