Causal rule ensemble method for estimating heterogeneous treatment effect with consideration of prognostic effects.

We propose a novel framework based on the RuleFit method to estimate heterogeneous treatment effect in randomized clinical trials. The proposed method estimates a rule ensemble comprising a set of prognostic rules, a set of prescriptive rules, as well as the linear effects of the original predictor variables. The prescriptive rules provide an interpretable description of the heterogeneous treatment effect. By including a prognostic term in the proposed model, the selected rule is represented as an heterogeneous treatment effect that excludes other effects. We confirmed that the performance of the proposed method was equivalent to that of other ensemble learning methods through numerical simulations and demonstrated the interpretation of the proposed method using a real data application.

Usefulness of Clinical Frailty Scale for Comprehensive Geriatric Assessment of Older Heart Failure Patients.

Background: Comprehensive geriatric assessment (CGA) is a multidisciplinary diagnostic process to identify the physical, psychological, and social functions of patients with frailty. The Clinical Frailty Scale (CFS) might aid in effectively identifying older patients with heart failure (HF) and frailty who would then reap maximum benefits from the CGA. Methods and Results: A single-centre prospective cohort study that enrolled consecutive hospitalised patients (age ≥75 years) with HF was conducted. The Barthel index (BI), Mini Mental State Examination (MMSE), the Charlson comorbidity index (CCI), and the COntrolling NUTritional (CONUT) for CGA was used. Among 190 enrolled patients (mean age, 85.4 years; 47.9% male), all-cause mortality (primary endpoint) occurred in 45 patients and HF-related rehospitalization (secondary endpoint) in 59 patients within 1 year. The cumulative incidence of all-cause mortality was significantly higher in the high CFS group (low 6.3%, high 30.5%, P<0.001). However, the cumulative incidence of HF-related rehospitalization was not significantly different (low 26.3%, high 32.0%, P=0.304). The multivariable analysis revealed that the CFS group was independently associated with the risk of all-cause mortality. CFS showed a strong correlation with the BI and moderate correlation with the MMSE. Conclusions: The CFS was associated with all-cause mortality within 1 year and was correlated with frailty domains of CGA.

Impact of Flagging/Risk Stratification System on Complications in Hospitalist Hip Fracture Co-management: Retrospective Cohort Study.

Purpose: Hip fractures are associated with high morbidity and mortality, the rates of which can be improved by comprehensive care. To improve hospitalist co-management of hip fractures, we designed and implemented hip fracture template (HFT), a flagging and risk stratification algorithm system. It includes consideration of perioperative management and preventative measures against hip fractures. We examined its effect on morbidity in patients with hip fractures and the factors associated with complications. Methods: We conducted a retrospective cohort study of patients who underwent surgery for hip fracture. The primary outcome was the perioperative complication rate, comparing patients managed with and without HFT. Multivariate analysis was adjusted for age, gender, and any significant variables shown in univariate analysis. Results: HFT was used in 121 patients and not used in 147 patients. In univariate analysis, patients were less likely to have complications if HFT was used (19.0% vs. 29.9%, P = 0.047), but there was no difference in length of stay (17 days vs. 17 days, P = 0.27) or in-hospital-mortality (0.8% vs. 0.7%, P = 1.00) between the groups. In adjusted analysis, patients managed by HFT had lower likelihood of complications (OR 0.55, 95% CI 0.31-0.98). Among patients managed by HFT, those with revised cardiac risk index (RCRI) ≥ 1 were more likely to have complications in both univariate (42.1% vs. 14.7%, P = 0.01) and adjusted analysis (OR 3.37, 95% CI 1.03-10.84). Conclusion: Patients with hip fractures managed with HFT were less likely to have complications, especially those with RCRI ≥ 1, suggesting benefits of using HFT.

Safety of skipping the tract dilation step for EUS-guided biliary drainage in patients with unresectable malignant biliary obstruction (with video).

