SBRT has non-inferior acute and late-term toxicity vs IMRT in two randomized clinical trials

(updated)

In October 2018, the American Society of Radiation Oncologists (ASTRO) strongly endorsed moderately hypofractionated IMRT (20/28 treatments) for primary radiation treatment (see this link). Since then, there has been another publication of a randomized clinical trial with ten years of follow-up (see this link).

The advantages for the patient are large: fewer visits than the conventional 38-44 treatments with a concomitant reduction in costs. Because there is now convincing proof that this can be accomplished without an increase in side effects and without loss of oncological effectiveness, there is no reason why any patient would suffer through the conventional regimen. The remaining question is whether the number of treatments (or fractions) can be reduced even further to only about 4 or 5. This kind of extreme hypofractionation is called stereotactic body radiation therapy or SBRT. This requires proof.

We have seen the results of a Scandinavian randomized clinical trial (RCT) that found that urinary, rectal, and sexual side effects were not inferior with extreme hypofractionation (see this link), and the oncological outcomes were about the same too (see this link).

Now two more RCTs have shown that the toxicity of SBRT is no worse than and possibly better than moderately hypofractionated or conventionally fractionated IMRT.

Van As et al. reported the acute toxicity results of the PACE-B RCT in the UK at the Genitourinary Conference of ASCO. 844 men with favorable risk prostate cancer were randomized to get SBRT (414 men) or conventionally fractionated/moderately hypofractionated  IMRT - "CFMHRT" (430 men). The qualifications were:

• localized, favorable risk prostate cancer (Gleason score ≤ 3+4, Stage T1 or T2, PSA ≤ 20 ng/ml)
• unsuitable for surgery or preferring radiation

The two groups were similar. The treatments were:

• SBRT: 36.25 Gy in 5 fractions over 1-2 weeks
• CFMHRT: 78 Gy in 39 fractions (conventional) or 62 Gy in 20 fractions (moderately hypofractionated)
• ADT was not permitted

At 12 weeks post-treatment, acute grade 2 or higher toxicity was:

• rectal: 10% for SBRT vs 12% for CFMHRT - difference was not statistically significant
• urinary: 23% for SBRT vs 27% for CFMRT - difference was not statistically significant

(Update 9/14/22)

Tree et al. reported the late-term toxicity results of the PACE-B RCT.

At 24 months post-treatment, the worst late-term grade 2 or higher toxicity (RTOG* criteria) was:

• rectal: 2% for SBRT vs 3% for CFMHRT - difference was not statistically significant
• Using CTCAE 4.0* criteria, patients treated on the CyberKnife platform had less toxicity (1%) vs CFMRT (4%) and were better off than patients treated with other linacs (5%)
• urinary: 3% for SBRT vs 2% for CFMRT - difference was not statistically significant
• CTCAE 4.0* urinary toxicity was worse vs. RTOG* urinary toxicity: 12% for SBRT vs 7% for CFMRT
• Patients treated on the CyberKnife platform had no difference in toxicity (6%) vs CFMRT (7%) and were much better off than patients treated in 5 treatments with other linacs (17%)
• Patient-evaluated (EPIC*) moderate/severe urinary bother was worse for SBRT (10%) than for CFMRT (5%)
• Grade 3 toxicity was <1% in all groups
• There was no difference in erectile dysfunction

By 24 months post-treatment, the cumulative incidence of late-term grade 2 or higher toxicity (RTOG* criteria) was:

• rectal: 8% for SBRT vs 8% for CFMHRT - difference was not statistically significant
• CTCAE 4.0* rectal toxicity was worse vs. RTOG* rectal toxicity: 12% for SBRT vs 12% for CFMRT
• urinary: 18% for SBRT vs 11% for CFMRT - difference was statistically significant
• CTCAE 4.0* urinary toxicity was worse vs. RTOG* urinary toxicity: 32% for SBRT vs 20% for CFMRT
• Increased urinary frequency was the type of urinary toxicity most often reported: 10% for SBRT vs 5% for CFMRT

*RTOG and CTCAE 4.0 have different criteria for physicians to evaluate toxicity. EPIC-26 is a questionnaire that patients fill out.

Patients treated on appropriate platforms in high-volume centers had equal or better outcomes. Toxicity was low.

 

(updated 9/30/23) After a median follow-up of 6 years of 874 predominantly (91%) intermediate-risk patients across 38 centers in the UK and Canada, van As et al. reported:

• 95% and 96% were free of biochemical (PSA) failure for SBRT and conventionally fractionated radiotherapy, respectively.
• Grade 2 or worse urinary toxicity was 5.5% and 3.2% (not significantly different) for SBRT and conventionally fractionated radiotherapy, respectively.
• Only 1 patient in each cohort had Grade 2 or worse rectal toxicity.

