Estimation of post-therapeutic liver reserve capacity using 99mTc-GSA scintigraphy prior to carbon-ion radiotherapy for liver tumors

There is currently no established imaging method for assessing liver reserve capacity prior to carbon-ion radiotherapy (CIRT) for liver tumors. In order to perform safe CIRT, it is essential to estimate the post-therapeutic residual reserve capacity of the liver. To evaluate the ability of pre-treatment 99mTc-galactosyl human serum albumin (99mTc-GSA) scintigraphy to accurately estimate the residual liver reserve capacity in patients treated with CIRT for liver tumors. This retrospective study evaluated patients who were performed CIRT for liver tumors between December 2018 and September 2020 and underwent 99mTc-GSA scintigraphy before and 3 months after CIRT, and gadolinium ethoxybenzyl diethylenetriamine pentaacetic acid (Gd-EOB-DTPA)-enhanced MRI within 1 month before CIRT were evaluated. The maximal removal rate of 99mTc-GSA (GSA-Rmax) was analyzed for the evaluation of pre-treatment liver reserve capacity. Then, the GSA-Rmax of the estimated residual liver (GSA-RL) was calculated using liver SPECT images fused with the Gd-EOB-DTPA-enhanced MRI. GSA-RL before CIRT and GSA-Rmax at 3 months after CIRT were compared using non-parametric Wilcoxon signed-rank test and linear regression analysis. Overall, 50 patients were included (mean age ± standard deviation, 73 years ± 11; range, 29–89 years, 35 men). The median GSA-RL was 0.393 [range, 0.057–0.729] mg/min, and the median GSA-Rmax after CIRT was 0.369 [range, 0.037–0.780] mg/min (P = .40). The linear regression equation representing the relationship between the GSA-RL and GSA-Rmax after CIRT was y = 0.05 + 0.84x (R2 = 0.67, P < .0001). There was a linear relationship between the estimated and actual post-treatment values for all patients, as well as in the group with impaired liver reserve capacity (y = − 0.02 + 1.09x (R2 = 0.62, P = .0005)). 99mTc-GSA scintigraphy has potential clinical utility for estimating the residual liver reserve capacity in patients undergoing carbon-ion radiotherapy for liver tumors. UMIN000038328, https://center6.umin.ac.jp/cgi-open-bin/ctr/ctr_view.cgi?recptno=R000043545.


Introduction
Primary liver cancer was the sixth most common type of cancer and the fourth most common cause of death worldwide in 2015, with the highest incidence of cases and deaths observed in East Asia [1]. Standard treatment strategies for localized hepatocellular carcinoma are surgical resection, radiofrequency ablation (RFA), transcatheter arterial chemoembolization (TACE), and radioembolization (TARE) [2][3][4][5][6]. Metastatic liver tumors are significantly more common than primary liver cancer and are difficult to cure despite various treatment approaches [7,8]. Although surgical resection is the current standard treatment for metastatic liver tumors, other treatments such as proton beam radiotherapy and immunotherapy have been applied for hepatocellular carcinoma and metastatic liver tumor [9][10][11][12][13]. However, treatment options are updated year by year, and the selection policy for newly developed treatments have not yet been established. Carbonion radiotherapy (CIRT) is another therapeutic option for primary and metastatic liver tumor. CIRT delivers high doses to the target tumors while sparing normal liver tissue, which allows reducing toxicity to normal tissues [14,15]. In addition, CIRT has biological advantages because of high linear energy transfer radiation compared to proton beam and X-ray. CIRT has showed efficacy and safety for the treatment of primary and metastatic liver tumors [14][15][16]. Because many patients with liver tumors have impaired liver function, it is essential to assess the reserve capacity of the liver prior to CIRT in order to accurately estimate the post-therapeutic residual reserve capacity.
There are some standard methods for evaluating preoperative liver reserve capacity, including indocyanine green (ICG) clearance [17,18], Child-Pugh classification [18,19], and model for end-stage liver disease (MELD) score [18,19]. These conventional methods reflect total liver function; therefore, they cannot always accurately assess or estimate the impact of treatments. 99m Tc-galactosyl human serum albumin ( 99m Tc-GSA), which binds specifically to the asialoglycoprotein receptor, is only expressed in normal hepatocytes and has been widely used to estimate the reserve capacity of the liver [20][21][22][23]. As Mizutani et al. and other investigators reported previously, 99m Tc-GSA liver scintigraphy is clinically useful imaging evaluating liver function before surgery of liver tumors [24][25][26][27].
The maximal removal rate of 99m Tc-GSA (GSA-Rmax), which is calculated by a kinetic analysis using the compartment model developed by Kawa et al., is useful for the evaluation of preoperative liver reserve capacity and for the prediction of postoperative outcomes [23][24][25][26][27]. Based on these prior studies, we hypothesized that this technique has a potential utility of liver functioning evaluation in patients with liver tumors before undergoing CIRT. However, there has been no report evaluating pre-treatment liver function using 99m Tc-GSA scintigraphy in patients with various liver tumors who are going to receive CIRT.
The aim of this study was to evaluate the ability of 99m Tc-GSA scintigraphy to accurately estimate the residual reserve capacity of the liver in patients treated with CIRT for liver tumors.