BACKGROUND: Thin delivery system stents can be inserted directly without the need for a tract dilation step and are expected to reduce bile leakage during endoscopic ultrasound-guided biliary drainage (EUS-BD). The present study retrospectively compared the safety and efficacy of EUS-BD using a thin metal stent (< 7.5 Fr) with those of EUS-BD using a conventional stent (≥ 7.5 Fr). METHODS: The present study enrolled 112 patients who underwent EUS-BD using metal stents for unresectable malignant biliary obstruction between April 2016 and July 2022. The primary endpoint was the rate of adverse events (AEs). The secondary endpoints were clinical success rate, procedure time, procedure success rate in the absence of the tract dilation step, recurrent biliary obstruction rate, time to biliary obstruction, and overall survival. Risk factors associated with early AEs were also evaluated. RESULTS: The rate of early AEs was significantly lower (12% vs. 35%, P = 0.013) and the procedure success without the tract dilation step was significantly higher (82% vs. 33%, P < 0.001) in the thin than in the conventional delivery system stent group. None of the other secondary endpoints differed significantly between the two groups. Multivariate analysis showed that employing the tract dilation step during EUS-BD was a significant independent risk factor for early AEs (skipping vs. employing; HR, 9.66; 95% CI, 1.13-83.0, P = 0.028). CONCLUSION: Employing the tract dilation step during EUS-BD was a significant risk factor for early AEs. Metal stents with a delivery diameter < 7.5 Fr can be inserted directly without the tract dilation step, resulting in lower early AE rates.

Mannitol versus furosemide in patients with thoracic malignancies who received cisplatin-based chemotherapy using short hydration: A randomized phase II trial.

BACKGROUND: Mannitol is exclusively recommended in the National Comprehensive Cancer Network guidelines for diuresis in cisplatin (CDDP)-based chemotherapy. The utility of furosemide, a widely used and convenient diuretic, thus requires clarification. METHODS: This is a prospective, single-centered, open-label, noninferiority phase II study. Patients with thoracic malignancies who planned to receive CDDP-based chemotherapy were randomly assigned to receive either mannitol (arm A) or furosemide (arm B). The primary end point was set as the proportion of patients who experienced any grade of "creatinine (Cr) increased" based on the upper limit of the normal range (ULN) during the first cycle as assessed by Common Terminology Criteria for Adverse Events Version 4.0. Secondary end points were Cr increased based on the baseline value during the first cycle, Cr increased after the completion of CDDP, and the proportion of patients with phlebitis. RESULTS: Between April 2018 and March 2022, 115 patients were enrolled and 106 were analyzed. Any grade of Cr increased based on the ULN during the first cycle was 17.3% (arm A) and 24.1% (arm B), respectively (p = 0.34). Therefore, the primary end point was not met. After completion of chemotherapy, any grade of Cr increased was observed in 23.1% (arm A) and 31.5% (arm B), respectively. However, the actual serum Cr level and Cr clearance during the courses were not different between the arms. Phlebitis occurred more frequently in arm A (28.8%) than arm B (16.7%). CONCLUSIONS: Mannitol should remain the standard diuresis in CDDP-based chemotherapy assessed by conventional CTCAE grading, but furosemide can be room for consideration when assessed by actual serum Cr level and Cr clearance.

Efficacy of the traction method in gastric endoscopic submucosal dissection: A randomized controlled trial.

BACKGROUND/AIMS: To overcome the technical difficulties associated with gastric ESD, a novel traction device that can alter the direction of traction was developed. Here, we compared the efficacy and safety of conventional ESD with that of traction-assisted gastric ESD. METHODS: Patients with a single gastric epithelial neoplasm were randomized to receive conventional (n=75) or traction-assisted (n=73) gastric ESD. The primary outcome was ESD procedure time. RESULTS: There were no differences between the conventional and traction-assisted groups with respect to treatment results or complications. The mean procedure time was similar for both groups (78.9 vs. 88.3 min, respectively; p=0.3); however, times for the traction device tended to be shorter for lesions in the lesser curvature of the upper or middle stomach (84.6 vs. 123.2 min, respectively; p=0.057). CONCLUSIONS: Traction-assisted ESD for lesions in the lesser curvature of the upper or middle stomach were shorter, thereby reducing the procedure time of conventional ESD.

