Earlier this year we reported on a paper by Zelefsky et al. suggesting lower levels of prostate cancer-specific morbidity and mortality after surgery than after radiation therapy. Now there’s similar data from the CaPSURE registry.
A recent paper by Cooperberg et al. actually suggests that the comparative risk for prostate cancer-specific mortality (PCSM) in patients after treatment for localized prostate cancer is very low — but is still about twice as high in patients treated with external beam radiation therapy (EBRT) as it is in patients treated with radical prostatectomy (RP).
Cooperberg and his colleagues were able to identify 8,982 men with localized prostate cancer (i.e., clinical stage T3a or lower) who were enrolled in the CaPSURE database between 1995 and 2008, and who received one of three types of first-line treatment for their prostate cancer: RP, EBRT, or androgen deprivation therapy (ADT). All the patients were followed for at least 6 months. However, 1,444 of these patients were missing key data and had to be excluded from this analysis.
Careful scrutiny and analysis of these data showed the following:
- 1,293/7,538 patients (17.2 percent) died.
- 226/7,538 patients (3.0 percent) died of prostate cancer.
- There were statistically significant differences between the clinical and sociodemographic factors for the patients in the different treatment groups.
- The mean and median times to death following first-line treatment were 6.8 ± 4.0 years an 6.4 years, respectively.
- The mean and median follow-up of surviving patients were 4.2 ± 3.3 years and 3.9 years, respectively.
- Relative to RP, the unadjusted hazard ratios (HRs) for prostate cancer-specific mortality were 2.46 for EBRT and 4.36 for ADT.
- When adjusted for age and case mix using the patients’ Kattan scores, the HRs for prostate cancer-specific mortality relative to RP were 2.21 for EBRT and 3.22 for ADT.
This paper goes into exhaustive detail with respect to the analysis of the differing patient subsets, and we strongly recommend careful reading of the entire paper to those who want to ensure a complete appreciation of the degree of this detail. However, the following are clearly worthy of note:
- When adjusted for age and case mix using the patients’ CAPRA scores instead of their Kattan scores, the HRs for prostate cancer-specific mortality relative to RP were 1.63 for EBRT and 2.65 for ADT
- Excluding the 136 men who received adjuvant EBRT after RP had no effect on the mortality data.
- Relative to RP, the adjusted HRs for all-cause mortality (based on age and case mix and using the patients’ Kattan scores) were 1.58 for EBRT and 2.25 for ADT, and the same results were observed using the CAPRA scores instead of the Kattan scores.
- If the researchers restricted their analysis to the 67/5,143 patients who died of prostate cancer but were treated after 1998, then, relative to RP, the adjusted HRs for prostate cancer-specific mortality (based on age and case mix and using the patients’ CAPRA scores) were 2.7 for EBRT and 6.5 for ADT.
- Absolute differences between RP and EBRT were small for men with low-risk disease but larger for men with intermediate- and high-risk disease
There is no doubt that many will use these data to claim that surgery is twice as good as radiotherapy at being able to offer cancer-specific survival for patients with localized prostate cancer, and there is presently just one large, randomized clinical trial with multiple treatment arms (the ProtecT study) that is actually studying this question. But the ProtecT trial won’t report results for many years to come.
Cooperberg et al. spend several paragraphs discussing the various reasons why the results of this analysis have to be interpreted with great caution. They are careful to point out that their findings need to be verified against data from randomized clinical trials (when those are available) and also through longer-term follow-up of patients in the CaPSURE database. However, it is hard to dismiss their results completely. They conclude that:
- “In a multi-institutional, prospective cohort of men with prostate cancer, we observed a low overall risk of [prostate cancer-specific mortality].”
- “We identified roughly 2-fold and 3-fold increases in the risk of cancer mortality among those who received external beam radiation or primary androgen deprivation” compared to those who received RP.
- “[T]he greatest differences were observed for higher risk patients.”
If one were to accept the premise that RP does, in fact, have better oncologic outcomes over time than EBRT, how does one “weight” the relative risks of the adverse effects of the two treatments. It is certainly the case that even modern surgery (with RALP or LRP) has a higher level of risk for side effects that modern radiation therapy.
Filed under: Diagnosis, Management, Risk, Treatment | Tagged: ADT, andogen deprivation therapy, EBRT, external beam radiation therapy, localized, outcome, surgery |
THE "NEW" PROSTATE CANCER INFOLINK 
Turn of the century EBRT was definitely less effective than current technology. No inclusion of HDR/LDR brachytherapy data is also a huge missing data set. Seems to me anyway that 8-10 yr data from radiation specialists like Dattoli, RCOG, CET, is at odds with this review. It’s an very interesting result, however, in that it flies in the face of the common opinion that radiation is better for higher risk patients. Information doesn’t always make a treatment decision easier, does it?
SELECTION BIAS HIGHLY LIKELY AS CAUSE FOR FINDING
This finding is so obvious: in the period studied, there was a strong bias toward surgery, and higher risk patients were steered to radiation, with even higher risk patients (like me) heading straight to hormonal blockade. The study did not control for Gleason score or do a disciplined D’Amico risk group analysis, which would have been impractical with the data sets. Even if it had, unless it stuck to clinical biopsy Gleasons vice pathology Gleasons, surgery would have had an unfair advantage by kicking higher risk cases seen only from surgery pathology to a higher risk category – impossible for radiation.
Of course, the earlier point about the change in radiation dosing is also well taken.
‘If one were to accept the premise that RP does, in fact, have better oncologic outcomes over time than EBRT, how does one “weight” the relative risks of the adverse effects of the two treatments. It is certainly the case that even modern surgery (with RALP or LRP) has a higher level of risk for side effects that modern radiation therapy.’
I’d start with the Gleason grade, stage, age of the patient, quality of a selected surgeon, and acceptance of the fact that increased death rate by prostate cancer is a side effect of radiation therapy over longer periods of time. Still I understand your point. But we are also missing longer-term unbiased studies on the efficacy and morbidity of the various radiation techniques, too. Thus a younger man with Gleason 7 or higher and stage T3 disease might feel differently than a 64-year-old man with Gleaosn 6 disease.
This is just one more study that shows that radical prostatectomy improves survival, especially in higher risk patients, over other techniques …. I certainly am glad I had an opportunity to know that I could benefit from combined therapy. Might have missed it if I had chosen radiotherapy first.
I do not believe that CaPSURE permits any valid conclusions about the relative effectiveness of a particular treatment. The self-selection of the data set means this is yet another example of trying to compare apples to oranges and tell which is sweeter. The only valid conclusion is that no conclusion is possible. Although we know many were missing data, we do not know how many did not sign up to participate. Even so, the lack of standardized treatment and the failure to offer hormone therapy to men with intermediate- or high-risk disease, which has been shown to improve survival, means this comparison should be disregarded.