What is the role of liver SBRT in the setting of an excellent interventional radiology program?

An excellent question, there is certainly a significant role for a liver SBRT program in the setting of an excellent and talented interventional radiology group.  The important hurdle that must be overcome for this to take hold is for everyone to view this as a collaboration rather than a competition. 

Selection of modality for treatment of HCC and liver metastasis is particularly challenging as there is very limited comparative modality data available.  When treating small lesions in healthy patients in the first line setting all modalities look quite good (surgery, RFA/microwave, TACE, and SBRT), though with SBRT being the newcomer there is overall far less data on this in the first-line setting.  Population based comparisons (SEER and NCDB) do not offer good data on one modality versus another, as these are very influenced by the selection criteria of institutions for lesions that can typically be treated by many modalities (such that one would expect outcomes of a modality that is typically the second or third choice to look worse as there is something about cases that fall through to this modality which makes them unsuitable for other therapies, and thus likely more complicated).  As such one must wait for randomized data to offer definitive comparisons between modalities, and for the most part this is still off in the future (though there is a small Loma-Linda trial comparing protons to TACE for bridging in small-medium HCC demonstrating less toxicity and a trend to improved local control and PFS in the proton group, PMID 27084661). 

Given a very limited volume of high-quality comparative data we have chosen to focus on areas of collaboration. 

For HCC patients undergoing transplant evaluation there is some concern about track-seeding, and as such our program typically has avoided ablation in these cases.  TACE as the first-line standard, though several patients (typically with lesions on the liver periphery) will have feeding vessels from the gastric, phrenic, or other vascular supply which are identified by our interventional radiology providers at the time of angiogram, and rather than increase risk/complications by attempting to selectively embolize braches of these vessels these patients are referred for SBRT after initial TACE.

For patients undergoing y90, the use of post y90 PET scan (which can estimate a sense of post-y90 dose-delivery based on brehmsstrahlung emission detection) identifies a significant sub-set of patients who obtain only partial coverage from y90.  While this might be addressed in a second procedure the same factors which lead to partial dose delivery in the first procedure often prevail, and the correlation of a sense of y90 delivery dosimetry and the location of viable residual disease will identify patients who might benefit from additional treatment with SBRT (and also allow you as the radiation oncologist to understand risks and benefits/method of combining SBRT with previous y90 dose). 

For patients with HCC and macroscopic PVTT, Sorafenib (or now newer agents) and y90 are common first line therapies, though again post y90 dosimetry will often demonstrate difficulty dosing the gross extent of the tumor thrombus with radio-embolization (logical given lack of small vessels to capture microspheres here), and this offers another group of patients that might benefit from fractionated XRT or SBRT (depending upon target volume).  While this is a high-risk group of patients and must be selected carefully for aggressive therapy, there is a randomized trial of Sorafenib +/- TACE and EBRT (PMID 29543938) which demonstrates disease control and survival benefits to the local therapy (TACE and EBRT) containing group, including approximately 20% of patients who attain long-term disease control despite very advanced presentation, lending additional ground for multi-disciplinary collaboration. 

While this data focuses on HCC, similar though-processes can be applied to select metastasis and cholangiocarcinoma patients.  Additionally in early un-resectable peri-hilar cholangiocarcinoma (<3cm, no extra-hepatic disease) the Nebraska/Mayo neoadjuvant chemoRT and brachytherapy followed by transplant data offers potential for excellent outcomes through intense multi-disciplinary collaboration (PMID22504095). 

I think the key to establishing a successful radiation program in a center with an excellent interventional radiology program is ongoing communication with the IR group to help them understand the situations in which radiation might offer an outlet in the care of patients whose disease they may be struggling with, as no physician wants to repeat strategies which are not bringing success and there is always room for an alternative option.  That being said this does create a dilemma for the radiation literature, as data for heavily pre-treated patients will often be more of a reflection of the patients history and condition at the time of treatment as much as the efficacy of the treatment itself. 

Do you mean interventional oncology program? Remember that interventional radiology has re-branded themselves as interventional oncology.  I agree that the best way to build a program is to show up consistently at the multidisciplinary conference, interact with your colleagues and challenge them publicly about the evidence that they're basing their treatments on. IO papers don’t report local tumor control  it’s typically only progression free survival and overall survival.  Remember they don't have oncology training. SBRT and embolus treatments are very much complementary. The results with ablative thermal ablation for small metastases are excellent and it's more convenient and cost-effective than SBRT. However, it also does not have a survival benefit in general. Their complications tend to be disasters such as bilomas and abscesses when they happen, so the best opportunities to treat are the ones that they cannot treat which are near the biliary tree and near the confluence of the hepatic veins into the IVC. These are also the ones most likely to have a survival benefit because to prevent venous and biliary obstruction secondary to hepatic decompensation and failure.  Also I believe in general we have better local tumor control for the large solitary tumors with ablative doses. We can be more selective yet should be aggressive because there could be a survival benefit in some cases. There are many in our field that prescribe attenuated doses of SBRT that it is below and ablative threshold. I see people giving 9Gy x3, 5-8Gyx5 and similar doses.  These are not ablative and will result in tumor progression between 1 and 2 years. Is obviously a complex issue because there are differences in the a/b ratio between tumors. However, referring physicians will eventually stop sending the patients if they see tumors growing again. So IO will continue to creep into this area if we continue this practice. It makes no sense to me to treat  asymptomatic oligo metastatic disease with a palliative dose. In general, a BED of 100Gy is necessary to ablate solid tumors, the use of lower doses is just palliation of asymptomatic metastasis with no OS benefit. That is not a good way to persuade IO colleagues that your modality is as good as theirs. Sometimes that means you have to give more fractions in order to ablate the tumor without causing biliary stricture. Typically we would use of SBRT technique and give 6Gy x10 or 4.5Gy x15 in those cases.

Ablative doses of radiation for intrahepatic cholangiocarcinoma have 90-100% local tumor control, and MS not reached around 40mo in our hands. Interventional oncologists have no data anywhere close to that. They have reported numerous small studies with Y90 showing 12-15mo MS. So you should pick your battles but this would be one to  stand your ground on.

Also for HCC remember that there 11 randomized trials from Asia showing that lower dose 2D and 3D radiation therapy improves overall survival over TACE alone in the larger tumors. You can use this information to partner with your neighborhood interventional oncologist as well.

Agree with all of the above - SBRT should optimally be used​ to augment (not replace) existing treatment strategies in a healthy locoregional therapy program.  In the context of a planned transplant it is important to remember that TACE is primarily used as a bridging therapy.  Median time to local recurrence is often right in the window of the average time to transplant (i.e. pmid 29066120​ Fig 1).  There is clear biologic synergy between TACE and SBRT (chemoembolization effects are generally most robust in the central part of the tumor and least robust at the periphery).  As such, I think it is important to consider SBRT as an adjunctive therapy after TACE in the bridging context.  It improves local control and is generally safe is planned carefully (pmid 25186290​).  As an added benefit, if timed appropriately, lipiodol can serve as an excellent fiducial.