DeepDyve requires Javascript to function. Please enable Javascript on your browser to continue.
The Predictive Value of SPECT/CT imaging in colorectal liver metastases response after 90Y-radioembolization The Predictive Value of SPECT/CT imaging in colorectal liver metastases response after...
Piasecki, Piotr;Narloch, Jerzy;Brzozowski, Krzysztof;Zięcina, Piotr;Mazurek, Andrzej;Budzyńska, Anna;Korniluk, Jan;Dziuk, Mirosław;
2018-07-10 00:00:00
99m 90 The aim of this study was to evaluate a modified method of calculating the Tc/ Y tumor- to-normal-liver uptake ratio (mT/N) based on SPECT/CT imaging, for use in predicting the OPENACCESS overall response of colorectal liver tumors after radioembolization. A modified phantom- based method of tumor-to-normal-liver ratio calculation was proposed and assessed. In Citation: Piasecki P, Narloch J, Brzozowski K, Zięcina P, Mazurek A, Budzyńska A, et al. (2018) contrast to the traditional method based on data gathered from the whole tumor, gamma The Predictive Value of SPECT/CT imaging in counts are collected only from a 2D region of interest delineated in the SPECT/CT section colorectal liver metastases response after 90 with the longest tumor diameter (as specified in RECIST 1.1). The modified tumor-to-nor- Y-radioembolization. PLoS ONE 13(7): mal-liver ratio (mT/N1) and Y predicted tumor absorbed dose (PAD) were obtained based e0200488. https://doi.org/10.1371/journal. 99m pone.0200488 on Tc-MAA SPECT/CT, and similarly the modified tumor-to-normal-liver ratio (mT/N2) 90 90 and Y actual tumor absorbed dose (AAD) were calculated after Y-SPECT/CT. Tumor Editor: Qinghui Zhang, North Shore Long Island Jewish Health System, UNITED STATES response was assessed on follow-up CTs. Using the newly proposed method, a total of 103 liver colorectal metastases in 21 patients who underwent radioembolization (between June Received: January 19, 2018 99m 2009 and October 2015) were evaluated in pre-treatment CT scans and Tc-MAA- Accepted: June 27, 2018 SPECT/CT scans and compared with post-treatment Y-SPECT/CT scans and follow-up Published: July 10, 2018 CT scans. The results showed that the mT/N1 ratio (p = 0.012), PAD (p< 0.001) and AAD Copyright:© 2018 Piasecki et al. This is an open (p< 0.001) were predictors of tumor response after radioembolization. The time to progres- access article distributed under the terms of the sion was significantly lengthened for tumors with mT/N1 higher than 1.7 or PAD higher than Creative Commons Attribution License, which permits unrestricted use, distribution, and 70 Gy. The risk of progression for tumors with mT/N1 lower than 1.7 or PAD below 70 Gy reproduction in any medium, provided the original was significantly higher. The mT/N2 ratio had no significant correlation with treatment author and source are credited. results. Data Availability Statement: All relevant data are within the paper and its Supporting Information file. Conclusion Funding: The authors received no specific funding for this work. The mT/N1 ratio, PAD, and AAD can be used as predictors of tumor response to SIRT treat- Competing interests: The authors have declared ment, and SPECT/CT imaging can be used for dosimetric assessment of radioembolization. that no competing interests exist. Abbreviations: mT/N1, modified tumor to normal 99m liver ratio in Tc-MAA SPECT/CT; mT/N2, PLOS ONE | https://doi.org/10.1371/journal.pone.0200488 July 10, 2018 1 / 15 90 The predictive value of SPECT/CT imaging in mCRC response after Y-radioembolization modified tumor to normal liver ratio in Y-SPECT/ Introduction CT (bremsstrahlung); PAD, predicted tumor Y 99m The main advantage of Yttrium-90 in palliative treatment of liver tumors is its pure beta-radia- absorbed dose calculated in TC-MAA SPECT/ tion emission with high energy and short half-life[1, 2]. The main limitation of radioemboliza- CT; AAD, actual tumor Y absorbed dose calculated in Y -SPECT/CT (bremsstrahlung); sT/ tion is the low tolerance of healthy liver tissue to radiation[3±5]. Therefore, proper calculation 99m N1, standard tumor to normal liver ratio in Tc- of the activity of Y-microspheres, which can destroy liver tumors while sparing healthy liver MAA SPECT/CT; sT/N2, standard tumor to normal tissue, is very important. The absence of primary gamma emission of Y leads to significant liver ratio in Y-SPECT/CT (bremsstrahlung); 90 problems not only with post-treatment imaging of therapy effects but also with calculation of sPAD, predicted tumor Y absorbed dose 99m dosimetry parameters for liver parenchyma[6±8]. Consequently, tumor dosimetry is not calculated in TC-MAA SPECT/CT based on sT/ N1; sAAD, actual tumor Y absorbed dose obtained in everyday practice. Dosimetry based on the MIRD equation (Medical Internal calculated in Y -SPECT/CT (bremsstrahlung) Radiation Dosimetry) and SPECT/CT imaging appears to be the best option for most centers based on sT/N2; CR, complete response performing radioembolization, but to attain wider use, it should be easier to calculate[7, 9, 10]. (according to RECIST 1.1); PR, partial response 99m Patients undergoing radioembolization typically undergo a pre-treatment Tc-MAA SPECT/ (according to RECIST 1.1); SD, stable disease CT scan to calculate the Y dose. The tumor-to-normal-liver uptake ratio (mT/N1) obtained in this (according to RECIST 1.1); PD, progressive procedure may serve as a basis for calculating further dosimetric parameters, including the predicted disease (according to RECIST 1.1). tumor absorbed dose (PAD) of Y. A second SPECT/CT (bremsstrahlung) scan is performed after radioembolization to assess the post-treatment Y-microsphere distribution. On this basis