prostate question

[yeasterbunny]

I have seen some variation between attendings regarding when to treat pelvic lymph nodes in prostate cancer, and I gather there isn't really a definite consensus. Looking at the ACR appropriateness criteria, they pretty much called it appropriate all the time. Just curious how that goes at other institutions. One high risk feature? Two? Two intermediates? Whatever the attending feels like that day?

---

Members don't see this ad.

 

[Palex80]

We treat prostate +/- seminal vesicles only in low+intermediate risk patients.

In patients with high risk features, as long as one criterium is fullfilled we offer larger volume irradiation to fit patients.
We generally do not treat the entire pelvis (contrary to the results of RTOG 9413 subanalysis), but rather do a mini-pelvis up to 46 Gy with 2 Gy/d (usually up until S2-S3). Then we boost prostate + seminal vesicles to 74 Gy.

We generally do not treat older patients or ones with serious comorbities that influence nerve + vessel function with larger volume irradiation (diabetes, arteriosclerosis, etc...)

We have to keep in mind that next to RTOG 9413 there is a randomized French study which showed no benefit through pelvic irradiation.

 

[radiaterMike]

The RTOG 9413 was re-analyzed. That analysis, in addition to being overly complicated, only muddies the waters in trying to answer this question.

 

[Palex80]

The RTOG 9413 was re-analyzed. That analysis, in addition to being overly complicated, only muddies the waters in trying to answer this question.

The problem with RTOG9413 was its power. It was a study designed to answer too many questions with a rather small number of patients.

 

[qwert]

we use cutoff 15% risk by roach formula
usually it means all NCCN high risk pts and some intermediate risk

I have seen some variation between attendings regarding when to treat pelvic lymph nodes in prostate cancer, and I gather there isn't really a definite consensus. Looking at the ACR appropriateness criteria, they pretty much called it appropriate all the time. Just curious how that goes at other institutions. One high risk feature? Two? Two intermediates? Whatever the attending feels like that day?

 

[TarHeelRadOnc]

The problem with RTOG 94-13 is that the analysis biochemical control using whole pelvis versus prostate only RT is stone cold negative (p=0.72 for protocol definition and p=0.99 for nadir + 2 definition). The speculation about an interaction between neoadjuvant and concurrent ADT and WPRT is thought provoking, but far from proven by the data in that study. Moreover, it is curious that the worst biochemical outcomes are in the WPRT and adjuvant ADT arm ???

There are also two other negative studies of WPRT versus prostate (+/-SVs) RT. GETUG 1 (Pommier et al. JCO 2008) was also completely negative, as was RTOG 77-06 (Asbell et al). The criticism of RTOG 77-06 is that it didn't include sufficiently high risk patients (T1b and T2 only) and didn't analyze PSA outcomes (because of the time period of treatment). But RTOG 94-13 included 67% of patients with T2c-T4 disease and GETUG 1 included 25% T3 patients and analyzed PSA outcomes.

In addition to an absence of clinical benefit, both RTOG 94-13 and GETUG 1 were associated with an increased risk of grade 2+ GI toxicity (16.3% increase in 94-13 and 6.9% in GETUG 1). I think that this data argues strongly against routine use of WPRT.

Regarding the Roach formula, there is an interesting article recently published from the Harvard group (Nguyen et al. Red J 2009) showing that the Roach formula overestimates the risk of LN involvement by 16 fold for patients wityh Roach formula <10%, 7 fold for Roach formula of 10-20% and 2.5 fold for Roach formula >20%. The Roach formula unfortunately does not include T stage in the calculation, which is the most important predictor of LN involvement in the updated Partin tables (Partin et al. Urology 2001).

In my opinion, WPRT should not be routinely used for T1c-T2 disease, even in high risk patients (Gleason 8-10, PSA >20, etc). There may be a role in the rare patient with T3 or T4 disease, but this question remains unanswered.

