Automated segmentation of lung, liver, and liver tumors from Tc-99m MAA SPECT/CT images for Y-90 radioembolization using convolutional neural networks.

PURPOSE: 90 Y selective internal radiation therapy (SIRT) has become a safe and effective treatment option for liver cancer. However, segmentation of target and organ-at-risks is labor-intensive and time-consuming in 90 Y SIRT planning. In this study, we developed a convolutional neural network (CNN)-based method for automated lungs, liver, and tumor segmentation on 99m Tc-MAA SPECT/CT images for 90 Y SIRT planning. METHODS: 99m Tc-MAA SPECT/CT images and corresponding clinical segmentations were retrospectively collected from 56 patients who underwent 90 Y SIRT. The collected data were used to train three CNN-based segmentation algorithms for lungs, liver, and tumor segmentation. Segmentation performance was evaluated using the Dice similarity coefficient (DSC), surface DSC, and average symmetric surface distance (ASSD). Dosimetric parameters (volume, counts, and lung shunt fraction) were measured from the segmentation results and were compared with clinical reference segmentations. RESULTS: The evaluation results show that the method can accurately segment lungs, liver, and tumor with median [interquartile range] DSCs of 0.98 [0.97-0.98], 0.91 [0.83-0.93], and 0.85 [0.71-0.88]; surface DSCs of 0.99 [0.97-0.99], 0.86 [0.77-0.93], and 0.85 [0.62-0.93], and ASSDs of 0.91 [0.69-1.5], 4.8 [2.6-8.4], and 4.7 [3.5-9.2] mm, respectively. Dosimetric parameters from the three segmentation networks show relationship with those from the reference segmentations. The overall segmentation took about 1 min per patient on an NVIDIA RTX-2080Ti GPU. CONCLUSION: This work presents CNN-based algorithms to segment lungs, liver, and tumor from 99m Tc-MAA SPECT/CT images. The results demonstrated the potential of the proposed CNN-based segmentation method for assisting 90 Y SIRT planning while drastically reducing operator time.

Diagnostic performance of 18F-fluciclovine PET/CT in prostate cancer patients with rising PSA level ≤ 0.5 ng/ml after multiple treatment failures.

This retrospective study is to assess the performance of 18F-Fluciclovine PET/CT in prostate cancer (PC) patients with multiple treatment failures and prostate-specific antigen (PSA) ≤ 0.5 ng/mL. PC patients with multiple treatment failures who had PSA level within 2-week interval of 18F-Fluciclovine PET/CT (PSAPET) ≤ 0.5 ng/mL were identified in retrospective review of our institution's database (n=28). Patient, tumor, treatment, PSA and castration characteristics as well as findings on 18F-Fluciclovine PET/CT were collected and compared between positive and negative 18F-Fluciclovine PET/CT subgroups by using Fisher's exact test. The overall detection rate of 18F-Fluciclovine PET/CT was 7 of 28 studies (25%). PSAPET > 0.2 ng/mL was associated with higher detection rates in all (33.3 vs 10%, P=0.172), castration-resistant (CR) (50 vs 20%, P=0.343) and castration-sensitive (CS) (28.6 vs 0%, P=0.179) patients. Sites of recurrence were local 42.9% (3/7), nodal 42.9% (3/7) and bone metastases 14.3% (1/7). Higher Gleason score (GS 8-10) (33.3 vs 14.5%, P=0.396), advanced tumor stage (T3-T4) (35.7 vs 20%, P=0.653), second-line androgen deprivation therapy (ADT) uses (66.7 vs 20%, P=0.145), chemotherapy uses (50 vs 23.1%, P=0.444) and CRPC (33.3 vs 21.1%, P=0.483) related to positivity of 18F-Fluciclovine PET/CT but none reached statistical significance. Performance of 18F-Fluciclovine PET/CT in prostate cancer patients with multiple treatment failures and PSAPET ≤ 0.5 ng/mL was acceptable particularly in patients with PSAPET ≥ 0.3 ng/mL, CRPC, initial GS ≥ 8 or T3-T4.

Disease control and failure patterns of unresectable hepatocellular carcinoma following transarterial radioembolization with yttrium-90 microspheres and with/without sorafenib.