the tumor-to-normal-liver ratio (mT/N2) and actual adsorbed dose of Y (AAD) can also be calculated and compared with pre-treatment and follow-up data [1, 7, 9, 10]. With early dosimetric evaluation, patients with high potential risk of radiation-induced liver disease or patients with non-curative tumor doses can be recognized shortly before or immediately after undergoing a selective internal radiation therapy (SIRT) procedure. In view of recent reports on SIRT from prospective trials (SIRFLOX/FOXFIRE), it is important to develop a method for selecting patients who would benefit from the therapy[11, 12]. In this study, we propose and assess a new approach involving SPECT/CT for predicting the effects of radioembolization in patients with liver tumors, based on a simplified method of calculating the tumor-to-normal-liver ratio. Materials and methods The study was approved by Ethical Committee Board of Military Institute of Medicine (deci- sion: 24/WIM/2009). All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and o/or national research commit- tee and with the 1964 Helsinki declaration and its later amendments or comparable ethical stan- dards. Informed consent was obtained from all individual participants included in the study. Phantom dosimetric study We first performed a phantom study to determine if gamma counts obtained only from the middle-of-sphere section could be used for dosimetric calculations in Y radioembolization. The IEC Body Phantom was used, filled with six different volumes of calibrated amounts of Y resin microspheres. The inner diameters of the fillable spheres were 10 mm, 13 mm, 17 mm, 22 mm, 28 mm and 37 mm. The activity in the spheres was 3.8 MBq/ml, while the back- ground activity was 0.038 MBq/ml. Gamma counts were collected from each phantom section containing spheres, and the sphere-to-background count ratio was calculated (phT/N). The two smallest spheres of the IEC Phantom were not clearly visible and were excluded from the study. Study population From June 2009 to December 2015 twenty-one patients with unresectable liver metastases of colorectal cancer were enrolled. A total of 103 tumors were selected and evaluated. PLOS ONE | https://doi.org/10.1371/journal.pone.0200488 July 10, 2018 2 / 15 90 The predictive value of SPECT/CT imaging in mCRC response after Y-radioembolization Study protocol Radioembolization was conducted in line with guidelines approved by a panel of experts and the local ethics committee. One session per patient of resin microsphere (Sirtex, Australia) 99m treatment was used for the study. As a pre-treatment procedure, 120-180 MBq of Tc-MAA was administered into the common hepatic artery. In the radioembolization procedure, resin microspheres were injected using the same microcatheter position. SPECT/CT was 99m 90 performed < 25 min after Tc/ Y administration. The Body Surface Area Method was used for calculation of Y activity[9, 10]. As the first step, a region-of-interest (ROI) analysis of the tumor-to-normal-liver ratio was used on the SPECT images to determine the standard ratio (sT/N)Ða detailed description is available else- where [7,9]. The modified tumor-to-normal-liver uptake ratio (mT/N) was then calculated using a new method proposed by the authors, as described below. The RECIST 1.1 criteria were used to evaluate target liver lesions. Correlations between 99m 90 dosimetric parameters of target liver lesions detected with Tc/ Y-SPECT/CT were investi- gated and compared with initial and follow-up CT results. The target liver lesions were those 99m 90 that had a longest diameter of at least 10 mm and were clearly visible on CT and Tc/ Y- SPECT/CT. Tumors with areas of necrosis were not excluded (9%)[13]. For SPECT/CT imaging the GE Infinia Hawkeye 4 was used, with 1-inch NaI crystal detec- 99m 90 tors and a low-dose four-slice CT protocol. The protocols for Tc-MAA imaging and Y bremsstrahlung image acquisition are shown in Table 1. For image post-processing, a Xeleris 3.0423 Workstation (GE, USA) with Volumetrix MI software was used. The ordered subset expectation maximization (OSEM) method of iterative reconstruction was applied, with 2 iterations and 10 subsets. The prefilter HANN 0.9 and the 3D postfilter BUTTERWORTH 0.48 were used. Patient dosimetric study The main steps in calculating dosimetric parameters were as follows: 1. Selection of target 99m 90 tumors in pre-treatment CT scans and identification of them with Tc/ Y-SPECT/CT. 2. Calculation of the modified tumor-to-normal-liver ratio (mT/N1 and mT/N2). 3. Calculation of the predicted and actual absorbed dose for the target liver tumors (PAD and AAD, respec- tively). 4. Evaluation of tumor responses using data obtained in CT scans and of tumor dosi- metric parameters [7, 14]. 99m To calculate mT/N1, the slice with the longest target tumor diameter was selected in Tc- MAA SPECT/CT. Its borders were delineated on CT using semi-automatic tools: a 2D ROI was drawn manually on a CT image generated with an abdominal CT window (Volumetrix MI application on Xeleris 3.0423 Workstation). In the next step, the ROI was automatically 99m duplicated for the fusion Tc-MAA scans, and counts of gamma emission from within it were obtained (see Fig 1). Another ROI was placed in an area of healthy liver tissue adjacent to the assessed liver tumor, within the same liver segment, and the gamma emission counts from 99m within this second ROI were determined. The mT/N1 ratio for Tc-MAA SPECT/CT was then calculated using the following formula: 1) [7, 14±16]. counts in tumor s ROI mT=N1 ˆ …1† counts in healthy liver s ROI 99m mT/N1 ± modified tumor to normal liver uptake ratio in Tc-MAA SPECT/CT To determine the mT/N2 ratio for Y-SPECT/CT, all the steps listed above were repeated. The liver absorbed dose (LAD1 or LAD2: predicted or actual, respectively), tumor predicted PLOS ONE | https://doi.org/10.1371/journal.pone.0200488 July 10, 2018 3 / 15 90 The predictive value of SPECT/CT imaging in mCRC response after Y-radioembolization 99m 90 Table 1. Details of Tc-MAA imaging and Y bremsstrahlung. 