Competing Risks Methods Research Articles

The impact of large aneurysm diameter on the outcomes of thoracoabdominal aneurysm repair by fenestrated and branched endografts.

Aim of the study was to analyze the impact of preoperative thoraco-abdominal aneurysm diameter on the outcomes of fenestrated/branched endografting. Patients underwent endovascular thoraco-abdominal repair at 2 European centers (2011-2021) were analyzed. Median diameter was calculated; third quartile was considered as cut-off. Outcomes were compared in two groups based on the diameter value. Primary end-points were technical success, spinal cord ischaemia and 30-day/in-hospital mortality. Survival, freedom-from-reintervention and target-visceral-vessels instability were follow-up outcomes. Out of 247 thoraco-abdominal aortic aneurysms, the median diameter was 65 mm, first quartile was 57 mm; third quartile was 80 mm, set as cut-off value. Fifty-nine(24%) patients had diameter ≥80mm. Custom-Made and off-the-shelf branched endograft were used in 160(65%) and 87(35%), respectively. Technical success was 93% (<80mm : 91% vs ≥80mm : 94%; P:0.47). Twenty-three(9%) patients had spinal injury (<80mm : 7% vs ≥80mm : 17%; P:0.03). Twenty-two(9%) patients died within 30-day/in-hospital (<80mm : 7%; vs ≥80mm : 15%; P:0.06). Multivariate analysis did not report pre-operative diameter ≥80mm as significant risk factor for primary end-points. The median follow-up was 13 (interquartile-range : 2-37) months and at 3-years survival and freedom from reintervention rates were 65% and 62%, respectively. After univariate and multivariate analysis preoperative diameter ≥80mm was considered an independent risk factor for reinterventions (HR : 1.9; 95% CI : 1.1-3.6; P:0.04), and for target visceral vessels instability (HR : 3.1; 95% CI : 1.3-5.1; P:0.04), occurred in 45(18%) cases. However after competing risk methods preoperative diameter did not show significance for follow-up results. A pre-operative thoraco-abdominal aortic aneurysm diameter greater than 80 mm has not a direct impact on early technical and clinical outcomes. Diameter ≥80mm is considered risk-factor for reinterventions and target-vessels instability considered separately during follow-up.

Heart failure and major haemorrhage in people with atrial fibrillation

BackgroundHeart failure (HF) is not included in atrial fibrillation (AF) bleeding risk prediction scores, reflecting uncertainty regarding its importance as a risk factor for major haemorrhage. We aimed to report...

Selection Bias in Reporting of Median Waiting Times in Organ Transplantation

Median organ waiting times published by transplant organizations may be biased when not appropriately accounting for censoring, death, and competing events. This can lead to overly optimistic waiting times for all transplant programs and, consequently, may deceive patients on the waiting list, transplant physicians, and health care policymakers. To apply competing-risk multistate models to calculate probabilities for transplantation and adverse outcomes on the Swiss national transplant waiting list. The WAIT (Waitlist Analysis in Transplantation) study was a retrospective cohort study of all transplant candidates in Switzerland listed from January 1, 2018, or later and observed until December 31, 2023. Transplant candidates were listed in 1 of the 6 transplant centers (Basel, Bern, Geneva, Lausanne, St Gallen, and Zurich) for heart, liver, lungs, kidney, or pancreas and/or islet transplant. A total of 4352 candidates were listed during the study period, of whom 709 (16.3%) were excluded due to living-donor transplant (691 in the kidney program and 18 in the liver program). Waiting for organ transplant. Time to transplantation, death, or delisting. Competing-risk multistate models were used to analyze time-to-event data from the national organ waiting list with the Aalen-Johansen estimator to compute probabilities for both transplant and adverse outcomes. Results were compared with the sample median among only those undergoing transplant and the Kaplan-Meier method with censoring of competing events. Data from 3643 transplant candidates (2428 [66.6%] male; median age, 56 [range, 0-79] years) were included in the analysis. The median time to transplantation (MTT) was 0.91 (95% CI, 0.83-1.07) years for heart, 3.10 (95% CI, 2.57-3.77) years for kidney, 1.32 (95% CI, 0.76-1.55) years for liver, 0.80 (95% CI, 0.37-1.12) years for lung, and 1.62 (95% CI, 0.91-2.17) years for pancreas and/or islet programs. Alternative estimation methods introduced bias to varying degrees: the sample median among only persons undergoing transplantation underestimated the waiting time by 38% to 61% and the Kaplan-Meier method by 2% to 12% compared with the MTT. In this cohort study of transplant candidates in Switzerland, the MTT, the duration at which the transplant probability is 0.50, was used as a measure of average waiting time. Suboptimal methods led to biased and overly optimistic waiting time estimations; thus, applying appropriate competing-risk methods to address censoring and competing events is crucial.

Fracture risk assessment in the presence of competing risk of death.

