Personalized Y-90 Radiation Therapy Outperforms Standard Dosing in Advanced HCC

Author(s):

Beau Toskich, MD, discusses key results that demonstrate the benefit of personalized dosimetry, and offered advice to fellow clinicians as the field moves toward adopting this radiotherapeutic modality.

Beau Toskich, MD

High-dose personalized dosimetry treatment with Yttrium-90 (Y-90) glass microspheres (TheraSphere) is a promising local therapy for patients with advanced hepatocellular carcinoma (HCC), said Beau Toskich, MD, who added that findings from the phase 2 DOSISPHERE-01 trial provide clinical evidence in support of the therapy as an ablative strategy.

“Y-90 can control liver disease to enable us to give the patient the benefit of the doubt when it comes to their tumor biology,” said Toskich. “Just because [a tumor] looks a certain way, doesn’t mean that it will behave a certain way. [Being able to] consolidate a patient to understand their disease process and to get them to definitive [are] important applications of this form of radioembolization.”

In the DOSISPHERE-01 study, patients were randomized 1:1 to receive personalized dosimetry with the goal of 205 Gy or more to the index lesion or standard dosimetry with the goal of 120±20 Gy to the perfused lobe. In the intention-to-treat (ITT) population, the median overall survival (OS) was 26.6 months with the personalized approach versus 10.7 months with the standard approach.

In an interview with OncLive^®, Toskich, an interventional radiologist at Mayo Clinic, shed light on the rationale for the DOSISPHERE-01 trial, discussed key results that demonstrate the benefit of personalized dosimetry, and offered advice to fellow clinicians as the field moves toward adopting this radiotherapeutic modality.

OncLive^®: What served as the basis for the DOSISPHERE-01 trial?

Toskich:This study came out of an era where we had 2 negative phase 3 studies with radioembolization in HCC. [DOSISPHERE-01] was trying to capture a parameter that would generate a positive result in patients who received [personalized] Y-90 versus binary treatment of Y-90.

Regarding the negative studies, patients were treated with the standard of care at the time. The patients had failed multiple lines of chemoembolization and had very extensive, typically bilobar disease. Patients received low-dose Y-90 to a large portion or all of the liver. The amount of liver that was left untreated, particularly in a cirrhotic patient, is important. That is not something that was captured in the design of these studies and, thus, they were negative. It was found that [Y-90] compared with sorafenib [Nexavar] didn’t improve survival but did improve quality of life. Post-hoc analyses showed that if we established a dose threshold of 100 Gy, Y-90 [did well in] targeting the tumor. Also, if the dose was at least 100 Gy, survival improved from 10 months to 24 months. Had [the investigators] screened accordingly, there is a chance that these would have been positive studies.

How did the DOSISPHERE-01 trial aim to improve upon those negative trials?

The DOSISPHERE-01 group has led the field in terms of breaking down what dose [of Y-90] is necessary to achieve a response in HCC. It is different for brachytherapy versus external beam radiation because [brachytherapy] uses a very heterogeneous, high dose [of Y-90].

Retrospectively, [the DOSISPHERE-01] group showed survival benefits if patients with portal vein thrombosis were treated at around 205 Gy. The study was opened evaluating the standard dose of approximately 120 Gy, administered to the whole liver. However, if [Y-90] targets the nuclear tumor well, we can [give doses] above 205 Gy. That is the difference between standard dosimetry and personalized dosimetry. Personalized dosimetry has the intention to add activity until the dose threshold that we know we can see benefit in [is reached].

The phase 2 [DOSISPHERE-01] trial was designed to question whether Y-90 was being used to our absolute maximum capabilities and whether it makes a difference. Could we prospectively capture what we were seeing clinically? That is really what drove DOSISPHERE-01.

What did the results of the DOSISPHERE-01 trial show and how does it add to the understanding of why the personalized dosimetry approach is effective?

The primary end point of the DOSISPHERE-01 trial was objective response rate [ORR] in the index lesion. A modified ITT patient population [was enrolled], encompassing patients who could receive personalized dosimetry. [The ORR was] 71.4% [with personalized dosimetry] versus 35.7% [with standard dosimetry].

Being able to establish with randomized phase 2 data that higher doses [yield] better response rates was an important landmark for patients with very advanced disease. The index tumors in DOSISPHERE-01 were over 10 cm, almost 65% of patients had portal vein thrombosis in the personalized treatment arm, 87% were Barcelona Clinic Liver Cancer stage C. Moreover, many patients had handed disease, where their disease was on just 1 side of the liver. That opened up an opportunity to [utilize] Y-90 because we know that if we can leave 30% or more of the liver alone in Child-Pugh A disease, patients can tolerate extremely high doses [of Y-90]. That is the HCC phenotype that we should be looking at. The study engaged a very difficult-to-treat population with HCC who had enormous tumors, were vascular invasive, and had entire lobes being replaced.

More importantly, if we look at the second end point of OS, patients lived over 26 months [with personalized dosimetry] compared with 10 months [with standard dosimetry].

The adverse effects were lower in the personalized dosimetry [arm versus the standard dosimetry arm]. Being able to do this for patients and not generate systemic therapy–limiting AEs, particularly when we know patients are at risk for metastatic disease or progression in the untreated liver, is important.

How can these findings be interpreted for clinical practice?

The next step for interventional radiology is to make these findings as reproducible as possible to make it a standard option for patients.

Making sure that we include all potential therapies as part of the trial design is a huge opportunity. Ultimately, I do think that DOSISPHERE-01 is a smart study [that enrolled] patients who simulate well and can [receive] impactful doses [of Y-90]. DOSISPHERE-01 helped to pin down and sharpen our understanding of what type of patients are going to best treated with Y-90 and [ensure] that AEs are not shutting any doors for systemic therapy.

What is your advice for fellow clinicians regarding adopting ablative radioembolization in the treatment paradigm for patients with HCC?

We know that a patient who doesn’t have portal hypertension, is not a transplant candidate, and has a lot of liver remnants is going to get a resection. Patients who have Milan [criteria]–staged disease with a high-MELD [Model for End-Stage Liver Disease] score need bridging therapy and a liver transplant. However, we know we can have an effect in patients with disease presentation, but we tend to use palliative therapies. That is one opportunity where Y-90 has been helpful.

Additionally, [Y-90 is important] as a definitive therapy because we are seeing control rates upward of 90%, pathologic necrosis rates over 50%, and higher dosing rates approaching 70%. I don’t see anything wrong with offering [Y-90] as a definitive form of radiation therapy. The formula is easy. We identify a volume of liver that is expendable for the patient and dose it above 400 Gy. I don’t think it excludes external-beam radiation therapy. Our group has found portions of tumors that have been large and complex that are amenable to Y-90 and other parts to radiation therapy.

Even though there is not a clinical trial currently, I encourage practices to do what is best within their local practice and experience level if it is safe. If you can enroll patients into some of the studies that are coming out, that is encouraged. Think of [TheraSphere] as an ablative modality that is tolerated with extremely large volumes of tissue.