99m 90 Tc-MAA imaging Y-SPECT/CT bremmstrahlung Energy window: 140 keV +/- 15%; collimator type: LEHR; TOMO heads in H mode Collimator type: HEGP, COR correction on Whole body imaging (WB) Energy window: 140 keV +/- 100% Scan mode: continuous; exposure time per pixel: 240 sec; Heads in H mode; start angle 0 ; body contour on speed: 10 cm/min; body contour on Scan mode: step and shoot, 30 sec per projection o o ANT-POST SPOT abdomen imaging Total angular range 360 , arc per detector 180 , view Stop on counts: 800 kcts, reached on each detector angle 6 , numer of views: 60 independently Matrix 128x128, zoom 1.0 Scan location: H mode, start angle 0 ; body contour off; Emission first, followed by CT/AC matrix 256x256, zoom 1.0 CT LLAT-RLAT SPOT abdomen imaging Scan type: axial; slice thickness: 5.0 mm Stop on counts: 800 kcts, reached on each detector X ray voltage 140 kV, current 2.5 mA; rotation independently velocity: 2.6 RPM Scan location: H mode, start angle 270 ; body contour off; CT reconstruction: matrix 512x512, pixel size 1.10 matrix 256x256, zoom 1.0 mm, filter stnd ANT-POST SPOT chest imaging Stop on time based on the ANT-POST SPOT abdomen imaging duration Scan location: H mode, start angle 0 ; body contour off; matrix 256x256, zoom 1.0 SPECT/CT abdomen imaging TOMO COR correction on; start angle 0 ; body contour on Scan mode: step and shoot, 30 sec per projection o o Total angular range 360 , arc per detector 180 , view angle 6 , numer of views: 60 Matrix 128x128, zoom 1.0 Emission first, followed by CT/AC CT Scan type: axial; slice thickness: 5.0 mm X ray voltage 140 kV, current 2.5 mA; rotation velocity: 2.6 RPM CT reconstruction: matrix 512x512, pixel size 1.10 mm, filter stnd https://doi.org/10.1371/journal.pone.0200488.t001 absorbed dose (PAD), and tumor actual absorbed dose (AAD) were calculated according to the following formulas, using the volume as a surrogate for the mass: 2-3) [7, 14±16]. 90Y A‰GBqŠ 49; 67…1 LB† LAD1=LAD2‰GyŠ ˆ …2† V ‡…mT=N V † Healthy liver tumor n 90 99m LAD1/LAD2- Y liver absorbed dose (Gy) calculated separately for TC-MAA and Y-SPECT/CT 90Y 90 A ± Y activity to administer (GBq) LB ± lung breakthrough mT/N ± modified-tumor-to-normal uptake ratio (mT/N1 or mT/N2 respectively) n ± number of liver tumors TD …Gy† ˆ mT=N LAD …3† PAD=RAD TD - Y tumor predicted or actual absorbed dose (Gy), respectively PAD/AAD PLOS ONE | https://doi.org/10.1371/journal.pone.0200488 July 10, 2018 4 / 15 90 The predictive value of SPECT/CT imaging in mCRC response after Y-radioembolization Fig 1. Gamma counts measurement in SPECT/CT after establishing ROI on CT scan with the longest tumor diameter. https://doi.org/10.1371/journal.pone.0200488.g001 Statistical analysis Continuous variables are reported as mean ± SD, and categorical variables as percentages. ANOVAs or Student's t-test were used for continuous variables if the distributions were suit- able; if not, the Kruskal-Wallace test or Mann-Whitney U test were used. For comparisons of categorical variables, the Wilcoxon signed-rank test was used. Median OS (overall survival) and PFS (progression-free survival) were computed using the Kaplan±Meier estimator. The log rank test was used for comparisons of time to progression (TTP ). Correlations tumor between variables were assessed with Spearman's correlation coefficient. Multivariate corre- spondence analysis was used for tumors to assess complete response (CR) and disease progres- sion (PD). Receiver operating characteristic (ROC) curves were generated for PAD and mT/ N1. We performed calculations for PAD based on the generally accepted threshold of 70 Gy [6]. P values less than 0.05 were considered significant. Statistical calculations were carried out using STATISTICA 12.0 software. Results Phantom data As shown in Table 2 and Fig 2, gamma counts from the middle (widest diameter) section closely matched those obtained from the entire sphere, indicating that they could be used to calculate the liver tumor absorbed dose. Table 2. Phantom data. Sphere No. Volume (ml) Diameter (mm) Y Activity (GBq) phT/N (±SD) total phT/N p median layer 1 25.52 37 100.76 37.33 (±2.8) 39.80 0.148 2 11.49 28 43.662 21.11 (±2.01) 22.15 0.261 3 5.58 22 21.204 14.36 (±0.78) 15.10 0.106 4 2.57 17 9.766 9.08 (±0.47) 9.40 0.335 https://doi.org/10.1371/journal.pone.0200488.t002 PLOS ONE | https://doi.org/10.1371/journal.pone.0200488 July 10, 2018 5 / 15 90 The predictive value of SPECT/CT imaging in mCRC response after Y-radioembolization Fig 2. ªTumor-to-normal-tissue-ratio in the phantom: phT/Nº in particular layers of sphere No.1 of IEC-phantom. There is no significant difference between T/N ratio of median layer (red) and mean phantom total T/N ratio of sphere No. 1. (p = 0.148, Student t-test). https://doi.org/10.1371/journal.pone.0200488.g002 Patient data The SIRT procedures were performed with technical success, and no serious adverse events were observed. Overall survival (OS) was 17.3 months and progression-free survival (PFS) was 6 months. The clinical and treatment data are summarized in Table 3. Tumor and dosimetric data Initially 112 liver tumors within 21 patients were found, but for 9 tumors the target lesion did not meet the inclusion criteria. Thus, data for 