Effect of istradefylline on postural abnormalities in patients with Parkinson's disease: An association study of baseline postural angle measurements with changes in Unified Dystonia Rating Scale total score.

In our previous study, istradefylline treatment in patients with Parkinson's disease (PD) improved postural abnormalities (PAs), as seen from a decrease in the mean Unified Dystonia Rating Scale (UDRS) total score from week 0 to week 24. A subgroup analysis based on baseline clinical characteristics investigated the association between improvement in the UDRS total score and istradefylline treatment. However, the association between an objective assessment of PAs and improvement in the UDRS total score is unclear. This ad hoc analysis investigated the association between improvement in the UDRS total score after istradefylline treatment and baseline trunk and neck angles, objective assessments of PAs, measured from patients' photographs taken in the previous study. The patients (n = 31) were stratified into groups based on the trunk forward flexion angle (TFFA), trunk lateral flexion angle (TLFA), and neck flexion angle (NFA) values at baseline. From week 0 to week 24, significant improvements in the UDRS total score were found in median percent change (-8.33% [interquartile range: -43.97, 0.00], P = 0.039) in patients with equal to or above the median TFFA values, and in median change (-|1.50 [-9.25, 0.00], P = 0.015) and median percent change (-13.33% [-50.47, 0.00], P = 0.009) in patients with equal to or above the median TLFA values. Patients with more advanced PAs showed more consistent improvements in the UDRS total score with istradefylline. Baseline TFFA and TLFA values, which are objective values, may be useful to assess the istradefylline effectiveness in patients with PD and PAs.

Effects of postoperative oral elemental nutritional supplement on skeletal muscle loss after gastrectomy for gastric cancer.

BACKGROUND: We previously showed that daily nutritional intervention with an oral elemental diet (ED) at 300 kcal/day for 6-8 weeks postoperatively decreased the percentage of body weight loss (%BWL), and that the effect was maintained for 1 year. This post hoc analysis aimed to determine whether this intervention decreased skeletal muscle mass loss 1-year post-gastrectomy. METHODS: Data from consecutive, untreated patients with histopathologically confirmed stage I-III gastric adenocarcinoma who planned to undergo total gastrectomy (TG) or distal gastrectomy (DG) and were enrolled in a previously published randomized trial were used. The primary endpoint was the percentage of skeletal muscle mass index (%SMI) loss from baseline at 1 year postoperatively, based on abdominal computed tomography images obtained preoperatively and at 1 year postoperatively. RESULTS: The overall median %SMI loss was lower in the ED versus control group, but the difference was not significant. The difference in %SMI loss in the ED and control groups was greater in patients with TG (10.1 vs. 13.0; P = 0.12) than in those with DG (5.5 vs. 6.8; P = 0.69). A correlation was observed between %BWL and %SMI loss in both groups (ED group, coefficient 0.591; control group, coefficient 0.644; P < 0.001 for both). Type of gastrectomy (coefficient 7.38; P = 0.001) and disease stage (coefficient - 6.43; P = 0.04) were independent predictors of postoperative skeletal muscle mass loss. CONCLUSION: ED administration for 6-8 weeks following gastrectomy had no inhibitory effect on skeletal muscle loss at 1 year postoperatively. CLINICAL TRIAL REGISTRATION: UMIN000023455.

Impact of sarcopenia on recurrent biliary obstruction after EUS-guided biliary drainage in patients with malignant biliary obstruction.

