Predictive Factors and Outcomes after Allogeneic Stem-Cell Transplantation for Adults with Acute Lymphoblastic Leukemia in Brazil

Introduction: Allogeneic stem-cell transplantation (HSCT) remains a potentially curative approach for acute lymphoblastic leukemia (ALL), especially for high-risk patients and those with relapsed/refractory disease, although its efficacy is offset by a not negligible toxicity. Adult patients with ALL fare worse in developing countries with low data about the HSCT in this setting. In this study, we aim to describe outcomes and examine risk factors for overall survival (OS), disease-free survival (DFS), cumulative incidence of relapse (CIR), non-relapse mortality (NRM) and graft-versus-host disease (GVHD) after HSCT for ALL in Brazilian centers. Methods: This is a retrospective registry study. Patients with ALL or ambiguous lineage leukemia above 16 years who underwent a first HSCT in 5 Brazilian centers between January 2007 and December 2017 were included. Kaplan-Meier method and competing risk analysis were used. Multivariable analysis (MVA) was performed using Cox regression and the Akaike's information criteria was used for model selection. Cut-offs for continuous variables were calculated through "findcut" R function. Center effect was evaluated by using frailty model. Results: Overall, 275 patients were included with a median age of 31y (range, 16-65). Philadelphia chromosome was found in 35%. Baseline characteristics are summarized in Table 1. Matched sibling donor (MSD), matched unrelated donor (MUD), mismatched unrelated donor (MMUD), haploidentical donor and umbilical cord were reported in 53%, 19%, 9%, 19%, and 5%, respectively. Total body irradiation (TBI) was used in 67% of myeloablative HSCT. Median time to HSCT in CR1 was 7.8 months. Engraftment failure rate was 1.5%. Median follow-up time was 6.4 y. Cumulative incidence of acute grade II-IV and chronic GVHD were 54.2% and 26.2%, respectively. In MVA, the use of MUD (HR=2.3) and increased donor age (HR=1.02) were associated with GVHD. Five-year CIR was 28.1% (95% CI 22.9-33.6) and 5-y NRM was 34.1% (95% CI 28.4-39.8). At D+100, NRM incidence was 22.6%. Central nervous system involvement at the diagnosis (HR=2.2), and disease status (HR 1.8 for CR2+, and HR 7.9 for refractory) increased relapse incidence, whereas the use of peripheral blood graft (HR=0.51) and haploidentical donor (HR=0.4) significantly decreased relapse incidence. In MVA, NRM was increased by patient's age (HR=1.04), refractory status (HR=4.2), MUD (HR=3.8) and donor age (HR=1.02). Center effect was significantly associated with relapse and NRM. Five-year OS and DFS were 40.7% (95% CI 35.1-47.1) and 37.8% (95% CI-32.3-44.1), respectively (Figure 1). Patient's age, donor age and disease status were independently associated with OS and DFS (Table 2). When GVHD (as a time-dependent variable) was introduced in the MVA for OS and DFS, it was associated with decreased OS (HR 4.2, p<0.001) but not with DFS. Pre-HSCT positivity of minimal residual disease (>0.01%) was associated with worse DFS in univariate analysis (HR=1.47) in available cases. Conclusions: This is the largest series of ALL adults receiving HSCT from Brazil. While OS and DFS were similar to published data, NRM was higher. Patient's age and donor age outweighed donor type or graft source in our analysis. Interestingly, haploidentical HSCT related to lower CIR, whereas the use of MUD was associated with higher NRM and GVHD rates. These results impact on donor selection strategy in our country, aiming to timely offer HSCT for high-risk ALL patients in our setting. Figure 1 Figure 1. Disclosures No relevant conflicts of interest to declare.

Adherence to European ovarian cancer guidelines and impact on survival: a French multicenter study (FRANCOGYN)

