LDR brachytherapy (LDRBT) monotherapy across risk groups

New registry data from the Cleveland Clinic shows good oncological control with low dose rate brachytherapy (LDRBT) monotherapy, at least for low risk and low-intermediate risk groups. This is the first time I’ve seen LDRBT monotherapy data for higher risk groups. We expect the lower dose/tighter margin monotherapy to reduce toxicity over combination treatment with external beam radiation and a brachy boost. The question is whether the trade-off with oncological control is worthwhile.

Recent clinical trials have focused on whether the addition of combination of EBRT with a brachy boost was better than EBRT alone. The ASCENDE-RT trial showed it did, at least for the higher risk groups. Other studies have looked at the benefit of adding ADT. They found that adding ADT conferred a bigger benefit than adding EBRT among “unfavorable intermediate risk” patients.  Herbert et al. found that 6 months of ADT with LDRBT on Gleason 7 patients led to 5-year “no biochemical evidence of disease” of 95% independent of the predominant Gleason pattern  (3 or 4) even without EBRT. But in the modern era of dose-escalated LDRBT, we do not yet have any randomized comparisons of combination therapy to monotherapy. RTOG 0232 will address this question, at least for intermediate risk patients, when we get its results in 2017. The Cleveland Clinic data do not directly answer that question, but by providing monotherapy data for high risk patients, they help us understand the trade-offs.

In the Cleveland Clinic series, analyzed by Kittel et al., ADT was included for many in the high risk group, but external beam was not. From 1996 to 2009, 1,989 patients were treated with LDR I-125 brachytherapy monotherapy. Treatment guidelines were as follows:

• ·      Prescribed dose was 144 Gy
• ·      18.2% received ADT:

o   10% among low risk, 20% among intermediate risk, 64% among high intermediate risk and 81% among high risk patients

o   ADT use decreased over time: 56% in 1997, 9% in 2004

• ·      Baseline urinary function and prostate volume were not used to exclude patients
• ·      Urethral dose <150%
• ·      Intra-operative planning
• ·      Cold spots allowed in the anterior/superior region
• ·      Margins were 3mm to 5mm
• ·      Wider margins increased to 5mm- 1 cm for high risk cases
• ·      Seeds avoided within 3-5mm of rectal wall
• ·      Stranded seeds used on the periphery
• ·      CT within 4 weeks to check dosimetry
• ·      F/U every 6 months for 3 years, then annually
• ·      Classified into NCCN risk groups
• ·      Intermediate risk sub-classified into “low intermediate” if only one intermediate risk factor (i.e., either PSA between 10 and 20 or Stage T2b/c or Gleason Score=7), otherwise “high intermediate.”
• ·      Median follow-up was 6.8 years

As an aside, it’s worth noting that the intermediate risk sub-stratification they used, sometimes known as the “Zelefsky stratification”, was developed in the 1990s when brachytherapy protocols and outcomes were considerably inferior to what they are today. Dr. Zelefsky now advocates the “Zumsteg stratification” of intermediate risk as we recently discussed here and here.

The 5- and 10-year biochemical relapse-free survival (bRFS) by risk group are summarized in the following table.

Risk Group	# of patients	5-yr bRFS (percent)	10-yr bRFS (percent)
Low Risk	1,219	95	87
Low Intermediate Risk	592	90	79
High Intermediate Risk	90	81	*
High Risk	88	68	*
TOTAL	1,989	92	82

* small sample

The authors reported acute toxicity in a previous report, which I do not have. They did not report late-term Grade 1 and Grade 2 toxicity. Genitourinary (GU) toxicity was Grade 3 or higher in 7.6%, while gastrointestinal (GI) toxicity was Grade 3 or higher in 0.8%. The part of the data I saw did not break out separately what the toxicity was for high risk patients, whose margins were wider. The authors note that the observed toxicity rates were lower than those reported for combination therapy in other series.

Men with prostate length of 5 cm or more were 2.4 times more likely to suffer serious GU complications, while the association with prostate volume was not clinically significant. Men aged 70 or over were 71% more likely to suffer from significant GU problems. These risk factors should be taken into account in counseling men considering LDRBT.

Unfortunately, the authors did not track patient-evaluated quality-of-life outcomes, and there is no data on potency preservation.

Low Risk

Because this was not a randomized comparative trial of monotherapy vs. combined therapy, it is impossible to draw conclusions as to its relative efficacy. The control among low risk patients is good, although based on recent findings about active surveillance, many might be better served by deferred treatment.

High Risk

The control among high risk patients seems to be somewhat less than in other recent trials where combination therapy with EBRT was used instead:

• ·      In the ASCENDE-RT trial, the 7-year biochemical control was 83%.
• ·      At Memorial Sloan Kettering, the 5-year biochemical control was 80%. At Schiffler Cancer Center and UMichigan, the 8-year biochemical control was 86%.
• ·      Liss et al. reported 5-year biochemical control was 85% among men with Gleason pattern 5.