103 liver tumors were analyzed. (Table 4) The overall tumor response rate (ORR ) was 36% (37 tumors). Their median values of tumor descriptive parameters were as follows: T0 = 19 mm, mT/N1 ratio = 2.1, mT/N2 ratio = 1.8, PAD = 96 Gy, AAD = 89 Gy. The median time to tumor progression (TTP ) obtained tumor using the Kaplan-Meier estimator was 7 months. Standard tumor to normal liver ratio According to the Kruskal-Wallis test, the ratios sT/N1 (mean 2.87 ± 0.54) and sT/N2 (mean 1.86 ± 0.39) and the derived parameters sPAD (mean 156.74 ± 40.39) and sAAD (mean 109.08 ± 44.46) were all non-significantly associated with tumor response, regardless of its extent (p = 0.053, p = 0.876, p = 0.23, p = 0.077, respectively). PLOS ONE | https://doi.org/10.1371/journal.pone.0200488 July 10, 2018 6 / 15 90 The predictive value of SPECT/CT imaging in mCRC response after Y-radioembolization Table 3. Patient clinical and treatment data. VARIABLE VALUE p-value GENDER (M/F) 17/4 nd AGE nd Mean 56 Range 35-67 Median 54 LIVER VOLUME (ml) ns Mean 1712 Range 920-2424 Median 1714 TUMOR VOLUME (ml) 0.03 Mean 276 Range 7-991 Median 208 % OF LIVER INFILTRATION 0.04 Mean 15 Range 1-52 Median 12 INJECTED 90Y ACTIVITY (GBq) ns Mean 1.8 Range 1.0-2.8 Median 1.8 PREDICTED LIVER ABSORBED DOSE (Gy) 0.04 Mean 43 Range 21-64 Median 48 ACTUAL LIVER ABSORBED DOSE (Gy) ns Mean 44 Range 21-63 Median 49 LUNG SHUNT (%) ns Mean 7.4 Range 3.5-13 Median 7 TREATMENT RESULTS CR 0 PR 4 SD 14 PD 3 - correlations between variable and tumor response (according to RECIST1.1) 2 , ± ANOVA ± U-Mann-Whitney test https://doi.org/10.1371/journal.pone.0200488.t003 Modified tumor-to-normal-liver ratio The ratio mT/N1 was higher for tumors that were responsive to treatment (2.4 vs 1.9, p = 0.003) and showed positive correlations with mT/N2 (R: 0.51, p < 0.001), PAD (R: 0.6, p < 0.001), AAD (R: 0.27, p = 0.004) and TTP . The ROC curve suggested a cutoff for mT/ tumor PLOS ONE | https://doi.org/10.1371/journal.pone.0200488 July 10, 2018 7 / 15 90 The predictive value of SPECT/CT imaging in mCRC response after Y-radioembolization N1 at the level 1.7 (p = 0.005). TTP was significantly longer for tumors with mT/N1 tumor higher than 1.7 than for tumors with lower ratios (6.9 vs 1.7 months, p < 0.001, log rank test). The risk of progression for tumors with mT/N1 lower than 1.7 was higher than for tumors with higher ratios HR: 2.1, CI:1.4±3.3, p = 0.001) (Fig 3) The ratio mT/N2 was also higher for tumors that were responsive to treatment (1.9 vs 1.8 ns) and showed positive correlations with mT/N1 (R: 0.51, p < 0.001), PAD (R: 0.2, p = 0.038), and AAD (R: 0.43, p < 0.001). Predicted and actual tumor Y absorbed dose Detailed information on PAD and AAD is presented in Table 4. For 37 tumors that were responsive to treatment, the mean PAD was 113 Gy and the mean AAD was 98 Gy (p < 0.001, Wilcoxon test). For 66 non-responding tumors the means were 82 Gy and 77 Gy, respectively (p<0.05). Negative correlations were found for PAD with T0 (R: - 0.45, p < 0.001), PAD with TUMOR VOLUME (R: - 0.43, p < 0.001), AAD with T0 (R: - 0.41, p < 0.001), and AAD with TUMOR VOLUME (R: - 0.4, p < 0.001). Both PAD and AAD showed significant correlations with tumor response after the SIRT procedure (Kruskal-Wallis test and ANOVA, respectively). In NIR ANOVA post-hoc tests performed for AAD, significant differences between CR, PR, SD and PD were revealed (e.g., 82 Gy in CR vs 50 Gy in PD, p = 0.031). Results of calculations conducted for tumors with absorbed dose higher or lower than 70 Gy and size smaller or larger than 20 mm are shown in Table 4 and Fig 4. TTP was significantly longer for tumors with tumor PAD higher than 70 Gy (6.9 vs 2.1 months, p = 0.004, log rank test). The risk of progression was elevated for tumors with PAD lower than 70 Gy (HR: 2.0, CI: 1.3±3.2, p = 0.001). The multivari- ate correspondence analysis performed for tumors with complete response (CR) and progres- sion disease (PD) revealed a large impact of PAD and AAD on these variables. (Fig 5) Discussion The aim of this study was to assess the value of SPECT/CT in predicting the effects of radioem- bolization for liver tumors, using a simplified method of calculation of the tumor-to-normal- Fig 3. Evaluation of time to tumor progression (TTPtumor) for the threshold of mT/N1 higher or lower than 1.7 (left) and PAD higher or lower than 70Gy (right). https://doi.org/10.1371/journal.pone.0200488.g003 PLOS ONE | https://doi.org/10.1371/journal.pone.0200488 July 10, 2018 8 / 15 90 The predictive value of SPECT/CT imaging in mCRC response after Y-radioembolization Table 4. Radiological response parameters and the dosimetric calculations for each tumor. VARIABLE Value for 103 tumors >70 Gy (75 tumors) <70 Gy (68 tumors) >20 mm (68 tumors) <20 mm (35 tumors) 1 3 3 3 3 2 INITIAL TUMOR SIZE (T0, mm) p = 0.001 ns ns p = 0.026 ns Mean 34 27 52 44 15 Range 10-117 10-94 15-117 20-117 10-19 Median 25 22 43 35 15 1 2 3 3 2 3 TUMOR VOLUME (mm) p = 0.01 ns ns p = 0.038 ns Mean 58 27 142 87 2.5 Range 1-847 1-296 1.6-847 4-847 1-18 Median 9 6 44 22 2 1 3 3 3 3 3 mT/N1 RATIO p = 0.012 ns ns ns p = 0.001 Mean 2.1 2.2 1.7 1.9 2.3 Range 1.0-6.8 1.1-6.8 1.0-3.2 1.0-5.0 1.1-6.8 Median 1.9 2.0 1.6 1.8 2.1 1 3 3 2 3 2 mT/N2 RATIO ns ns ns ns ns Mean 1.8 1.8 1.8 1.8 1.8 Range 1.0-3.4 1.0-3.4 1.0-3.1 1.0-3.4 1.1-2.8 Median 1.8 1.8 1.7 1.8 1.8 1 3 3 3 3 3 PAD (Gy) p<0.001 p = 0.005 ns ns p = 0.001 Mean 95 110 56 82 120 Range 33-376 71-376 36-69 33-248 61-376 Median 85 94 58 80 103 1 2 2 2 3 2 AAD (Gy) p<0.001 p = 0.008 ns p = 0.011 p = 0.028 Mean 84 94 56 76 98 Range 32-168 52-168 34-77 32-168 58-155 Median 78 89 56 71 94 RESPONSE