BACKGROUND: Impressive survival outcome of our previous study in unresectable hepatocellular carcinoma (HCC) patients undergoing yttrium-90 glass microspheres transarterial radioembolization (TARE) with/without sorafenib according to individuals' disease burden, i.e., intrahepatic tumor load (IHT) and adverse disease features (ADFs) might partly be confounded by other treatments and underlying hepatic function. Therefore, a dedicated study focusing on treatment response and assessment of failure patterns might be a way to improve treatment outcome in addition to patient selection based on the disease burden. AIM: To assess the tumor response, disease control and patterns of disease progression following TARE with/without sorafenib in unresectable HCC patients. METHODS: This retrospective study was conducted in successful TARE procedures with available pre- and post-treatment imaging studies (n = 169). Three treatment subgroups were (1) TARE only (TARE_alone) for IHT ≤ 50% without ADFs, i.e., macrovascular invasion, extrahepatic disease (EHD) and infiltrative/ill-defined HCC (n = 63); (2) TARE with sorafenib (TARE_sorafenib) for IHT > 50% and/or presence of ADFs (n = 81); and (3) TARE only for patients who could not receive sorafenib due to contraindication or intolerance (TARE_no_sorafenib) (n = 25). Objective response rate (ORR; consisted of complete response (CR) and partial response (PR)), disease control rate (DCR; consisted of CR, PR and stable disease) and failure patterns of treated, intrahepatic and extrahepatic sites were assessed using the modified response evaluation criteria in solid tumors. Time to progression (TTP) was calculated from TARE to the first radiologic progression at any site using Kaplan-Meier method. Identification of prognostic factors for TTP using the univariate Kaplan-Meier method and multivariate Cox proportional hazard model were performed in major population subgroups, TARE_alone and TARE_sorafenib. RESULTS: The median radiologic follow-up time was 4.4 mo (range 0.5-48.8). In treated area, ORR was highest in TARE_sorafenib (53.1%), followed by TARE_alone (41.3%) and TARE_no_sorafenib (16%). In intrahepatic area, DCR remained highest in TARE_sorafenib (84%), followed by TARE_alone (79.4%) and TARE_no_sorafenib (44%). The overall DCR was highest in TARE_alone (79.4%), followed by TARE_sorafenib (71.6%) and TARE_no_sorafenib (40%). Dominant failure patterns were intrahepatic for both TARE_alone (44.5%) and TARE_sorafenib (38.4%). Extrahepatic progression was more common in TARE_sorafenib (32%) and TARE_no_sorafenib (40%) than in TARE_alone (12.7%). TTP was longest in TARE_alone (8.6 mo; 95%CI: 3.4-13.8), followed by TARE_sorafenib (5.1 mo; 95%CI: 4.0-6.2) and TARE_no_sorafenib (2.7 mo; 95%CI: 2.2-3.1). Pre-existing EHD (HR: 0.37, 95%CI: 0.24-0.56, P < 0.001) was a sole prognostic factor for TTP in TARE_sorafenib with no prognostic factor for TTP in TARE_alone. CONCLUSION: TARE with/without sorafenib according to individuals' disease burden provided DCR approximately 70% with intrahepatic progression as dominant failure pattern. Extrahepatic progression was more common in procedures with initially high disease burden.

Survival Outcomes for Yttrium-90 Transarterial Radioembolization With and Without Sorafenib for Unresectable Hepatocellular Carcinoma Patients.

PURPOSE: To assess the overall survival (OS) and progression-free survival (PFS) of unresectable hepatocellular carcinoma (HCC) patients undergoing yttrium-90 glass-microsphere transarterial radioembolization (TARE) with and without concurrent sorafenib. METHODS: OS and PFS were analyzed in 55 patients with an intrahepatic tumor (IHT) ≤50% without advanced or aggressive disease features (ADFs), which was referred to presence of infiltrative/ill-defined HCC, macrovascular invasion, or extrahepatic disease treated with only TARE (TARE_alone) and in 74 patients with IHT ≤50% with ADFs or IHT >50% treated with TARE and sorafenib (TARE_sorafenib). Prognostic factors for OS and PFS were identified using univariate and multivariate analyses. RESULTS: Median OS and PFS of TARE_alone patients were 21.6 (95% CI 6.1-37.1) and 9.1(95% CI 5.2-13.0) months, respectively, and for TARE_sorafenib patients 12.4 (95% CI 9.1-15.6) and 5.1 (95% CI 2.6-7.5) months, respectively. Better OS was associated with serum AFP <400 (HR 0.27, p=0.02) in TARE_alone, and IHT ≤50% (HR 0.39, p=0.004) and AFP <400 (HR 0.5, p=0.027) in TARE_sorafenib. Unilobar involvement (HR 0.43, p=0.029) and AFP <400 ng/mL (HR 0.52, p=0.015) correlated with better PFS in TARE_alone and TARE_sorafenib, respectively. Adverse events (AEs) were more frequent in TARE_sorafenib than TARE_alone (92.4 vs 80.3%), but only 9.3% were grade 3 or higher AEs. CONCLUSION: TARE_alone provided the most prominent survival benefit in IHT ≤50%-without ADF patients who had unilobar HCC and serum AFP <400 ng/mL. TARE and sorafenib yielded the best outcomes in patients with IHT ≤50% and serum AFP <400 ng/mL, with some additional grade 1-2 AEs compared to TARE only.

