Extraprostatic extension (EPE) alone is not enough to justify adjuvant radiation

Patrick Walsh and Nathan Laurentschuk wrote an opinion piece in European Urology taking issue with the 2013 AUA/ASTRO recommendation that adjuvant radiation is indicated for men with a pathological finding of extraprostatic extension (EPE, stage pT3a) after surgery, regardless of the surgical margin status. The combination of EPE and negative surgical margins is the most common adverse finding, accounting for 60% of them. Therefore, the AUA/ASTRO guideline would lead to gross overtreatment if it were followed. They believe that it is fortunate that that guideline is increasingly ignored (see this commentary).

They looked at the three randomized clinical trials of adjuvant radiation vs. wait-and-see, for evidence that EPE alone justified adjuvant radiation.

• ·      Although it concludes that all patients with EPE should have adjuvant radiation, SWOG 8794 never looked at that subgroup separately.
• ·      In ARO 96-02, men with EPE and negative margins received no statistically significant benefit in terms of freedom from biochemical failure from adjuvant radiation.
• ·      Not only was there no benefit, but EORTC 22911 found a 78% increased risk of dying among men with EPE and negative margins who received adjuvant radiation.

They conclude with a set of recommendations about adjuvant radiation:

“Who should NOT receive it:

 • Men with extraprostatic extension (capsular penetration) with negative margins

 • Men aged >70 yr unless they are very healthy and have high grade or positive margins

 • Men with bladder neck contractures or significant incontinence who have marginal indications

Who should receive it:

 • Men with Gleason ≥7 with positive surgical margins

Marginal benefit:

 • Men with positive seminal vesicles

In a commentary published in Practice Update, Christopher King, a radiation oncologist at UCLA, takes tissue with their recommendations. He argues that until the findings of randomized clinical trials provide more reliable data, current evidence does not justify adjuvant radiation based only on adverse pathology. Instead, based on several retrospective studies (reviewed on this site), he advocates waiting for some evidence of measurable disease. He believes that early salvage (before PSA rises above 0.2 ng/ml) will have equivalent oncological outcomes to adjuvant radiation, but will avoid the toxicity of overtreatment.

Metastases after early vs. delayed salvage radiation

Until we have the results of randomized clinical trials on the relative efficacy of early salvage radiation, we have to look for other clues to inform the timing of that decision. Adjuvant radiation carries a high risk of overtreatment, whereas delayed salvage may preclude the window of opportunity during which salvage radiation might have been curative.

Den et al. posted the outcomes of their investigative analysis at the ASCO Genitourinary Conference (Abstract 12). Data on 422 patients treated at 4 institutions were retrospectively analyzed. All had adverse pathology (either stage T3 or positive margins) after RP. Patients were arbitrarily divided according to their PSA after surgery at the time they received radiation:

• ·      <0.2 ng/ml – “adjuvant RT” (111 patients)
• ·      >0.2 but <0.5 ng/ml – “early salvage RT” (70 patients)
• ·      >0.5 ng/ml – “delayed salvage RT” (83 patients)
• ·      No radiation received (157 patients)

CAPRA-S scores and Decipher genomic classifier scores were found to independently predict risk of metastatic progression. Adjusting for those scores:

• ·      Delayed salvage RT increased risk of metastases by 4.3 times over adjuvant RT
• ·      No radiation increased risk of metastases by 5.4 times over adjuvant RT
• ·      Early salvage and adjuvant RT had about the same risk of metastases
• ·      Men with low CAPRA-S and Decipher scores had low risk of metastases
• ·      Men with high CAPRA-S and Decipher scores benefit from adjuvant RT, but had high rates of metastases nonetheless.