CR 7 7 0 2 5 PR 30 26 4 16 14 SD 61 41 20 46 15 PD 5 1 4 4 1 1 - correlations between variable and tumor's response (according to RECIST1.1) 2 ± ANOVA 3 ±K-W test 99m 90 mT/N1 ± modified tumor to normal liver ratio in Tc-MAA SPECT/CT, mT/N2 ± modified tumor to normal liver ratio in Y-SPECT/CT (bremsstrahlung), PAD - 90 90 90 prognostic tumor Y absorbed dose calculated in 99mTC-MAA SPECT/CT (MIRD), AAD ± real tumor Y absorbed dose calculated in Y -SPECT/CT (bremsstrahlung) https://doi.org/10.1371/journal.pone.0200488.t004 liver ratio (mT/N). SPECT/CT scans are performed two times during the radioembolization 99m 90 procedure. The results showed that if Tc-MAA and Y-microspheres are similar in size, it is possible to predict the actual distribution of the latter in the liver [9, 10]. SPECT/CT imaging can give valuable dosimetric data, crucial for both planning and evalu- ating SIRT treatment. We have proposed a fast and easy method for dosimetric calculations [13]. The method relies on the RECIST 1.1 criteria, in which the longest tumor diameter (not the tumor volume) is used for further evaluation on follow-up CT scans. These criteria, adopted for SPECT/CT imaging, reflect the notion that counts from the largest lesion diameter accurately reflect the counts in the entire tumor. We confirmed in the SPECT phantom study that gamma counts from the section with the longest tumor diameter are usable to calculate the tumor absorbed dose. PLOS ONE | https://doi.org/10.1371/journal.pone.0200488 July 10, 2018 9 / 15 90 The predictive value of SPECT/CT imaging in mCRC response after Y-radioembolization Fig 4. Tumor prognosed absorbed dose (a) and tumor actual absorbed dose (b) distribution depending on tumor response. Tumor prognosed absorbed dose (a) and tumor actual absorbed dose (b) distribution depending on tumor size (threshold at 20mm). https://doi.org/10.1371/journal.pone.0200488.g004 99m The first step to achieving this goal in the liver is to measure the gamma counts in the Tc- MAA and Y-Bremsstrahlung SPECT/CT slices with the longest tumor diameters and calculate the tumor-to-normal-liver ratios (mT/N1 and mT/N2 respectively). Other dosimetric parame- ters, such as predicted tumor dose (PAD) and actual tumor dose (AAD), can then be calculated [7, 13±16]. The process of gathering dosimetric parameters is fast, taking only 1-2 minutes for liver tumors. From a dosimetric point of view, the tumor-to-normal-liver ratio is a very interest- ing parameter because it reflects blood flow in the tumor, as the hepatic artery is the main source of blood supply for liver lesions[1, 17, 18]. This value also forms a basis for calculating further dosimetric parameters, especially PAD for pre-treatment procedures and AAD after radioembolization [7, 14, 16]. As we observed, the mean mT/N1 was 2.1, which means two times greater blood flow in the tumor than in the healthy liver. The mT/N1 ratio was higher for responsive tumors and showed a positive correlation with TTP , thus it may be used as a predictor of tumor tumor response. This knowledge should give clinicians an opportunity to plan better treatments, by re-evaluation of the Y dose before radioembolization. We should remember, however, that calculations made using mT/N1 may sometimes be less accurate due to differences in the dis- 99m 90 tribution of Tc-MAA and Y-microspheres in the liver caused by various factors (e.g., in vivo blood flow variations)[10, 18]. If we take these limitations into account, the use of mT/N1 is simple and gives us the possibility of quick evaluation of conditions of planned Y treat- ment. The mT/N2 ratio, which is calculated after radioembolization, would potentially be much more helpful because it reflects the actual Y dose distribution within liver tumors. Unfortunately, mT/N2 (p = 0.057) is not adequate as a prognostic factor, although its p-value PLOS ONE | https://doi.org/10.1371/journal.pone.0200488 July 10, 2018 10 / 15 90 The predictive value of SPECT/CT imaging in mCRC response after Y-radioembolization Fig 5. The multivariate correspondence analysis (MCA) for tumors with complete response (CR) and disease progression (PD). https://doi.org/10.1371/journal.pone.0200488.g005 is close to statistical significance and a study with a larger population might give positive results. It is worth mentioning that AAD calculated using mT/N2 showed significant correla- tion with treatment results. Absorbed dose evaluation Based on mT/N1 and mT/N2 it is possible to calculate PAD and AAD for tumors and liver absorbed dose (LAD) [7, 14]. As we observed in the multiple correspondence analysis, the magnitude of predicted or actual absorbed dose is a major factor determining the tumor response (Fig 5). The mean PAD was 95 Gy, and PAD showed a strong correlation with tumor response. TTP was longer when PAD was higher than 70 Gy. There were 75 tumors with tumor PAD higher than 70 Gy (minimal absorbed dose for tumor destruction). Within this group, there were 7 lesions with CR (all in the entire study) and only one lesion with PD. Interest- ingly, the median PAD was higher in tumors with PR than with CR. Complete destruction of tumors with lower PAD may be partially explained by a more uniform distribution of micro- spheres in smaller lesions. Larger, partially necrotic tumors are more resistant, even if they absorb higher Y doses. The mean AAD was 84 Gy, and AAD was also a predictor of tumor response. Moreover, in terms of AAD all 5 tumors with PD were below the threshold of 70 Gy, whereas for PAD only 4/5 were below this threshold. It makes sense that AAD would better predict tumor response, because this value is calculated