Background After resection of colorectal cancer liver metastases (CRLM) two main histopathological growth patterns can be observed; a desmoplastic and a non-desmoplastic subtype. The desmoplastic subtype has been associated with superior survival. These findings require external validation. Methods An international multicenter retrospective cohort study was conducted in patients treated surgically for CRLM at three tertiary hospitals in the US and the Netherlands. Determination of histopathological growth patterns was performed on hematoxylin & eosin stained sections of resected CRLM according to international guidelines. Patients displaying a desmoplastic histopathological phenotype (only desmoplastic growth observed) were compared to patients with a non-desmoplastic phenotype (any non-desmoplastic growth observed). Cut-off analyses on the extent of non-desmoplastic growth were performed. Overall (OS) and disease-free (DFS) survival were estimated using Kaplan-Meier and multivariable Cox analysis. All statistical tests were 2-sided. Results In total 780 patients were eligible. A desmoplastic phenotype was observed in 19.1% and was associated with microsatellite instability (14.6% versus 3.6%, p = .01). Desmoplastic patients had superior 5-year OS (73.4% [95% CI = 64.1–84.0] versus 44.2% [95% CI = 38.9–50.2], p < .001) and DFS (32.0% [95% CI = 22.9–44.7] versus 14.7% [95% CI = 11.7–18.6], p < .001) compared to their non-desmoplastic counterparts. A desmoplastic phenotype was associated with an adjusted hazard ratio for death of 0.36 (95% CI = 0.23–0.58), and 0.50 (95% CI = 0.37–0.66) for cancer recurrence. Prognosis was independent of KRAS and BRAF status. The cut-off analyses found no prognostic relationship between either OS or DFS and the extent of non-desmoplastic growth observed (all p > .1). Conclusions This external validation study confirms the remarkably good prognosis after surgery for CRLM in patients with a desmoplastic phenotype. The extent of non-desmoplastic growth does not impact prognosis.

No abstract

Objectives To evaluate image quality and reconstruction times of a commercial deep learning reconstruction algorithm (DLR) compared to hybrid-iterative reconstruction (Hybrid-IR) and model-based iterative reconstruction (MBIR) algorithms for cerebral non-contrast CT (NCCT). Methods Cerebral NCCT acquisitions of 50 consecutive patients were reconstructed using DLR, Hybrid-IR and MBIR with a clinical CT system. Image quality, in terms of six subjective characteristics (noise, sharpness, grey-white matter differentiation, artefacts, natural appearance and overall image quality), was scored by five observers. As objective metrics of image quality, the noise magnitude and signal-difference-to-noise ratio (SDNR) of the grey and white matter were calculated. Mean values for the image quality characteristics scored by the observers were estimated using a general linear model to account for multiple readers. The estimated means for the reconstruction methods were pairwise compared. Calculated measures were compared using paired t tests. Results For all image quality characteristics, DLR images were scored significantly higher than MBIR images. Compared to Hybrid-IR, perceived noise and grey-white matter differentiation were better with DLR, while no difference was detected for other image quality characteristics. Noise magnitude was lower for DLR compared to Hybrid-IR and MBIR (5.6, 6.4 and 6.2, respectively) and SDNR higher (2.4, 1.9 and 2.0, respectively). Reconstruction times were 27 s, 44 s and 176 s for Hybrid-IR, DLR and MBIR respectively. Conclusions With a slight increase in reconstruction time, DLR results in lower noise and improved tissue differentiation compared to Hybrid-IR. Image quality of MBIR is significantly lower compared to DLR with much longer reconstruction times. Key Points • Deep learning reconstruction of cerebral non-contrast CT results in lower noise and improved tissue differentiation compared to hybrid-iterative reconstruction. • Deep learning reconstruction of cerebral non-contrast CT results in better image quality in all aspects evaluated compared to model-based iterative reconstruction. • Deep learning reconstruction only needs a slight increase in reconstruction time compared to hybrid-iterative reconstruction, while model-based iterative reconstruction requires considerably longer processing time.

Background The surgeons' estimate of the extent of resection (EOR) shows little accuracy in previous literature. Considering the developments in surgical techniques of glioblastoma (GBM) treatment, we hypothesize an improvement in this estimation. This study aims to compare the EOR estimated by the neurosurgeon with the EOR determined using volumetric analysis on the postoperative MR scan. Methods Pre-and post-operative tumor volumes were calculated through semi-automatic volumetric assessment by three observers. Interobserver agreement was measured using intraclass correlation coefficient (ICC). A univariate general linear model was used to study the factors influencing the accuracy of estimation of resection percentage. Results ICC was high for all three measurements: pre-operative tumor volume was 0.980 (0.969-0.987), post-operative tumor volume 0.974 (0.961-0.984), and EOR 0.947 (0.917-0.967). Estimation of EOR by the surgeon showed moderate accuracy and agreement. Multivariable analysis showed a statistically significant effect of operating neurosurgeon (p = 0.01), use of fluorescence (p < 0.001), and resection percentage (p < 0.001) on the accuracy of the EOR estimation. Conclusion All measurements through semi-automatic volumetric analysis show a high interobserver agreement, suggesting this to be a reliable assessment of EOR. We found a moderate reliability of the surgeons' estimate of EOR. Therefore, (early) postoperative MRI scanning for evaluation of EOR remains paramount.