Study patients
This study was approved by our institutional review board as a single-center clinical trial (#19-027). This study is registered in the University Hospital Medical Information Network Clinical Trials Registry (#UMIN000038328), and written informed consent was obtained from each patient. This study was conducted in accordance with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards. And this study has been reported based on the Standards for Reporting of Diagnostic Accuracy guidelines. Data generated or analyzed during this study are available from the corresponding author by request.
Subjects were considered eligible for inclusion in this study if they (1) were clinically diagnosed with a liver tumor and scheduled to undergo CIRT, (2) underwent 99m Tc-GSA scintigraphy and gadolinium ethoxybenzyl diethylenetriamine pentaacetic acid (Gd-EOB-DTPA)enhanced MRI within 1 month before CIRT, and (3) underwent 99m Tc-GSA scintigraphy 3 months after CIRT. Patients were excluded from this study if they (1) refused to participate or (2) did not undergo 99m Tc-GSA scintigraphy from 3 months after CIRT. 99m Tc-GSA dynamic scintigraphy and SPECT procedure All patients underwent 99m Tc-GSA scintigraphy using a gamma camera (E.CAM, Canon Medical, Tokyo, Japan) equipped with a low-medium-energy general purpose collimator centered over the liver and the precordium. Dynamic images were acquired immediately after intravenous injection of 185 MBq of 99m Tc-GSA (Nihon Medi-Physics, Tokyo, Japan) at a rate of 30 s per frame for the first 30 min. Singlephoton emission computed tomography (SPECT) images of the liver were acquired in 90 steps, with a 360-degree rotation at 3.3 s per view and a matrix size of 128 × 128 pixels. SPECT image reconstruction was performed with an ordered subset expectation maximization method.

MRI procedure
For reference with SPECT-MRI fusion images, hepatobiliary phase images acquired 20 min after contrast agent administration were selected. We used Gd-EOB-DTPA-enhanced MRI as reference images, not treatment planning computed tomography (CT), because MRI showed higher sensitivity and diagnostic accuracy than CT for the detection and the determination of the margin of liver tumors [28,29]. In addition, because CIRT-treatment planning CT had not been obtained at the time of evaluation of reserve capacity of the liver in all patients, we could not use the planning CT images as a reference of fusion images. Therefore, using SPECT-MRI fusion images combined with contrast-enhanced MRI, which provides high contrast between the tumor and the normal liver tissue, we simulated the estimated post-treatment liver reserve capacity prior to CIRT.
All MRI examinations were performed before CIRT using the 1.5-T Achieva dStream (Philips Healthcare, Best, the Netherlands). The hepatobiliary phase of Gd-EOB-DTPA-enhanced MRI was acquired using axial three-dimensional gradient-echo imaging with breath-hold examination. The MRI protocol for the spectral attenuated inversion recovery T1-weighted sequence was as follows: repetition time ms/echo time ms, 5.0/2.4; echo train length, 38; 12° flip angle; 5-mm slab thickness; 256 × 256 matrix; and a 350 × 350-mm field of view.

