Pt Selection Key to Radioembolization of Liver Ca's

SEATTLE—Therapy via radioembolization improves outcomes in some patients with primary or metastatic liver tumors that are unresectable or refractory to chemotherapy, and research is helping to better identify these patients upfront, according to a set of studies presented at the 32nd Annual Meeting of the Society of Interventional Radiology.

Unresectable HCC

In the first study (abstract 220), Riad Salem, MD, MBA, of Northwestern University Lurie Comprehensive Cancer Center, and colleagues assessed long-term outcomes of radioembolization in 233 patients with unresectable hepatocellular carcinoma (HCC) treated between 2001 and 2007. Tumors were embolized with yttrium-90 (90Y) glass microspheres (TheraSphere, MDS Nordion) at a dose of 100 to 150 Gy, with either a lobar or a segmental protocol (see Figure 1).

Median overall survival, assessed from the first of two radioembolizations, was better in patients with Okuda class 1 disease than in those with class 2-3 disease (530 vs 236 days) and better in low-risk than in high-risk patients (476 vs 139 days), Dr. Salem said. Survival decreased across the Child-Pugh A, B, and C classes (512, 231, and 136 days, respectively).

In subset analyses among patients without portal vein thrombosis (PVT), median survival was 574 days with noninfiltrative Child-Pugh A disease, 672 days with noninfiltrative Okuda class 1 disease, and 920 days with nonmetastatic Child-Pugh A disease. In similar subset analyses of patients with PVT, median survival was 308 days with Child-Pugh A disease, 428 days with Child-Pugh A disease with lobar PVT, 231 days with Child-Pugh A or B disease, and 284 days with Child-Pugh A or B disease with lobar PVT.

Final subset analyses focused on the impact of the location of PVT and of cirrhosis. Among all patients, median survival was 467, 304, and 134 days with no PVT, with branch PVT, and with main PVT, respectively (P = .005), Dr. Salem said. Among patients with cirrhosis, the corresponding values were 352, 234, and 118 days (P = .002), whereas among their counterparts without cirrhosis, they were 813, 323, and 134 days (P = .03).

Further subset analyses are forthcoming and will hopefully lead to controlled or randomized trials, Dr. Salem said. "Patient selection is critical," he concluded.

Cholangiocarcinoma

In a retrospective study (abstract 222), Robert J. Lewandowski, MD, of Northwestern University, and colleagues assessed the safety and efficacy of radioembolization with 90Y glass microspheres (TheraSphere) in patients who had progressive cholangiocarcinoma despite chemotherapy and who were treated in a prospective, open-label phase II trial. The procedure was performed using a lobar protocol with a target dose of 120 Gy.

Eight patients underwent 11 radioembolizations (mean dose, 100 Gy). The 90-day CT response was a partial response in 9% of treated lobes, stable disease in 73%, and progressive disease in 9% according to WHO criteria (one patient was lost to follow-up). Median overall survival was 532 days from the time of treatment and was better in patients with an initial ECOG score of 0 vs 1 (616 vs 204 days). "From our preliminary results of treating patients with unresectable cholangiocarcinoma progressing despite chemotherapy, radioembolization may be a potential therapeutic option," Dr. Lewandowski said. "The toxicity profile was acceptable, and once again, selection criteria are important."

In another study (abstract 223), Dr. Lewandowski and colleagues assessed the safety and efficacy of 90Y glass microsphere (TheraSphere) radioembolization in women with liver metastases of breast cancer that had progressed despite polychemotherapy and hormonal therapy, using data also obtained from the phase II trial (see above). A total of 27 patients were treated; the median dose was 123 Gy to the left lobe and 121 Gy to the right lobe. On 90-day CT images, 39% of patients had a complete or partial response, 52% had stable disease, and 9% had progressive disease by WHO criteria. On PET images, 63% of patients had a positive response (see Figure 2).

Median overall survival from the first radioembolization was better, although not significantly so, for patients with an ECOG score of 0 vs those with a score of 1, 2, or 3 (6.8 vs 2.6 months) and for patients whose tumor burden was less than 25% of the liver vs those whose tumor burden was greater (9.4 vs 2.0 months).

"Patients with progressive breast cancer liver metastases despite polychemotherapy may benefit from 90Y radioembolization. The toxicity profile is acceptable," Dr. Lewandowski said.He pointed out that, once again, benefit appears to vary with patient- and tumor-related factors.

Liver Mets of Colon Cancer

In a prospective, multicenter phase II clinical trial presented by Giuseppe Pizzi, MD, of the Regina Elena Cancer Institute in Rome, Italy (abstract 224), investigators assessed the safety and efficacy of 90Y radioembolization for unresectable liver metastases of colorectal cancer that had progressed despite two lines of chemotherapy (including either FOLFOX or FOLFIRI) in patients who did not have metastases elsewhere.

In the trial, which opened in 2005, radioembolization was performed by selective hepatic transarterial delivery of 90Y resin microspheres (SIR-Sphere, Sirtex Medical), also called selective internal radiation therapy (SIRT). The median target dose was 1.6 GBq.

Among the 32 evaluable patients assessed at 6 weeks, 22% had a partial response, 72% had stable disease, and 6% had progressive disease according to RECIST criteria, Dr. Pizzi said. The corresponding values among the 28 evaluable patients assessed at 12 weeks were 18%, 39%, and 43% (see Figure 3). Median overall survival was 9 months in responders and 5 months in nonresponders.

Relative to patients who had progressive disease, patients who had a response tended to have a smaller maximal tumor diameter (median, 3.5 vs 5.1 cm) and fewer metastases (median, 3.5 vs 5.2). In a comparison of tumor biopsies obtained before and 6 weeks after radioembolization, the treatment was associated with decreased cell proliferation (as assessed by Ki-67 protein level) and increased apoptosis (as assessed by p53, bcl-2, and survivin protein levels).

"In this study, we have to consider that SIRT was performed only in 'nonresponding' patients with advanced liver disease who therefore had a very poor prognosis, and this is, of course, a limiting factor," Dr. Pizzi commented.

Although the toxicity profile was acceptable, the tumor response was unsatisfactory, he said. "Moreover, SIRT does not influence the unavoidable progression of extrahepatic disease that often occurs. Therefore, we think that phase II randomized trials focusing on targeted therapy with or without SIRT in patients undergoing second-line therapy could be very useful," he concluded.