ABSTRACT
Objectives:
Metastases and primary malignancies are common in the liver. Local ablative applications such as transarterial chemoembolization (TACE), and transarterial radioembolization (TARE) provide minimally invasive and safe treatment in unresectable liver tumors. Early detection of response to treatment prevents unnecessary toxicity and cost in non-responder patients and provides an earlier use of other options that may be effective. This study aimed to identify the role of 18F-fluorothymidine (FLT) positron emission tomography/computed tomography (PET/CT) in the assessment of early response to TACE and TARE treatments in patients with unresectable primary and metastatic liver tumors
Methods:
This single-center study included 63 patients who underwent 18F-FLT PET/CT for response evaluation after TACE and TARE. After excluding 20 patients whose data were missing 43 TARE-receiving patients were analyzed. The compatibility of change in semi-quantitative values obtained from the 18F-FLT PET/CT images with the treatment responses detected in 18F-fluorodeoxyglucose PET/CT, CT, and MR images and survival was evaluated.
Results:
There was no correlation between early metabolic, morphological response, and 18F-FLT uptake pattern, and change in standardized uptake values (SUV) which were ΔSUVmax, ΔSUVmean, ΔSUVpeak., ΔSUVmean, Δ SUVpeak values. There was no significant correlation between 18F-FLT uptake pattern, ΔSUVmax, ΔSUVmean, ΔSUVpeak, and overall survival, progression-free survival (PFS) for the target lobe PFS for the whole-body. The survival distributions for the patients with >30% change in Δ SUVmax and ΔSUVpeak values were statistically significantly longer than the patients with <30% change (p<0.009 and p<0.024, respectively).
Conclusion:
There was significant longer PFS for target liver lobe in patients with more than 30% decrease in 18F-FLT SUVmax and SUVpeak of the liver lesion in primary and metastatic unresectable liver tumors undergoing TARE.
Introduction
Both metastases and primary malignancies, such as hepatocellular carcinoma (HCC) and cholangiocellular cancer, are common in the liver. Metastases are the most common liver malignancy, and leading tumors metastasize to the liver are colorectal cancer, neuroendocrine tumors, other gastrointestinal cancers, and breast cancer. HCC is the sixth common cause of cancer and the third common cause of cancer-related deaths worldwide (1,2). Since the liver involvement is effective on survival, curative surgical applications are the first-line therapy, either with adjuvant chemotherapy or alone, providing the most significant survival advantage. However, surgery cannot be applied to most patients at diagnosis or tumor recurrence due to advanced-stage disease or inappropriate clinical status (1,3). Local ablative applications such as radiofrequency, microwave, and cryo-ablation, irreversible electroporation (IRE), endovascular transarterial chemoembolization (TACE), and transarterial radioembolization (TARE) provide minimally invasive and safe treatment (4,5,6,7,8). It has been reported that TARE and TACE provide long-term survival advantage and low toxicity, especially in patients with good performance and low tumor burden (5,7,8,9,10,11).
The prediction or early detection of response to therapy prevents unnecessary toxicity and cost that may be life-threatening in non-responder patients and provides an earlier use of other treatment options that may be effective. The morphological response evaluation with computed tomography (CT) and magnetic resonance (MR) requires a relatively long period and tumor shrinkage. Positron emission tomography (PET)/CT or PET/MR hybrid imaging, based on metabolic processes, provides earlier response assessment and concurrent anatomical information. 18F-fluorodeoxyglucose (FDG) is the most commonly used agent in PET imaging (12,13,14,15,16). However, since tumors with low-glucose metabolism and low cellularity, small-sized and well-differentiated tumors show low 18F-FDG uptake, alternative agents such as thymidine analog 3’-deoksy-(F-18)-3´-fluorotimidin (18F-FLT), which reflects the proliferation of cells, 11C-acetate, which reflects hypoxia, C11-choline or 18F-choline, which reflects aerobic metabolism (fatty acid synthesis) are being investigated (15,17,18,19,20,21,22,23).