I think most men diagnosed with high risk prostate cancer would be willing to pay the cost of increased toxicity to get the additional oncological control from added EBRT, but that is certainly a matter of personal preference. The brachy monotherapy biochemical control for high risk is similar to what we might expect from surgery; however, the toxicity is quite a bit lower with LDRBT.

Intermediate Risk

The real controversy is in the intermediate risk category. The patient diagnosed with “favorable intermediate risk” prostate cancer is faced with a bewildering array of alternatives, including active surveillance, any of several kinds of focal ablation, surgery, SBRT, IGRT/IMRT, LDRBT monotherapy, HDRBT monotherapy, PBT,  (LDR or HDR)BT+IMRT, LDRBT+SBRT,  or PBT+IMRT. Piling on the radiotherapies has the potential to pile on side effects as well. Intermediate risk readers interested in brachytherapy will be interested in reading Spratt and Zelefsky’s argument for combined therapy, Stone’s counterargument, and Spratt and Zelefsky’s rebuttal. There are points of agreement. They agree that “favorable intermediate risk” patients do not need combined therapy. They also agree that treatment with high enough radiation dose is critical to success.

Aside from the references they cited, here are a few more for intermediate risk patients treated with combination therapy (including ADT):

• ·      ASCENDE-RT 7-yr bRFS: 94%
• ·      CALGB 99809 6-yr bRFS: 85%
• ·      RTOG 00-19  8-yr bRFS: 82% 

While the control rates look excellent, none of those studies divide the intermediate risk group into separate sub-categories. The Cleveland Clinic outcomes for “low-intermediate risk” are certainly within this range (the weighted average bRFS for the entire intermediate risk group was 89% at 5 years), but were accomplished without the potential for extra toxicity from the added external beam radiation.

The Cleveland Clinic data demonstrates the very disparate outcomes within the intermediate risk sub-groups. It would be interesting to see their outcomes, as well as the outcomes of the other cited studies, stratified according to the Zumsteg criteria. Lacking that, and pending the definitive randomized clinical trial data from RTOG 0232 in 2017, the decision about whether to add EBRT or ADT to LDRBT for intermediate risk patients should involve a close analysis of his individual risk factors and his attitudes about potential side effects.

 note: thanks to Dr. Jay Ciezki  for making the full text of the article available to me.

External beam radiation therapy (EBRT) with a low dose rate brachytherapy (LDRBT) boost provides superior cancer control compared to EBRT alone.

Numerous retrospective analyses have suggested that the combination of external beam radiation therapy (EBRT) with a low dose rate brachytherapy (LDRBT) boost is highly effective in controlling prostate cancer in unfavorable risk patients. For the first time, to my knowledge, we have a randomized comparative trial confirming that. An abstract was presented at the GU Conference and there is a press release about it.

ASCENDE-RT was a randomized clinical trial among 122 intermediate and 276 high risk patients treated in 6 Canadian centers from 2002-2011. The treatment specifications were:

• All patients received:
• Whole pelvis EBRT of 46 GY
• 8 months of neoadjuvant ADT + 4 months of concurrent and adjuvant ADT
• The EBRT-only group of 200 patients received an additional 32 Gy to the prostate (total = 78 Gy)
• The LDRBT-boost group of 198 patients received an additional boost of 115 Gy I-125 seeds in the prostate.
• Median follow up was 6.5 years, and was as long as 9 years for 65 patients.

The researchers found:

• After 9 years, the biochemical progression-free survival  (bPFS) was 83% for the LDRBT-boost group compared to 62% for the EBRT-only group.
• bPFS deteriorated by about 6% per year for the EBRT-only group.
• bPFS was fairly stable for the LDRBT-boost group after reaching 89% at 5 years.
• After 7 years, LDRBT-boost had better bPFS than EBRT-only both among intermediate risk men (94% vs. 80%), and among high risk men (83% vs. 72%).
• Median PSA at latest follow up was 0.02 ng/ml for the LDRBT-boost group and 0.24 ng/ml for the EBRT-only group.
• Reflecting the long natural history of disease progression, there were no significant differences in metastasis-free survival, prostate cancer-specific survival, or overall survival. Differences may emerge with longer follow up.

The improved oncological control came at the expense of increased toxicity for the combination therapy.

• Late term Grade 2 or higher genitourinary (GU) toxicity was higher for the LDRBT-boost group. Late term Grade 3 GU toxicity reached 19% for the LDRBT-boost group vs. 5% for the EBRT-only group.
• Late term gastrointestinal (GI) toxicity was similarly mild for both groups
• This early report did not include an analysis of acute toxicity, or an analysis of erectile function.