BACKGROUND AND AIMS: Sarcopenia is an important prognostic factor for cancer patients. The aim of this study was to assess the ability of sarcopenia to predict recurrent biliary obstruction (RBO) in patients with unresectable cancer after EUS-guided biliary drainage (EUS-BD). METHODS: The study enrolled 113 patients who underwent EUS-BD using the self-expandable metal stent (SEMS) for unresectable malignant biliary obstruction (MBO) between April 2016 and December 2021 at Wakayama Medical University Hospital. The skeletal muscle index at the third lumbar spine level (L3) was calculated from computed tomography images. We analyzed the cumulative incidence of RBO at 180 days after stent insertion. Univariate and multivariate analyses were performed to identify variables significantly associated with RBO. RESULTS: Seventy-six patients were assigned to the sarcopenia group, and 37 were assigned to the non-sarcopenia group. The 180-day cumulative incidence of RBO was 11% in the non-sarcopenia group and 29% in the sarcopenia group (p = 0.034). The time to RBO was significantly shorter for the sarcopenia group (p = 0.028; Gray's test). Multivariate analyses identified sarcopenia as an independent prognostic factor for RBO (present vs absent; HR 4.61; 95% CI 1.76-12.10, p = 0.001). The rates of biliary sludge/food impaction were significantly higher in the sarcopenia group for the causes of RBO (p = 0.048). There were no significant differences between the sarcopenia and the non-sarcopenia groups with respect to related EUS-BD adverse events. CONCLUSION: Sarcopenia is an independent indicator of RBO in patients with MBO who receive EUS-BD with SEMS.

Protocol for a Multi-Center Confirmatory Trial to Evaluate the Differential Diagnostic Performance of Contrast-Enhanced Ultrasonography Using Perflubutane in Patients with a Pancreatic Mass: A Multicenter Prospective Study.

For pancreatic masses, an evaluation of their vascularity using contrast-enhanced ultrasonography can help improve their characterization. This study was designed to evaluate the utility and safety of contrast-enhanced transabdominal ultrasonography (CE-TUS) and endoscopic ultrasonography (CE-EUS) in the diagnosis of pancreatic masses including solid or cystic masses. This multi-center comparative open-label superiority study is designed to compare Plain (P)-TUS/EUS alone with P-TUS/P-EUS plus CE-TUS/CE-EUS. Three hundred and one patients with a total of 232 solid pancreatic masses and 69 cystic masses were prospectively enrolled. The primary endpoints are to compare the diagnostic accuracy between P-TUS/P-EUS alone and P-TUS/P-EUS plus CE-TUS/CE-EUS for both the TUS and EUS of solid pancreatic masses, and to compare the diagnostic accuracy between P-EUS alone and P-EUS plus CE-EUS in cystic pancreatic masses. The secondary endpoints are to compare the diagnostic sensitivity and specificity of P-TUS/P-EUS alone and P-TUS/P-EUS plus CE-TUS/CE-EUS for pancreatic solid/cystic masses, and the accuracy of P-TUS alone and P-TUS plus CE-TUS for pancreatic cystic masses. Other secondary endpoints included comparing the diagnostic sensitivity, specificity, and accuracy of CE-TUS, CE-EUS and CE-computed tomography (CT) for solid/cystic pancreatic masses. The safety, degree of effective enhancement, and diagnostic confidence obtained with CE-TUS/CE-EUS will also be assessed.

Phase II study of S-1 plus docetaxel as first-line treatment for older patients with advanced gastric cancer (OGSG 0902).

BACKGROUND: Although there is insufficient evidence for the treatment of older patients with advanced gastric cancer, fluorouracil combined with platinum chemotherapy has been recognized as a standard first-line treatment for such populations in Japan despite the lack of efficacy and toxicity data. METHODS: Patients aged 75 years or older with advanced gastric cancer were enrolled. S-1 plus docetaxel (docetaxel: 40 mg/m2, day 1; S-1: 80 mg/m2, days 1-14; q21 days) was repeated every 3 weeks. The primary endpoint was overall response rate. Secondary endpoints were safety, progression-free survival, time to treatment failure, and overall survival. The sample size was calculated as 30 under the hypothesis of an expected response rate of 40% and a threshold response rate of 20%, at a power of 90% and a two-sided alpha value of 5%. RESULTS: From February 2010 to January 2015, 31 patients were enrolled and assessed for efficacy and toxicity. The response rate was 45.2% (95% CI 27.3%-64.0%; p = 0.001) and it exceeded the expected response rate set at 40%. Median progression-free survival was 5.8 months, the 1-year survival rate was 58.1%, and the median survival time was 16.1 months. The major grade 3/4 adverse events were neutropenia (58%), febrile neutropenia (13%), anemia (10%), anorexia (10%), and fatigue (6%). CONCLUSIONS: These findings indicate that S-1 plus docetaxel as first-line treatment for older patients is feasible and that it has promising efficacy against advanced gastric cancer.

