Hypofractionated radiation therapy with IMRT: same results in less time

The largest-yet randomized clinical trial comparing hypofractionated (fewer treatments or fractions) to normally fractionated intensity-modulated radiation therapy (IMRT) has proved that oncological outcomes and late-term toxicities were the same for both treatment schedules.

The results of the CHHiP study were reported in a presentation by Dearnaly et al. delivered at the European Cancer Congress this week (click here for the abstract). There was also an interim toxicity analysis in 2012. There were 3,216 patients treated at 71 centers in the UK between 2002 and 2011. All patients were stage T1b to T3a, with < 30 percent probability of seminal vesicle involvement. Patients were randomly assigned to one of three IMRT treatment schedules, for which I also show the relative biologically effective dose (BED) for oncological control compared to normal fractionation:

  1. 74 Gy =2 Gy × 37 fractions (normal fractionation)
  2. 60 Gy = 3 Gy × 20 fractions; relative BED: +4 percent
  3. 57 Gy = 3 Gy × 19 fractions; relative BED: −1 percent

ADT began 3 months prior to the start of IMRT, and continued through treatment. The 2012 analysis showed that 93 percent of patients in each group received ADT.

Patient characteristics were as follows:

  • NCCN risk groups:
    • Low risk: 15 percent
    • Intermediate risk: 73 percent
    • High risk: 12 percent
  • Median age: 69 years
  • PSA: 10.1 ng/ml

After a median follow-up of 63.2 months, the 5-year progression-free (either biochemical or clinical) survival (PFS) rates were:

  • 74 Gy: 88 percent
  • 60 Gy: 91 percent
  • 57 Gy: 85 percent

The difference in PFS for the 57 Gy vs. the 60 Gy schedule was statistically significant; the other differences were not statistically significant.

The toxicity outcomes reported as those with RTOG toxicity grades of 2 or higher were as follows:

  • Acute gastrointestinal (GI) toxicity was lower with normal fractionation:
    • 74 Gy: 25 percent
    • 60 Gy: 39 percent
    • 57 Gy: 38 percent
  • Acute genitourinary (GU) toxicity was not significantly different among groups.
  • 2-year GI toxicity was lower with hypofractionation in the 57 Gy group:
    • 74 Gy: 4 percent
    • 60 Gy: 3 percent
    • 57 Gy: 2 percent
  • 5-year GI toxicity was not significantly different among groups.
    • 74 Gy: 1 percent
    • 60 Gy: 2 percent
    • 57 Gy: 2 percent
  • Neither 2-year nor 5-year GU toxicities were different among groups.

It is important to note that the normal fractionation schedule used in this trial (2 Gy × 37 fractions) is low compared to the current standard of care (2 Gy × 40 fractions), but was standard when this trial began in 2002. The 60 Gy schedule comes close to the current standard of care in terms of its biologically effective dose. Given this, it is not surprising that only the 60 Gy schedule achieved 5-year progression-free survival levels over 90 percent. The lowest dose schedule is on the steep part of the dose/response curve where even small increases in dose achieve large increases in cancer control.

While acute GI toxicity was higher at first with hypofractionation, the effect was transient, and had disappeared by 2 years. Lasting GI toxicity was negligible, and there were no differences at any time in GU toxicity.

Based on all this, the authors state, “Modest hypofractionated RT using 60 Gy/20 f appears effective and safe and may be recommended as a new standard of care.”

We should be clear that this is not SBRT; it is only IMRT with an accelerated dosing schedule, and there are some important differences. SBRT typically uses doses of 6 to 8 Gy per fraction and just four or five fractions. Because of the extreme hypofractionation, it becomes critical to track prostate motion during each fraction and not just between fractions. Treatment margins are typically narrower with SBRT and may be as low as zero on the rectal side. These differences are what make SBRT safe. In the current study, there was no allowance made for intra-fractional prostate motion and the margins were not altered, so it is not very surprising that rectal toxicity was higher at first, but it was perhaps surprising that there were no lasting differences.