Analysis of reserve capacity of the liver
The GSA-Rmax was calculated by two nuclear medicine physicians (with 12 and 9 years of experience), with application of 99m Tc-GSA dynamics of the heart and liver to a radiopharmacokinetic model composed of five compartments, including the extrahepatic blood, hepatic blood, hepatocytes, interstitial fluid, and urine, as developed by Ha-Kawa [23]. Liver SPECT images were then fused with the hepatobiliary phase of Gd-EOB-DTPAenhanced MRI using the "Fusion Viewer" image analysis software (AZE, Kawasaki, Japan). Rectangular volumes of interest (VOIs) were drawn on the fusion images to include the whole liver. Elliptical VOIs were then drawn around the tumor and an approximately 15-mm margin based on CIRT planning. Kasuya et al. have previously described that the planning target volume can be defined as the gross tumor volume plus a 10-mm margin, with the whole planning target volume covered by at least 95% of the prescribed irradiation dose [14]. Therefore, we determined VOIs of the CIRT planning area by estimating the gross tumor volume plus a 15-mm margin as the area in which hepatocytes lose liver function after CIRT. At the time of this analysis, SPECT and MR images were obtained, but planning CT was not yet obtained. For this reason, estimation was used in the manner described above. Using radioisotope counts within these VOIs, the GSA-Rmax of the estimated residual liver (GSA-RL) was calculated as follows, with reference to the modified method reported by Shuke et al. [30]: All fusion images were analyzed by a consensus of two nuclear medicine physicians (with 10 and 24 years of experience, respectively), a radiologist specializing in body imaging (with 12 years of experience), and two radiation oncologists (with 9 and 5 years of experience in radiation oncology). Then, 99m Tc-GSA scintigraphy 3 months after CIRT was also performed to evaluate the accuracy of the pre-treatment estimated values. It has been reported that patients with classic radiationinduced liver disease usually have symptoms 1-3 months after liver radiotherapy [31], and follow-up 99m Tc-GSA scintigraphy after 3 months of treatment were performed for many patients at our hospital. Therefore, the data of 99m Tc-GSA scintigraphy 3 months after CIRT was used for comparison. Values before CIRT and at 3 months after CIRT were compared for all eligible patients stratified by both GSA-RL group and Child-Pugh classification. Each patient was classified into the following three groups based on their GSA-RL value, with a GSA-RL ≥ 0.50 considered "Excellent," a GSA-RL < 0.50 and ≥ 0.25 considered "Fair," and a GSA-RL < 0.25 considered "Poor." A GSA-RL of < 0.25 corresponds to the value that Mizutani et al. suggested should cause patients to forego liver surgery [25]. Validation of all procedures in terms of 99m Tc-GSA-related parameters were done by consensus of two nuclear medicine physicians (with 10 and 24 years of experience).

Statistical analysis
Normally distributed continuous variables are expressed as mean ± standard deviation, and non-normally distributed variables are expressed as median and range. The variables were compared using non-parametric Wilcoxon signed-rank test, with linear regression analysis of the relationship between GSA-RL values before CIRT and GSA-Rmax values 3 months after CIRT. P values <.05 were considered statistically significant. Sample size estimations were not performed because no previously published applicable data were available. Statistical analysis was performed using JMP statistical software (SAS Institute, Cary, NC, USA).

Patients' characteristics
This retrospective study analyzed 114 potential patients (with a total of 179 99m Tc-GSA scintigraphy examinations) who were scheduled to undergo CIRT for various liver tumors between December 2018 and September 2020. As Fig. 1 shows, of the 114 potential patients, 2 refused to participate in the study, 5 did not undergo MRI before CIRT, 60 did not undergo 99m Tc-GSA scintigraphy 3 months after CIRT, and 4 of them did not undergo both. One patient had circulatory failure due to cardiac disease and multiple aneurysms. As Kawa et al. discussed in relation to the cardiac blood pool and systemic blood or hepatic blood volume [20,23], we determined that an adequate analysis could not be performed because of unbalance of the blood volume of each compartment in this patient. Therefore, this patient was excluded from this study. Ultimately, 50 patients were included in this study (mean age ± standard deviation, 73 years ± 11; range, 29-89 years), with 35 men (74 years ± 8) and 15 women (71 years ± 15). All 50 patients underwent 99m Tc-GSA scintigraphy 3 months (median, 96.5 days; range, 81-126 days) after CIRT. As Fig. 1 shows, characteristics Tc-galactosyl human serum albumin; Gd-EOB-DTPA, gadolinium ethoxybenzyl diethylenetriamine pentaacetic acid of 50 patients treated with CIRT were as follows. Thirtysix patients were diagnosed with hepatocellular carcinoma, 2 with cholangiocellular carcinoma, 1 with mixed hepatocholangiocellular carcinoma, 11 with metastatic liver tumor, and another with giant hemangioma. The total exceeds 100% because one patient had two types of tumors. All patients were classified by liver function according to Child-Pugh classification previously reported [32]. Regarding the history of liver treatment before this study, twelve patients had undergone surgery, eight had undergone radiofrequency ablation, six had undergone transcatheter arterial chemoembolization, five had undergone CIRT, and another had undergone microwave ablation. Some patients had undergone more than one treatment. Patients' demographics and characteristics are presented in Table 1, and a summary of clinical information is in Table 2.

