People are jumping to conlusions way too soon...
I understand the "hype" with the final results of RTOG 9601 being published, but:
1. It's called RTOG 9601, for a reason. The protocol was written in 1996. I was in high school back then...
2. Patient in late 90s, early 00s eligible for salvage RT often:
- had inadequate staging (as in RTOG 9601) by today's standards. No MRI done before salvage RT, no PET-scan (as practiced in many places today)
- were treated with 3D-conformal techniques without adequate daily controls of hitting the target. What we see today as an increased in local control with the same dose of radiation treatment is also part of actually hitting the (whole) target every day, since we can perform CBCTs. We certainly occasionally missed the target in the 90s and early 00s. Thus the additional effect of a systemic treatment in terms of local control can also be attributed to us being unable to safely hit the target every day.
- were treated with rather low doses of salvage RT, in the case of RTOG 9601 with 64.8 Gy (with 1.8 Gy/d). We generally give 66 Gy + today, I know several colleagues who give 70 Gy to all and go even higher in high-risk scenarios.
The 1.8 Gy/d actually mean that the equivalent dose with 2 Gy/fraction is probably even lower than 62 Gy (depending on which a/b you use, something which has led to the whole hypofractionation hype in the past years, since many prostate cancers probably have a rather low a/b). Thus 64.8 Gy delivered in 1.8 Gy/d are probably quite a low dose by today's standards.
- had very little other than hormonal treatment in case of progression. The same applies to the RTOG 9601-patients. Failure in the non-hormones arm meant they could get hormones, then they died. Few probably got chemo, but options were limited. Todays patients have 5 lines of possible systemic treatment in the castration-refractory disease setting (enzalutamid, abiraterone, docetaxel, cabazitaxel, alpharadin) and more are being tested (metformin, immunotherapy, etc...). All these treatment options would make a survival benefit less apparent was the trial run again today.
Now let's a look at the subgroups of RTOG 9601:
I generally dislike subgroup analyses, but if you take a look at the full paper you will see that ONLY patients with a PSA > 0.7 ng/ml profitted from the hormonal treatment. The ones that had a PSA <0.7 ng/ml before salvage RT did not profit at all. In fact, it seems placebo was even better for them (although not statistically significant, but provoking in terms of possible increased cardiovascular events etc in the group with hormomes, which may lead to increased mortality).
This finding was not highlighted enough, in my opinion, by the authors. Post-hoc subgroup analyses are not ideal, but if you divide the whole group of patients to PSA < or > 0.7 ng/ml at salvage RT, you end up with two groups of 405 and 355 patients respectively. One group is not profitting, the other is. This is not a "small" subgroup analysis.
Interestingly if you look at the full paper you will see that 48 patients in the placebo-group progressed during the 24-month-period of placebo. If you exclude all the other patients that were allocated to that group but did not complete treatment you end up with 188 patients, out of which apparently 41 progressed over the 24 month period.
Now, from my experience, salvage-RT patients
a) either progress immediately after salvage-RT because of a microscopic tumor deposit outside of the fields (microscopic lymph node metastasis or bone metastasis), some call them the "primary failures", meaning salvage-RT missed the target.
b) or they progress locally several years after treatment, because the dose was apparently inadequate to control the disease.
In RTOG 9601 these "primary failures" accounted for over 25% of all patients in the placebo group. This means that quite a lot of the patients harbored disease outside the prostatic fossa at the time of inclusion. In my experience not that many patients are "primary failures" in today's practice. Perhaps rather 10%.
Thus when you try to extrapolate this into today's practice you can actually say that hormonal treatment is good in addition to salvage RT in case your patients comes for salvage RT with a PSA >0.7 ng/ml.
Which means that either a) you urologist b) your patient c) you, yourself FAILED to get the patient to RT at a prior timepoint, which we already know is ESSENTIAL for long lasting remissions (PSA <0.5 ng/ml, some claim even lower).
Now from the 15 or so salvage RT-patients I saw last year, only 2 had a PSA >0.7ng/ml. And both of them had macroscopic recurrent disease when staged (MRI and PET-CT). One had a local recurrence (which neither I nor the urologist noticed on DRE) and one had lymph node metastases in the pelvis. I treated both with high-dose RT (70 Gy) + hormones.
Now take a look at the PSMA-data (Eiber, et al NEJM).
In patients with PSA 0.5-1.0 ng/ml detection rates of macroscopic disease with PSMA-PET are >70%. This means that in more than half of the patients in RTOG 9601 (if the trial was done again today and PSMA-PET-CT performed at inclusion) the PET result would come out positive. Would you still be comfortable in treating a patient with a macroscopic recurrence in the prostate or a lymph node metastasis or a bone metastasis with 64.8/1.8 to the prostatic fossa?
Now, I am not saying that we did not learn anything from RTOG 9601. I am not saying that the trial is rubbish. I am simply arguing that you should NOT offer hormones to ALL your patients based on this trial.
You can safely ommit it in patients coming to you with a low PSA (I would probably say <0.5 ng/ml, but based on the trial you could also argue for <0.7 ng/ml). And you should (if you can) perform staging before you deliver treatment.
We generally perform MRI and choline-PET-CT fpr patients with a PSA >0.5 ng/ml. I generally do these exams in lower PSA-levels too in case the patient had a rather low PSA compared to his disease extent before surgery (so if I get a patient with a big pT3b who only had a PSA of 8 ng/ml before surgery, I will also order the exams at lower PSA-levels than 0.5 ng/ml, since this is obviously a low-PSA-producing disease)
We recently gained access to PSMA-PET-CT and are now sending all are patients with PSA >0.2 ng/ml for PSMA-PET scans prior to salvage RT. My experience is rather limited yet.
If you see macroscopic local recurrence treat it aggressively with 70Gy, go higher if you want to. Give hormones. Give them as long as you would give them in the primary setting (so if it was a Gleason 7 on surgery, give it 4-6 months. If it was a Gleason 9, give it for 3 years). I'd also consider chemo in young, high-risk patients with GS 9-10. It seems to work in the primary setting, why shouldn't it work in the macroscopic-recurrence setting? Treat nodes in the pelvis, if any pop-up in the MRI or PET scan. I also occasionally treat cN0 patients with pelvic RT, if I feel the lymphadenectomy was not sufficient and it's a high-risk patient.
If you don't see macroscopic disease and the PSA is >0.5/0.7 ng/ml, then sure, RTOG 9601 makes sense. You can give hormones then.