after SIRT treatment and reflects the actual Y-microspheres deposition. We should, however, remember that AAD is determined by the mT/N2 ratio, which was not statistically significant in our study. PAD remains only value calculated before treatment that may predict the response to treatment. As shown in the Results section, the difference between PAD and AAD in non-responding tumors was not statistically significant. A possible explanation is that non-responders fail to PLOS ONE | https://doi.org/10.1371/journal.pone.0200488 July 10, 2018 11 / 15 90 The predictive value of SPECT/CT imaging in mCRC response after Y-radioembolization 99m 90 absorb sufficient doses in both the preparatory ( Tc) and treatment phases ( Y). In our opinion, the prognostic evaluation of SIRT treatment should be undertaken using all dosimet- ric parameters, because there is no single prognostic factor that is strongly and reliably predic- tive of treatment results. Other factors, for example tumor size or volume, may also be important. Consistent with previous reports, our study shows that tumors smaller than 20 mm gather higher mean absorbed doses than larger tumors and show a better response to radioem- bolization [2, 19, 20]. Information about Y-microsphere distribution within the liver may be a key to improve- ment of SIRT treatment results. Using this information, patients prone to potential radiation- induced liver disease (RILD) or patients who accumulate non-curative tumor doses may be identified immediately after the SIRT procedure. Moreover, these data may be useful during treatment planning for much more selective tumor destruction. In other words, mT/N1 and PAD are important, because they allow changes in radioembolization before the procedure itself. Several papers have shown the usefulness of the T/N ratio in prediction of SIRT treat- ment results. This applies especially in hepatocellular carcinoma, which is known for its devel- oped tumor vasculature and high T/N ratio [21±23]. In comparison to hepatocellular carcinoma, colorectal metastases are characterized by low blood flow and a lower T/N ratio, so there is no easy way of applying dosimetric parameters for assessing tumor response after treatment [18, 24]. As we can observe in previous reports, there are several methods of measuring dosimetric 99m data in SPECT/CT. In the partition model, analysis of Tc-MAA distribution between healthy liver and tumor compartments is performed on SPECT images (2D slices recon- structed from 3D projections of the organ), and the activity of Y is calculated [1, 7, 10, 18, 22, 25]. In some cases, it is difficult to establish proper tumor ROIs (especially in liver with multi- ple lesions). Moreover, the presence of ªhot spotsº does not mean there is a tumor in each case; these may be healthy liver areas with higher macroalbumin accumulation. Some authors 2 3 have used ROIs (or VOIs) with strictly defined size (e.g., 1 cm or cm , smaller than the tumor size) and defined them on the basis of the researcher's impressions), while others have admin- istered angiotensin shortly before radioembolization, which should have a significant influence on microsphere deposition in the liver and would considerably alter the pre-treatment proce- dure [18]. In another study, the authors used only a visual (qualitative) scale of Y deposition [26]. An interesting new proposal is to use functional dual-tracer SPECT imaging with fusion 99m 99m Tc-MAA and Tc-Sulphur colloid with semiautomatic tumor segmentation. 99m A study performed by Lam on 122 patients proved the usefulness of Tc-MAA-based dosimetry in prediction of radioembolization results [27]. Within this group, there were 29 patients with metastatic colorectal cancer (which corresponds to 58 tumors investigated based on the RECIST 1.1 criteria). The mean tumor absorbed dose in this study was 30 Gy, although the generally accepted therapeutic dose is much higher, i.e., 70 Gy. This discrepancy may be explained by differences between anatomo-functional and functional methods of gamma count measurement [27, 28]. This implies that there is no simple and direct comparison between these methods. Moreover, the highly sophisticated and time-consuming process of data gathering is in our opinion a barrier to the widespread use of this method. The above-mentioned methods are based on a standard approach to calculating the tumor- to-normal-liver ratio. The sT/N1 and sT/N2 ratios were proved to be of limited predictive value of tumor response in our study, as they were in majority of previous reports [1, 7, 10, 18, 22, 25]. We have presented a simpler method of dosimetric calculation on a similar cohort of patients (21 patients, 103 tumors). In our opinion, the advantages of gamma count measure- ment from the scan section with the largest diameter are simplicity, repeatability and represen- tativeness. Our method for calculating the tumor to liver ratio proved to be a valuable tool for PLOS ONE | https://doi.org/10.1371/journal.pone.0200488 July 10, 2018 12 / 15 90 The predictive value of SPECT/CT imaging in mCRC response after Y-radioembolization predicting treatment results and could serve as a basis for further dosimetric calculations. The resulting information about tumors dosimetric parameters obtained prior to treatment should improve treatment planning, and may help in selecting tumors with insufficient PAD. At this stage, it may be possible to calculate a new slightly higher level of Y-microsphere activity that may be sufficient to destroy all tumors and still be safe for liver tissue. The alternative approach is to plan much more selective Y-microsphere injection within tumor