A Phase II Study of High-Flow Nasal Cannula for Relieving Dyspnea in Advanced Cancer Patients.

CONTEXT: The efficacy and tolerability of high-flow nasal cannula (HFNC) for relieving dyspnea in advanced cancer patients with limited prognosis requires elucidation. OBJECTIVES: The primary aim of this trial was to assess the efficacy and tolerability of HFNC regarding dyspnea including severe as well as moderate for longer durations in patients under palliative care. METHODS: In this prospective study, hospitalized patients with advanced cancer who had dyspnea at rest (numeric rating scale, NRS≥3) and hypoxemia were enrolled. They were treated with HFNC for five days in the respiratory unit. Primary endpoint was mean change of modified Borg scale at 24 hours. Key secondary endpoints consisted of mean changes in modified Borg scale during the study period and feasibility (Trial Identifier, UMIN000035738). RESULTS: Between February 2019 and February 2022, 25 patients were enrolled and 21 were analyzed. Twenty patients used inspired oxygen and the mean fraction of inspired oxygen (FiO2) was 0.34 (range, 0.21-1.0). At baseline, mean NRS (dyspnea) was 5.9 (range, 3-10). Median survival time was 19 days (range, 3-657). The mean change of modified Borg scale was 1.4 (80% confidence interval [CI]: 0.8-1.9) at 24 hours, 12 patients (57%) showed 1.0 points improvement of modified Borg scale. Within two hours, 15 patients showed 1.0 points improvement of modified Borg scale and such early responders were likely to maintain dyspnea improvement for 24 hours. Nineteen patients could continue HFNC for 24 hours and 11 patients completed five days of HFNC. CONCLUSION: To our knowledge, this trial is the first prospective study to assess the five-day efficacy and tolerability of HFNC for dyspnea in patients under palliative care. Although this did not reach the prespecified endpoint, about half of the patients showed 1.0 point improvement, a minimally clinically important difference (MCID) in the chronic lung disease. HFNC can be a palliative treatment option in advanced cancer patients with dyspnea.

K-ras gene mutation analysis to diagnosis pancreatic adenocarcinoma from endoscopic ultrasound-guided tissue acquisition; a systematic review and meta-analysis.

BACKGROUND: Endoscopic ultrasound-guided tissue acquisition (EUS-TA) has high sensitivity for the pathological diagnosis of pancreatic masses, but also a high false-negative rate. K-ras gene mutations occur in over 75 % of pancreatic ductal adenocarcinomas (PDAC), and this meta-analysis evaluated the utility of detecting K-ras gene mutations from EUS-TA specimens for the diagnosis of PDAC. METHODS: Relevant studies in PubMed, the Cochrane Library, and Web of Science were systematically searched. Meta-analysis was performed on data from the selected studies using a bivariate model to provide pooled values of sensitivity, specificity, and their 95 % confidence intervals (CIs). RESULTS: This meta-analysis included 1521 patients (from 10 eligible studies) who underwent EUS-TA with K-ras gene mutation analysis for diagnosis of pancreatic solid masses. The pooled estimates of sensitivity and specificity were 76.6 % (95 % CI, 70.9-81.5 %) and 97.0 % (95 % CI, 94.0-98.5 %), respectively, for pathological diagnosis, 75.9 % (95 % CI 69.5-81.4 %) and 95.3 % (95 % CI, 92.3-97.2 %) for K-ras gene mutation analysis, and 88.7 % (95 % CI 87.1-91.7 %) and 94.9 % (95 % CI, 91.5-97.0 %) for pathological diagnosis in combination with K-ras gene mutation analysis. The sensitivity for diagnosis of PDAC was significantly higher for pathological diagnosis in combination with K-ras gene mutation analysis than for pathological diagnosis or K-ras gene mutation analysis alone (both, p < 0.001). There was no difference in specificity between pathological diagnosis in combination with K-ras gene mutation analysis and both either (p = 0.234, 0.945, respectively). CONCLUSIONS: K-ras gene mutation analysis in combination with to pathological diagnosis of EUS-TA increases the accuracy of differential diagnosis of PDAC.