We should also note a few similar randomized comparative trials in the last year:

  • A study at Fox Chase Cancer Center (n = 333), looked at 76 Gy delivered in 38 fractions of 2 Gy each (normal fractionation) compared to 70 Gy delivered in 26 fractions of 2.7 Gy each (hypofractionated) among intermediate- and high-risk patients. The 5-year biochemical and/or clinical disease failure rate was the same — 21 percent for normal fractionation, 23 percent for hypofractionation — and there was no difference in late term toxicity, except among men with compromised urinary function.
  • An M. D. Anderson Cancer Center study (n = 203) compared the late toxicity of a normally fractionated schedule (76 Gy in 1.8 Gy fractions) to a hypofractionated schedule (72 Gy in 2.4 Gy fractions). As in the UK study, there were no differences in GU toxicity. There was an increase in GI toxicity in the hypofractionated group, although it was not statistically significant. Unsurprisingly, the authors found it was related to the volume of the rectum that received high doses.
  • A multi-institutional study from the Netherlands among intermediate- and high-risk men (n = 820) reported on the acute toxicity of a normally fractionated schedule (78 Gy in 2 Gy fractions) compared to a hypofractionated schedule (65 Gy in 3.4 Gy fractions). At 3 months, there was no difference in GI or GU toxicity. At 6 months, there was no difference in GU toxicity, but GI toxicity was higher in the hypofractionated group (42 percent vs. 31 percent).

We also saw recently (see Can salvage radiation therapy be safely and effectively completed in less time?) that a shortened treatment schedule appeared to be safe and effective for salvage IMRT. However, this new UK study is more compelling because it is a randomized comparative trial of great size.

The only impediment seems to be the higher rate of acute rectal side effects. The patient will have to decide if it is worth accepting those transient symptoms in exchange for the convenience of a 4-week treatment schedule. Given the lower rate of patient-reported adverse outcomes and the high rate of oncological control with SBRT, however, it would seem that the 10-day schedule (five fractions, every other day) is a better alternative on all counts.

Editorial note: This commentary was written for The “New” Prostate Cancer InfoLink by Allen Edel.








15 Responses

  1. Your sitemaster would add that, at least in his opinion, the results of the CHHiP study are clearly “game changing” in the use of radiation therapy for the treatment of localized prostate cancer.

    Quite how fast the results of this study will be accepted by the US radiation oncology community is a whole other issue. If you are only giving radiation therapy for 4 weeks as opposed to 8 weeks, payers may be of the opinion that they should be paying less for the services of the radiation oncologist and his team.

  2. It is important to recognize that the study still has no mortality outcomes, and is unlikely to be stopped until these are available — meaning it is still too early to know if the mortality rates are comparable. An important randomized radiation study from New Zealand some years ago found that one arm of the therapy showed better biochemical/failure results but had inferior mortality.

  3. Thanks Allen for a great review of this important research! Bravo! While the whole review was helpful, I particularly appreciated the discussion of the difference in the dose/response curve between the two hypofractionated groups, an area that is obscure to many of us.

    I too was struck by the likely further improvement over these already impressive results that modern imaging would enable. Back in the early 2000s when many of these men would have been treated (start of study 2002), imaging to support radiation was pretty crude compared to what we have now, a little more than a decade later. It wasn’t just the lack of imaging during the session (“intrafraction”), but also there was likely a lack of imaging on a daily basis (“interfraction”), with reliance back then just on the somewhat crude imaging used for planning the whole course of treatment. The turning point at centers of excellence in the US was likely centered in the period from 2003 to 2006. I’m basing that on talks given by Dr. Michael Dattoli to the national conferences of patients. In 2003, he was still leaning on color Doppler as the most advanced imaging, perhaps with some daily CT support. By 2005 he was discussing his use of interfraction imaging, and by 2006 he was emphasizing the importance of such imaging and discussing intrafraction imaging. My impression is that imaging each day during the course of treatment to check for changes in target location is now typical, at least at major centers, with systems like CyberKnife using intrafraction imaging to make sure each of the handful of fractions used is well targeted. I suspect the centers in this study are now using advanced imaging.