areas with PAD lower than 70 Gy to achieve a curative Y dose. Knowing about insufficient AAD on the day of radioembolization may also allow immediate planning of treatment using other methods (especially radiofrequency ablation), focused only on resistant tumors. In this study, when we focus on the liver absorbed dose, both predicted and actual doses were almost the same and no serious adverse events were observed. As we can see in the stan- dard guidelines, the limit of liver tolerance is 30 Gy, although some authors have reported that higher doses (up to 70 Gy) were well tolerated by liver tissue [4, 5, 14]. Our study has several limitations. First, we assumed uniform distributions of macroalbu- min and microspheres within the liver and tumor compartments. Some tumors had central necrosis areas that did not accumulate the tracer, but these were not excluded from the analy- sis. Second, we hypothesized that if the method proved to be applicable for varied tumor mor- phologies, then in a more uniform group, the prognostic value of the results would be greater. Third, even though ROC curve analysis suggested that the cutoff point for PAD should be 87 Gy, we performed calculations using the generally accepted threshold of 70 Gy. Fourth, in accordance with our assumptions regarding the uniformity of dose distribution, gamma counts from the middle (widest diameter) section closely matched those obtained from the entire sphere. Conclusions 99m The mT/N1 ratio and PAD based on Tc-MAA SPECT/CT and AAD calculated from Y-SPECT/CT can be used as predictors of radioembolization results. TTP was signifi- tumor cantly longer with mT/N1 greater than 1.7 and for tumors with PAD greater than 70 Gy. The risk of progression was elevated for tumors with mT/N1 lower than 1.7 and PAD lower than 70 Gy. The mT/N2 ratio had no significant correlation with treatment results. Supporting information S1 Table. Raw data for the analysis. (XLSX) Author Contributions Conceptualization: Piotr Piasecki. Data curation: Piotr Piasecki, Jerzy Narloch, Krzysztof Brzozowski, Piotr Zięcina, Andrzej Mazurek, Anna Budzyńska. Formal analysis: Piotr Piasecki, Krzysztof Brzozowski, Andrzej Mazurek, Jan Korniluk. Investigation: Piotr Piasecki, Piotr Zięcina. Methodology: Piotr Piasecki, Andrzej Mazurek, Anna Budzyńska, Mirosøaw Dziuk. Project administration: Piotr Piasecki. Software: Piotr Piasecki, Jerzy Narloch, Piotr Zięcina, Andrzej Mazurek, Anna Budzyńska. PLOS ONE | https://doi.org/10.1371/journal.pone.0200488 July 10, 2018 13 / 15 90 The predictive value of SPECT/CT imaging in mCRC response after Y-radioembolization Supervision: Mirosøaw Dziuk. Validation: Jerzy Narloch, Krzysztof Brzozowski, Andrzej Mazurek, Jan Korniluk, Mirosøaw Dziuk. Visualization: Piotr Piasecki, Jerzy Narloch. Writing ± original draft: Piotr Piasecki. Writing ± review & editing: Jerzy Narloch, Andrzej Mazurek, Mirosøaw Dziuk. References 1. Gulec SA, Mesoloras G, Dezarn WA, McNeillie P, Kennedy AS. Safety and efficacy of Y-90 micro- sphere treatment in patients with primary and metastatic liver cancer: the tumor selectivity of the treat- ment as a function of tumor to liver flow ratio. J Transl Med. 2007; 5:15. https://doi.org/10.1186/1479- 5876-5-15 PMID: 17359531; PubMed Central PMCID: PMCPMC1845138. 2. Lewandowski RJ, Memon K, Mulcahy MF, Hickey R, Marshall K, Williams M, et al. Twelve-year experi- ence of radioembolization for colorectal hepatic metastases in 214 patients: survival by era and chemo- therapy. Eur J Nucl Med Mol Imaging. 2014; 41(10):1861±9. https://doi.org/10.1007/s00259-014-2799- 2 PMID: 24906565. 3. Dawson LA, Ten Haken RK, Lawrence TS. Partial irradiation of the liver. Semin Radiat Oncol. 2001; 11 (3):240±6. PMID: 11447581. 4. INGOLD JA, REED GB, KAPLAN HS, BAGSHAW MA. RADIATION HEPATITIS. Am J Roentgenol Radium Ther Nucl Med. 1965; 93:200±8. PMID: 14243011. 5. Gray BN, Burton MA, Kelleher D, Klemp P, Matz L. Tolerance of the liver to the effects of Yttrium-90 radiation. Int J Radiat Oncol Biol Phys. 1990; 18(3):619±23. PMID: 2318695. 6. Kennedy AS, Nutting C, Coldwell D, Gaiser J, Drachenberg C. Pathologic response and microdosimetry of (90)Y microspheres in man: review of four explanted whole livers. Int J Radiat Oncol Biol Phys. 2004; 60(5):1552±63. https://doi.org/10.1016/j.ijrobp.2004.09.004 PMID: 15590187. 7. Gulec SA, Mesoloras G, Stabin M. Dosimetric techniques in 90Y-microsphere therapy of liver cancer: The MIRD equations for dose calculations. J Nucl Med. 2006; 47(7):1209±11. PMID: 16818957. 8. Kennedy A. Selective internal radiation therapy dosimetry. Future Oncol. 2014; 10(15 Suppl):77±81. https://doi.org/10.2217/fon.14.230 PMID: 25478774. 9. Kennedy A, Nag S, Salem R, Murthy R, McEwan AJ, Nutting C, et al. Recommendations for radioembo- lization of hepatic malignancies using yttrium-90 microsphere brachytherapy: a consensus panel report from the radioembolization brachytherapy oncology consortium. Int J Radiat Oncol Biol Phys. 2007; 68 (1):13±23. https://doi.org/10.1016/j.ijrobp.2006.11.060 PMID: 17448867. 10. Salem R, Lewandowski RJ, Gates VL, Nutting CW, Murthy R, Rose SC, et al. Research reporting stan- dards for radioembolization of hepatic malignancies. J Vasc Interv Radiol. 2011; 22(3):265±78. https:// doi.org/10.1016/j.jvir.2010.10.029 PMID: 21353979; PubMed Central PMCID: PMCPMC4337835. 11. van Hazel GA, Heinemann V, Sharma NK, Findlay MP, Ricke J, Peeters M, et al. SIRFLOX: Random- ized Phase III Trial Comparing First-Line mFOLFOX6 (Plus or Minus Bevacizumab) Versus mFOL- FOX6 (Plus or Minus Bevacizumab) Plus Selective Internal Radiation Therapy in Patients With Metastatic Colorectal Cancer. J Clin Oncol. 