Adverse events of self-expandable metal stent placement for malignant distal biliary obstruction: a large multicenter study.

BACKGROUND AND AIMS: Endoscopic placement of self-expandable metal stents (SEMSs) for malignant distal biliary obstruction (MDBO) may be accompanied by several types of adverse events. The present study analyzed the adverse events occurring after SEMS placement for MDBO. METHODS: The present study retrospectively investigated the incidence and types of adverse events in patients who underwent SEMS placement for MDBO between April 2018 and March 2021 at 26 hospitals. Risk factors for acute pancreatitis, cholecystitis, and recurrent biliary obstruction (RBO) were evaluated by univariate and multivariate analyses. RESULTS: Of the 1425 patients implanted with SEMSs for MDBO, 228 (16.0%) and 393 (27.6%) experienced early adverse events and RBO, respectively. Pancreatic duct without tumor involvement (P = .023), intact papilla (P = .025), and SEMS placement across the papilla (P = .037) were independent risk factors for acute pancreatitis. Tumor involvement in the orifice of the cystic duct was an independent risk factor for cholecystitis (P < .001). Use of fully and partially covered SEMSs was an independent risk factor for food impaction and/or sludge. Use of fully covered SEMSs was an independent risk factor for stent migration. Use of uncovered SEMSs and laser-cut SEMSs was an independent risk factor for tumor ingrowth. CONCLUSIONS: Pancreatic duct without tumor involvement, intact papilla, and SEMS placement across the papilla were independent risk factors for acute pancreatitis, and tumor involvement in the orifice of the cystic duct was an independent risk factor for cholecystitis. The risk factors for food impaction and/or sludge, stent migration, and tumor ingrowth differed among types of SEMSs.

Efficacy and safety of neoadjuvant nab-paclitaxel plus gemcitabine therapy in patients with borderline resectable pancreatic cancer: A multicenter single-arm phase II study (NAC-GA trial).

Background: Nab-paclitaxel plus gemcitabine is a standard treatment for metastatic/locally advanced pancreatic cancer. The effectiveness of neoadjuvant therapy with nab-paclitaxel plus gemcitabine (GnP-NAT) in patients with borderline resectable pancreatic cancer (BRPC) remains unclear. Patients and Methods: This single-arm phase II trial included 61 patients with BRPC that were treated with two cycles of GnP-NAT, (nab-paclitaxel 125 mg/m2 and gemcitabine 1000 mg/m2), on days 1, 8, and 15 over a 4-week period, which comprised one cycle. The primary endpoint was overall survival time. In the absence of disease progression, patients underwent planned pancreatectomy. Results: Median overall survival, the primary endpoint, was 25.2 months, and the median recurrence-free survival was 12.3 months. The overall rate of grade 3/4 events was 73.8%. One patient, who had a history of radiation therapy for past esophageal cancer, died from exacerbation via pneumonia. The overall resection rate was 73.8% (n = 45), and the R0 resection rate was 63.9% (n = 39). Overall, postoperative complications were found in 19 patients (42%) with 24 events, and nine patients (20%) with nine events ≥ grade IIIa, based on Dindo's classification. Conclusions: This protocol treatment is thought to be a feasible, safe, and promising treatment regimen, but we caution against its use in patients with a history of interstitial lung disease and/or prior pulmonary irradiation. The survival data from this study suggest the need for further investigations of GnP-NAT efficacy in patients with BRPC, as well as prospective evaluation of adverse events. Clinical Trial Registration: UMIN Clinical Trials Registry, UMIN000024154 and ClinicalTrials.gov, NCT02926183.

Virtual Reality Images of the Home Are Useful for Patients With Hospital-Based Palliative Care: Prospective Observational Study With Analysis by Text Mining.

