TheraSphere Y-90 glass microspheres offer a precision medicine technique aimed at delivering high-dose radiation directly to tumors.
TheraSphere Y-90 glass microspheres offer a precision medicine technique aimed at delivering high-dose radiation directly to tumors. Most recently, investigators examined data from the retrospective TARGET study (NCT03295006), which aimed to provide a validated consensus between absorbed doses, adverse effects (AEs), and objective response rates (ORR) of the therapy in patients with hepatocellular carcinoma (HCC).1
Specifically, the primary objective of the study was to use an alternative 2-compartment TheraSphere dosimetry methodology to calculate normal tissue-absorbed dose to then establish a relationship with grade 3 or higher hyperbilirubinemia without disease progression. Secondary measures included establishing a relationship between tumor-absorbed dose and objective response, overall survival (OS), and probability of α-fetoprotein (AFP) response.
Principal investigator Riad Salem, MD, walked through the latest data from the TARGET study as part of an OncLive® Rapid Readouts video program. Salem is vice chair for image guided therapy, chief of vascular and interventional radiology, and professor of radiology, medicine, and surgery at the Feinberg School of Medicine at Northwestern University in Chicago, Illinois.
The Y-90 glass microspheres are manufactured during a process in which inactive yttrium-89 is combined with aluminum and silicon into glass microspheres. After a heating process, the neutron bombardment transforms the inactive yttrium-89 trapped in the spheres into highly concentrated radioactive yttrium-90.2
TheraSphere, a personalized selective internal radiation therapy for local tumor control in patients with HCC whose tumors measure 1 to 8 cm in diameter, was approved by the FDA in March3 based on results of the retrospective LEGACY study. At a median follow-up of 29.9 months, the therapy elicited an ORR of 72.2% (95% CI, 64.9%-78.5%) among 162 patients, the majority of whom (76.1%; 95% CI, 67.6%-82.9%) experienced a duration of response lasting at least 6 months.4
In TARGET, the treated population consisted of 209 patients with a median age of 66 years (range, 27-87). A majority of patients has Barcelona Clinic Liver Cancer C status (54.5%), Child-Pugh A status (89.5%), and unilobar disease (70.8%). The target lesion size ranged from 3 cm to at least 8 cm per RECIST 1.1 criteria. Forty-one patients (19.6%) had tumors measuring between 3 cm and less than 5 cm, 72 had tumors between 5 cm and less than 8 cm, and 96 patients had tumors at least 8 cm in diameter. Of note, 69.5% of patients had a total of 1 lesion, 21.5% had 2, 6.7% had 3, and 2.4% had between 4 and 10 target and nontarget lesions.
In an analysis of the correlation between grade 3 hyperbilirubinemia and normal tissue-absorbed dose, only 4.8% of patients experienced the AE. “When you look at the [data for the] first primary end point, we did not find a relationship,” Salem said. “Less than 5% of the patients experienced [the AE], so that [relationship] was actually not found, which is a good thing, because it shows that there’s room for optimization and even increasing of the dose when performing this kind of therapy.”
AEs of any grade were reported for 62.7% of the TARGET study population: 24.9% were grade 1, 17.2% were grade 2, 16.3% were grade 3, and 4.3% were grade 4. “If you look at all of the AEs that were identified, 63% of patients had AEs that [are] what you would expect in [patients with] HCC with some fatigue, pain, weakness, asthenia, and some very minor nausea,” Salem said. Specifically, all-grade AE rates were under 20% for fatigue (16.3%), abdominal pain (12.4%), asthenia (9.1%), and nausea (5.3%). Reported grade 4 AEs included ascites (0.5%), lymphocyte count decrease (1.4%), and decreased appetite (0.5%).
“As we hypothesized, the higher the delivered dose, the higher the response rate and, in particular, the complete response rate,” Salem noted on the response data. “You can stratify this patient population by absorbed dose. If you look at the entire patient population, 71% had an objective response, and interestingly, there was a significantly higher dose delivered to the tumor in responders compared with the nonresponders.”
Specifically, data from TARGET demonstrated an ORR of 70.8% (95% CI, 64.3%-76.6%) among all treated patients. The responders had a significantly higher tumor-absorbed dose (22.5.5 Gy; 95% CI, 201.0-253.0) compared with nonresponders (188.3 Gy; 95% CI, 164.6-215.3). In terms of complete responses, the rates increased from approximately 10% at a dose less than 225 Gy to over 30% with a dose of at least 300 Gy.
“The higher the dose, the more likely you are to achieve a response, which is something that you might expect and now that’s something that we are now demonstrating,” Salem said.
The OS data—“the gold standard in oncologic analyses,” Salem noted—also demonstrated a relationship with dosage. The median OS was 20.3 months (95% CI, 16.7-26.4) with an HR that corresponded to every 100 Gy change in tumor-absorbed dose of 0.826 (95% CI, 0.71-0.95; P = .009). Salem said that increase in tumor-absorbed dose is independently associated with increased OS.
Finally, the end point of the association between absorbed dose and the probability of AFP response 90 days post treatment was reported. In 71 patients with AFP level of at least 200 ng/mL at baseline, 38.0% had a response at 90 days posttreatment. In 107 patients who had an AFP of at least 20 ng/mL at baseline, 38.3% had a response in the same time frame. Investigators noted that the absorbed dose to total perfused tumor was significantly associated with AFP response at 90 days (P < .05).
“In some cases, AFP is a surrogate of OS, a surrogate of response…if you pick a strict AFP level of 200, nearly 40% of patients showed a response at day 90. So again, the absorbed dose was related to the AFP response,” Salem said.
Investigators have demonstrated that personalized dosimetry is associated with significant response rates when compared with standard dosimetry and should be used for the development of trials using selective internal radiation approaches, particularly those with the Y-90 glass microspheres. “[The TARGET findings] are consistent with [results of] other clinical trials that have been published historically, in particular the DOSISPHERE-01 study that showed that patients survive longer when you perform individualized patient dosimetry,” Salem said.
Data from the phase 2 DOSISPHERE-01 trial (NCT02582034) showed that optimized treatment with TheraSphere was significantly more efficient than a standard dose. The study examined patients with unresectable locally advanced HCC across 4 health centers in France who had at least 1 measurable lesion measuring at least 7 cm in diameter. Patients were randomly assigned to either standard dosimetry (120 ± 20 Gy) or personalized dosimetry (≥ 205 Gy).5
In total, 28 patients were assessed in each group as a modified intention-to-treat population. The ORR in the personalized therapy group was 71% (95% CI, 51%-87%) compared with 36% (95% CI, 19%-56%) in the standard therapy group (P = .0074).