Medical Imaging Biomarker Discovery and Integration Towards AI-Based Personalized Radiotherapy

Intensity-modulated radiation therapy (IMRT) has been used for high-accurate physical dose distribution sculpture and employed to modulate different dose levels into Gross Tumor Volume (GTV), Clinical Target Volume (CTV) and Planning Target Volume (PTV). GTV, CTV and PTV can be prescribed at different dose levels, however, there is an emphasis that their dose distributions need to be uniform, despite the fact that most types of tumour are heterogeneous. With traditional radiomics and artificial intelligence (AI) techniques, we can identify biological target volume from functional images against conventional GTV derived from anatomical imaging. Functional imaging, such as multi parameter MRI and PET can be used to implement dose painting, which allows us to achieve dose escalation by increasing doses in certain areas that are therapy-resistant in the GTV and reducing doses in less aggressive areas. In this review, we firstly discuss several quantitative functional imaging techniques including PET-CT and multi-parameter MRI. Furthermore, theoretical and experimental comparisons for dose painting by contours (DPBC) and dose painting by numbers (DPBN), along with outcome analysis after dose painting are provided. The state-of-the-art AI-based biomarker diagnosis techniques is reviewed. Finally, we conclude major challenges and future directions in AI-based biomarkers to improve cancer diagnosis and radiotherapy treatment.

Individual Brain Metabolic Connectome Indicator Based on Jensen-Shannon Divergence Similarity Estimation Predicts Seizure Outcomes of Temporal Lobe Epilepsy

Objective: We aimed to use an individual metabolic connectome method, the Jensen-Shannon Divergence Similarity Estimation (JSSE), to characterize the aberrant connectivity patterns and topological alterations of the individual-level brain metabolic connectome and predict the long-term surgical outcomes in temporal lobe epilepsy (TLE).Methods: A total of 128 patients with TLE (63 females, 65 males; 25.07 ± 12.01 years) who underwent Positron emission tomography (PET) with 18F-fluorodeoxyglucose (FDG) imaging were enrolled. Patients were classified either as experiencing seizure recurrence (SZR) or seizure free (SZF) at least 1 year after surgery. Each individual’s metabolic brain network was ascertained using the proposed JSSE method. We compared the similarity and difference in the JSSE network and its topological measurements between the two groups. The two groups were then classified by combining the information from connection and topological metrics, which was conducted by the multiple kernel support vector machine. The validation was performed using the nested leave-one-out cross-validation strategy to confirm the performance of the methods.Results: With a median follow-up of 33 months, 50% of patients achieved SZF. No relevant differences in clinical features were found between the two groups except age at onset. The proposed JSSE method showed marked degree reductions in IFGoperc.R, ROL. R, IPL. R, and SMG. R; and betweenness reductions in ORBsup.R and IOG. R; meanwhile, it found increases in the degree analysis of CAL. L and PCL. L, and in the betweenness analysis of PreCG.R, IOG. R, PoCG.R, PCL. L and PCL.R. Exploring consensus significant metabolic connections, we observed that the most involved metabolic motor networks were the INS-TPOmid.L, MTG. R-SMG. R, and MTG. R-IPL.R pathways between the two groups, and yielded another detailed individual pathological connectivity in the PHG. R-CAU.L, PHG. R-HIP.L, TPOmid.L-LING.R, TPOmid.L-DCG.R, MOG. R-MTG.R, MOG. R-ANG.R, and IPL. R-IFGoperc.L pathways. These aberrant functional network measures exhibited ideal classification performance in predicting SZF individuals from SZR ones at a sensitivity of 75.00%, a specificity of 92.79%, and an accuracy of 83.59%.Conclusion: The JSSE method indicator can identify abnormal brain networks in predicting an individual’s long-term surgical outcome of TLE, thus potentially constituting a clinically applicable imaging biomarker. The results highlight the biological meaning of the estimated individual brain metabolic connectome.

Maximum 11C-methionine PET uptake as a prognostic imaging biomarker for newly diagnosed and untreated astrocytic glioma

This study aimed whether the uptake of amino tracer positron emission tomography (PET) can be used as an additional imaging biomarker to estimate the prognosis of glioma. Participants comprised 56 adult patients with newly diagnosed and untreated World Health Organization (WHO) grade II–IV astrocytic glioma who underwent surgical excision and were evaluated by 11C-methionine PET prior to the surgical excision at Osaka City University Hospital from July 2011 to March 2018. Clinical and imaging studies were retrospectively reviewed based on medical records at our institution. Preoperative Karnofsky Performance Status (KPS) only influenced progression-free survival (hazard ratio [HR] 0.20; 95% confidence interval [CI] 0.10–0.41, p < 0.0001), whereas histology (anaplastic astrocytoma: HR 5.30, 95% CI 1.23–22.8, p = 0.025; glioblastoma: HR 11.52, 95% CI 2.27–58.47, p = 0.0032), preoperative KPS ≥ 80 (HR 0.23, 95% CI 0.09–0.62, p = 0.004), maximum lesion-to-contralateral normal brain tissue (LN max) ≥ 4.03 (HR 0.24, 95% CI 0.08–0.71, p = 0.01), and isocitrate dehydrogenase (IDH) status (HR 14.06, 95% CI 1.81–109.2, p = 0.011) were factors influencing overall survival (OS) in multivariate Cox regression. OS was shorter in patients with LN max ≥ 4.03 (29.3 months) than in patients with LN max < 4.03 (not reached; p = 0.03). OS differed significantly between patients with IDH mutant/LN max < 4.03 and patients with IDH mutant/LN max ≥ 4.03. LN max using 11C-methionine PET may be used in prognostic markers for newly identified and untreated WHO grade II–IV astrocytic glioma.