Background: Malignancy patients who need long-term hospitalization can feel loneliness affecting their quality of life. The global COVID-19 pandemic has caused visiting restrictions that could mean patients who might be missing out on family support and palliative care, therefore, need to adapt and change. We used virtual reality (VR) technology with the aim of reducing feelings of loneliness among these patients. Objectives: In a small cohort setting, we aimed to clarify the usefulness of VR viewing for this purpose by text mining interviews with the patients in palliative care after their VR experience, and to clarify the feasibility of this program. Design and Setting/Subjects: Four consecutive Japanese patients in the palliative care unit viewed personalized familiar persons or places through VR goggles, while communicating by telephone. After the VR experience, text mining of the patients' interviews was used to extract the words for the frequency count and co-occurrence analysis. Results: Four clusters were extracted: "relief from the pain of hospitalization by feeling safe and secure with family members nearby," "using VR to regain daily life," "immersive feeling of being in the same space as family," and "loneliness due to the realistic feeling of separation from the family through VR experience." There were no cases of VR sickness. Conclusion: Our results attained by text mining suggest the promising potential of VR imaging of familiar surroundings for patients in palliative care.

Meta-analysis of three randomized trials of capecitabine plus cisplatin (XP) versus S-1 plus cisplatin (SP) as first-line treatment for advanced gastric cancer.

BACKGROUND: S-1 plus cisplatin (SP) and capecitabine plus cisplatin (XP) are standard first-line regimens for advanced gastric cancer (AGC) worldwide. We conducted a meta-analysis using individual participant data (IPD) to investigate which is more suitable. METHODS: IPD from three randomized trials were collected. In these trials, patients with AGC were randomly allocated to SP (S-1 80-120 mg for 21 days plus cisplatin 60 mg/m2 (q5w)) or XP (capecitabine 2000 mg/m2 for 14 days plus cisplatin 80 mg/m2 (q3w)). RESULTS: In 211 eligible patients, median overall survival (OS) for SP versus XP was 13.5 and 11.7 months (hazard ratio [HR], 0.787; p = 0.114), progression-free survival (PFS) was 6.2 and 5.1 months (HR, 0.767; P = 0.076), and TTF was 5.1 and 4.0 months (HR, 0.611; P = 0.001). The most common grade ≥ 3 adverse events with SP or XP were neutropenia (18% vs. 29%) and anorexia (16% vs.18%). Subgroup analysis demonstrated significant interaction between treatment effect and performance status > 1 (HR, 0.685; P = 0.036), measurable lesion (HR, 0.709; P = 0.049), primary upper third tumor (HR, 0.539; P = 0.040), and differentiated type (HR, 0.549; interaction, 0.236; P = 0.019). For the differentiated type, OS was significantly longer in the SP group (13.2 months) than in the XP group (11.1 months) (HR, 0.549; P = 0.019). For the undifferentiated type, OS was similar in the SP group (14.2 months) and in the XP group (12.4 months) (HR, 0.868; P = 0.476). CONCLUSIONS: SP and XP were both effective and well tolerated. SP might be suitable for the pathological differentiated subtype of AGC. CLINICAL TRIAL REGISTRATION: The HERBIS-2, HERBIS-4A, and XParTS II trials were registered with UMIN-CTR as UMIN000006105, UMIN000006755, and UMIN000006045, respectively.

Novel Technique of Endoscopic Ultrasonography for the Differential Diagnosis of Gallbladder Lesions and Intraductal Papillary Mucinous Neoplasms: A Single-Center Prospective Study.