This study once again underscores the importance of early salvage radiation for curative therapy after failed surgery when there is adverse pathology. They didn’t investigate the use of ultrasensitive PSA to determine what the lowest level that avoids overtreatment might be. Adverse pathology and PSA are important to consider, but other clinical/genomic factors can contribute to the decision-making process as well. Low Decipher scores can help rule out those cancers that are unlikely to metastasize in the next 5-10 years. However, it is less useful at indicating those cancers that will metastasize.  And there are no good tests for determining if the cancer is already systemic and micrometastatic, in which case salvage radiation would be futile. This remains a challenging situation for discussion between the patient and radiation oncologist.

Declining use of RT in treating clinical stage T3 patients and those with adverse pathology after surgery

Patients clinically diagnosed with prostate cancer outside of the prostate capsule (stage cT3), are increasingly treated with radical prostatectomy (RP) rather than with primary radiation therapy (RT). In addition, patients who have adverse pathological features after first-line surgery (stage pT3 and/or positive margins) are increasingly not receiving either adjuvant or early RT.

Nezolosky et al. looked at the SEER database records of 11,604 patients clinically diagnosed with stage T3 prostate cancer from 1998 to 2012. They found:

• ·      RP use increased from 12.5% to 44.4%.
• ·      RT use decreased from 55.8% to 38.4%
• ·      “No treatment” decreased from 31.7% to 17.2%
• ·      For extracapsular extension (stage T3a), RP use was 49.8% vs. 37.1% for RT in 2012.
• ·      For seminal vesicle invasion (stage T3b), RP use was 41.6% vs. 42.1% for RT in 2012.
• ·      RT use exceeded RP by 59% if the biopsy Gleason score was 8-10.
• ·      RT use exceeded RP by 3% among those with higher PSA, and by 7% among older patients.

This trend is troubling because RP for cT3 is often not curative. The following biochemical recurrence-free survival rates have been reported and are very consistent:

• ·      Mitchell et al. (Mayo Clinic): 41% after 20 years for cT3 patients.
• ·      Freedland et al. (Johns Hopkins): 49% at 15 years for cT3a patients.
• ·      Carver et al. (Memorial Sloan Kettering): 44% at 10 years for cT3 patients.
• ·      Hsu et al. (Leuven, Belgium): 51% at 10 years for cT3a patients.
• ·      Xylinas et al. (Paris, France): 45% at 5 years for cT3 patients.

The rates are similar among those diagnosed with stage T3 at pathology. Hruza et al. reported bRFS of 47% and 50% for those staged pT3a and pT3b respectively. Pagano et al. reported bRFS of 40% for those staged pT3b. Watkins et al. found that 40% of pT3 surgical patients had already biochemically relapsed after a median of 18 months.

There are other factors that affect recurrence prognosis after surgery. Age, a high pre-treatment PSA, high Gleason score, positive surgical margin (including its size and Gleason score at the margin), and the length of extraprostatic extension (EPE) are all risk factors (see Fossati et al., Djaladat et al., Ball et al., Jeong et al.). In the Watkins et al. study, patients with EPE and negative surgical margins biochemically relapsed at the rate of 0%, 28% and 63% for Gleason scores of 6, 7 and 8-10, respectively. However, if the surgical margins were also positive, the relapse rates were significantly worse: 33%, 50%, and 71% for Gleason scores of 6, 7 and 8-10, respectively. Briganti et al. found that the 5-year bRFS was 55.2% among surgical patients categorized as high risk, which includes stage T3, Gleason score 8-10 or PSA>20 ng/ml.

Can primary radiation alone do any better? I haven’t seen breakdowns for stage cT3 patients specifically, but we have long-term follow up in many clinical trials where high-risk patients were treated with radiation and ADT. Here are some bRFS results we discussed recently:

• ·      HDR brachy monotherapy: 77 – 93% (3-8 years)
• ·      HDR brachy boost + EBRT: 66 - 96% (5-10 years)
• ·      LDR brachy monotherapy: 68% (5 years)
• ·      LDR brachy boost + EBRT:  83% (9 years)
• ·      EBRT monotherapy: 71 - 88% (5 years)

