In this interview we discuss results of the HYBRID trial, which tested a hypofractionated regimen of 36 Gy over six fractions in elderly muscle-invasive bladder cancer patients who were ineligible for standard treatments.
Robert A. Huddart, MB, PhD, BS
In this interview with Robert A. Huddart, MB, PhD, BS, of the Institute of Cancer Research in the United Kingdom, we discuss results of the HYBRID trial, which tested a hypofractionated regimen of 36 Gy over six fractions in elderly muscle-invasive bladder cancer patients who were ineligible for standard treatment. Dr. Huddart presented results of the trial at the 2017 American Society for Radiation Oncology (ASTRO) Annual Meeting in San Diego.
-Interviewed by Ian Ingram
Cancer Network: Could you discuss some of the obstacles in treating bladder cancer?
Dr. Huddart: Bladder cancer is common in older patients, so when we’re treating patients with bladder cancer we have a challenge. A significant portion of patients are in their late 70s and 80s when they come through our clinic-often many of them have comorbidities related to the fact that they are smokers. This means they’re often not fit enough to tolerate treatments that we use standardly in the management of bladder cancer-chemotherapy, radiation therapy, radical surgery, radical radiotherapy, chemoradiotherapy. Their cancer is often contained within the bladder, so it’s a dilemma as to what to do and how to manage those patients.
One of the approaches we’ve been exploring is whether hypofractionated radiotherapy-giving one treatment a week-is a feasible approach. A previous trial from the Medical Research Council (MRC) in the United Kingdom showed that this approach was as good as doing daily treatment for the palliation of symptoms, but the actual biological dose used in that study was relatively low and the local control rate was relatively low, so we’re trying to find something that works a bit better.
We, among others, have explored the idea of doing 6 Gy fractions once a week over 6 weeks. There’s some retrospective data suggesting that this achieves quite good local control and is well tolerated. One goal of the HYBRID trial was to test this formally in a prospective multicenter format.
The second issue we have in treating bladder cancer is that the bladder itself can be quite mobile and deformable and can vary a lot from day to day-even when you’re trying to set everything up the same way-which has meant that, generally speaking, we’ve had to use large safety margins. This increases the treatment volume and risk of side effects and toxicity, which is obviously an issue for an elderly group of patients. Also, if you miss on one fraction that’s a significant portion (15%) of the treatment, so accuracy is important.
One route around this is image-guided radiotherapy using an adaptive approach we call “plan of the day.” With this approach we design-before the treatment starts-a series of plans where each day we can select the best plan to get the best coverage with a minimal amount of excess normal tissue. We’ve done some pilot studies that have shown that this reduces geographical miss as well as the treatment volume. But whether this results in a benefit to patients is still a question.
Cancer Network: How is HYBRID designed and what were you looking to discover?
Dr. Huddart: HYBRID is a non-comparative randomized phase II study where we took muscle-invasive bladder cancer patients (T2–T4aN0M0) who aren’t fit enough for standard radical treatment approaches and treated them with 36 Gy in six fractions, randomizing them 1:1 between standard planning or the adaptive “plan of the day” approach. In the standard arm the bladder was treated with a 1.5-cm safety margin. In the adaptive arm we made three plans-a medium plan that was a little bit smaller than the standard plan, a small plan with a 0.5-cm safety margin, and a large plan for days when the bladder is particularly large.
We’ve recruited 65 patients in just over 2 years at 12 sites in the United Kingdom and now have over 2 years of follow-up, which is the analysis we presented at ASTRO.
Cancer Network: What was the primary endpoint of the study, and what were the results?
Dr. Huddart: Most patients got through the treatment, with 90% receiving all six fractions. Among those who didn’t get all six fractions, most received five fractions. Slightly more patients stopped treatment for toxicity in the standard arm than in the adaptive arm (4 vs 2 patients), but this difference was not significant.
When we looked at the question of whether we needed to do adaptive planning, we found that 85% of patients used a plan other than medium plan at least once during the course of treatment, so it did seem like there was a need for this sort of adaptive approach. In total, 40% of the fractions used one of the adaptive plans-25% for the smaller plan and 15% the larger plan.
