A matched, case-control study of EBRT + PBRT vs brachytherapy

It would be all too easy to get the idea — from the title and abstract of a recently published study — that proton beam radiation therapy (PBRT) was more effective than permanent seed brachytherapy in the treatment of men with low- and some intermediate-risk prostate cancers. This would be a very inappropriate conclusion.

Coen et al. have carried out a retrospective, case-control analysis in which they compare outcomes of carefully selected patients from two cohorts of patients. Patients from Cohort A participated in the Proton Radiation Oncology Group (PROG) 95-09 trial. This trial enrolled patients between 1996 and 1999 at Massachusetts General Hospital and at Loma Linda University Medical Center. Patients in Cohort B were all treated by a single brachytherapist at Massachusetts General Hospital between 1997 and 2002. While the data generated by this retrospective analysis may offer us some guidance as to the relative merits of the two types of treatment, one cannot use an analysis like this to make any absolute decisions about the relative merits of PBRT and permanent seed brachytherapy in similar groups of patients.

In the first place, it is important to understand that the PROG 95-09 trial was not a trial in which anyone got only PBRT. It was actually a trial of photon-based external beam radiation therapy (EBRT, at a dose level of 50.8 Gy) together with one or other of two “boost” doses of PBRT (at 19.8 Gy or 28.8 Gy). So this was actually a trial of two combinations of EBRT and PBRT. In carrying out the current, matched, case-control analysis, Coen et al. selected patients exclusively from the “high-dose” arm of PROG 95-09.

The high-dose arm of PROG 95-09 enrolled 196 men with clinical T1-2 disease and PSA levels of ≤ 15 ng/ml. In carrying out their analysis, Coen et al. excluded all men with a Gleason score of 7 or higher. Once they had done this, 177 patients were available for case matching. At Massachusetts General Hospital, 203 similar patients were treated by a single brachytherapist from 1997 to 2002. Minimum follow-up of all patients was 3 years.

Case matching was carefully carried out based on clinical T stage, Gleason score, PSA level, and age. This resulted in 141 case-control matches involving 282 patients. The primary study endpoint was biochemical progression.

The results of the analysis by Coen et al. show the following:

  • The average (median) follow-up of patients in the original studies was 8.6 years for the patients in the high-dose arm of PROG 95-09 and 7.4 years for the patients treated by brachytherapy.
  • Based on use of the Phoenix definition of biochemical failure (a PSA level equal to or higher than the PSA nadir + 2 ng/ml)
    • The 8-year biochemical progression rate was 7.7 percent for the men who received EBRT + high-dose PBRT.
    • The 8-year biochemical progression rate for the men who received brachytherapy was 16.1 percent.
    • The difference between these two rates of biochemical progression was not statistically significant (p = 0.42).
  • A stratified analysis was performed by risk group.
    • In the EBRT + high-dose PBRT group, 113 patients were low risk and 28 patients were intermediate risk.
    • In the brachytherapy group, 118 patients were low risk and 23 were intermediate risk.
    • The rates of biochemical progression were similar by either technique.

Coen et al. conclude from these data that, because the combination of EBRT with high-dose PBRT has similar rates of biochemical progression at 8 years to brachytherapy, the conduct of comparative quality-of-life and cost-effectiveness studies can be justified.

The “New” Prostate Cancer InfoLink does not agree with this conclusion. We consider that what is essential is a head to head comparison of treatment of men with low- and intermediate-risk prostate cancer using differing types of radiation therapy. Possible types of radiation therapy could include intensity-modulated radiation therapy (IMRT), PBRT, permanent seed brachytherapy, and stereotactic body radiation therapy (SBRT). This does not have to be a randomized clinical trial. Patients who met eligibility criteria for any and all of these types of radiation therapy could simply elect the therapy of their choice, and then be treated and followed according to a set of standardized protocols. There is no good reason why such a trial could not be carried out today, and it is high time that we at least tried to resolve whether any of the multiple types of radiation therapy was better or worse than any other for the treatment of low- and intermediate-risk prostate cancer.

4 Responses

  1. The question will remain, regardless of seemingly how well the groups are matched is whether any valid conclusion can ever be made from a non-randomized study. If that were to become acceptable it would totally change the entire paradigm for clinical research but I sincerely doubt that anyone at the FDA would agree to such an approach.

  2. Actually, it is my understanding that the FDA has been willing to discuss clinical trials that compare data from prospectively enrolled patients to highly characterized patient data from historic databases under certain circumstances. Whether they would approve a new therapy solely on the basis of such a study, I doubt. However, it is feasible that the FDA might consider such a study as one of two Phase III trials if the other study was a prospective randomized trial.


    Apart from the focus of the article on comparison of proton and brachytherapy, I’m glad to see proton results, even if on a mix of photon and proton and published in an abstract, with follow-up beyond the 5-year point. In fact, the median follow-up, as noted above, was nearly 8.6 years for the higher dose arm of the photon plus proton group.

    I’m impressed that the biochemical progression rate at 8 years was only 7.7%, meaning that 92.3% of these predominantly low-risk men in that group were free of progression! That strikes me as an excellent success rate for external beam radiation at an extended follow-up point. This research documents a benchmark for proton beam (combo) with modern dosing at around 80 Gy or higher instead of lower, probably inadequate doses, used in some earlier proton studies. Patients will be grateful for that clarification.

    I’m not so impressed with the brachytherapy success rate of (100% -16.1% biochemical progression) 83.9% for a predominantly low-risk group of patients. There is a striking contrast between that rate of success and several rates within a percent or two of 100% for a number of brachytherapy studies at centers of excellence with longer follow-up. One likely explanation of the difference is that the patients in the Mass General brachytherapy groups were treated in the years 1997 to 2002, a period when many brachytherapy doctors were not using advanced imaging techniques. Undetected “cold spots” were not rare during that period, in my layman’s recollection of talks that I’ve heard.

    I’m glad to read Sitemaster’s comment about the willingness of the FDA to consider non-randomized studies. Randomization for early stage patients poses severe cultural and ethical problems in our society, of course, and we need practical, sound alternatives.

  4. I see that the proton results were formerly published in 2010 by Zeitman et al.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: