Hotly debated across the Society of Interventional Oncology (SIO; 25–29 January 2024, Long Beach, USA) annual conference programme, speakers took to the stage to contest the survival of transarterial chemoembolization (TACE) in the age of yttrium-90 (Y90) transarterial radioembolization (TARE) and immunotherapy, their compelling arguments making some audience members “more confused and others happy”.
Riad Salem (Northwestern University, Chicago, USA) began by iterating that TACE will undoubtedly remain the global standard in terms of availability and feasibility. A better, “spicier” question, he posed, is what happens when both TACE and Y90 are widely available. “In reality TACE is already being replaced by Y90, based on curative data and therapeutic narratives,” Salem averred.
Despite his pro-Y90 position, Salem acknowledged that he is a “big fan” of TACE and has conducted the two largest TACE trials for hepatocellular carcinoma (HCC) in the USA—the first in 2010, and the second in 2016, evaluating response, toxicity and overall survival of TACE for HCC—alongside his mentor Michael Soulen (Abramson Cancer Center, Philadelphia, USA).
Yet, aside from his own contributions to the literature, Salem scrutinised the level of data that has solidified TACE in practice today, regarding the two seminal 2002 trials which found TACE to be beneficial in patients with limited disease and without portal vein thrombosis (PVT), when compared to no treatment in a small sample size randomised controlled trial (RCT). However, when compared to larger analyses of Y90, Salem highlighted that Y90 outperforms chemoembolization in four variables, namely: reproducibility/standardisation, precision, intentional threshold dosimetry, and clinical benefit.
A new narrative
Presenting data from individual cases of his colleagues and well-known trials to date, Salem attested to the “compelling” narrative that is being created by Y90, that provides high-risk patients with the same survival rates as low-risk patients, as well as improvements in quality-of-life measurements. “Who cares about quality of life?” he asked the audience. “Patients care about quality of life. They want to feel good, they want fewer treatments, they don’t want to feel sick”.
“You have the Swiss army knife of locoregional therapy. That’s the way I see it”, Salem continued, asserting that when reviewing the “scorecard”, Y90 far outperforms chemoembolization in areas such as downstaging, bridging, PVT management and immune therapies, but lacks the global accessibility of chemoembolization due to the greater infrastructure needs of a Y90 programme.
With a focus on patient experience, Salem wrapped up his presentation with a frank appeal to the “pragmatists” in the audience. He called to those that “actually have to talk to patients and manage humans” in the hopes that his argument has resonated, while acknowledging the “purists”, who must have had “little-to-no interest” in his point of view. To the latter he remarked “I wish you well and good luck”, but concluded that he is optimistic about how Y90— building upon what has been learned through TACE—is creating new treatment paradigms to improve cancer care.
Key trials backing TACE
Next, moderator Sarah White (Medical College of Wisconsin, Wauwatosa, USA) invited Soulen to the podium to “slap his mentee” and present in favour of TACE’s survival in the age of radioembolization and immunotherapy.
Sparing no time, Soulen prefaced his presentation with a direct response to this question, and remarked: “Hell yeah, [TACE] is still the only evidence-based treatment in the space for unresectable hepatocellular carcinoma [HCC]”. Yet he spoke of the lack of a comprehensive regulatory pathway for procedures and for medical devices, which enter clinical practice without compelling evidence. “This is the Achilles’ heel of interventional radiology [IR]” Soulen said, in contrast to the highly regulated phase 3 trials involving thousands of patients that must be carried out for anti-cancer drugs to gain approval.
Referencing new data however, Soulen contended that there are level 1 data indicating that TACE works and level 1 data that Y90 does not in unresectable disease; notably the RCT of TACE +/- brivanib which prospectively analysed >500 TACE patients, the Japanese-Korean cooperative group study of TACE for HCC, and the Precision Italia trials, all of which showed two to three-year median overall survival after TACE versus the universally negative RCT’s for TARE.
Yet, he maintained that Y90 does have a valuable role in a curative intent setting. Published in January 2024, he offered the updated analysis of the DOSISPHERE-01 trial which evaluated overall survival using Y90 microspheres in patients with inoperable locally advanced HCC. Principal investigator Etienne Garin (Rennes 2 University, Rennes, France) et al reported that, at long-term follow up, a meaningful improvement in overall survival was sustained after applying personalised dosimetry in patients who were successfully downstaged and resected, but consistent with prior studies that showed no survival benefit in unresectable patients.
Soulen then postulated on whether immunotherapy will change the landscape, considering the positive immunostimulatory effects of embolization using a large series of matched cohort studies from the 1970s concerning patients with Stage 4 renal cell carcinoma who had nephrectomy—“and there weren’t any good drugs back then”, he added.
Additionally, looking at the recent release of EMERALD-1 data, Soulen asserted that the benefits of combining of TACE and immunotherapy are apparent. “We now have data showing better outcomes with the combination— so where’s the data for radiation and immunomodulation?” Posing this rhetorical question in front of a slide that read ‘crickets chirping’, Soulen finalised his claim, acknowledging that there are trials coming in this area with much still yet to be learned.
Mechanistic data are crucial
With a focus on immunotherapy, Rony Avritscher (MD Anderson Cancer Center, Houston, USA) spoke next, and first pointed out the lack of mechanistic data offered by Salem and Soulen. He noted that collecting standardised data for interventional oncology (IO) therapies is one of the “biggest challenges” interventional oncologists face, due to the complex nature of their procedures and equipment, as well as the specialists needed to a dminister treatments.
Without crucial mechanistic and therapy standardisation, scalability for IO procedures will continue to be an area of “struggle” Avritscher claimed, placing the field at a “disadvantage”. In order to make these gains, the speaker stated that the development of tools which include innovative technologies such as artificial intelligence (AI) should be sought, with the capability of prescribing dose for complex tumours.
Following Avritscher’s position on gaining better data, the panel were then invited to discuss; Salem first stated that new ways of collecting meaningful analyses must be devised, as RCTs—although ideal—are not constructive or representative of most of the procedures carried out in clinics. Developing this point, Soulen reiterated the difficulty they face in gaining data for both devices and drugs, and how making headway with the former is one of the “biggest challenge[s] in IR”. “We have no data because our devices don’t have to have data to be sold, and the companies won’t pay for trials because they don’t have to”.
With the information they have now, Soulen added, there will never have been a situation where TARE will replace TACE. As every patient is fundamentally different, individualised treatment is key. “It’s not a black and white decision in everyday life,” he contended, “there are other clinical factors besides their cancer and you’re going to initialise your care using whichever tool in your toolbox is suited to that so that [the patient] can have the best quality of life and outcomes can be improved”.
Avritscher added that even different types of embolization procedures have “fundamentally different” effects on the human body which the community must seek to understand in order to pinpoint which patients would benefit from what combination of agents to specialise their care. Avritscher averred however that the only way this can be achieved on IO’s “turf” is to invest in basic science and translational studies looking at patient subsets and their molecular profiles.
The three speakers agreed that every treatment option to their disposal is valuable. To Soulen, these options give patients agency to select a route that best suits them in terms of quality of life: “TACE is quick and sick, Y90 is slow and glow. They work similarly well and some patients want the slow, gentle route, because they don’t want to be sick, but others say ‘No, I want this tumour dead tomorrow’ and will tolerate the subsequent sickness.”
Wrapping up, White asked the room to raise their hands if their institution offers and performs TACE or Y90, showing a visibly divided audience. Of the panel’s most salient conclusory points, consensus on TACE versus TARE may never be reached for good reason, so that case-by-case the most applicable treatment can be deployed for individual patient needs. Reaching agreement on this, Salem resolved that rather than focus being placed on one winning out over the other, attention should be paid to the new narrative that is being created by Y90 that is crucially expanding the dialogue for IO treatment.
