ASTRO 2015
Individualised stereotactic body radiotherapy of liver metastases
Dr Theodore Hong - Massachusetts General Hospital, Boston, USA
We’re presenting on a prospective phase II study evaluating proton-based SBRT for liver metastases. The general idea is that we know that liver metastases can be well treated with SBRT but there’s always an issue of can you get adequate dose due to the exit dose from standard photon-based SBRT. A lot of people have speculated that protons would allow for a higher dose of radiation; we wanted to see if that would then lead to improved local control.
Could you tell us about the proton-based SBRT technique used?
What we do is a proton-based SBRT. So we treat in five fractions using a respiratory guided technique, double passively scattered protons. What that means is we’re using conventional proton therapy but with respiratory management to treat in these five high dose fractions. The way that we prescribe the radiation dose is based on the dose that the non-involved liver receives so because of the lack of exit dose from the protons it generally allows us to give a high dose of radiation.
What was the study design and what methods were used?
It was a single arm phase II study; it allowed for any patient with a solid tumour primary with metastasis to the liver, up to four liver metastases. 700ml of the liver had to be uninvolved to be enrolled on the trial and the extra-hepatic disease had to be either controlled or non-existent and stable for at least three months.
What were the main findings of the phase II study?
We found that we were trying to evaluate if the one year local control would exceed 70% and in fact it did; the one year local control exceeded 75% so we were quite pleased with that. Additionally we did find that the average dose of radiation that we were able to deliver with this strategy was higher than we would have expected with a photon-based strategy, recognising that it’s not a randomised trial. However, the most interesting finding is we did have a number of local failures as a number would indicate, and when we looked at characteristics including the dose of radiation we delivered, the primary tumour location or the size of tumour, none of those things predicted an increased risk of local failure. What ended up predicting an increased risk of local failure was actually if the cancer of origin was KRAS mutant or not. So it was very interesting for us to have a potential biomarker that I don’t think is proton specific but suggests that maybe we need to be looking beyond clinical characteristics in following the trend of medical oncology and looking at the genetic mutation in terms of whether or not we will have success with an SBRT based approach.
Is the association between the KRAS mutation and local failure something that is new to radiation oncology?
People have reported this, there’s a study from Ray Mak that looked at this issue in non-small cell lung cancer and SBRT where he found that KRAS mutant tumours were harder to control than KRAS wildtype. I’ve done studies at Mass. General suggesting that patients with rectal cancer with KRAS mutant tumours are far less likely to have a pathological complete response to neoadjuvant chemoradiation compared to KRAS wildtype tumours. So this has very much been an interest of ours and that’s why we looked at it. There’s actually a long laboratory history suggesting that KRAS mutant tumours are substantially more radiation resistant than KRAS wildtype tumours but to our knowledge this is the first prospective dataset to evaluate this in the setting of liver SBRT.
What is the next step for building on these data?
That’s a topic of substantial conversation, as you can imagine. In my heart I don’t believe that simple dose escalation with radiation is going to be the answer and I think that we are going to need to look at ways to modify the radiation with biologic agents to try to restore the radiosensitivity of KRAS mutant tumours. The doses that we gave are quite high and the doses certainly that they give in inoperable lung cancer are extremely high. If we can’t control it with a high percentage rate with these types of doses, it does make me think that the next step is a radiation modifier to specifically alter the natural history of KRAS mutant tumours.
Is there a particular radiation modifier that you might try to look at?
A lot of people have looked at this. We have looked at it as well and I currently actually have a rectal study that’s using an analogue to restore radiation sensitivity to KRAS mutant rectal cancer but we’re still quite early in that, we’ve just completed the phase I portion and we’re on the dose expansion of that. So I don’t know if that strategy is going to work. It’s also unclear whether simply using classic cytotoxic chemotherapies with the radiation or even DNA damaging agents like PARP inhibitors would be an effective strategy. This is an area that is certainly ripe for investigation and I’m collaborating with my colleagues at Mass. General in trying to identify candidate drugs to combine with radiation for KRAS mutant tumours.
Is there anything else on the study that you would like to highlight?
A lot of people come looking for proton therapy and a lot of that is going to be amplified by a paper that will be coming out in The Journal of Clinical Oncology in the next couple of months where there was a separate phase II study evaluating protons in primary liver cancer, so either hepatocellular carcinoma or cholangiocarcinoma where we demonstrate a two year local control rate exceeding 90% for both HCC or intrahepatic cholangiocarcinoma. When patients see that data I think that it’s a very natural question to say shouldn’t protons then work for my metastatic tumour that has gone to the liver? Our local control is substantially lower in the metastatic population than it was in the primary tumour population but one important difference is that in the primary tumours we treat with fifteen fractions as opposed to five fractions with liver metastases. So it’s unclear to me still whether or not there’s a difference in terms of just the radiosensitivity of those other tumours versus metastatic tumours. It’s also unclear to me what role the fractionation schedule we use plays as well. But certainly I would say that our liver metastasis study is hypothesis generating. I don’t personally feel that there’s enough there for me to move to a randomised trial of proton- versus photon-based SBRT at this time but rather we’re interested in following up on the KRAS observation. In contrast, in hepatocellular carcinoma our next step is actually a proton versus photon randomised trial because of the high level of local control that we’ve seen compared to historical controls with photon-based radiation. I think that will be a point of confusion for a lot of people wondering if proton therapy is right for them.
What is your take-home message from the study?
We’re able to demonstrate that proton-based SBRT for liver metastases are feasible and we achieved our local control goal and it was extremely well tolerated. The KRAS story is yet to be seen.