Analysis of reserve capacity of the liver
These results are shown in Table 3 (Fig. 3).
The linear regression equation representing the relationship between the GSA-RL before CIRT and the GSA-Rmax 3 months after CIRT was y = 0.05 + 0.84x (R 2 = 0.67, P < .0001). Focusing on the GSA-RL "Poor" group with a GSA-RL of < 0.25 only, the linear regression equation was y = − 0.02 + 1.09x (R 2 = 0.62, P = .0005). It was possible to estimate post-treatment liver reserve capacity even in Poor group patients. There was a positive correlation between the estimated and actual post-treatment values for all patients, as well as in the group with impaired liver reserve capacity (Fig. 4). One patient had a more than twofold elevation in alkaline phosphatase than that of normal levels and was diagnosed with classic radiation-induced liver disease 2 months after CIRT, with a GSA-RL of 0.090 mg/min and a GSA-Rmax 3 months after CIRT of 0.037 mg/min [31]. This patient was alive and being followed up with imaging and blood test without any treatment during the observation period till 3 months after CIRT. No other patients had apparent worsening of the liver function by serological testing. Representative case is shown in Fig. 5. This case was an 80-year-old man (patient 1) with mixed hepatocholangiocellular carcinoma at S8 and hepatocellular carcinoma at S5/8. A rectangular VOI was drawn on the SPECT-MRI fusion image to include the whole liver. Elliptical VOIs were then drawn around the tumor plus an approximately 15-mm margin to represent the CIRT planning area, demonstrating GSA-RL was 0.297 mg/min. SPECT-MRI fusion image at 3 months after CIRT showed that hepatic uptake of 99m Tc-galactosyl human serum albumin after CIRT was reduced or absent at the irradiated site (GSA-Rmax = 0.316 mg/min). On the other hand, uptake outside the irradiation site seemed to be preserved.