I look forward to the full analysis of the data when published. I hope they will break out the results separately for favorable and unfavorable intermediate risk patients to the extent that sample size may allow. Perhaps the favorable intermediate risk patients can be spared the extra toxicity of the LDRBT-boost treatment while still enjoying oncological control.

For the high risk patients especially, this study establishes LDRBT-boost therapy as the preferred treatment compared to EBRT-only, unless pre-existing urinary issues rule it out. It is unclear whether the 12 months of ADT and the whole-pelvis radiation would be necessary for all patients.

In an earlier randomized clinical trial (Sathya et al.), high dose rate brachytherapy (HDRBT) boost was shown to reduce the biochemical and clinical failure rate by 50% compared to EBRT-only (66 Gy). Other randomized clinical trials of HDRBT-boost (Hoskin et al., Guix et al.) also found that the boost improved outcomes. It is unclear whether boost with HDRBT, LDRBT, or treatment with SBRT alone will eventually emerge as the preferred treatment for unfavorable risk patients, or whether it will make a difference.

Can invasive procedures spread prostate cancer?

Prostate cancer is seldom spread by invasive procedures such as biopsies, prostatectomy, TURP, LDR brachytherapy, HDR brachytherapy, or insertion of fiducials for image-guided radiotherapy. We know this because those procedures have high cure rates. Nevertheless, there have been isolated case reports of such inadvertent cancer dissemination occurring.

Mechanism of unintended cancer dissemination

Two direct mechanisms have been proposed as ways in which invasive procedures may facilitate the spread of cancers:

(1) by direct implantation from invasive instruments like biopsy needles or surgical knives, and

(2) by release of tumor cells into the bloodstream or lymph.

It is likely that only a few, less prevalent types of prostate cancer cells are amenable to spreading by invasive procedures. The most prevalent types of prostate cancer are incapable of survival outside of the prostatic environment. Several studies have now shown that true Gleason 6 tumors have never been known to metastasize. However, that does not preclude their eating into adjoining tissue, or possibly evolving to higher Gleason grade. Only cells that have some major alterations in their genetic structure are capable of moving through and beyond the prostate. Cancer stem cells spread readily yet are not always detected.

Detection

Until very recently, we lacked the technology to detect the very small foci of cancer cells that may have been accidentally seeded. Those foci have been found only when they grew much larger. That has occurred as long as 19 years later in one case, and then suspicion was raised because it recurred in such an unusual spot (the perineum). The advent of multiparametric MRIs has enabled us to see much smaller foci of recurrences than we have before, with the potential to see foci as small as 4 mm in length (Barchetti and Panebianco). The limit of detection may be even lower for the new generation of PSMA-antibody-based radiotracers coupled with the new PET/MRI scanners. Even so, if the cancer is in a more usual place, such as the anastomosis, how do we distinguish a cancer placed by instrumentation from one that grew there naturally?

Circulating tumor cells may be found with CellSearch® or ADNA® tests.

Biopsy

Biopsies can break off rogue cancer cells and plant them along the needle tract. This has been observed in breast cancer, liver cancer, and rarely in other cancer biopsies as well (Shyamala et al.) Although the cells are planted there, they do not remain viable for long outside of their host environment (Loughran et al.) and usually do not produce tumors. There have been isolated cases of tumors produced by needle-tracking of prostate cancer biopsies. In 1987, Haddad and Somain found 15 such cases of prostate tumors that had to have arisen following transperineal biopsies.

In 1991, Bastacky, Walsh and Epstein at Johns Hopkins found that tumor growth along the needle track was evident in the periprostatic soft tissue in 2% of the prostatectomy specimens they examined. Unlike earlier reports that only found needle tracking in transperineal biopsies of high-grade tumors, they found it in transrectal biopsies of Gleason 7 tumors as well. A recent literature review by Volanis et al. found 42 case reports of needle-tract seeding.

Another way that biopsies can potentially spread cancer is by release of isolated cells into systemic circulation. Tumor cells are less sticky than healthy cells, so they may be more easily dislodged by invasive procedures. In a recent study by Ladjevardi et al., the researchers looked at the peripheral blood (from their arms) of 38 men (23 patients with PC, 15 patients without PC) before and after prostate biopsy. They examined the blood for presence of epithelial cells that might have become dislodged by the biopsy. They found cellular material in 83% of the men who had PC, but only in 13% of the men who did not have PC. This does not mean that the epithelial cells were tumor cells, or if there were, that they were viable. The most viable kinds of tumor cells are mesenchymal rather than epithelial, but those were not searched for. It would be interesting to see this experiment repeated with CellSearch® or ADNA® technology, which can detect and distinguish circulating tumor cells.