Toxicity assessment was the primary endpoint of the study. Because of the nature of the group of patients, we looked at acute toxicity (within the first 3 months) as the primary endpoint rather than long-term toxicity. It gets quite tricky-as these patients are old-so we were particularly looking at treatment-emergent adverse reactions (those directly related to radiation therapy) as the principal endpoint. In particular we were looking at non-genitourinary (GU) events, as that’s where we thought we might see a difference between the standard and adaptive plans.
We wanted to rule out a greater than 30% incidence of non-GU grade 3 events-quite a high level-and we achieved that. The rate of non-GU grade 3 toxicity was 6% in the adaptive arm (upper confidence was well below 30%) and 13% in the standard arm (upper confidence was below 30%). Looking at all events, the rate of grade 3 or greater toxicity was much higher. A number of patients died within this 3-month period-from sepsis, pneumonia, heart disease, and renal failure (one patient each)-but we don’t think those were directly related to treatment.
As a non-comparative study, we haven’t done any formal toxicity analysis between the two groups, but there was a suggestion in the data that more patients in the adaptive arm had grade 1 toxicity as their maximum toxicity compared to the standard group, and perhaps a slightly lower rate of grade 3 toxicity as well-that applies when looking at both GU or non-GU events. I find it quite striking that there were 10 non-GU grade 3 events in the standard arm but only 4 in the adaptive arm. We can’t test this for significance, but it is supportive that the adaptive planning might actually be doing something and is promising for future work.
Late toxicity rates were pretty low as well.
Cancer Network: What were some other important or interesting findings from the study?
Dr. Huddart: The other thing that we wanted to show was that the hypofractionated regimen of 36 Gy and six fractions, across both arms, actually achieves a good level of local control. We expected that the local control rate would be about 60% and wanted to rule out a rate lower than 40%.
Of the patients we were able to assess (47/65 patients), we had a local control rate of over 80%. Assessing this is a little bit tricky, because not every patient can have the ideal assessment modality, but that’s our best estimate. Even if you take all the patients that weren’t able to be assessed and assume they all failed, we get a local control rate of 58%-that’s still a pretty satisfactorily level of control.
We also looked at survival data. It’s quite difficult to know what you would expect from this elderly population, but we had over 80% of patients alive at 6 months post-treatment and a median overall survival of 18 months. At 2 years, over one-third of patients were still alive. Given that that these patients had muscle-invasive bladder cancer and are elderly, I think these are actually very promising results.
I think the data we have from this study-and from a previous pilot study from Royal Marsden that included 50 patients and showed very similar results-suggest that if you’ve got a patient in front of you who looks like they would be eligible for curative treatment, but for whatever reason (social, medical) are not fit enough to come every day for treatment, this regimen of 36 Gy over six weekly fractions is a reasonable treatment approach. The regimen was well tolerated, with some acute toxicity. The control rates were not that dissimilar from what you’d expect from a standard course of radiotherapy, and the survival is respectable. I think some form of image guidance will improve the quality of treatment. It then depends on your healthcare system whether you want to do the more adaptive approach, which needs more investment in terms of money and resources.
Cancer Network: Could you discuss the extra time involved in the adaptive treatment planning? How is it more resource intensive, and does it require more time from patients as well?
Dr. Huddart: In this study, we based our plans on a single CT scan, so the CT scan that the patient attends for planning is exactly the same as if they were receiving standard of care. Because the bladder is fairly round, a pretty straightforward conformal plan was used. The physics department and dosimetrist, however, are making three plans rather than a single plan. The dosimetrists who do the work with me say it probably takes twice as long-not three times as long-as they’re using the same type of beam arrangements and just making the Gy volumes different. So it does double that planning input.
In terms of the time for the patient, most of our work suggests you can select the plan in about 5 minutes. Most patients completed treatment in about 12–15 minutes. Given older patients take a bit of extra setting up, this is pretty reasonable and achievable in most radiation departments.
Cancer Network: Are there plans to continue following these patients or study this approach in other patient populations?
Dr. Huddart: The data we have here on adaptive planning is promising in terms of supporting that this approach may offer benefit for patients. A formal comparative study is needed, though, and I think this data should encourage us to carry on this approach in the future, in either this context or other contexts.
One of the reasons we want to prove that adaptive planning actually makes things better is because that time investment is required, but that will require further study. We have been back to the funders to ask whether they would support a larger continuation study, but unfortunately we have been unsuccessful so far. It’s disappointing that we’ve not yet been able to support that larger confirmatory study, but some of this data has only recently come through so there may still be a chance.