Remembering Old Lessons in Prostate Cancer: Avoiding Errors 'in PARPituity'

July 15, 2013

The “experts” should maintain a stringent standard regarding what merits further development and reconsider carefully and critically the available data before committing to PARP inhibition, attacks on the PI3K pathway, and vasoactive agents in prostate cancer.

In this issue of ONCOLOGY, two interesting papers set directions for the management of prostate cancer.[1,2] In a thoughtful review, Mitin and colleagues,[1] have addressed an important topic-patients who present with clinically or pathologically evident lymph node metastases. The authors weave a logical tale, reviewing the current NCCN guidelines,[3] and suggesting that the American Joint Committee on Cancer should revisit its aggregation of M+ and N+ disease into a single prognostic group. They also propose a schema for the management of these patients based on their review of the literature.

I marvel at the penchant of our professional societies for issuing “guidelines” in the absence of level 1 evidence; it often confuses clinicians who struggle with the morass of conflicting data, only to be perturbed at a more sophisticated level by a series of partly supported opinions!! Unfortunately some of the extant guidelines have acquired a cachet that convinces payers to reimburse for treatments that are actually supported only by opinion rather than hard data. I would have preferred that the authors’ treatment recommendation be based on level 1 evidence-but it may well be that this proposal will be taken up by one of the cooperative groups, leading to a well-powered, randomized clinical trial. Perhaps if such a trial were to use a 2×2 factorial design, it could meaningfully address the role of locoregional radiotherapy, optimization of radiation dose, duration of androgen deprivation therapy (ADT), or even late toxicity.

Lending credence to this proposal is an important Intergroup study led by the National Cancer Institute of Canada, which compared ADT vs ADT plus radiation (65–69 Gy to the prostate and seminal vesicles, 45 Gy to the pelvic nodes).[4] The population of this study does not equate to the clinically or pathologically node-positive group under the spotlight in the review. However, it is clear that the majority of patients with T3/T4 disease, with high Gleason score and prostate-specific antigen (PSA) level > 20 ng/mL, will, in fact, have occult lymph node metastases. The overall survival difference at 7 years favored ADT plus locoregional radiotherapy, with a hazard ratio of 0.77 (95% confidence interval, 0.61–0.98, P = .033). These data are too early to be definitive, but the final report will be of interest, as the results will relate to the spectrum of disease discussed by Mitin et al.

Before leaving this topic, I do think it important to emphasize that this review hasn’t really addressed the multiple observations (summarized in its Table) that radical prostatectomy followed by immediate adjuvant hormonal therapy (without radiotherapy) appears to yield the best reported long-term survival figures, and one could still question whether radiation is necessarily a crucial element in the treatment of node-positive disease.

The other thought-provoking offering in this issue of the journal is the elegant term paper from Derleth and Yu,[2] which reviews the various mechanisms underlying potentially “druggable” targets in advanced prostate cancer. This is authoritative work, but I would offer a few cautionary notes.

First, the authors seem to provide apparent endorsement for virtually all the promising regimens that they review.[2] I suppose that I sometimes see the glass as half empty, but I do think that we need to avoid abandoning the more conventional data-driven criteria of success that have served us well in traditional anticancer drug development in prostate cancer.[5,6] The FDA Oncology Drug Advisory Committee has struggled for years to define reliable surrogate endpoints to facilitate early introduction of novel therapies, but it has resisted draconian changes because of the absence of data.

We must avoid over-interpreting xenograft studies, small absolute PSA responses, changes of PSA slope, 50% reductions of pretreatment PSA that start at < 10 ng/mL, and waterfall plots that exaggerate the importance of minor changes in PSA.[7] While novel drugs may constitute tempting new tools, they are expensive, and many are investigated as part of a corporate strategy rather than because they offer real benefit to patients. The “experts” should maintain a stringent standard regarding what merits further development and reconsider carefully and critically the available data before committing to PARP inhibition, attacks on the PI3K pathway, and vasoactive agents in prostate cancer!



1. Mitin T, Blute M, Lee R, Efstathiou J. The role of surgery and radiation therapy in the management of lymph-node positive patients with prostate cancer. Oncology. 2013;27:647-55.

2. Derleth C, Yu E. Targeted therapy in the treatment of castration-resistant prostate cancer. Oncology. 2013;27:620-8.

3. Mohler J, Armstrong AJ, Bahnson RR, et al. Prostate cancer, version 3.2012; featured updates to the NCCN Guidelines. J Natl Compr Cancer Netw. 2012;10:1081-7.

4. Warde P, Mason M, Ding K, et al. Combined androgen deprivation therapy and radiation therapy for locally advanced prostate cancer: a randomized, phase 3 trial. Lancet. 2011;378:2104-11.

5. Raghavan D. Non-hormone chemotherapy for prostate cancer: principles of treatment and application to the testing of new drugs. Semin Oncol. 1988;15:371-89.

6. Petrylak D, Ankerst D, Jiang C, et al. Evaluation of post treatment PSA declines for surrogacy using Prentice's criteria in patients treated on SWOG 99-16. J Natl Cancer Inst. 2006;98:516-21.

7. Raghavan D. Novel therapies for advanced prostate cancer – have we widened the goal posts too far? Ann Oncol. 2012;23:2473-5.