 

[Gfunk6]

Great discussion. I don't think anybody would argue for WPRT in low-risk or most intermediate-risk patients. The real question is whether to use WPRT for high-risk disease. Obviously, our institutional bias (Roach is the chair) is to pretty much use WPRT uniformly in high-risk patients. The toxicity from this treatment, using IMRT, is very low. To argue that WPRT is toxic by using RTOG 94-13 as an example is an apples to oranges comparison since nobody uses open fields like that any more.

Ngyuen's paper had problems with sufficient LN sampling and general lack of sophisticated approaches to detect nodal disease (RT-PCR etc.). This underestimates nodal risk IMO. Granted this was the 80's and 90's where such approaches were not available but that's the problem with prostate CA in general . . . by the time the data matures the conclusions are frequently of less relevance.

Nevertheless, I agree that WPRT is controverisal and I certainly would not fault anyone for doing a mini-pelvic field or, situtationally, omitting WPRT altogether.

 

[TarHeelRadOnc]

Totally agree about IMRT reducing pelvic toxicity... but that being said, IMRT does introduce increased dose inhomogeneity into the plan and with a pelvic LN prescription dose of 50.4Gy, hot spots are frequently in the 55-56Gy range for whole pelvic treatment. Since the small bowel is a mobile structure, it is conceivable, and even likely, that the bowel moves into the 55-56Gy isodose during fractions (intra-fraction motion) or between fractions (inter-fraction motion). Most institutions using cone-beam IGRT perform image guidance to the prostate, not the pelvic LN regions, so image guidance is not actually helping reduce the risk of focal high dose regions within the bowel. In short, I agree that grade II GI toxicity is significantly reduced with IMRT, but the rare grade III-IV bowel toxicity (strictures/perforations/etc) may not be reduced, and may even be increased, for the reasons described above. We did a study of whole pelvis IMRT with SIB to the prostate (Pelvic LN treated to 50.4Gy in 1.8Gy/fx and Prostate/prox SVs to 70Gy in 2.5Gy/fx) and had one case of high grade small bowel toxicity which has heightened our awareness of this issue.

I also agree about the limitations of the Nyugen study, but even if those estimates (16 fold, 7 fold and 2.5 fold) are too high you have to admit that the Roach formula still overestimates pelvic LN risk by some degree. The Roach formula is based on empiric observations correlating the Partin tables with clinical variables (Gleason and PSA). The problem is that this formula was developed using the original Partin tables, from the pre/early PSA era, and clearly overestimates the incidence of Pelvic LN involvement observed in the updated Partin tables, obtained in the PSA era.

Whole pelvic LN treatment is an area of continued interest for a selec group of patients IMO, but for patients with low volume disease (T1c-T2) there doesn't appear to be a benefit, even in the setting of high PSA or gleason grade.

 

[medgator]

Some may even question the concept of WPRT in the first place. As per the AJCC staging, nodal mets is Stage IV, and surgeons will stop a RP if nodal mets are found intra-op.

Is there any reason to think that RT is going to change the course of what (some would argue) is systemic disease?

 

[TarHeelRadOnc]

Ah... the issue of treatment for pathologically node-positive prostate cancer.

In spite of my previous posts against WPRT for T1c-T2, high risk, clinically node negative prostate cancer, I am a proponent of WPRT for pathologically node positive disease. The AJCC does classify patients with positive pelvic LN as stage IV, however, both prospective and retrospective data suggest that a proportion of these patients can have prolonged disease-free survival with a combination of local and systemic therapy. Of course, the best known example of this is the Messing trial (NEJM 1999 and Lancet Oncol 2006). With 7 year follow-up in patients with pathologically positive nodes at prostatectomy, immediate long-term ADT was associated with 77% RFS. These results were later confirmed with 12 year FU, suggesting that a substantial proportion of patients can have prolonged DFS despite positive LN. This trial most likely represents an outlier, overestimating the true RFS rate with surgery and long-term ADT, but is the only level I evidence addressing this concept. Unfortunately, no phase III trials of whole pelvis RT exist in this setting. There is, however, good retrospective data from MDACC (Zagars et al. Urology 2001) and UPenn (Whittington et al. Red J 1997) evaluating long-term ADT and pelvic RT in patients with pathologically node positive disease. In the MDACC study 10yr bFFF was 80% and in the UPenn study 8yr bDFS was 47%. Moreover, among patients with pathologically positive LN treated on in RTOG 85-31, 9yr bFFF was 33% in those randomized to RT plus life-long ADT. Of course, the benefit of pelvic RT in node positive patients is still debatable and a SEER study showed no benefit of post-op RT in N+ patients (Johnstone et al. Prostate Cancer Prostatic Diseases 2006). Additionally, the role of definitive RT (without lymphadenectomy) in patients with clinically positive nodes on CT/MRI/Combidex/etc has not been adequately studied to date.