18F-FLT, an analog of thymidine, is phosphorylated with thymidine kinase-1 (TK1) and is converted to 18F-FLT-monophosphate, which cannot penetrate DNA and is trapped in the cytosol. 18F-FLT is a TK1-specific substrate that increases in proliferating cells while not found in silent cells and correlates with a proliferation marker Ki-67 index (24,25). Imaging with 18F-FLT has advantages such as non-invasive quantitation of cell proliferation, three-dimensional tumor imaging, and evaluating the whole tumor proliferation heterogeneity in multiple tumor areas simultaneously. Studies show that tumor proliferation changes can be detected early with 18F-FLT PET/CT after radiotherapy (1,2,3). Knowing that TARE is an internal radiotherapy method, this study aimed to describe the role of 18F-FLT PET/CT in assessing the early response to TARE and TACE in patients with primary and metastatic liver tumors.
Materials and Methods
Ankara University Faculty of Medicine Human Research Ethics Committee Approval (İ3-117-19) was obtained for this single-center study with prospective and retrospective components and was performed under the Helsinki Directive and Good Clinical Practices Guidelines. Informed consent was obtained from all volunteers included in the study.
Results
Discussion
This study assessed the role of PET/CT with 18F-FLT, a radiopharmaceutical reflecting cell proliferation, in response evaluation after TARE and found significant longer PFS for the target liver lobe in patients with more than 30% decrease in 18F-FLT SUVmax and SUVpeak of the target liver lesion. There was no significant relationship between SUV values and treatment response.
Although there are metabolic and morphological techniques used for assessing treatment response, there is no standard response evaluation and follow-up protocol for TARE. Response evaluation after TARE is performed at different times with PET/CT, CT, or MR depending on the center’s practice. Since response assessment with CT and MRI takes a longer time and has their limitations, PET/CT and PET/MR, functional, molecular and anatomical imaging techniques, are used for early response evaluation with agents that reflect tumor-specific metabolism (13,14,15,16,18,21,22,23). 18F-FDG PET/CT is the most common metabolic imaging method due to increased glucose metabolism in many types of cancer. 18F-FDG PET/CT can be used to assess treatment response in poorly differentiated and high-grade tumors. However, since small and well-differentiated tumors (such as HCC, NET) show low or no 18F-FDG uptake due to low glucose metabolism and cellularity, imaging with new-tumor-specific agents is needed (13,16,21,22,23). PET/CT imaging with 18F-FLT, which reflects cell proliferation, is a non-invasive imaging method and has been used for the response evaluation (24,25,26,27). In addition to complex and competing factors in the FLT uptake mechanism, there are notable differences between patient preparation, imaging time after injection, protocol, amount of injected activity, reconstruction method, analysis techniques, timing before and after treatment, patient numbers, and disease groups in studies with F-18FLT PET/CT (24,25,26,27,28,29,30,31,32).
As far as it is known, this is the first study to investigate the role of 18F-FLT PET/CT in the early response evaluation after TARE. There are few studies investigating the role of FLT PET/CT in evaluating the liver-specific treatment response, considering high background liver uptake especially in HCC patients that hamper the detection of liver/lesions. Studies evaluated therapy of TACE-receiving HCC patients and systemic chemotherapy-receiving liver metastatic colorectal cancer patients (28,29,32).
Sharma et al. (32) investigated the role of 18F-FLT PET/CT in assessing treatment response to TACE in HCC patients. They used temporal-intensity voxel clustering [kinetic spatial filtering (KSF)] in lesion detection to overcome high background liver signal and thus 18F-FLT uptake but they could not achieve improvement in lesion detection by applying it. They reported 73% detection rate for pretreatment 18F-FLT PET, and 30% reduction in mean 18F-FLT PET uptake after TACE. In the current study, KSF could not be used due to unavailability, target lesion based detection rate for pretreatment 18F-FLT PET/CT was 53% (9/17) for HCC patients. In our study, although the change in 18F-FLT SUVmax, SUVmean and SUVpeak values (Table 3) had no significant relationship with treatment response; patients with more than 30% decrease in 18F-FLT SUVmax and SUVpeak of the target lesion had significant longer PFS for target liver lobe after TARE.