Surgery

Some cancers are easily spread through inoculation by surgical instruments. For this reason, surgeons try to avoid cutting into the tumor. With unifocal tumors (e.g., breast cancer) the surgeon cuts a margin around the tumor. But with prostate cancer, where tumors are almost always multifocal and can be anywhere in the prostate, surgeons try to remove the entire prostate in one piece. Sometimes, surgeons slice through the tumor at the margin, leaving behind a positive surgical margin (PSM). Sometimes this is inevitable, but experienced surgeons typically have a lower PSM rate. At Johns Hopkins, for example, the PSM rate is as low as about 10%. The cancer left behind may continue to grow, may become non-viable after detachment, or may get cleaned up by the immune system. It is unknown at this time whether tumor cells detached by the cut at the PSM seed new tumors.

A similar effect may occur when an attempt is made to spare neurovascular bundles. In a study of 9,915 patients treated at Memorial Sloan Kettering and Ottawa Hospital from 1985-2010, 6% had prostate incision. Those who had bilateral nerve-sparing had incision rates over twice as high as those who did not have nerve-sparing surgery, after adjustment for confounders. Patients who had robotic surgery had incision rates almost twice as high as those who had open or laparoscopic surgery. Risk of prostate incision has decreased over time, presumably with surgeon’s experience.

Another difficulty arises where the surgeon must detach the prostate from the urethra, which runs right down the middle. The surgeon scrapes prostate tissue away from the urethra, and cuts it as far away as he can from the bladder neck on top, and the urethral sphincter, on the bottom. He then joins the two ends together, which is called an anastomosis. This procedure may leave cancerous tissue behind. In a recent CT/MRI study of post-prostatectomy tissue, 76% of recurrences after surgery were found to occur at the anastomosis. How many of those were from cancerous tissue that was left behind, and how many from contamination of the surgical blade?

Sometimes, especially with laparoscopic procedures on large prostates, the surgeon is forced to cut the prostate up into smaller pieces that he can remove through the port – a process known as morcellation. This may be especially risky for releasing cancer cells into systemic circulation. In April 2014, the FDA discouraged the use of morcellation on the uterus or uterine fibroid tumors because of the high risk of cancer spread associated with the process. Sometimes surgeons will recommend hormone therapy to their patients with especially large prostates in order to perform robotic surgery without morcellation. To my knowledge, there have been no studies of the effect of morcellation on prostate cancer spreading.

Spread of cancer at the laparoscopic port site is exceedingly rare. A 2004 study looked at 10,912 urologic laparoscopic procedures across 50 different treatment centers, and found only 10 cases of port seeding and 3 cases of peritoneal spread from the procedure. There have been only a handful of cases reported since then. Robotic laparoscopic surgery is responsible for only 3 documented cases of port site and/or peritoneal spread: one case in Japan, one case in Korea, and one case in Turkey.

Cancer cells may be released into systemic circulation by surgery. A study of circulating epithelial tumor cells in breast cancer patients found that the serum-detected cell numbers did increase in some patients following surgery, and the increase was sustained in some, indicating viability. A study of bladder cancer circulating tumor cells using CellSearch® found an increase following transurethral bladder resection. Eschwège et al. found increased numbers of prostate epithelial cells in the serum after surgery, but found no association with metastatic progression or survival. To my knowledge, there has not yet been a study specifically of circulating tumor cells pre- and post-prostatectomy.

There is not enough documented proof that the magnitude of cancer spread by surgery is large enough to be of concern. However, the potential for cancer spreading by poor surgical technique is one more reason to find the most experienced surgeon possible.

Low Dose Rate (LDR) Brachytherapy or “Seeds”

Implanting seeds is a highly invasive procedure, with 70 or more radioactive seeds injected into the prostate. In a Japanese study among 616 consecutive patients receiving LDR brachytherapy between 2003-2010, 5 patients had a pulmonary metastasis after clinical recurrence. Pulmonary metastases are rare, but they were hormone-responsive, which suggests a prostate cancer origin. The authors note that they may have been caused by seed migration to the lungs. All of those 5 had high Gleason scores, and only one had neoadjuvant hormone therapy.

High Dose Rate (HDR) Brachytherapy

Raleigh et al. at UCSF recently reported the first case of prostate cancer seeding following HDR brachytherapy treatment for a man with high risk PC treated with a combination of HDR brachytherapy and EBRT. The cancer recurred at the site where an HDR brachytherapy catheter was known to have touched the patient's bladder. The authors conclude:

"This case is the first report of prostate cancer recurrence in the bladder wall after brachytherapy and raises questions about prostate cancer biology, brachytherapy technique, and the timing of brachytherapy boost relative to whole pelvic radiotherapy for prostate cancer."
 

I hope that readers will not be dissuaded by these reports from seeking diagnostics and therapies they may be considering. There are risks with any invasive procedure, but it is important to keep the relative magnitude of those risks in perspective. I think these case studies are useful insofar as they are generative of hypotheses. It is clearly an area ripe for further scientific inquiry.