ObjectiveThe primary objective of the study was to validate the European Society for Medical Oncology (ESMO)–European Society of Gynecologic Oncology (ESGO) ovarian cancer guideline as a method of assessing quality of care, and to identify patient characteristics predictive of non-adherence to European guideline care. The secondary objectives were to analyze the evolution of practices over the years and to evaluate heterogeneity between centers.MethodsThis retrospective multicenter cohort study of invasive epithelial ovarian cancer reported to the FRANCOGYN database included data from 12 French centers between January 2000 and February 2017. The main outcome was adherence to ESMO–ESGO guidelines, defined by recommended surgical procedures according to the International Federation of Gynecology and Obstetrics (FIGO) stage and appropriate chemotherapy. Mixed multivariable logistic regression analysis with a random center effect was performed to estimate the probability of adherence to the guidelines. Survival analysis was carried out using the Kaplan–Meier method and a mixed Cox proportional hazards model.Results1463 patients were included in the study. Overall, 317 (30%) patients received complete guideline adherent care. Patients received appropriate surgical treatment in 69% of cases, while adequate chemotherapy was administered to 44% of patients. Both patient demographics and disease characteristics were significantly associated with the likelihood of receiving guideline adherent care, such as age, performance status, FIGO stage, and initial burden of disease. In univariate and multivariate survival analysis, adherence to the guidelines was a statistically significant and independent predictor of decreased overall survival. Patients receiving suboptimal care experienced an increased risk of death of more than 100% compared with those treated according to the guidelines (hazard ratio 2.14, 95% confidence interval 1.32 to 3.47, p<0.01). In both models, a significant random center effect was observed, confirming the heterogeneity between centers (p<0.001).ConclusionsAdherence to ESMO–ESGO guidelines in ovarian cancer was associated with a higher overall survival and may be a useful method of assessing quality of care.

Center Effects: Counseling Center Variables as Predictors of Psychotherapy Outcomes

Treatment context may have a unique impact on psychotherapy outcomes, above and beyond client, therapist, and therapy process variables. University counseling centers represent one such treatment context facing increasing treatment demands. This study examined the role of counseling centers and center variables in explaining differences in psychotherapy outcomes. The Center for Collegiate Mental Health, a large practice–research network, contained data from 116 counseling centers, 2,362 therapists, and 58,423 clients. Multilevel modeling tested if some counseling centers systematically achieved better outcomes than others (a “center effect”). Outcome was operationalized as clients’ magnitude and rate of change in distress across treatment. Results showed a relatively small “center effect” for both outcomes. Analyses sought to explain that center effect through administrative policies and characteristics. As a group, these variables partially explained the center effect. None explained a large portion of total outcome variance. Potential future implications for policy and advocacy efforts are discussed.

A transfer learning approach to facilitate ComBat-based harmonization of multicentre radiomic features in new datasets

Purpose To facilitate the demonstration of the prognostic value of radiomics, multicenter radiomics studies are needed. Pooling radiomic features of such data in a statistical analysis is however challenging, as they are sensitive to the variability in scanner models, acquisition protocols and reconstruction settings, which is often unavoidable in a multicentre retrospective analysis. A statistical harmonization strategy called ComBat was utilized in radiomics studies to deal with the “center-effect”. The goal of the present work was to integrate a transfer learning (TL) technique within ComBat—and recently developed alternate versions of ComBat with improved flexibility (M-ComBat) and robustness (B-ComBat)–to allow the use of a previously determined harmonization transform to the radiomic feature values of new patients from an already known center. Material and methods The proposed TL approach were incorporated in the four versions of ComBat (standard, B, M, and B-M ComBat). The proposed approach was evaluated using a dataset of 189 locally advanced cervical cancer patients from 3 centers, with magnetic resonance imaging (MRI) and positron emission tomography (PET) images, with the clinical endpoint of predicting local failure. The impact performance of the TL approach was evaluated by comparing the harmonization achieved using only parts of the data to the reference (harmonization achieved using all the available data). It was performed through three different machine learning pipelines. Results The proposed TL technique was successful in harmonizing features of new patients from a known center in all versions of ComBat, leading to predictive models reaching similar performance as the ones developed using the features harmonized with all the data available. Conclusion The proposed TL approach enables applying a previously determined ComBat transform to new, previously unseen data.

Peritoneal dialysis and center effect

Outcomes of patients treated by peritoneal dialysis depends on their characteristics, which generally cannot be modified. Conversely, center characteristics or practices having an effect on the outcomes of patients and may be subject to change. Study the center effect and identify "center" variables associated with the patient's future is therefore necessary. Taken as part of a science thesis, the importance of the center effect and the role and importance of nursing home visits in preventing peritoneal infections. We also observed that there was a disparity between centers in the use and allocation of home assistance by a nurse for carrying out dialysis. In a context where organizations favor the versatility of health professionals, our work underlines the importance of the experience and expertise of peritoneal dialysis nurses in the success of the method. Increasing the size of the centers and optimizing resources could have a beneficial effect on the survival of the method by improving the experience of the centers.