Quantitative washout in patients with hepatocellular carcinoma undergoing TACE: an imaging biomarker for predicting prognosis?

Background The delayed percentage attenuation ratio (DPAR) was recently identified as a novel predictor of an early complete response in patients with hepatocellular carcinoma (HCC) undergoing transarterial chemoembolization (TACE). In this study, we aimed to validate the role of DPAR as a predictive biomarker for short-, mid-, and long-term outcomes after TACE. Methods We retrospectively reviewed laboratory and imaging data for 103 treatment-naïve patients undergoing initial TACE treatment at our tertiary care center between January 2016 and November 2020. DPAR and other washin and washout indices were quantified in the triphasic computed tomography performed before the initial TACE. The correlation of DPAR and radiologic response was investigated. Furthermore, the influence of DPAR on the 6-, 12-, 18-, and 24-month survival rates and the median overall survival (OS) was compared to other established washout indices and estimates of tumor burden and remnant liver function. Results The DPAR was significantly of the target lesions (TLs) with objective response to TACE after the initial TACE session was significantly higher compared to patients with stable disease (SD) or progressive disease (PD) (125 (IQR 118–134) vs 110 (IQR 103–116), p < 0.001). Furthermore, the DPAR was significantly higher in patients who survived the first 6 months after TACE (122 vs. 115, p = 0.04). In addition, the number of patients with a DPAR > 120 was significantly higher in this group (n = 38 vs. n = 8; p = 0.03). However, no significant differences were observed in the 12-, 18-, and 24-month survival rates after the initial TACE. Regarding the median OS, no significant difference was observed for patients with a high DPAR compared to those with a low DPAR (18.7 months vs. 12.7 months, p = 0.260). Conclusions Our results confirm DPAR as the most relevant washout index for predicting the short-term outcome of patients with HCC undergoing TACE. However, DPAR and the other washout indices were not predictive of mid- and long-term outcomes.

Granzyme B PET Imaging in Response to In Situ Vaccine Therapy Combined with αPD1 in a Murine Colon Cancer Model

Immune checkpoint inhibitors (ICIs) block checkpoint receptors that tumours use for immune evasion, allowing immune cells to target and destroy cancer cells. Despite rapid advancements in immunotherapy, durable response rates to ICIs remains low. To address this, combination clinical trials are underway assessing whether adjuvants can enhance responsiveness by increasing tumour immunogenicity. CpG-oligodeoxynucleotides (CpG-ODN) are synthetic DNA fragments containing an unmethylated cysteine-guanosine motif that stimulate the innate and adaptive immune systems by engaging Toll-like receptor 9 (TLR9) present on the plasmacytoid dendritic cells (pDCs) and B cells. Here, we have assessed the ability of AlF-mNOTA-GZP, a peptide tracer targeting granzyme B, to serve as a PET imaging biomarker in response to CpG-ODN 1585 in situ vaccine therapy delivered intratumourally (IT) or intraperitoneally (IP) either as monotherapy or in combination with αPD1. [18F]AlF-mNOTA-GZP was able to differentiate treatment responders from non-responders based on tumour uptake. Furthermore, [18F]AlF-mNOTA-GZP showed positive associations with changes in tumour-associated lymphocytes expressing GZB, namely GZB+ CD8+ T cells, and decreases in suppressive F4/80+ cells. [18F]AlF-mNOTA-GZP tumour uptake was mediated by GZB expressing CD8+ cells and successfully stratifies therapy responders from non-responders, potentially acting as a non-invasive biomarker for ICIs and combination therapy evaluation in a clinical setting.

MRI-assessed tumor-free distance to serosa predicts deep myometrial invasion and poor outcome in endometrial cancer