While primary radiation typically does about 50-100% better than primary surgery at controlling the cancer, urologists often argue that adjuvant or salvage RT will bring the numbers into line. There is an ongoing randomized clinical trial (NCT02102477) among men diagnosed with stage T3 comparing initial radiation treatment to prostatectomy plus salvage radiation. While we wait for those results, we have to rely on retrospective studies. In many of the studies cited above, about a quarter of the patients received salvage/adjuvant RT following surgery. In the Mayo study, 72% were recurrence-free after 20 years, which does bring the combination close to what radiation alone often delivers. However, that comes at a cost. Adjuvant and salvage RT usually has worse quality-of-life outcomes than the patient would have suffered had he had radiation to begin with.

This brings us to the second alarming trend: adjuvant and early salvage RT rates have been declining among men with adverse pathology after prostatectomy. We discussed this previously (see this link). So not only are T3 patients receiving a therapy upfront that is less likely to control their cancer, they also may not be receiving the adjuvant or salvage RT that might control it if used early enough.

It is especially troubling that there has been no corresponding shift to later salvage RT. Sineshaw et al. conjecture as to the reasons for the trend:

“This pattern of declining use could be due to multiple factors, including patient preference, physician and referral bias, concern about toxicity, lack of a consistent survival benefit seen in the updated randomized trials, or a growing preference for salvage radiation at time of biochemical failure, rather than immediate adjuvant RT. With respect to the last point, our data did not show a rise in RT use after 6 mo and within the first 5 yr post-RP, suggesting that a shift to salvage RT does not likely entirely explain the declining use of immediate (within 6 mo) postoperative RT.” [emphasis added]

I’d like to believe that the decline in salvage radiation utilization is attributable to better selection of patients. Utilization was higher in those with positive surgical margins and those with Gleason scores 8-10. However, Dr. Sandler may very well be right in attributing the drop-off to urologists who don’t immediately refer patients with adverse pathology to radiation oncologists. In my experience, many patients making the primary therapy decision also never consult with a radiation oncologist. High-risk patients are especially needful of guidance from the first doctor they see – almost always a urologist – to seek second opinions. It would be unconscionable if they are not receiving that guidance.

Declining trend in the utilization of adjuvant radiation.

A recent report in European Urology found that in spite of three randomized clinical trials (RCTs) that proved the efficacy of immediate or adjuvant radiation following surgery with adverse pathology results compared to a wait-and-see approach, a lower%age of such patients are getting adjuvant treatment. Why should this be?

I refer readers to a recent discussion of the issues involved, which I won’t fully reiterate here. First, let’s look at the report by Sineshaw et al. The authors examined the records of 97,270 patients in the National Cancer Database where patients were found to have adverse pathological features (pT3/4 or positive surgical margins) in the period from 2005-2011. What they found is this:

·      Postoperative RT utilization declined from 9.1% to 7.3%.

·      Utilization declined with age: 8.5% in patients aged 18–59 to 6.8% in patients aged 70–79. 

·      Utilization was 14% at community cancer programs compared to 7% at teaching/research centers.

·      Among those with stage pT3/4, utilization was 17% if they had positive margins, but 7% if they had negative margins.

·      Utilization was 17% among those with pathology Gleason score of 8-10 compared to 4% among those with Gleason score  of 6 or less.

First, a note about the timeframe examined in their study: only one of the three RCTs  (Thompson et al. 2009) was published in that timeframe. The Bolla et al. study was not published until 2012, and the Wiegel et al. study was not presented until 2013. The AUA/ASTRO guidelines advocating adjuvant radiation were not issued until 2013. So in the timeframe examined in their study, we would not expect to see the full impact of those three studies and the new guidelines. This conflicts with the statement made in the publication:

“In a retrospective analysis of 97 270 patients with prostate cancer, we showed that use of postoperative radiotherapy for adverse pathologic features has declined over time after the publication of findings from major randomized clinical trials and consensus guidelines supporting consideration of such therapy.”

