A Phase I Dose-Escalation Trial of Single-Fraction Stereotactic Radiation Therapy for Liver Metastases

This phase I clinical trial was approved by the UT Southwestern institutional review board. Adult patients with Zubrod performance status of 2 or less, with 5 or fewer liver metastases (non-germ cell or hematologic origin) were eligible for enrollment. Patients had to have an expected life span of at least 6 months and could not have received prior liver radiation, which would lead to exceeding protocol-defined constraints for the liver and other normal tissues. Patients underwent multidisciplinary evaluation for consideration of liver-directed therapies. The treated tumor(s) had to be located outside of the central liver zone, defined as a 2-cm expansion around the course of the portal vein contoured to its bifurcation in the liver. Patients were not eligible if they had significant and uncontrolled active comorbidities, including any or all of the following: unstable angina and/or congestive heart failure or myocardial infarction within the preceding 6 months; COPD exacerbation or acute infection at the time of registration; or active hepatitis or Child’s-Pugh class B or C cirrhosis. Patients’ laboratory values had to meet these restrictions: hemoglobin ≥10 g/dL, Platelets ≥100,000 per microliter, ANC ≥1000 per microliter; albumin >3 g/dL, alkaline phosphatase, ALT, AST, total bilirubin, and PT/INR ≤1.5 times the upper limit of normal. A critical liver volume constraint also had to be met with the single-fraction radiation treatment plan: at least 700 cc of normal liver had to receive <9.1 Gy. Steroid premedication was encouraged.

The radiation prescription dose was escalated in increments of 5 Gy. The starting dose was 35 Gy. Dose-limiting toxicity (DLT) was defined according to the National Cancer Institute Common Toxicity Criteria for Adverse Events (CTCAE). Treatment-related grade 3, 4, or 5 gastrointestinal, hepatobiliary/pancreatic, renal/genitourinary, or neurologic toxicity was considered dose-limiting toxicity, as were any grade 4 or 5 toxic events in these categories: blood/bone marrow, cardiac, pulmonary/upper respiratory, metabolic/laboratory, musculoskeletal/soft tissue, and skin. Any other grade 4 or 5 events that were treatment-related also constituted DLT.

Patients were enrolled in cohorts of 7–15. If zero of the first 7 patients, 2 or fewer of the first 9, 3 or fewer of the first 12, or 4 or fewer of the first 15 experienced DLT within the first 90 days following treatment, the dose level was escalated. The maximum tolerated dose level was established as the tolerated dose level below which excessive DLTs (at a rate ≥33 %) had been reached.

Patients were immobilized in a full-body, vacuum-lock mold, which was placed in a stereotactic body frame (Elekta, Stockholm, Sweden). Respiratory motion was assessed with fluoroscopy and motion of the diaphragm was limited to <1 cm with the use of abdominal compression as needed. Placement of fiducials to assess motion and guide radiation delivery was allowed but not required. A four-dimensional computed tomography (CT) scan was obtained to aid in delineation of the motion envelope of the gross tumor volume. A magnetic resonance imaging (MRI) scan of the abdomen, unless contraindicated, was recommended to aid in delineation of the tumor for treatment planning. This scan could serve as a baseline image set for assessment of treatment response. Expansions of 5 mm in the axial plane and 5–10 mm in the cranio-caudal plane around the GTV were made to generate a planning target volume (PTV). Both three-dimensional (3D) conformal radiation treatments and intensity-modulated radiation therapy with megavoltage photons could be used for treatment planning. For the 3D conformal treatments, multiple noncoplanar beams were used, with minimal block margin surrounding the PTV. As such, prescription isodose lines for these plans could be to the 60–90 % isodose line. Planning guidelines called for 95 % of the PTV to be covered by the prescription dose and 99 % of the PTV to receive a minimum of 90 % of the prescription dose. Normal tissue dose delivery guidelines for bowel, skin, lung, and kidney were provided to facilitate planning but could be exceeded at the discretion of the treating physician. The liver critical volume and spinal cord constraints had to be respected in all cases. Normal tissue constraints are shown in Table 1.

Patients were followed at 4–6 weeks and at 3 months following treatment, with subsequent follow-up visits every 3 months through 1 year. Then, patients were followed every 6 months through 5 years of follow-up, and annually thereafter. At the time of follow-up, imaging of the abdomen, preferably with MRI, was obtained, as were the following: serum chemistries and PT/INR.

Local control was defined as lack of tumor growth on imaging studies within the treated tumor volume. Pre- and post-SABR MRI scans were reviewed to assess control on a per-lesion basis. T2-weighted axial images were followed as these images most consistently and clearly delineated the target tumor for radiation treatment planning. Follow-up and survival were calculated from the day of radiation treatment.

Treatment with systemic therapy was allowed both before and after the single-fraction SABR treatment. However, it had to be held for 6 days before, during, and 5 days after the delivery of the radiation.

The primary end point of this study was to establish a maximum tolerated dose for single-fraction irradiation of liver metastases. The initial goal was to escalate to a dose of 50 Gy if dose-limiting toxicities were not prohibitive at lower doses. However, with feasibility issues, including patient accrual as well as the local control results obtained at 40 Gy, further escalation beyond 40 Gy was not undertaken. Secondary endpoints included local control of the treated lesions and overall survival.