Mogensen et al. (29) investigated the role of 18F-FLT PET/CT in patients with at least one measurable colorectal cancer liver metastasis and received first-line chemotherapy. They reported a reduction in 18F-FLT uptake in 85% patients, whereas there was no relationship between the early change in measured 18F-FLT SUVmax, and RECIST 1.1 based response. In this study, similar to their study, there was no relationship between the change in SUV values (ΔSUVmax, ΔSUVmean, and ΔSUVpeak) and RECIST 1.1 and PERCIST-based responses. Contractor et al. (28) investigated the role of 18F-FLT PET/CT in evaluating the treatment response of breast and colorectal cancer liver metastases. They reported that SUVave and SUVmax showed a significant decrease in responders two weeks after the first-line chemotherapy, and the change in FLT uptake can distinguish those who responded to the treatment from non-responders. In our study, 18F-FLT PET/CT was evaluated for the treatment response after TARE, an locoregional therapy (LRT), not a systemic treatment and there was no significant difference in the change in SUVmax, SUVmean, and SUVpeak values among responder and non-responders.
The key point in the early evaluation of the treatment response is to distinguish non-responder to discontinue unnecessary treatment, thus avoid toxicity and cost. It is critical to distinguish the resectable disease from those who require more aggressive treatment. Patients with shrinkage of tumors up to 30% are considered to have stable disease, according to RECIST 1.1, and are unresponsive to treatment (12,13,15,16). In this study, tumor sizes of patients with stable disease decreased, reflecting the beneficial effect of the treatment. However, since this decrease in size remained below the RECIST 1.1 response criteria, it was accepted as a stable disease and unresponsive to treatment. It should be recognized that patients with stable disease, especially with colon cancer, are accepted as responders and continue to receive systemic treatment in clinical practice (33). Generally, chemotherapy-refractory liver metastases are referred for LRTs such as TARE. Thus, even defined stable disease can provide longer survival and can be accepted as responsive. If patients with stable disease are accepted as responders to therapy, statistical analysis can be found significantly in long-term follow-up results. Because liver resection was not performed on any patient after radioembolization, except for the transplantation patient, post-treatment histopathological tumor changes, background of persistent 18F-FLT hypometabolism and correlation of histopathology with 18F-FLT values could not be evaluated.
It can be argued that the timing of the 18F-FLT was not right. But, TARE is an internal radiotherapy procedure, and response to radiotherapy is generally evaluated later than chemotherapy/selective systemic therapies (27,30). 18F-FDG PET/CT and 18F-FLT PET/CT imaging were done approksimately ≥4 weeks after the procedure. Studies evaluating radiotherapy response revealed a significant relationship between 18F-FLT PET/CT and response or survival in patients with head-neck, esophageal, breast, lung, rectal, etc., cancer (30). This study distinguished real responders from non-responders who were grouped based on post-radiotherapy response assessment techniques 18F-FDG PET/CT and CT or MR. No correlation was found between the semi-quantitative values such as ΔSUVmax, ΔSUVmean, ΔSUVpeak, SUVmaxTBR, SUVmeanTBR, and SUVpeakTBR values calculated from 18F-FLT PET/CT images. There was a significant relationship with PFS for target liver lobe and >30% decrease in 18F-FLT SUVmax and SUVpeak of the target lesion.
Conclusion
This study found significantly longer PFS for the target liver lobe in patients with more than 30% decrease in 18F-FLT SUVmax and SUVpeak of the liver lesion in patients with primary and metastatic unresectable liver tumors undergoing TARE. The changes in 18F-FLT PET/CT SUVmax, SUVmean, SUVpeak, SUVmaxTBR, SUVmeanTBR, and SUVpeakTBR values had no significant relationship with response in 18F-FDG PET/CT or in contrast-enhanced CT/MR after TARE. 18F-FLT PET/CT can be used as an alternative/complementary imaging method to 18F-FDG PET/CT in the early evaluation of the treatment response in patients undergoing TARE for primary or secondary liver tumor.