Discussion
The present study clearly showed the clinical utility of precarbon-ion radiotherapy (CIRT) 99m Tc-galactosyl human serum albumin ( 99m Tc-GSA) scintigraphy to estimate the residual liver reserve capacity of the liver in patients who were scheduled to receive CIRT for liver tumors. There has been no direct comparison study between the predicted values with semi-quantitative method such as GSA-RL and the actual measured values after treatment as GSA-Rmax, in terms of predicting post-hepatectomy liver failure and radiation-induced liver disease development after liver tumor treatment. In the present study, we have collected preand post-treatment data from a total of 50 patients. Our data demonstrated pre-treatment GSA-RL calculated from 99m Tc-GSA scintigraphy was consistent with post-treatment liver reserve capacity. In the surgical treatment of liver tumors, investigators have reported that GSA-RL calculated by 99m Tc-GSA scintigraphy is useful for predicting post-hepatectomy    liver failure by comparison of GSA-RL before surgery and clinical course after treatment of liver tumors [24,25]. And there is another report describing that 99m Tc-GSA SPECT image-guided inverse planning may reduce hepatic toxicities after radiation therapy for HCC [33]. To the best of our knowledge, our present study would be the first report  SPECT-MRI fusion image at 3 months after CIRT. Hepatic uptake of 99m Tc-galactosyl human serum albumin after CIRT was reduced or absent at the irradiated site (GSA-Rmax = 0.316 mg/min). On the other hand, uptake outside the irradiation site seemed to be preserved (D) describing a detailed comparison of preoperative prediction and postoperative residual liver function by 99m Tc-GSA scintigraphy. For assessing the relationship between pre-treatment estimated residual liver reserve capacity and actual post-treatment liver reserve, there was a linear relationship between the maximal removal rate of 99m Tc-GSA (GSA-Rmax) of the estimated residual liver (GSA-RL) and the GSA-Rmax after CIRT (R 2 = 0.67, P < .0001). Our estimations were moderately consistent with the actual liver reserve capacity 3 months after CIRT.
When examined by classifying by pre-treatment liver function status, there was a linear relationship with the estimated value between actual values not only in the group with a good liver reserve capacity of 0.50 ≤ GSA-RL, but also in the group with poor liver reserve capacity with a GSA-RL of < 0.25 (R 2 = 0.62, P = .0005), corresponding to the group that Mizutani et al. suggested should avoid liver surgery [25]. We consider that even linear regression in the "Poor" group alone shows that the estimation of post-CIRT liver reserve capacity is promising, although the small sample size may result in a large variability in the estimation. Even inoperable patients with low liver reserve capacity were able to receive CIRT safely without acute radiation-induced liver disease except for 1 patient. As for acute radiation-induced liver disease after CIRT, Komatsu and Kasuya reported they observed no severe adverse events in the liver, only required controllable management in a few patients [14,34]. Combining post-treatment estimation with 99m Tc-GSA scintigraphy has the advantage of expanding indications for CIRT, which is less invasive than other treatments. Besides, the classification based on patients' pre-treatment GSA-RL could contribute arrangement of the margin of irradiation area and adjustment of the beam angle to reduce the damage to normal liver parenchyma in CIRT, and to the prediction of patients who would require careful medical management after treatment. Thus, 99m Tc-GSA scintigraphy could accurately estimate liver reserve capacity after CIRT regardless of the degree of liver function, and our proposed method may facilitate pre-emptive identification of patients who will need careful follow-up. In this preliminary study, no threshold value for GSA-RL in estimating residual liver capacity was identified, since only one case was diagnosed radiationinduced liver disease during the 3-month observation period. Further study discussing on assessing the threshold value by comparing the GSA-Rmax at a certain point after CIRT with longer follow-up would be needed.
The advantage of this proposed method compared with the conventional liver function test should be discussed. Our method could evaluate the regional liver function, which is difficult to assess with the Child-Pugh classification and the ICG clearance test. Previously reported liver function assessments have included serological tests, Child-Pugh classification system, and ICG clearance, all of which are widely used techniques in clinical practice. These methods reflect total liver function; therefore, they cannot always accurately assess or estimate the impact of CIRT, which provides localized treatment for individual lesions. Assessment of the liver and tumor volumes by CT (CT volumetry) is known to be another option for pre-treatment [35]. However, an accurate assessment of liver reserve capacity is difficult to obtain using this strategy because it cannot reflect the heterogeneity of normal and dysfunctional hepatocytes in the liver [25,33,36]. 99m Tc-GSA kinetic analyses of HH15 (retention rate in blood) and LHL15 (hepatic uptake rate) are other widely used indices of liver function [20]; however, these values are obtained from planar images and do not allow for the accurate assessment of anatomy [37]. In clinical practice of radiation therapy, it is desirable to have radiation dose on tumor as high as possible to the extent that damage of surrounding normal liver parenchyma is kept to the minimal.
The present study demonstrated that a SPECT-MRI fused imaging method could easily estimate the liver reserve capacity in each lobe of liver while also taking into account the anatomy of individual patients. This can be accomplished by combining the advantages of SPECT and Gd-EOB-DTPA-enhanced MRI. The former can assess regional liver reserve capacity, while the latter can distinguish tumors from non-tumor areas with high contrast. Therefore, this method might have a clinical utility for patients who receive CIRT for various liver tumors. 99m Tc-mebrofenin has also been clinically used as a radiopharmaceutical similar to 99m Tc-GSA, although there are no head-to-head studies comparing 99m Tc-GSA with 99m Tcmebrofenin. Several trials have been reported for both radiopharmaceuticals to predict clinical outcomes before treatment in various liver diseases, but as Espersen et al. point out, direct comparison of the two radiopharmaceuticals in a multicenter study is needed [38].
In ongoing study with longer observation period, we are planning to determine clear indication criteria of CIRT for liver tumor by a combination of this GSA-RL method with other important prognostic factors such as liver function blood test (AST, ALT, ALP, etc.) and Gd-EOB-DTPAenhanced MRI.
Our study had several limitations. First, we did not compare GSA-RL and post-treatment GSA-Rmax with ICG clearance, which was considered to be one of the standards and the most accurate liver function test [17,18]. Some patients have checked ICG clearance in the other hospital before coming to our institution. However, ICG clearance was not a routine examination in our institution. The accuracy and feasibility of 99m Tc-GSA-related parameters such as GSA-RL should be confirmed for further study. Second, there are technical limitations with the use of SPECT, including errors in radioisotope counting between superficial and deep areas of the liver due to the large size of this organ. Technical verification is needed using a scanner with a high spatial resolution, such as digital SPECT/CT, which will facilitate accurate attenuation correction. Third, VOIs for the CIRT planning area might be partially overlapped on the SPECT-MRI fusion images in patients with multiple liver tumors scheduled for CIRT. In these cases, estimations of liver reserve capacity might be inappropriate. It is necessary to perform validation using the planning CT for CIRT to estimate more precisely. Finally, misregistration of SPECT and MRI image might influence the calculation of the GSA-related parameters. We analyzed using the rigid registration method when fused SPECT and MRI. These images could not be perfectly matched in some cases. Even if we analyze using non-rigid registration, which is considered a more precise method, it may influence the calculation of true values. More accurate methods of image analysis are needed to be validated.
In conclusion, this study demonstrated that 99m Tc-galactosyl human serum albumin scintigraphy has potential clinical utility for estimating the residual reserve capacity of the liver in patients undergoing treatment with carbon-ion radiotherapy for liver tumors.