Detective flow imaging endoscopic ultrasonography (DFI-EUS) is an innovative imaging modality that was developed to detect fine vessels and low-velocity blood flow without contrast agents. We evaluate its utility for the differential diagnosis of gallbladder lesions and intraductal papillary mucinous neoplasms (IPMNs). We enrolled patients who underwent DFI-EUS, e-FLOW EUS, and contrast-enhanced EUS for gallbladder lesions or IPMNs. The detection of vessels using DFI-EUS and e-FLOW EUS was compared with that via contrast-enhanced EUS and pathological findings. The vessel pattern was also categorized as regular or irregular. Of the 33 lesions included, there were final diagnoses of 13 IPMNs and 20 gallbladder lesions. DFI-EUS was significantly superior to e-FLOW EUS for discriminating between mural nodules and mucous clots and between solid gallbladder lesions and sludge using the presence or absence of vessel detection in lesions (p = 0.005). An irregular vessel pattern with DFI-EUS was a significant predictor of malignant gallbladder lesions (p = 0.002). DFI-EUS is more sensitive than e-FLOW-EUS for vessel detection and the differential diagnosis of gallbladder lesions and IPMNs. Vessel evaluation using DFI-EUS may be a useful and simple method for differentiating between mural nodules and mucous clots in IPMN, between solid gallbladder lesions and sludge, and between malignant and benign gallbladder lesions.

Phase I dose-escalation study on irinotecan, cisplatin, and S-1 combination in chemotherapy-naïve patients with HER2-negative advanced gastric cancer (HERBIS-4B, OGSG 1106).

BACKGROUND: The development of triplet regimens for advanced gastric cancer is challenging. The aim of this phase I dose-escalation study was to determine the maximum tolerated dose and recommended dose of the combination of irinotecan, cisplatin, and S-1 in chemotherapy-naïve patients with HER2-negative advanced gastric cancer. METHODS: The 3 + 3 design was adopted. Every 4 weeks, patients received an escalating dose of intravenous irinotecan (100-150 mg/m2) on day 1 and fixed doses of intravenous cisplatin (60 mg/m2) on day 1 and oral S-1 (80 mg/m2) on days 1 to 14. RESULTS: Twelve patients were enrolled in two dose level cohorts. In the level 1 cohort (irinotecan 100 mg/m2, cisplatin 60 mg/m2, and S-1 80 mg/m2), dose-limiting toxicity including grade 4 neutropenia and febrile neutropenia occurred in one of six patients, whereas in the level 2 cohort (irinotecan 125 mg/m2, cisplatin 60 mg/m2, and S-1 80 mg/m2), dose-limiting toxicities including grade 4 neutropenia developed in two of six patients. Thus, the level 1 and 2 doses were determined to be the recommended and maximum tolerated doses, respectively. Common grade 3 or higher adverse events were neutropenia (75%; n = 9), anemia (25%; n = 3), anorexia (8%; n = 1), and febrile neutropenia (17%; n = 2). Irinotecan, cisplatin, and S-1 combination therapy achieved an overall response rate of 67% with a median progression-free survival and overall survival of 19.3 and 22.4 months, respectively. CONCLUSIONS: The potential treatment efficacy of this triplet regimen in HER2-negative advanced gastric cancer warrants further evaluation, especially in patients requiring intensive chemotherapy.

Rule ensemble method with adaptive group lasso for heterogeneous treatment effect estimation.

The increasing scientific attention given to precision medicine based on real-world data has led to many recent studies clarifying the relationships between treatment effects and patient characteristics. However, this is challenging because of ubiquitous heterogeneity in the treatment effect for individuals and the real-world data on their backgrounds being complex and noisy. Because of their flexibility, various machine learning (ML) methods have been proposed for estimating heterogeneous treatment effect (HTE). However, most ML methods incorporate black-box models that hamper direct interpretation of the relationships between an individual's characteristics and treatment effects. This study proposes an ML method for estimating HTE based on the rule ensemble method RuleFit. The main advantages of RuleFit are interpretability and accuracy. However, HTEs are always defined in the potential outcome framework, and RuleFit cannot be applied directly. Thus, we modified RuleFit and proposed a method to estimate HTEs that directly interpret the relationships among the individuals' features from the model. Actual data from an HIV study, the ACTG 175 dataset, was used to illustrate the interpretation based on the ensemble of rules created by the proposed method. The numerical results confirm that the proposed method has high prediction accuracy compared to previous methods, indicating that the proposed method establishes an interpretable model with sufficient prediction accuracy.