IMO, treatment of microscopic disease in the pelvic LNs is a reasonable strategy. However, in patients treated with definitive radiotherapy (without pathologic evidence of LN involvenment), I think that whole pelvis radiation is controversial and, outside of a clinical trial, should be limited to patients with high volume (T3-4) disease based on the results of three negative randomized clinical trials.

 

[Palex80]

In my opinion, WPRT should not be routinely used for T1c-T2 disease, even in high risk patients (Gleason 8-10, PSA >20, etc). There may be a role in the rare patient with T3 or T4 disease, but this question remains unanswered.

Thats's quite easy to say, but you know that we often have a problem of underestimating cT-Stage (we also have a problem of overestimating it sometimes).
Therefore I would recommend WPRT to a fit 65 year old with cT2c cN0 cM0 GS 4+4=8 PSA 23 ng/ml with bilateral 8/12 positive biopsies.

The T-Stage may be a very important factor in the Partin Tables, but we only have cT-status before definitive radiotherapy not pT-status.

In the high-risk patients some view pelvic lymphadenectomy before radiotherapy (ideally laparoscopically done) as the gold standard. If this is available and the patient suitable, then you have an approach that may very well help you a lot in deciding whether to do WPRT or not.

One last thought:
Through the widespreading in image modalities availanle nowadays, like pelvic MRI, Cholin-PET-CT or perhaps soon USPIO-MRI I presume that we identify more patients today with cN1 than we did 10 years ago. Therefore studies performed 10 years ago may have actually treated a lot more patients in cN1 but which were not identified back then, but would be identified in daily practice with todays available modalities.
One could indeed speculate that the impact shown by previous studies through WPRT was because a lot more patients were actually curatively treated for then undiscovered macroscopic lymph node metastasis through WPRT.

 

[TarHeelRadOnc]

Fortunately, our ability to accurately stage the local extent of prostate cancer is improving with MRI and particularly with endorectal coil MRI. I would treat the pelvis in a patient with ECE or SVI on EC-MRI.

"The T-Stage may be a very important factor in the Partin Tables, but we only have cT-status before definitive radiotherapy not pT-status."

The Partin tables actually use the CLINICAL stage, not the PATHOLOGIC stage, to predict the risk of ECE, SVI and Pelvic LN involvement. That is why they are such a valuable resource for radiation oncologists, because they quantify the relationship between clinical and pathologic variables.

"One could indeed speculate that the impact shown by previous studies through WPRT was because a lot more patients were actually curatively treated for then undiscovered macroscopic lymph node metastasis through WPRT"

That is the whole point, previous studies have shown NO impact of WPRT, even in a group of patients with higher risk of having cN1 disease due to older staging modalities. In the setting of improved pelvic LN staging, one could argue that the rationale for routine use of WPRT in cN0 patients is even less sound...

 

[Palex80]

"The T-Stage may be a very important factor in the Partin Tables, but we only have cT-status before definitive radiotherapy not pT-status."

The Partin tables actually use the CLINICAL stage, not the PATHOLOGIC stage, to predict the risk of ECE, SVI and Pelvic LN involvement. That is why they are such a valuable resource for radiation oncologists, because they quantify the relationship between clinical and pathologic variables.

I am aware that the Partin tables use the cT-Status.
The problem is that the data, the Partin tables are based on, are quite old and most cT-estimations were done without MRI or modern techniques. This is the point I was trying to make, I am sorry you misunderstood me.