2016; 34(15):1723±31. Epub 2016/02/22. https://doi.org/ 10.1200/JCO.2015.66.1181 PMID: 26903575. 12. Wasan HS, Gibbs P, Sharma NK, Taieb J, Heinemann V, Ricke J, et al. First-line selective internal radiotherapy plus chemotherapy versus chemotherapy alone in patients with liver metastases from colorectal cancer (FOXFIRE, SIRFLOX, and FOXFIRE-Global): a combined analysis of three multicen- tre, randomised, phase 3 trials. Lancet Oncol. 2017; 18(9):1159±71. Epub 2017/08/03. https://doi.org/ 10.1016/S1470-2045(17)30457-6 PMID: 28781171; PubMed Central PMCID: PMCPMC5593813. 13. Eisenhauer EA, Therasse P, Bogaerts J, Schwartz LH, Sargent D, Ford R, et al. New response evalua- tion criteria in solid tumours: revised RECIST guideline (version 1.1). Eur J Cancer. 2009; 45(2):228± 47. https://doi.org/10.1016/j.ejca.2008.10.026 PMID: 19097774. 14. Bernardini M, Smadja C, Faraggi M, Orio S, Petitguillaume A, Desbre  e A, et al. Liver Selective Internal Radiation Therapy with (90)Y resin microspheres: comparison between pre-treatment activity calcula- tion methods. Phys Med. 2014; 30(7):752±64. https://doi.org/10.1016/j.ejmp.2014.05.004 PMID: 15. Dieudonne  A, Garin E, Laffont S, Rolland Y, Lebtahi R, Leguludec D, et al. Clinical feasibility of fast 3- dimensional dosimetry of the liver for treatment planning of hepatocellular carcinoma with 90Y- PLOS ONE | https://doi.org/10.1371/journal.pone.0200488 July 10, 2018 14 / 15 90 The predictive value of SPECT/CT imaging in mCRC response after Y-radioembolization microspheres. J Nucl Med. 2011; 52(12):1930±7. https://doi.org/10.2967/jnumed.111.095232 PMID: 16. Ho S, Lau WY, Leung TW, Chan M, Ngar YK, Johnson PJ, et al. Partition model for estimating radiation doses from yttrium-90 microspheres in treating hepatic tumours. Eur J Nucl Med. 1996; 23(8):947±52. PMID: 8753684. 17. Ackerman NB, Lien WM, Kondi ES, Silverman NA. The blood supply of experimental liver metastases. I. The distribution of hepatic artery and portal vein blood to "small" and "large" tumors. Surgery. 1969; 66(6):1067±72. PMID: 5402533. 18. Dhabuwala A, Lamerton P, Stubbs RS. Relationship of 99mtechnetium labelled macroaggregated albu- min (99mTc-MAA) uptake by colorectal liver metastases to response following Selective Internal Radia- tion Therapy (SIRT). BMC Nucl Med. 2005; 5:7. https://doi.org/10.1186/1471-2385-5-7 PMID: 16375764; PubMed Central PMCID: PMCPMC1360059. 19. Anderson JH, Angerson WJ, Willmott N, Kerr DJ, McArdle CS, Cooke TG. Is there a relationship between regional microsphere distribution and hepatic arterial blood flow? Br J Cancer. 1992; 66 (2):287±9. PMID: 1503900; PubMed Central PMCID: PMCPMC1977807. 20. Campbell AM, Bailey IH, Burton MA. Analysis of the distribution of intra-arterial microspheres in human liver following hepatic yttrium-90 microsphere therapy. Phys Med Biol. 2000; 45(4):1023±33. PMID: 21. Gales S, Peters S, Salehi N, Better N. Tailoring 99mTc Macroaggregated Albumin administration to optimize patient dose reduction. J Nucl Med Technol. 2015. https://doi.org/10.2967/jnmt.114.151811 PMID: 26271803. 22. Garin E, Lenoir L, Rolland Y, Edeline J, Mesbah H, Laffont S, et al. Dosimetry based on 99mTc-macro- aggregated albumin SPECT/CT accurately predicts tumor response and survival in hepatocellular carci- noma patients treated with 90Y-loaded glass microspheres: preliminary results. J Nucl Med. 2012; 53 (2):255±63. https://doi.org/10.2967/jnumed.111.094235 PMID: 22302962. 23. Knesaurek K, Machac J, Muzinic M, DaCosta M, Zhang Z, Heiba S. Quantitative comparison of yttrium- 90 (90Y)-microspheres and technetium-99m (99mTc)-macroaggregated albumin SPECT images for planning 90Y therapy of liver cancer. Technol Cancer Res Treat. 2010; 9(3):253±62. https://doi.org/10. 1177/153303461000900304 PMID: 20441235. 24. Wondergem M, Smits ML, Elschot M, de Jong HW, Verkooijen HM, van den Bosch MA, et al. 99mTc- macroaggregated albumin poorly predicts the intrahepatic distribution of 90Y resin microspheres in hepatic radioembolization. J Nucl Med. 2013; 54(8):1294±301. https://doi.org/10.2967/jnumed.112. 117614 PMID: 23749996. 25. Sarfaraz M, Kennedy AS, Lodge MA, Li XA, Wu X, Yu CX. Radiation absorbed dose distribution in a patient treated with yttrium-90 microspheres for hepatocellular carcinoma. Med Phys. 2004; 31 (9):2449±53. https://doi.org/10.1118/1.1781332 PMID: 15487724. 26. Ulrich G, Dudeck O, Furth C, Ruf J, Grosser OS, Adolf D, et al. Predictive value of intratumoral 99mTc- macroaggregated albumin uptake in patients with colorectal liver metastases scheduled for radioembo- lization with 90Y-microspheres. J Nucl Med. 2013; 54(4):516±22. https://doi.org/10.2967/jnumed.112. 112508 PMID: 23447653. 27. Lam MG, Wondergem M, Elschot M, Smits ML. Reply: A clinical dosimetric perspective uncovers new evidence and offers new insight in favor of 99mTc-macroaggregated albumin for predictive dosimetry in 90Y resin microsphere radioembolization. J Nucl Med. 2013; 54(12):2192±3. https://doi.org/10.2967/ jnumed.113.132852 PMID: 24198388. 28. Lam MG, Banerjee A, Goris ML, Iagaru AH, Mittra ES, Louie JD, et al. Fusion dual-tracer SPECT- based hepatic dosimetry predicts outcome after radioembolization for a wide range of tumour cell types. Eur J Nucl Med Mol Imaging. 2015; 42(8):1192±201. Epub 2015/04/28. https://doi.org/10.1007/s00259- 015-3048-z PMID: 25916740; PubMed Central PMCID: PMCPMC4480819. PLOS ONE | https://doi.org/10.1371/journal.pone.0200488 July 10, 2018 15 / 15
http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png
PLoS ONE
Unpaywall
http://www.deepdyve.com/lp/unpaywall/the-predictive-value-of-spect-ct-imaging-in-colorectal-liver-BAh9iImgvM