Poon et al. reported the one year late-term toxicity results of a RCT in Hong Kong. 64 low- and intermediate-risk patients were randomized to get SBRT (31 patients) or conventionally fractionated IMRT - "CFIMRT" (33 patients). The qualifications were: Stage T1 or T2, Gleason score ≤ 7, and PSA < 20 ng/ml.

The treatments were:

• SBRT: 36.25 Gy in 5 fractions over 2 weeks
• IMRT: 76 Gy in 38 fractions
• Intermediate risk patients could optionally have ADT before their radiation.

at 1 year post treatment:

• one grade 3 (serious) urinary side effect was reported in each arm
• rectal grade 1 (mild) or higher: 64% for SBRT vs 84% for CFIMRT - significantly different
• urinary grade 1 (mild) or higher: 93% for SBRT vs 100% for CFIMRT - not significantly different

It is too early to assess if there are any differences in oncological outcomes in these two RCTs.

ASTRO, ASCO, & AUA strongly endorse a shortened course of IMRT for primary therapy

It will come as no surprise to my readers that moderately hypofractionated IMRT (first-line radiation delivered in 20-26 treatments or fractions instead of the conventional 40-44 fractions) received strong endorsement from all of the major US organizations of physicians who treat prostate cancer. The American Society for Radiation Oncology (ASTRO), in collaboration with the American Society of Clinical Oncology (ASCO) and the American Urological Association (AUA) issued the new guidelines, which are also supported by the Society of Urologic Oncology (SUO), European Society for Radiotherapy & Oncology (ESTRO), and Royal Australian and New Zealand College of Radiologists.

A hypofractionation task force issued the new evidence-based guidelines. They divided their guidelines into two parts: (1) moderately hypofractionated IMRT (20-26 fractions); (2) ultrahypofractionated IMRT (4-5 fractions), usually called SBRT, SABR, SHARP, or CyberKnife (I will refer to it as SBRT). They strongly support moderate hypofractionation. They conditionally support SBRT, because of the moderate degree of evidence published by their cut-off date of March 31, 2017. They may revisit those guidelines after further review.

The following guidelines were strongly endorsed based on high quality evidence with strong consensus:

1A: Low risk men who refuse active surveillance should be offered moderately hypofractionated IMRT.

1B: Intermediate risk men should be offered moderately hypofractionated IMRT.

1C: High risk men should be offered moderately hypofractionated IMRT.

1D: Moderate hypofractionation should be offered regardless of patient age, comorbidity, anatomy, or urinary function. However, physicians should discuss the limited follow-up beyond five years for most existing RCTs evaluating moderate hypofractionation. *

1E: Men should be counseled about the small increased risk of acute gastrointestinal (GI) toxicity with moderate hypofractionation. Moderately hypofractionated EBRT has a similar risk of acute and late genitourinary (GU) and late GI toxicity compared to conventionally fractionated EBRT. However, physicians should discuss the limited follow-up beyond five years for most existing RCTs evaluating moderate hypofractionation.*

The following guidelines were strongly endorsed based on moderate quality evidence with strong consensus:

7A: Image guidance (e.g., fiducials, transponders, cone beam CT, etc.) should be used for both moderate hypofractionation and SBRT.†

8A: 3D-CRT should not be used with hypofractionation.§

The following guidelines were conditionally endorsed based on moderate quality evidence with strong consensus:

2A: 60 Gy in 20 fractions or 70 Gy in 28 fractions are suggested for moderate hypofractionation.

2B: No variation in treatment regimen by patient age, comorbidity, anatomy, or urinary function.

3A: Low risk men who refuse active surveillance should be offered SBRT

4A: The SBRT dose for low and intermediate risk men should be 35 Gy - 36.25 Gy in 5 fractions.**

4B: SBRT doses of 36.25 Gy in 5 fractions should not be exceeded outside of a clinical trial or registry.**

5A: At least two dose-volume constraint points for rectum and bladder should be used for moderate hypofractionation or SBRT: one at the high-dose end (near the total dose prescribed) and one in the mid-dose range (near the midpoint of the total dose).


The following guidelines were conditionally endorsed based on low quality evidence with strong consensus:

3B: Intermediate risk men should be offered SBRT, but should be encouraged to do so in a clinical trial or registry.**

3C: High risk men should be not be offered SBRT outside of a clinical trial or registry.

4C: Daily SBRT treatment is not recommended due to increased risk of toxicity.

5B: Normal dose/volume constraints used in the reference study should be adhered to for both moderate hypofractionation and SBRT


The following guideline was strongly endorsed based on low quality evidence with strong consensus:

6A: Planned target volume definition of the reference study should be adhered to for both moderately hypofractionated IMRT and SBRT.††


* While most of the hypofractionation trials did not report beyond 5 years of follow-up (see Table at this link), some did. The Archangeli et al. trial reported survival outcomes out to ten years. (I believe the guideline authors erred about this.) M.D. Anderson published an eight-year update after the close of the task force review. As we saw in our review of RTOG 0126, survival does not become a useful endpoint for perhaps 15-20 years for men with localized prostate cancer, and surrogate endpoints, such as 5-year recurrence-free survival or metastasis-free survival must be used instead. Kishan et al. proposed that for ultrahypofractionated regimens, 3-year PSA may be an excellent surrogate endpoint. The ProtecT clinical trial showed that adverse effects of radiation almost always show up in the first two years.

† For the disaster that can ensue when fiducials are not used with SBRT, see this link. The guidelines should state that intra-fractional motion tracking should be used with SBRT.

§ In the recently presented (not published in time for these guidelines) randomized clinical trial of ulrahypofractionated RT vs conventionally fractionated RT, they did use 3D-CRT in both arms. There was no difference in 5-year biochemical recurrence-free survival or 6-year toxicity.

** In a large, multi-institutional clinical trial (too late to make it into these guidelines), Meier et al. reported excellent 5-year oncological and toxicity outcomes using 40 Gy in 5 fractions. In SBRT dose escalation trials, both Zimmerman at UT Southwestern (reported here) and Zelefsky at MSKCC (I've heard from his patients) found that 45 Gy in 5 fractions gave excellent oncological and toxicity outcomes. The task force neglected the fact that prescribed doses are reported differently by different ROs. Alan Katz, for example, reports a prescribed dose of 35 Gy to the planned target volume (the prostate plus the margin around it), but the clinical target volume (the prostate itself) gets about 38 Gy, while the margin gets considerably less.

†† Smaller margins are possible when fiducials are used for intra-fractional tracking. Tighter margins cause less toxicity to organs at risk.

Sadly, the effect of hypofractionation on erectile function was seldom reported and was not part of the task force's analysis.

It is worth noting that conventionally fractionated IMRT became the standard of care without any comparative clinical trials. The longest running single institution dose-escalated IMRT trial (at MSKCC) had 10 years of follow-up on a small sample size (n=170). By contrast, Alan Katz is expected to report 10-year SBRT outcomes this year on 515 patients. The task force is holding SBRT to a higher standard that by this time next year, it should have the published results to meet.

While the task force endorsed moderate hypofractionation, we will have to see whether radiation oncologists (ROs) follow their guidelines. Because ROs are reimbursed by the number of fractions they give, they will be understandably reluctant to reduce the number of fractions. It remains to be seen whether insurance companies will enforce a limit. It is a clear benefit to the patient in terms of convenience and cost.

Eighth randomized clinical trial of hypofractionated radiation therapy

We now have an eighth randomized clinical trial of hypofractionated radiation therapy. There are no surprises: it showed that oncological and toxicity outcomes were not significantly different between the two regimens. We last looked at it here. This trial is unusual because of the length of follow-up.

Arcangeli et al. report the 10-year outcomes of their study covering 168 high risk patients treated using 3D-CRT (not IMRT) at the Regina Elena Cancer Institute in Rome between 2003 and 2007. The details of the treatments were as follows:

• Half (85 patients) received conventionally fractionated (Conv)  80 Gy in 40 fractions
• Half (83 patients) received hypofractionated (Hypo) 62 Gy in 20 fractions

After a median of 9 years of follow-up:

• 10-year freedom from biochemical failure was 72% for the Hypo group vs. 65% for the Conv group.(no statistically significant difference)
• 10-year prostate cancer -specific survival was 95% for the Hypo group vs. 88% for the Conv group (no statistically significant difference)
• 10-year overall survival was 75% for the Hypo group vs. 64% for the Conv group (no statistically significant difference)
• Hypofractionation was a significant variable in determining prostate cancer-specific survival in multivariate analysis
• There were no differences in late-term grade 2 or higher urinary or rectal toxicity between the 2 groups.

There are a couple of caveats. For those who insist on rigorous analysis, the Hypo group had worse oncological and toxicity outcomes on an intention-to-treat basis. It was only after the patients were analyzed according to the treatment they actually received that the lack of statistically significant difference became apparent. James Yu, in an accompanying editorial, points out that blood in urine was 16.5% for the Hypo group vs. 3.6% for the Conv group. This may be a caution that hypofractionation should not be attempted using 3D-CRT. In the US, where IMRT is widely available, this should not be an issue.