A report in Medscape included comments from some illustrious radiation oncologists that are worth noting:

·      Jeffrey Michalski  (Washington University, St. Louis) echoed the authors’ anachronistic lament that doctors were not following the evidence in the RCTs and guidelines.

·      Anthony D’Amico (Dana-Farber and Brigham and Women’s Hospital) pointed out that only one of the RCTs showed an advantage in metastasis-free and overall survival. He further explained that multiple risk factors may be a better indication for adjuvant radiation.

·      Michael Zelefsky (MSKCC) noted that we don’t yet know if waiting for rising PSA would have any worse outcomes.

·      Howard Sandler (Cedars-Sinai) blamed low utilization on urologists who don’t immediately refer adverse pathology patients to radiation oncologists. They are not given options or provided with expertise.

Until the results of ongoing clinical trials on the benefit of early salvage radiation become available, this remains a difficult decision. A patient with adverse pathology should immediately begin discussions with a radiation oncologist, preferably at a teaching/research hospital, so that he fully understands what the risks and benefits are of waiting.

What should one do if one’s PSA remains detectable after radical prostatectomy?

After surgery, PSA should become undetectable on a normal PSA test within a month or two, but sometimes it remains elevated. The purpose of the ARO 96-02 randomized clinical trial was to determine whether there was an advantage to treating stage T3 patients while PSA was still undetectable, or whether they could wait to be treated. Waiting has the advantage of allowing better healing of recently cut and handled tissues, and avoiding overtreatment. In that study, there was a significant advantage to immediate treatment in preventing eventual clinical recurrence. However, the study also included 74 patients whose PSA never became undetectable, and they all received immediate radiation therapy.  Wiegel et al. did a 10-year follow-up analysis of that group to see how they fared.

Among the 74 patients:

• ·      Their median PSA after surgery was 0.6 ng/ml (range 0.05-5.6 ng/ml)
• ·      They were checked for distant metastases with a bone scan and X-rays.
• ·      They all received 66 Gy of 3D CRT to the prostate fossa at a median of 86 days after surgery. 58% received adjuvant hormone therapy.
• ·      Compared to those who reached undetectable PSA, they had higher pre-surgery PSA, stages, Gleason scores, and incidence of positive surgical margins.
• ·      Among 48 patients for whom data was available, only 15% achieved undetectable PSA following radiotherapy
• ·      Their 10-year clinical relapse-free survival was 63%.
• ·      Their 10-year metastasis-free survival was 67%, compared to 83% among patients who had undetectable PSAs initially.
• ·      Their 10-year overall survival was 68%, compared to 84% among patients who had undetectable PSAs initially.
• ·      There were no Grade 3 or higher acute toxicities, but 7% experienced Grade 3 late urinary toxicity.

Wiegel concludes:

“A persisting PSA after prostatectomy seems to be an important prognosticator of clinical progression for pT3 tumors. It correlates with a higher rate of distant metastases and with worse overall survival. A larger prospective study is required to determine which patient subgroups will benefit most from which treatment option.”

This seems a reasonable conclusion. Some patients with persistent PSA after surgery will enjoy a long-term survival benefit from adjuvant radiation aimed at the prostate fossa, but a third will develop metastases and die in spite of such treatment. It seems that only 15% were actually cured, or at least became undetectable, by the therapy. It was not the purpose of their study to detect a survival benefit in this subset of patients who had persistent PSA, so there are no conclusions that can be drawn about the strategy of immediate treatment. We cannot yet reliably identify through imaging or biochemical tests those men who will benefit from immediate radiation, although some men with high or quickly rising PSA may have PET-detectable metastases.

In spite of this, Wiegel is quoted in the ASTRO press release as saying:

“Our analysis demonstrates that patients who have detectable PSA post-prostatectomy may benefit from more aggressive, early and uniform treatment that could improve survival outcomes.”

This conclusion seems unwarranted from the data. It is certainly a reasonable decision, and one that many patients will make in consultation with their radiation oncologists.