Diagnostic performance of F-18 fluciclovine PET/CT in post-radical prostatectomy prostate cancer patients with rising prostate-specific antigen level ≤0.5 ng/mL.

PURPOSE: The aim of the study was to assess the diagnostic performance of fluciclovine positron emission tomography (PET)/computerized tomography (CT) in post-radical prostatectomy prostate cancer patients with rising prostate-specific antigen (PSA) ≤0.5 ng/mL, and identify the associated predictive factors of positive studies. PATIENTS AND METHODS: From 30 June 2017 to 9 August 2019, patients with post-radical prostatectomy prostate cancer who underwent F-18 fluciclovine PET/CT and had PSA level within 2-week interval (PSAPET) ≤0.5 ng/mL were enrolled into this single-institution retrospective study. Data on tumor characteristics, including Gleason scores, extra-prostatic extension, seminal vesicle invasion, surgical margin and nodal metastasis, PSA after radical prostatectomy, previous hormonal therapy, PSA doubling time (PSADT), scanner type, PSAPET and site of recurrence were collected. Comparison of these factors between groups of positive and negative fluciclovine PET/CT was done by using Mann-Whitney U-test and Fisher's exact test. RESULTS: Of 94 eligible patients with post-radical prostatectomy prostate cancer, 10 patients had positive studies (10.6%). Detection rate at PSAPET 0.1, 0.2, 0.3, 0.4 and 0.5 ng/mL were 0% (0/11), 0% (0/15), 20% (6/30), 4% (1/25) and 23.1% (3/13), respectively. Upon multivariate analysis of clinical factors, only a PSADT <3 months (P = 0.023) was shown to have a statistically significant correlation with a positive study. CONCLUSION: In post-radical prostatectomy prostate cancer patients with rising PSA 0.1-0.5 ng/mL, the sensitivity of F-18 fluciclovine PET/CT for identifying tumor recurrence/metastases is poor with an overall detection rate of 10.6%. Larger prospective studies are required to validate these findings.

Is There Any Role for 18F-Fluciclovine PET/CT in the Presence of Undetectable PSA in Prostate Cancer Patients After Definitive Treatment?

PURPOSE: The aim of this study was to investigate the role of F-fluciclovine PET/CT in the evaluation of prostate cancer (PC) patients after definitive treatment in the presence of undetectable prostate-specific antigen (PSA). PATIENTS AND METHODS: This retrospective study was conducted in PC patients who had undetectable PSA level and underwent fluciclovine PET/CT within a 2-week interval of PSA examination and without interval treatment or other cancer. Patient and tumor characteristics at initial diagnosis, treatment regimens, and findings on fluciclovine PET/CT were collected. Comparisons between groups of positive and negative fluciclovine PET/CT were done by using descriptive statistics. RESULTS: A total of 34 fluciclovine PET/CTs from 34 patients met the inclusion criteria. There were 4 positive (11.8%) and 30 negative fluciclovine PET/CTs (88.2%). All of the patients with positive results had an initial Gleason score of 7 or higher and locally advanced tumor (T3-T4). More common features at the time of diagnosis among positive study patients as compared with negative ones were atypical histologic variants (25% vs 0%) and very high-risk PC (50% vs 30%). Most of the patients with positive study received second-line hormonal therapy (HT) (50%), whereas patients with negative results received first-line HT (53.3%). Chemotherapy naivety was less common among positive patients (75% vs 96.7%). Sites of involvement on positive fluciclovine PET/CTs were pelvic lymph nodes (2/4, 50%), lung and mediastinal lymph node (1/4, 25%), and prostatectomy bed (1/4, 25%). CONCLUSIONS: In the presence of undetectable PSA in PC patients after definitive treatment, fluciclovine PET/CT would benefit most to patients with Gleason score of 7 or higher, high disease burden (T3-T4), and atypical histologic variants at the time of diagnosis, and the ones who have history of second-line HT and/or chemotherapy.

