Validation study on accuracy of our newly proposed methods for post-therapeutic liver reserve capacity estimation utilizing 99 mTc-GSA scintigraphy prior to carbon-ion radiotherapy.

Abstract

Objective: Carbon-ion radiotherapy (CIRT) is known as a promising treatment for liver tumors. However, no method has been established for estimating post-therapeutic residual liver reserve capacity (pRLRC). We previously introduced the estimation method of pRLRC using ^{99 m}Tc-galactosyl human serum albumin (^{99 m}Tc-GSA) scintigraphy (Yamazaki method). In this study, we developed "Yamazaki-variant" method for pRLRC using dose distribution data of CIRT planning. The purpose of this study was to compare pRLRC calculated by Yamazaki method and Yamazaki-variant in CIRT patients for liver tumors, and to provide the clinical advantages in both methods in terms of estimating ability of pRLRC.

Methods: Patients who received CIRT for liver tumors in our hospital and underwent ^{99 m}Tc-GSA scintigraphy before and 3 months after CIRT, and contrast-enhanced liver MRI within 1 month before CIRT were included. A total of 71 patients, 21 additional to the 50 previously reported, were evaluated. The maximal removal rate of ^{99 m}Tc-GSA (GSA-Rmax) was analyzed and the GSA-Rmax of the estimated residual liver (GSA-RL) was calculated using fusion images of MRI and SPECT by Yamazaki method. In Yamazaki-variant, multiple simulations were performed using the dose distribution of the CIRT planning and SPECT fusion images to obtain higher diagnostic accuracy of GSA-RL. Two of these estimates, Validation 1 and Validation 2 by Yamazaki-variant, were used as validation data. GSA-RL and Validation 1 and 2 were compared with GSA-Rmax 3 months after CIRT (post-GSA-Rmax) using linear regression analysis.

Results: GSA-RL, Validation 1 and 2 were calculated as 0.448 ± 0.214, 0.413 ± 0.199 and 0.435 ± 0.208 mg/min, respectively. Post-GSA-Rmax was 0.428 ± 0.220 mg/min. The relationship between post-GSA-Rmax and each parameter was y = 0.02 + 0.90x (R² = 0.78) for GSA-RL, y = 0.02 + 0.94x (R² = 0.79) for Estimation 1, y = 0.02 + 0.99x (R² = 0.80) for Estimation 2, respectively (P <.0001). Both Yamazaki method and Yamazaki-variant showed comparable accurate estimation ability for post-GSA-Rmax.

Conclusions: The estimation of pRLRC by Yamazaki method and Yamazaki-variant was in good agreement with post-therapeutic liver reserve capacity after CIRT, and there was no difference in the accuracy of the estimation. The usefulness of Yamazaki method, which is simpler and more clinically applicable, was confirmed.

Trial registration: This study is registered in UMIN Clinical Trials Registry (UMIN-CTR) as UMIN study ID: UMIN000038328 and UMIN000049770, UMIN000038174.