Here's the table summarizing all 8 randomized clinical trials:

Randomized Clinical Trial	Risk Groups	Fractionation	5-yr bPFS	Urinary toxicity Grade 2+	Rectal toxicity Grade 2+	Ref.
PROFIT	100% intermediate	60 Gy/20fx 78 Gy/39fx	85% 85%	22% 21%	8% 14%	1
Fox Chase	67% Intermediate, 33% high	70.2 Gy/26fx 76 Gy/38fx	77% 79%	22% 13%	18% 23%	2
CHHiP	73% intermediate, 15% low, 12% high	60 Gy/20fx 74 Gy/37fx	91% 88%	12% 9%	12% 14%	3
MD Anderson	71% intermediate, 28% low, 1% high	72 Gy/30fx 75.6 Gy/42fx	89%† 85%†	16% 17%	10% 5%	4
RTOG 0415	100% low risk	70 Gy/28fx 73.8 Gy/41fx	94% 92%	30% 23%	22% 14%	5
HYPRO	>70% high, <30% intermediate	64.6 Gy/19fx 78 Gy/39fx	81% 77%	41% 39%	22% 18%	6, 7
Cleveland Clinic	49% low, 51% intermediate	70 Gy/28fx 78 Gy/39fx	94% 88%	1% 2%	5% 12%	8
Regina Elena	100% high risk	62 Gy/20 fx 80 Gy/40 fx	72%* 65%*	21% 14%	NA NA	9

*10-year figures for the Regina Elena trial
† 8-yr failure-free survival update for MD Anderson

No need to go through 38-44 treatments with IMRT anymore

There have been several hypofractionation trials maturing in the last couple of years. With minor exception, they all tell the same story: external beam radiation therapy (EBRT) can be completed in less time without loss of efficacy or increase in toxicity. Hypofractionation means completing EBRT in fewer treatments or fractions using higher doses per fraction.

Catton et al. now report the 5-year outcomes of a multi-institutional, multinational (27 centers in Canada, Australia and France) randomized clinical trial (called the “PROFIT” trial) among 1,206 intermediate-risk patients treated from 2006 to 2011. All patients received radiation doses now considered curative: 78 Gy in 39 fractions (conventional fractionation - CFN) or 60 Gy in 20 fractions (hypofractionation - HFN). The doses are biologically equivalent for cancer control, no ADT was allowed. After median follow-up of 6.0 years:

• 5-year freedom from biochemical or clinical failure was 85% in both groups
• Acute urinary toxicity, grade 2: 27% in both groups; grade 3:4% in both groups 
• Acute rectal toxicity, grade 2: 16% for HFN*, 10% for CFN; grade 3: <1% in both groups 
• Late term urinary toxicity, grade 2: 20% for HFN, 19% for CFN; grade 3+:2% for HFN, 3% for CFN Late term rectal toxicity, grade 2: 7% for HFN, 11% for CFN*; grade 3+: 1% for HFN, 3% for CFN 

*Difference between arms was statistically significant, but not meaningful

The table below summarizes the key oncological and late-term toxicity outcomes of the various hypofractionation trials:

Randomized Clinical Trial	Risk Groups	Fractionation	5-yr bPFS	Urinary toxicity Grade 2+	Rectal toxicity Grade 2+	Ref.
PROFIT	100% intermediate	60 Gy/20fx 78 Gy/39fx	85% 85%	22% 21%	8% 14%	1
Fox Chase	67% Intermediate, 33% high	70.2 Gy/26fx 76 Gy/38fx	77% 79%	22% 13%	18% 23%	2
CHHiP	73% intermediate, 15% low, 12% high	60 Gy/20fx 74 Gy/37fx	91% 88%	12% 9%	12% 14%	3
MD Anderson	71% intermediate, 28% low, 1% high	72 Gy/30fx 75.6 Gy/42fx	96% 92%	16% 17%	10% 5%	4
RTOG 0415	100% low risk	70 Gy/28fx 73.8 Gy/41fx	94% 92%	30% 23%	22% 14%	5
HYPRO	>70% high, <30% intermediate	64.6 Gy/19fx 78 Gy/39fx	81% 77%	41% 39%	22% 18%	6, 7
Cleveland Clinic	49% low, 51% intermediate	70 Gy/28fx 78 Gy/39fx	94% 88%	1% 2%	5% 12%	8

Hypofractionation has demonstrated equal efficacy and side effects compared to conventional fractionation. Hypofractionation requires greater care on the part of the radiation oncologist. He must use advanced image guidance with placement of fiducials or radio transponders and localization with cone beam CT, set tighter margins, lower dose constraints for organs at risk, assure adequate bladder filling and lack of bowel distension at each treatment, use fused MRI/CT images if possible, and have very rapid linacs to minimize intrafractional motion. With this much cumulative level 1 evidence, it is hard to justify the use of conventionally fractionated EBRT anymore. Patients should not have to endure 38 or more treatments, and pay the extra cost of that, even if insurance or Medicare is willing to pay. Patients should shop for radiation oncologists who have experience with hypofractionation, or preferably, with extreme hypofractionation (SBRT).