    These results are especially impressive at the important 5-year point for median follow-up because the study population was strongly skewed toward the higher risk side: 85% of patients were intermediate or high risk. That means the impressive figures for progression-free survival were not artificially raised by having a large portion of low-risk patients, many of whom would have likely not needed any treatment. Also, for radiation patients, results at 5 years in many studies seem fairly robust as time goes by, so we can have a good measure of confidence in these high levels of success. I’m also impressed that 71 centers were involved, which suggests that the results are a fairly typical level of success rather than success that is achievable only at a few centers of excellence.

    While not new, the low levels of long-term GI toxicity for all three groups should add to confidence for patients. As for the acute toxicity, having gone through that myself, it’s not fun but quite tolerable for the typically short term. To me it was more a nuisance than a burden.

  4. Whether a Game Changer

    Thanks for your comment, Sitemaster.

    My respected local hospital has been moving steadily toward hypofractionation, even using CyberKnife SBRT for some patients. They were not the first in the Washington, DC, metro area. This institution is somewhat on the reasonably conservative side, though keeping well aware of the state of the art and innovation, so I take their adoption of hypofractionation as a sign that many US centers will be moving in this direction.

    If I were now choosing a treatment, I would be glad to see this study, and I believe I would want the hypofractionated course rather than the regular fractionation. That said, I’d like to see a few more years of follow-up with the five session program.

    As for decreased reimbursement, that seems likely, but it may not be a problem for adoption of hypofractionation as staff time should decrease substantially, making the “product” less expensive to deliver. I realize that producing the radiation itself is a big cost item, but the treatment is also staff intensive, so there may be some large cost savings that can be passed on.

  5. The primary outcome measures were toxicity and freedom from recurrence. It is important in studies like this that such surrogate endpoints be used. Considering that most of the men were intermediate risk, it could take 20 years before differences in cause-specific survival are detectable, and most of the men in the study would be lost to other causes. Also by then, the results will be irrelevant because of technological changes and better understanding of radiobiology.

  6. Just curious about the cost of a radiation session and its length

    I’m wondering how the cost of an image-guided external beam radiation session relates to the length of the session, as I understand hypofractionated sessions tend to be longer to deliver a greater dose in the session. Obviously the one-time set-up charges per session, such as time to prep the patient and do imaging would not vary with the dosage, but the costs for the radiation itself and for staff time should be greater when the dose is greater. Are there any rules of thumb for the different types of radiation that would be understandable to us laymen?

  7. Dear Jim:

    There is very little logic to any of this. The price for exactly the same form of radiation treatment can vary from state to state, from institution to institution, and from provider to provider depending upon all sorts of factors (such as who is actually paying). This is true even for Medicare, with different reimbursement levels in different zip codes.

  8. Jim,

    ROs are reimbursed by the number of sessions, not by the length of each. The length of each session varies by the kind of machine, and not very much by the amount of radiation delivered. So, for example, the Varian Truebeam with RapidArc machine I was treated with delivered 8 Gy in about 5 minutes. Most of that time was taken for calibration of fiducials and the arc swing of the gantry. By contrast, a step-and-shoot IMRT can take 30 minutes to deliver 2 Gy. Radiation oncologists like faster machines so they can have more billable sessions per day, but patients prefer it as well, and toxicity is reduced by speed, making everyone happy.

  9. Thanks to both of you. I’m amazed that 8 Gy can be delivered so quickly! My own TomoTherapy sessions took about 15 minutes of time on the table to deliver 2 Gy, and much of that was for checking the pre-delivery CT and making minor adjustments.

  10. Interesting study. Too early to deduce much apart from 60/20 is a reasonable treatment. Any failures in the first 5 years are likely to be predominantly from occult metastasis at time of treatment. Results at 10 years will give much more information but ultimately hypofractionation is being compared with 74/37 which does not appear to be as good as 78 or 80 and is used by very few centres. In addition the 74 Gy in this study was prescribed to the prostate plus a 0.5 cm margin not the PTV as would be standard now. Therefore the dose hypofractionation is being compared to is somewhere closer to 72 Gy, barely dose escalated.

    To call is a “gamechanger” is hyperbole. If you want a shortened treatment time and are prepared to accept a possibility of worse long-term control from all available evidence then this study supports this approach.