"One could indeed speculate that the impact shown by previous studies through WPRT was because a lot more patients were actually curatively treated for then undiscovered macroscopic lymph node metastasis through WPRT"

That is the whole point, previous studies have shown NO impact of WPRT, even in a group of patients with higher risk of having cN1 disease due to older staging modalities. In the setting of improved pelvic LN staging, one could argue that the rationale for routine use of WPRT in cN0 patients is even less sound...

You are not getting what I am trying to say.
Some people often look back at older RTOG trials that tested hormone therapy duration and argue that WPRT is important because those older trials were performed with WPRT, so this should be standard of care.
Some other people demonstrate new data with prostate only radiotherapy and say that the control rates with prostate only radiotherapy are the same with WPRT, so we could ommit WPRT.
My point is:
One could speculate that the following two mechanisms led to same control rates in older studies with WPRT and in modern studies without WPRT:
1. The older studies treated patients with cN1 (mistakenly) with WPRT , because imaging was not sophisticated and did not recognize cN1. This lowered the progression free survival in the entire population.
2. In the newer studies patients with cN1 were identified more often than in the provious trials and eliminated from the studies. Patients with cN0 were then treated with prostate only RT, however some of them did harbor micrometastatic disease in the lymph nodes. Therefore some of these patients later recurred regionally. This lowered again the progression free survival in the entire population.
There two effects brought the progression free survival rates of WPRT and prostate only RT close eo each other and led to the argument, that one can ommit WPRT. What people did not think about was, that the patients included in the studies 15 years ago were staged in a different way than patients are staged nowadays.
Therefore one can speculate that WPRT is still useful in patients with high risk of harboring micrometastasis and staged correctly to rule out cN1.

Do you understand what I mean? I am sorry, but I am not a native speaker.

 

[TarHeelRadOnc]

My appologies for misunderstanding you in the previous post. I definitely agree with your point about the Partin tables and their limitations when compared to pelvic staging with modern imaging modalities.

My whole point regarding whole pelvic treatment is this: There are now three RCTs spanning three different eras (pre-PSA: RTOG 77-06; early-PSA era: RTOG 94-13 and mid-PSA era: GETUG 1) that show no advantage with whole pelvic treatment, regardless of the rationale. The best argument against the data from these trials IMHO is that each of these trials delivered low doses to the prostate (65Gy in 77-06, 70Gy in 94-16 and 66-70Gy in GETUG 1). I would argue that whole pelvic treatment may become necessary to improve biochemical DFS in high risk patients in the setting of improved local control with prostate dose escalation (76-81Gy EBRT; EBRT+ LDR/HDR brachy boost), but we just don't have any data supporting routine use of this at the present time.

I have also heard the argument for using WPRT because it was the RT method used in the trials establishing the standard of care for patients with high risk disease (RTOG 85-31, RTOG 92-02, Bolla 1, Bolla 2). However, I think that this argument is flawed. This would be similar to arguing that every patient with early stage (T1-2N0M0) breast cancer needs a complete level I-II axillary lymph node dissection becuase that is the axillary evaluation/treatment technique used in the trials establishing breast conservation therapy as an alternative standard of care (NSBAP B06, Veronesi, EORTC, Gustave-Roussy, Dutch, NCI, etc). More recently, studies have shown that SLN Bx is feasible and associated with False negative rates of <10% and negative predicted values of >92% and, is therefore considered to be a standard method of evaluating the axilla for patients with primary breast tumors <4cm in size in this country. Similarly, after the completion of the previoulsy listed trials establishing the standard of care (RT + long-term ADT) for high risk prostate cancer, trials have shown that whole pelvic treatment increases toxicity and dose not influence DFS. As a result, I would argue that whole pelvic treatment is at the present time controversial at best and, IMO, not supported by the available data.

Another point that I agree with from your prior post is that pelvic lymphadenectomy may be a reasonable standard staging modality for high risk patients prior to definitive EBRT + ADT... Unfortunately, I don't think that Urologists in this country would buy into that approach...