Patients
The inclusion criteria of this study were TACE or TARE therapy for histologically/cytologically or radiologically diagnosed primary (HCC, cholangiocellular carcinoma) or metastatic liver tumor; staged with CT/MR, 18F-FDG PET/CT, or PET/MR; Eastern Cooperative Oncology Group performance score ≤2; over 18 years of age; follow-up more than three months; available data. Patients with claustrophobia and pain that prevent imaging and patients who did not want to participate in the study were excluded. There was no intervention in the treatment selection or management of the patients. According to the standard evaluations, the relevant specialist (medical oncology, gastroenterology specialist, or general surgeon) chose the treatment.
18F-FLT and 18F-FDG PET/CT Imaging
The presence or history of systemic or local ablative therapy, chronic disease, etc., can affect the evaluation was questioned and noted. To reduce the total body radiation dose and increase the image quality, oral hydration and emptying of the bladder before imaging was provided. Approximately 60 min after the 18F-FDG and 18F-FLT were given intravenously, the whole-body PET/CT imaging was performed starting 60 min after injection. Following at least 6 hours of fasting, when blood glucose level was <150 mg/dL, ~4-5 MBq/kg 18F-FDG was administered. Approximately 60 min after the administration of radiopharmaceutical, whole-body PET/CT images were obtained. FLT was synthesized in-house according to standard procedures (25). After administration of 3.4-9.3 mCi 18F-FLT intravenously, the whole-body PET/CT imaging was performed starting 60 min after injection (26,27,28,29,30). Following CT for attenuation correction, and anatomical correlation, whole-body PET images were obtained, in the supine position, from the vertex to the middle thigh, and 3 min per bed. PET/CT Discovery ST (GE Healthcare Waukesha, Wisconsin, USA) was used for PET/CT hybrid imaging. After assessing maximum intensity projection, cross-sectional and fusion images, areas with high, mixed (heterogeneous), equal and low uptake from adjacent liver parenchyma were noted. The same parameters and assessments were used for 18F-FDG and 18F-FLT imaging, which were performed twice, before the treatment as baseline and for response evaluation after therapy.
The target lesion was defined as sole or the largest lesion in the target lobe. Standardized uptake values (SUV): SUVmax, SUVmean, SUVpeak were calculated automatically for hypermetabolic and heterogeneous (mixed) target lesions on a workstation by using PET software (GE Healthcare). A 2 cm region of interest was manually defined for isometabolic, and hypometabolic target lesions on the summed images by using the same software. Since the reference (non-tumoral) liver parenchymal SUV values of the patients showed a significant difference both between the patients and the baseline and post-treatment images of the same patient, the target background ratio (TBR) of the target lesions were calculated by proportioning the SUV values of the target lesion to reference values and were evaluated separately. Reference SUV values were calculated by manually placing a 2 cm region of interest in the liver in a tumor-free area to measure background liver activity (26,28,29,30). Patients were divided into groups with and without the change of SUV values calculated from the difference between the target lesion’s post-treatment and pre-treatment SUV values, which were calculated and referred to as delta (Δ) SUV values.
Statistical Analysis
The changes between baseline and post-treatment 18F-FLT PET/CT images were compared to the responses detected with 18F-FDG PET/CT and CT/MRI, evaluated according to the PERCIST and RECIST 1.1 criteria, respectively, and progression-free survival (PFS) and overall survival (OS). All statistical analyses were performed using IBM SPSS for Windows, version 25.0 (SPSS, Armonk, NY: IBM Corp.). Kolmogorov-Smirnov test was used to assess the assumption of normality. The continuous variables that did not have a normal distribution were expressed as medians (minimum-maximum). For non-normally-distributed continuous variables, differences between groups were tested using Mann-Whitney U test and Kruskal-Wallis test. Lastly, Pearson chi-square analysis and Fisher’s Exact test determined associations between categorical variables, while Pearson and Spearman correlation analysis determined associations between continuous variables. The survival times of groups were obtained using Kaplan-Meier analysis and the difference in survival times between groups were compared with the Log Rank test. A two-sided p value <0.05 was considered as statistically significant.
Patients
Sixty-three consecutive patients were included in the study between December 2018 and January 2020, who underwent pre- and post-treatment 18F-FLT PET/CT to evaluate their response to TARE and TACE treatments. Although all patients underwent baseline imaging, 4 of the TACE-receiving patients and 16 of the TARE-receiving patients could not undergo 18F-FLT PET/CT or other imaging for response evaluation either due to decreased performance status that hindered further procedure or death. Since the patients who received TACE did not undergo PET/CT or CT/MRI to evaluate the response to treatment, and most of their data were missing TACE-receiving patients were excluded from the analysis. Forty-three TARE-receiving patients were analyzed to have a homogenous population and statistical analysis. Detailed patient characteristics are listed in Table 1.
18F-FLT, 18F-FDG PET/CT and CT/MRI
Other than one patient who did not undergo 18F-FDG PET/CT scanning for response evaluation because of the tumor’s 18F-FDG non-avidity at the baseline, all remaining patients underwent 18F-FLT PET/CT, 18F-FDG PET/CT, and CT/MR before and after TARE. The morphological response evaluation was performed with contrast-enhanced CT for 2 patients and with contrast-enhanced MR for 41 patients. Imaging characteristics of 18F-FLT PET/CT are given in Tables 2, 3; characteristics of 18F-FDG PET/CT and CT/MRI are given in Table 2. 18F-FLT PET/CT, 18F-FDG PET/CT, and contrast-enhanced liver MR of a patient with 18F-FDG non-avid, persistent 18F-FLT avid lesions and progressive disease are presented in Figure 1.
Correlation between the diagnosis, longest diameter of the target lesion, volume and percentage of tumors in the target lobe, age, the number of lesions in the target lobe, early metabolic, morphological response and 18F-FLT visual change, ΔSUVmax, ΔSUVmean, ΔSUVpeak, ΔSUVmaxTBR, ΔSUVmeanTBR, and ΔSUVpeakTBR values were not significant. Calculated p values from statistical analyses are presented in Table 4.
Survival
During 18.4 months follow-up, 22 patients died. OS was median 7.0 (3.3-17.4) months, PFS was median 3.4 (1.3-17.4) months for the target lobe; and median 3.2 (1.3-17.4) months for whole-body. There was no significant correlation between 18F-FLT visual change, ΔSUVmax, ΔSUVmean, ΔSUVpeak, ΔSUVmaxTBR, ΔSUVmeanTBR, and ΔSUVpeakTBR and OS, PFS for target lobes, and PFS for whole-body (Table 4). A log-rank test was run to determine whether there were differences in the target lobe’s PFS distribution for the ΔSUVmax and ΔSUVpeak groups when the cut-off >30% change was applied. The target lobe’s PFS for the patients with a >30% decrease in SUVmax was significantly longer than those without [350±57 days 95% confidence interval (CI) 238-463 vs. (130±21 days 95% CI 90-171 (χ2 (1): 6.774) p=0.009]. The target lobe’s PFS for the patients with more than 30% change in SUVpeak was statistically significantly longer than the patients with <30% change [338±59 days 95% CI 222-453 vs. 1730±38 days 95% CI 98-247 χ2 (1): 5.095, p=0.024]. Estimated survival chance at 209th day was 0.549±0.129 for 17 patients with no change in SUVmax, while the estimated survival chance at 92nd day was 0.500±0.098 in patients with more than 30% decrease in SUVmax. Estimated survival proportion at 209th days were 0.514±0.134 in 16 patients without change in SUVpeak value; while this proportion was 0.519±0.096 at 90th day for the patients with more than 30% decrease in SUVpeak value (Figure 2, Table 4).
Study Limitations
The most significant limitations of this study are the small sample size, consequent heterogeneous patient population, and the small number of patients who responded to the therapy. Therefore, in statistical analysis, results reaching a significant degree could not be obtained for SUV parameters. TARE candidate patients have different clinical scenarios, such as highly variable liver lesion number and size, disease stage, history of single or multi-step systemic treatment, liver resection, transplant, and LRT’s. Also, since there is a clear difference in disease etiologies, clinical and radiological status, it was not possible to standardize the patient group. Reproducible and re-applicable clinical data from a larger and standardized patient population are required to assess the role of 18F-FLT PET/CT in the evaluation of response to TARE treatment.