Objectives To explore the diagnostic accuracy of preoperative magnetic resonance imaging (MRI)-derived tumor measurements for the prediction of histopathological deep (≥ 50%) myometrial invasion (pDMI) and prognostication in endometrial cancer (EC). Methods Preoperative pelvic MRI of 357 included patients with histologically confirmed EC were read independently by three radiologists blinded to clinical information. The radiologists recorded imaging findings (T1 post-contrast sequence) suggesting deep (≥ 50%) myometrial invasion (iDMI) and measured anteroposterior tumor diameter (APD), depth of myometrial tumor invasion (DOI) and tumor-free distance to serosa (iTFD). Receiver operating characteristic (ROC) curves for the prediction of pDMI were plotted for the different MRI measurements. The predictive and prognostic value of the MRI measurements was analyzed using logistic regression and Cox proportional hazard model. Results iTFD yielded highest area under the ROC curve (AUC) for the prediction of pDMI with an AUC of 0.82, whereas DOI, APD and iDMI yielded AUCs of 0.74, 0.81 and 0.74, respectively. Multivariate analysis for predicting pDMI yielded highest predictive value of iTFD < 6 mm with OR of 5.8 (p < 0.001) and lower figures for DOI ≥ 5 mm (OR = 2.8, p = 0.01), APD ≥ 17 mm (OR = 2.8, p < 0.001) and iDMI (OR = 1.1, p = 0.82). Patients with iTFD < 6 mm also had significantly reduced progression-free survival with hazard ratio of 2.4 (p < 0.001). Conclusion For predicting pDMI, iTFD yielded best diagnostic performance and iTFD < 6 mm outperformed other cutoff-based imaging markers and conventional subjective assessment of deep myometrial invasion (iDMI) for diagnosing pDMI. Thus, iTFD at MRI represents a promising preoperative imaging biomarker that may aid in predicting pDMI and high-risk disease in EC.

Combination of extramural venous invasion and lateral lymph node size detected with magnetic resonance imaging is a reliable biomarker for lateral lymph node metastasis in patients with rectal cancer

Purpose Pathological extramural venous invasion (EMVI) is defined as the active invasion of malignant cells into veins beyond the muscularis propria in colorectal cancer. It is associated with poor prognosis and increases the risk of disease recurrence. Specific findings on MRI (termed MRI-EMVI) are reportedly associated with pathological EMVI. In this study, we aimed to identify risk factors for lateral lymph node (LLN) metastasis related to rectal cancer and to evaluate whether MRI-EMVI could be a new and useful imaging biomarker to help LLN metastasis diagnosis besides LLN size. Methods We investigated 67 patients who underwent rectal resection and LLN dissection for rectal cancer. We evaluated MRI-EMVI grading score and examined the relationship between MRI-EMVI and LLN metastasis. Results Pathological LLN metastasis was detected in 18 cases (26.9%), and MRI-EMVI was observed in 32 cases (47.8%). Patients were divided into two cohorts, according to LLN metastasis. Multivariate analyses demonstrated that higher risk of LLN metastasis was significantly associated with MRI-EMVI (P = 0.0112) and a short lateral lymph node axis (≥ 5 mm) (P = 0.0002). The positive likelihood ratios of MRI-EMVI alone, LLN size alone, and the combination of both factors were 2.12, 4.84, and 16.33, respectively. Patients negative for both showed better 2-year relapse-free survival compared to other patients (84.4% vs. 62.1%, P = 0.0374). Conclusions MRI-EMVI was a useful imaging biomarker for identifying LLN metastasis in patients with rectal cancer. The combination of MRI-EMVI and LLN size can improve diagnostic accuracy.

Association of myocardial fibrosis detected by late gadolinium-enhanced MRI with clinical outcomes in patients with diabetes: a systematic review and meta-analysis

ObjectiveThis meta-analysis assessed the associations of myocardial fibrosis detected by late gadolinium-enhanced (LGE)-MRI with the risk of major adverse cardiac and cerebrovascular events (MACCEs) and major adverse cardiac events (MACEs) in patients with diabetes.DesignSystematic review and meta-analysis reported in accordance with the guidelines of the Meta-analysis of Observational Studies in Epidemiology statement.Data sourcesWe searched the Medline, Embase and Cochrane by Ovid databases for studies published up to 27 August 2021.Eligibility criteriaProspective or respective cohort studies were included if they reported the HR and 95% CIs for MACCEs/MACEs in patients with either type 1 or 2 diabetes and LGE-MRI-detected myocardial fibrosis compared with patients without LGE-MRI-detected myocardial fibrosis and if the articles were published in the English language.Data extraction and synthesisTwo review authors independently extracted data and assessed the quality of the included studies. Pooled HRs and 95% CIs were analysed using a random effects model. Heterogeneity was assessed using forest plots and I2 statistics.ResultsEight studies with 1121 patients with type 1 or type 2 diabetes were included in this meta-analysis, and the follow-up ranged from 17 to 70 months. The presence of myocardial fibrosis detected by LGE-MRI was associated with an increased risk for MACCEs (HR: 2.58; 95% CI 1.42 to 4.71; p=0.002) and MACEs (HR: 5.28; 95% CI 3.20 to 8.70; p<0.001) in patients with diabetes. Subgroup analysis revealed that ischaemic fibrosis detected by LGE was associated with MACCEs (HR 3.80, 95% CI 2.38 to 6.07; p<0.001) in patients with diabetes.ConclusionsThis study demonstrated that ischaemic myocardial fibrosis detected by LGE-MRI was associated with an increased risk of MACCEs/MACEs in patients with diabetes and may be an imaging biomarker for risk stratification. Whether LGE-MRI provides incremental prognostic information with respect to MACCEs/MACEs over risk stratification by conventional cardiovascular risk factors requires further study.