Comparison of Diagnostic Utility of Fluciclovine PET/CT Versus Pelvic Multiparametric MRI for Prostate Cancer in the Pelvis in the Setting of Rising PSA After Initial Treatment.

PURPOSE: The aim of this study was to investigate the imaging diagnostic performance of F-fluciclovine PET/CT and pelvic multiparametric MRI (mpMRI) for prostate cancer in the setting of rising PSA after initial treatment, with a focus on detection of recurrent and metastatic prostate cancer in the pelvis. METHODS: Patients with prostate cancer who had fluciclovine PET and pelvic mpMRI between October 2017 and October 2018 in our center were retrospectively reviewed. Patients were included if they had fluciclovine PET/CT and mpMRI within a 3-month interval. Patients were excluded if they had separate concurrent cancer or if the PSA were more than 2-fold difference with an absolute difference more than 1 ng/mL between the 2 image studies. For each eligible patient, we compared all abnormal lesions identified on either scan. The findings were verified by pathology or other imaging techniques within minimal 10-month clinical follow-up. RESULTS: A total of 129 patients with 129 paired tests were included in this study. Fluciclovine PET/CT and pelvic MRI had a high degree of concordance (121/129, 93.8%). The sensitivity, specificity, positive predictive value, and negative predictive value for fluciclovine PET/CT and mpMRI were 96.6%, 94.3%, 93.4%, and 97%, and 91.5%, 95.7%, 94.7%, and 93%, respectively. There were no statistical significant differences in diagnostic performance between the 2 imaging tests. Among the 8/129 discordant cases, although fluciclovine PET/CT provided definitive diagnosis when mpMRI was equivocal due to paramagnetic artifacts from fiducial markers and detected normal-sized regional lymph nodes, mpMRI detected subcentimeter periurethral recurrence and clarified physiological urinary artifacts that was not appreciated on fluciclovine PET/CT. CONCLUSIONS: Our single-center study demonstrated that fluciclovine PET/CT has similar diagnostic performance with pelvic mpMRI in detecting recurrent/metastatic prostate disease in the pelvis in the setting of rising PSA after initial treatment. Moreover, fluciclovine PET/CT and mpMRI have different implications in different clinical scenario; each test has its own limitation and pitfalls, but can be complementary to each other.

A Comparison of Accuracy of Planar and Evolution SPECT/CT Bone Imaging in Differentiating Benign from Metastatic Bone Lesions.

Objective: To compare sensitivity, specificity, accuracy and diagnostic confidence in the differentiation between benign and metastatic bone lesions on whole body planar bone scintigraphy and Evolution SPECT/CT. Material and Method: Eighty diagnosed or suspected cancer patients with indeterminate lesions on planar scintigraphy were recruited in the present prospective study. Additional whole body Evolution SPECT/CT was performed after whole body planar scintigraphy. All lesions on both imagings were categorized into 5 categories; definitely metastasis, probably metastasis, indeterminate, probably benign and definitely benign. The diagnosis of each lesion was confirmed by follow-up imaging, pathological findings or clinical follow-up for at least 6 months. Results: Detected lesions on planar scintigraphy and Evolution SPECT/CT imaging were 442 and 477 lesions, respectively.The sensitivity, specificity and accuracy of planar scintigraphy and Evolution SPECT/CT imaging in the diagnosis of metastatic lesions were 27% (95% CI: 13.8, 44.1), 63.2% (95% CI: 58.5, 67.7), 60%, and 97.3% (95% CI: 85.8, 99.9), 100% (95% CI: 96.4, 100) and 99.8%, respectively. Indeterminate lesions on planar scintigraphy were 34.2% (151 lesions from total 442 lesions, which 135 of these 151 indeterminate lesions or 89.4% were located in axial skeleton). Evolution SPECT/CT images were able to characterize all indeterminate lesions. Conclusion: Differentiation of benign and metastatic lesions by Evolution SPECT/CT images has superior diagnostic performance and diagnostic confidence over the planar scintigraphy. Thus, Evolution SPECT/CT images should be considered in characterization of indeterminate lesions on planar scintigraphy, especially in the axial skeleton.