  11. Prostate,

    All of the treatment schedules had approximately equal BEDs — the 60/20 was approximately 4% higher, the 57/39 was 1% lower. Thus all the schedules were lower than what would be given now. Clearly, the cancer control will be better across the board — hypofractionated or normal fractionated — with dose escalation, but that is irrelevant in a comparative study like this one. So the fact remains that the 60/20 schedule gave approximately the same cancer control as the normally fractionated schedule. So why would anyone go for the longer treatment schedule?

    I have no idea why you assume that all failures in the first 5 years were occult and later failures would be local — there are no studies to support such a hypothesis.

    In this study, the prostate + 0.5 cm received 74 Gy, with an additional 0.5 cm receiving 71 Gy, and 59 Gy to the base + 1 cm around the seminal vesicles. Although the doses are now higher, this is very similar to PTV definitions now in use for IMRT, and were used for all schedules in this study. It was not 72 Gy, as you assert.

  12. Allen,

    Thanks for your comment.

    (1) The study, as you mention, was set up to compare similar BED regimes. The point is 74/37 would be considered low now by most groups. The study does not support moving from, for example, 78/39 to 60/20. You might make an argument to move to 62/20 or 63/21 but that has potential issues for toxicity.

    (2) IGRT in this study was not of sufficient quality to compare to modern IMRT with a 0.5 cm margin. Due to geographical miss there will be a drop off giving a lower dose giving around 72 equivalent — although this is a rough estimate.

    As I say it’s a good, interesting study. It doesn’t show 60/20 should be the new standard of care.

  13. Prostate,

    As we raise dose rates, normal fractionated and hypofractionated both, we risk increasing toxicity. Taking advantage of the low alpha/beta ratio of prostate cancer, and the large differences in the alpha/beta ratio between early-responding healthy tissue and prostate cancer tissue, hypofractionation offers equal cancer control without raising toxicity. That’s what was proven in this study.

    This study used state-of-the-art IGRT techniques at the time to deliver fractions accurately to within 3 mm. I know machines now can reduce that error to within 1 mm — but how does that change the finding? It really doesn’t matter that we have improved methods now — that will always be true — what matters is that the treatment groups were all given the same treatment except for the treatment schedule. This was not a dose escalation trial, and it was not a trial of alternative IGRT techniques. The only thing varied was the fractionation schedule.

    The conclusion is unequivocal — given the same method of treatment, the patient is better off with the shorter treatment schedule.

  14. Thanks for your comment Allen, I think we largely agree the findings but disagree on interpretation.

    The point I keep making is that this study doesn’t show equivalence between 60/20 and any dose greater than 74/37 at 5 years follow up. Patients should not be encouraged to have this hypofractionation regime in place of higher standard fractionation doses unless they understand that, based on available evidence, they may be trading convenience for an inferior outcome.

    What would you advise someone offered 80/40 or 60/20?

  15. Dear Prostate:

    I am the one who stated that these results are (in my humble opinion) “game changing”, not Allen.

    The reason I stated that they were “game changing” is that — because of the sheer size of the study — they call into further, very serious question the historic belief that widely used forms of external beam radiation therapy still need to be given over a 7- to 8-week period (which is a serious inconvenience for the majority of patients) as opposed to a shorter period. That was certainly appropriate prior to the availability of sophisticated systems to target radiation therapy with a high degree of accuracy, but we do now have much better and more sophisticated systems, as I know you appreciate.

    You are raising a different issue, which is whether these data can be extrapolated to the use of shorter treatment times for men being given doses > 60 Gy. That is a perfectly reasonable question, and it needs an answer. I would hope that the radiation oncology community would set out to answer it as soon and as fast as possible. But we already have the data from M. D. Anderson suggesting that we can shorten the time frame for delivery of 72 Gy.

    In my view, the real question that we don’t have a good answer for yet, based on this paper and other available data is not, “What would you advise someone offered 80/40 or 60/20?” Rather, it is, “What would you advise someone today offered 80/40 or something like 72/30?” based on the additional data from the M. D. Anderson study. We all recognize that higher dose levels are known to be associated with better outcomes … but do we still need to take 8 weeks to deliver those dose levels?

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: