Prognostic value of deep learning-derived body composition in advanced pancreatic cancer-a retrospective multicenter study

BACKGROUND

Despite the prognostic relevance of cachexia in pancreatic cancer, individual body composition has not been routinely integrated into treatment planning. In this multicenter study, we investigated the prognostic value of sarcopenia and myosteatosis automatically extracted from routine computed tomography (CT) scans of patients with advanced pancreatic ductal adenocarcinoma (PDAC).

PATIENTS AND METHODS

We retrospectively analyzed clinical imaging data of 601 patients from three German cancer centers. We applied a deep learning approach to assess sarcopenia by the abdominal muscle-to-bone ratio (MBR) and myosteatosis by the ratio of abdominal inter- and intramuscular fat to muscle volume. In the pooled cohort, univariable and multivariable analyses were carried out to analyze the association between body composition markers and overall survival (OS). We analyzed the relationship between body composition markers and laboratory values during the first year of therapy in a subgroup using linear regression analysis adjusted for age, sex, and American Joint Committee on Cancer (AJCC) stage.

RESULTS

Deep learning-derived MBR [hazard ratio (HR) 0.60, 95% confidence interval (CI) 0.47-0.77, P < 0.005] and myosteatosis (HR 3.73, 95% CI 1.66-8.39, P < 0.005) were significantly associated with OS in univariable analysis. In multivariable analysis, MBR (P = 0.019) and myosteatosis (P = 0.02) were associated with OS independent of age, sex, and AJCC stage. In a subgroup, MBR and myosteatosis were associated with albumin and C-reactive protein levels after initiation of therapy. Additionally, MBR was also associated with hemoglobin and total protein levels.

CONCLUSIONS

Our work demonstrates that deep learning can be applied across cancer centers to automatically assess sarcopenia and myosteatosis from routine CT scans. We highlight the prognostic role of our proposed markers and show a strong relationship with protein levels, inflammation, and anemia. In clinical practice, automated body composition analysis holds the potential to further personalize cancer treatment.

Impact of encorafenib on survival of patients with BRAFV600E-mutant metastatic colorectal cancer in a real-world setting

PURPOSE

Patients with BRAFV600E-mutant metastatic colorectal cancer (mCRC) have a dismal prognosis. The best strategies in these patients remain elusive. Against this background, we report the clinical course of patients with BRAFV600E-mutant mCRC to retrieve the best treatment strategy.

PATIENTS AND METHODS

Clinico-pathological data were extracted from the electronic health records. Kaplan-Meier method was used to estimate overall (OS) and progression-free survival (PFS). Objective response rate (ORR) was assessed according to RECIST 1.1.

RESULTS

In total, 51 patients were enrolled. FOLFOXIRI was administered to 12 patients; 29 patients received FOLFOX or FOLFIRI as first-line treatment. Median OS was 17.6 months. Median PFS with FOLFOXIRI (13.0 months) was significantly prolonged (HR 0.325) as compared to FOLFOX/FOLFIRI (4.3 months). However, this failed to translate into an OS benefit (p = 0.433). Interestingly, addition of a monoclonal antibody to chemotherapy associated with superior OS (HR 0.523). A total of 64.7% patients received further-line therapy, which included a BRAF inhibitor in 17 patients. Targeted therapy associated with very favourable OS (25.1 months).

CONCLUSION

Patients with BRAFV600E-mutated mCRC benefit from the addition of an antibody to first-line chemotherapy. Further-line treatment including a BRAF inhibitor has a dramatic impact on survival.

Smaller panel, similar results: genomic profiling and molecularly informed therapy in pancreatic cancer

BACKGROUND

Pancreatic cancer has a dismal prognosis. One reason is resistance to cytotoxic drugs. Molecularly matched therapies might overcome this resistance but the best approach to identify those patients who may benefit is unknown. Therefore, we sought to evaluate a molecularly guided treatment approach.

MATERIALS AND METHODS

We retrospectively analyzed the clinical outcome and mutational status of patients with pancreatic cancer who received molecular profiling at the West German Cancer Center Essen from 2016 to 2021. We carried out a 47-gene DNA next-generation sequencing (NGS) panel. Furthermore, we assessed microsatellite instability-high/deficient mismatch repair (MSI-H/dMMR) status and, sequentially and only in case of KRAS wild-type, gene fusions via RNA-based NGS. Patient data and treatment were retrieved from the electronic medical records.

RESULTS

Of 190 included patients, 171 had pancreatic ductal adenocarcinoma (90%). One hundred and three patients had stage IV pancreatic cancer at diagnosis (54%). MMR analysis in 94 patients (94/190, 49.5%) identified 3 patients with dMMR (3/94, 3.2%). Notably, we identified 32 patients with KRAS wild-type status (16.8%). To identify driver alterations in these patients, we conducted an RNA-based fusion assay on 13 assessable samples and identified 5 potentially actionable fusions (5/13, 38.5%). Overall, we identified 34 patients with potentially actionable alterations (34/190, 17.9%). Of these 34 patients, 10 patients (10/34, 29.4%) finally received at least one molecularly targeted treatment and 4 patients had an exceptional response (>9 months on treatment).

CONCLUSIONS

Here, we show that a small-sized gene panel can suffice to identify relevant therapeutic options for pancreatic cancer patients. Informally comparing with previous large-scale studies, this approach yields a similar detection rate of actionable targets. We propose molecular sequencing of pancreatic cancer as standard of care to identify KRAS wild-type and rare molecular subsets for targeted treatment strategies.

Multimodal survival prediction in advanced pancreatic cancer using machine learning

Background

Existing risk scores appear insufficient to assess the individual survival risk of patients with advanced pancreatic ductal adenocarcinoma (PDAC) and do not take advantage of the variety of parameters that are collected during clinical care.

Methods

In this retrospective study, we built a random survival forest model from clinical data of 203 patients with advanced PDAC. The parameters were assessed before initiation of systemic treatment and included age, CA19-9, C-reactive protein, metastatic status, neutrophil-to-lymphocyte ratio and total serum protein level. Separate models including imaging and molecular parameters were built for subgroups.

Results

Over the entire cohort, a model based on clinical parameters achieved a c-index of 0.71. Our approach outperformed the American Joint Committee on Cancer (AJCC) staging system and the modified Glasgow Prognostic Score (mGPS) in the identification of high- and low-risk subgroups. Inclusion of the KRAS p.G12D mutational status could further improve the prediction, whereas radiomics data of the primary tumor only showed little benefit. In an external validation cohort of PDAC patients with liver metastases, our model achieved a c-index of 0.67 (mGPS: 0.59).

Conclusions

The combination of multimodal data and machine-learning algorithms holds potential for personalized prognostication in advanced PDAC already at diagnosis.

•

We developed a machine-learning-based prediction model that outperforms the AJCC staging system and mGPS.

•

Applying our model to an external validation cohort demonstrates generalizability.

•

Explainable machine learning enables to understand the decision making of our model and identifies relevant parameters.

•

Combining clinical, imaging and genetic data holds potential for personalized prognostication in advanced PDAC.

Improving interinstitutional and intertechnology consistency of pulmonary SBRT by dose prescription to the mean internal target volume dose

Purpose

Dose, fractionation, normalization and the dose profile inside the target volume vary substantially in pulmonary stereotactic body radiotherapy (SBRT) between different institutions and SBRT technologies. Published planning studies have shown large variations of the mean dose in planning target volume (PTV) and gross tumor volume (GTV) or internal target volume (ITV) when dose prescription is performed to the PTV covering isodose. This planning study investigated whether dose prescription to the mean dose of the ITV improves consistency in pulmonary SBRT dose distributions.

Materials and methods

This was a multi-institutional planning study by the German Society of Radiation Oncology (DEGRO) working group Radiosurgery and Stereotactic Radiotherapy. CT images and structures of ITV, PTV and all relevant organs at risk (OAR) for two patients with early stage non-small cell lung cancer (NSCLC) were distributed to all participating institutions. Each institute created a treatment plan with the technique commonly used in the institute for lung SBRT. The specified dose fractionation was 3 × 21.5 Gy normalized to the mean ITV dose. Additional dose objectives for target volumes and OAR were provided.

Results

In all, 52 plans from 25 institutions were included in this analysis: 8 robotic radiosurgery (RRS), 34 intensity-modulated (MOD), and 10 3D-conformal (3D) radiation therapy plans. The distribution of the mean dose in the PTV did not differ significantly between the two patients (median 56.9 Gy vs 56.6 Gy). There was only a small difference between the techniques, with RRS having the lowest mean PTV dose with a median of 55.9 Gy followed by MOD plans with 56.7 Gy and 3D plans with 57.4 Gy having the highest. For the different organs at risk no significant difference between the techniques could be found.

Conclusions

This planning study pointed out that multiparameter dose prescription including normalization on the mean ITV dose in combination with detailed objectives for the PTV and ITV achieve consistent dose distributions for peripheral lung tumors in combination with an ITV concept between different delivery techniques and across institutions.

Supplementary Information

The online version of this article (10.1007/s00066-021-01799-w) contains supplementary material, which is available to authorized users.

Endoscopic therapy for infected pancreatic necrosis using fully covered self-expandable metal stents: combination of transluminal necrosectomy, transluminal and percutaneous drainage

BACKGROUND/AIMS

Endoscopic transluminal therapy has become the standard of care as a less invasive alternative to surgery. In a retrospective case series of two tertiary referral centers we report on an individualized concept combining EUS-guided drainage with self-expanding metal stents, direct transluminal debridement und percutaneous drainage.

METHODS

We treated 13 patients with infected pancreatic necrosis. Initially in all patients an EUS-guided drainage with plastic stents was performed under antibiotic protection (transduodenal: 2, transgastral: 11). After clinical consolidation (after 9.6 ± 9.4 days) a covered self-expanding metal stent (Niti-S, Taewoong medical Co., Seoul, Korea) was inserted by performing direct endoscopic necrosectomy in 2.9 ± 1.7 sessions through the stent. In cases of disrupted duct syndromes a pancreatic plastic stent was inserted (5 of 13 patients). In 5 of 13 cases additional percutaneous drainage was applied because of extended necrosis. In one patient percutaneous endoscopic drainage using the percutaneous access was needed.

RESULTS

A sustained clinical success was achieved in 12 of 13 cases (CRP before therapy 23.5 ± 14.4 mg/L, after 3.1 ± 2.6 mg/lL). Discharge occurred after 2.5 ± 22.4 days. The self-expanding metal stent was extracted after 82.5 ± 56.6 days. Mean follow up was 8.5 ± 5.9 months.

CONCLUSION

Our concept of combining transluminal drainage, direct endoscopic necrosectomy and percutaneuos drainage offers a safe and reliable alternative to surgery, even in case of extended necrosis.

Visualizing Validation of Protein Surface Classifiers

Many bioinformatics applications construct classifiers that are validated in experiments that compare their results to known ground truth over a corpus. In this paper, we introduce an approach for exploring the results of such classifier validation experiments, focusing on classifiers for regions of molecular surfaces. We provide a tool that allows for examining classification performance patterns over a test corpus. The approach combines a summary view that provides information about an entire corpus of molecules with a detail view that visualizes classifier results directly on protein surfaces. Rather than displaying miniature 3D views of each molecule, the summary provides 2D glyphs of each protein surface arranged in a reorderable, small-multiples grid. Each summary is specifically designed to support visual aggregation to allow the viewer to both get a sense of aggregate properties as well as the details that form them. The detail view provides a 3D visualization of each protein surface coupled with interaction techniques designed to support key tasks, including spatial aggregation and automated camera touring. A prototype implementation of our approach is demonstrated on protein surface classifier experiments.

Standard value of ultrasound elastography using acoustic radiation force impulse imaging (ARFI) in healthy liver tissue of children and adolescents

PURPOSE

Ultrasound elastography by acoustic radiation force impulse imaging (ARFI) is used in adults for non invasive measurement of liver stiffness, indicating liver diseases like fibrosis. To establish ARFI in children and adolescents we determined standard values of healthy liver tissue and analysed potentially influencing factors.

MATERIALS AND METHODS

132 patients between 0 and 17 years old were measured using ARFI. None of them had any liver disease or any other disease that could affect the liver secondarily. All patients had a normal ultrasound scan, a normal BMI and normal liver function tests. The mean value of all ARFI measurements was calculated and potentially influencing factors were analysed.

RESULTS

The mean value of all ARFI elastography measurements was 1.16 m/sec (SD ± 0.14 m/sec). Neither age (p = 0.533) nor depth of measurement (p = 0.066) had no significant influence on ARFI values, whereas a significant effect of gender was found with lower ARFI values in females (p = 0.025), however, there was no significant interaction between age groups (before or after puberty) and gender (p = 0.276). There was an interlobar difference with lower values in the right liver lobe compared to the left (p = 0.036) and with a significantly lower variance (p < 0.001). Consistend values were measured by different examiners (p = 0.108), however, the inter examiner variance deviated significantly (p < 0.001).

CONCLUSION

ARFI elastography is a reliable method to measure liver stiffness in children and adolescents. In relation to studies which analyse liver diseases, the standard value of 1.16 m/sec (± 0.14 m/sec) allows a differentiation of healthy versus pathological liver tissue.

Dosimetric evaluation of a 2D pixel ionization chamber for implementation in clinical routine

In this paper we present the results of a dosimetric evaluation of a 2D ionization chamber array with the objective of its implementation for quality assurance in clinical routine. The pixel ionization chamber MatriXX (Scanditronix Wellhofer, Germany) consists of 32x32 chambers with a distance of 7.6 mm between chamber centres. The effective depth of measurement under the surface of the detector was determined. The dose and energy dependence, the behaviour of the device during its initial phase and its time stability as well as the lateral response of a single chamber of the detector in cross-plane and diagonal directions were analysed. It could be shown, that the detector's response is linear with dose and energy independent. Taking the lateral response into account, two different dose profiles, for a pyramidal and an IMRT dose distribution, were applied to compare the data generated by a treatment planning system with measurements. From these investigations it can be concluded that the detector is a suitable device for quality assurance and 2D dose verifications.

Design of a computer-controlled multileaf collimator for advanced electron radiotherapy

A multileaf collimator for electrons (eMLC) has been designed that fulfils the technical requirements for providing advanced irradiation techniques with electrons. In the present work, the basic design parameters of leaf material, leaf height, leaf width and number of leaves as well as leaf overtravel and leaf shape were determined such that an eMLC with motorized leaves can be manufactured by a company specialized in MLC technology. For this purpose, a manually driven eMLC with variable source-to-collimator distance (SCD) was used to evaluate the chosen leaf specification and investigate the impact of the SCD on the off-axis dose distribution. In order to select the final SCD of the eMLC, a compromise had to be found between maximum field size, minimum beam penumbra and necessary distance between eMLC and isocentre to eliminate patient realignments during gantry rotation. As a result, the eMLC is placed according to the target position at 72 and 84 cm SCD, respectively. This feature will be achieved by interchangeable distance holders. At these SCDs, the corresponding maximum field sizes at 100 cm source-to-isocentre distance are 20 x 20 cm and 17 x 17 cm, respectively. Finally, the off-axis dose distribution at the maximum opening of the eMLC was improved by fine-tuning the settings of the accelerator jaws and introducing trimmer bars above the eMLC. Following this optimization, a prototype eMLC consisting of 2 x 24 computer-controlled brass leaves is manufactured by 3D Line Medical Systems.

Precision branching ratio measurement for the superallowed beta+ emitter 62Ga and isospin-symmetry-breaking corrections in A>or=62 nuclei

A high-precision branching ratio measurement for the superallowed beta+ decay of 62Ga was performed at the Isotope Separator and Accelerator radioactive ion beam facility. Nineteen gamma rays emitted following beta+ decay of 62Ga were identified, establishing the dominant superallowed branching ratio to be (99.861+/-0.011)%. Combined with recent half-life and Q-value measurements, this branching ratio yields a superallowed ft value of 3075.6+/-1.4 s for 62Ga decay. These results demonstrate the feasibility of high-precision superallowed branching ratio measurements in the A>or=62 mass region and provide the first stringent tests of the large isospin-symmetry-breaking effects predicted for these decays.

Nuclear charge radii of 9,11Li: the influence of halo neutrons

The nuclear charge radius of 11Li has been determined for the first time by high-precision laser spectroscopy. On-line measurements at TRIUMF-ISAC yielded a 7Li-11Li isotope shift (IS) of 25 101.23(13) MHz for the Doppler-free [FORMULA: SEE TEXT]transition. IS accuracy for all other bound Li isotopes was also improved. Differences from calculated mass-based IS yield values for change in charge radius along the isotope chain. The charge radius decreases monotonically from 6Li to 9Li, and then increases from 2.217(35) to 2.467(37) fm for 11Li. This is compared to various models, and it is found that a combination of halo neutron correlation and intrinsic core excitation best reproduces the experimental results.

Performance of electronic portal imaging devices (EPIDs) used in radiotherapy: Image quality and dose measurements

The aim of our study was to compare the image and dosimetric quality of two different imaging systems. The first one is a fluoroscopic electronic portal imaging device (first generation), while the second is based on an amorphous silicon flat-panel array (second generation). The parameters describing image quality include spatial resolution [modulation transfer function (MTF)], noise [noise power spectrum (NPS)], and signal-to-noise transfer [detective quantum efficiency (DQE)]. The dosimetric measurements were compared with ionization chamber as well as with film measurements. The response of the flat-panel imager and the fluoroscopic-optical device was determined performing a two-step Monte Carlo simulation. All measurements were performed in a 6 MV linear accelerator photon beam. The resolution (MTF) of the fluoroscopic device (f 1/2 = 0.3 mm(-1)) is larger than of the amorphous silicon based system (f 1/2 = 0.21 mm(-1)), which is due to the missing backscattered photons and the smaller pixel size. The noise measurements (NPS) show the correlation of neighboring pixels of the amorphous silicon electronic portal imaging device, whereas the NPS of the fluoroscopic system is frequency independent. At zero spatial frequency the DQE of the flat-panel imager has a value of 0.008 (0.8%). Due to the minor frequency dependency this device may be almost x-ray quantum limited. Monte Carlo simulations verified these characteristics. For the fluoroscopic imaging system the DQE at low frequencies is about 0.0008 (0.08%) and degrades with higher frequencies. Dose measurements with the flat-panel imager revealed that images can only be directly converted to portal dose images, if scatter can be neglected. Thus objects distant to the detector (e.g., inhomogeneous dose distribution generated by a modificator) can be verified dosimetrically, while objects close to a detector (e.g., a patient) cannot be verified directly and must be scatter corrected prior to verification. This is justified by the response of the flat-panel imaging device revealing a strong dependency at low energies.

Dynamics of esophageal bolus transport in healthy subjects studied using multiple intraluminal impedancometry

The dynamics of a bolus transport through the esophagus are largely unexplored. To study this physiological process, we applied multiple intraluminal impedancometry in 10 healthy subjects. Three different protocols were used: 1) liquid bolus administered with subject supine, 2) liquid bolus with subject upright, or 3) semisolid bolus with subject supine. Transit of different parts of a bolus (bolus head, body, and tail) was analyzed at different anatomic segments, namely the pharynx and the proximal, middle, and distal thirds of the esophagus. A characteristic pattern of bolus transport was seen in all subjects. Impedance changes related to air were observed preceding the bolus head. The bolus head propelled significantly faster than did the bolus body and tail. Pharyngeal bolus transit was significantly faster than esophageal bolus transit. Within the esophagus, bolus propulsion velocity gradually decreased. Bolus transport was significantly accelerated in the upright position and delayed with increase of bolus viscosity. In conclusion, the dynamics of a bolus transport from the pharynx into the stomach are complex. It varies within both different anatomic segments and different parts of the bolus and depends on bolus characteristics and test conditions. The spatial and temporal resolution of a bolus transport can be obtained by the impedance technique.

Criminality and mental illness: a study of arrest rates in a rural state

The authors examined cumulative arrests for the period of 1964-1973 for psychiatric patients admitted to a state hospital in 1969 (286 men and 175 women) and for the general population in Wyoming. Male patients were arrested as often as or more often than individuals in the general population. No men diagnosed as schizophrenic were arrested for crimes against persons; one-fourth of arrests among individuals diagnosed as alcoholics were for such crimes. Arrest rates among women patients seemed to parallel those of women in the general population, although the sample was too small for definite conclusions. Male patients were arrested significantly less often following dischanrge than before hospitalization.