To identify the optimal statistical approach for predicting the risk of fragility fractures in the presence of competing event of death. We used real-world data from the Dubbo Osteoporosis Epidemiology Study that has monitored 3035 elderly participants for bone health and mortality. Fragility fractures were ascertained radiologically. Mortality was confirmed by the State Registry. We considered four statistical models for predicting fracture risk: (i) conventional Cox's proportional hazard model, (ii) cause-specific model, (iii) Fine-Gray sub-distribution model, and (iv) multistate model. These models were fitted and validated in the development (60% of the original sample) and validation (40%) subsets, respectively. The model performance was assessed by discrimination and calibration analyses. During a median follow-up of 11.3 years (IQR: 7.2, 16.2), 628 individuals (34.5%) in the development cohort fractured, and 630 (34.6%) died without a fracture. Neither the discrimination nor the 5-year prediction performance was significantly different among the models, though the conventional model tended to overestimate fracture risk (calibration-in-the-large index = - 0.24; 95% CI: - 0.43, - 0.06). For 10-year risk prediction, the multistate model (calibration-in-the-large index = - 0.05; 95% CI: - 0.20, 0.10) outperformed the cause-specific (- 0.23; - 0.30, - 0.08), Fine-Gray (- 0.31; - 0.46, - 0.16), and conventional model (- 0.54; - 0.70, - 0.39) which significantly overestimated fracture risk. Adjustment for competing risk of death has minimum impact on the short-term prediction of fracture. However, the multistate model yields the most accurate prediction of long-term fracture risk and should be considered for predictive research in the elderly, who are also at high mortality risk. Fracture risk assessment might be compromised by the competing event of death. This study, using real-world data found a multistate model was superior to the current competing risk methods in fracture risk assessment. A multistate model is considered an optimal statistical method for predictive research in the elderly.

Cause-specific and all-cause mortalities in vegetarian compared with those in nonvegetarian participants from the Adventist Health Study-2 cohort

BackgroundThere have been mixed results reported internationally when associating vegetarian dietary patterns with all-cause and cause-specific mortalities. ObjectivesThis study aimed to extend our previous results by evaluating, with a larger number of deaths (N = 12,515), cause-specific mortalities comparing different vegetarian types with nonvegetarians. MethodsThis prospective study used data from the Adventist Health Study-2 cohort. Mortality was ascertained between study baseline, 2002–2007, and follow-up through 2015. Dietary data were collected at baseline using a validated quantitative food frequency questionnaire and then categorized into 5 dietary patterns: nonvegetarian, semivegetarian, pescovegetarian, lacto-ovovegetarian, and vegan. Main outcomes and measures include all-cause and cause-specific mortalities using Cox proportional hazards regression models and competing risk methods. ResultsThe analytic sample included 88,400 participants who provided 971,424 person-years of follow-up. We report results pairwise as estimated at ages 65 and 85 y owing to age dependence of many hazard ratios (HRs). Compared with nonvegetarians, vegetarians had lower risks of mortality, overall (HR: 0.89; 95% confidence interval [CI]: 0.83, 0.95; HR: 0.98; 95% CI: 0.91, 1.04), from renal failure (HR: 0.52; 95% CI: 0.38, 0.70; HR: 0.65; 95% CI: 0.55, 0.76), infectious disease (HR: 0.57; 95% CI: 0.40, 0.82; HR: 0.90; 95% CI: 0.70, 1.17), diabetes (HR: 0.51; 95% CI: 0.33, 0.78; HR: 0.69; 95% CI: 0.53, 0.88), select cardiac (HR: 0.75; 95% CI: 0.65, 0.87; HR: 0.89; 95% CI: 0.83, 0.95), and ischemic heart disease causes (HR: 0.73; 95% CI: 0.59, 0.90; HR: 0.84; 95% CI: 0.75,0.94). Vegans, lacto-ovovegetarians, and pescovegetarians were also observed to have lower risks of total mortality and several similar cause-specific mortalities. However, higher cause-specified neurologic mortalities were observed among older vegetarians (estimated at age 85 y), specifically stroke (HR: 1.17; 95% CI: 1.02, 1.33), dementia (HR: 1.13; 95% CI: 1.00, 1.27), and Parkinson's disease (HR: 1.37; 95% CI: 0.98, 1.91). Results in Black subjects for vegetarian/nonvegetarian comparisons largely followed the same trends, but HRs were less precise owing to smaller numbers. ConclusionsVegetarian diets are associated with lower risk for all-cause and many cause-specific mortalities, especially among males and in younger subjects. However, higher risks are observed among older vegetarians for stroke and dementia. These results need further support and investigation.

Modeling home property listings’ time-on-market duration and listing outcome using copula-based competing risk method

Modeling home property listings’ time-on-market duration and listing outcome using copula-based competing risk method

The conditional recurrence-free survival after R0 hepatectomy for locally advanced intrahepatic cholangiocarcinoma: A competing risk analysis based on inflammation-nutritional status

BackgroundConditional survival analysis can serve as a dynamic prognostic metric, which helps to estimate the real-time survival probability over time. The present study conducted a conditional recurrence-free survival (CRFS) analysis for locally advanced intrahepatic cholangiocarcinoma (ICC) after R0 hepatectomy from an inflammatory-nutritional perspective using the competing risk method. MethodsWe extracted the medical data of 164 locally advanced ICC patients after R0 resection from Sun Yat-sen University Cancer Center. The calculation formula of the CRFS rate is CRFS(y/x) = RFS(y + x)/RFS(x). Univariable and multivariable COX regression analysis and competing risk analysis were conducted to identify RFS indicators. ResultsConsidering death before recurrence as a competing risk factor, the conditional RFS rates every 6 months gradually increased over time. The 24-month RFS rate increased from 29.2 % to 49.9 %, 68.5 %, and 85.1 % given 6, 12, and 18-month already recurrence-free survival, respectively. Both in multivariate COX regression analysis and competing risk analysis, tumor diameter and number, lymph node metastasis, aggregate systemic inflammation index score (AISI), and albumin-bilirubin score (ALBI) all remained significant. For both AISI and ALBI variables, the CRFS rates in the low-value set were higher than those of the high-value set. ConclusionsConditional RFS rates of locally advanced ICC after R0 hepatectomy dynamically increased over time, which contributed to reducing survivors' psychological distress and facilitating personalized follow-up schedules. In addition, a person's inflammatory and nutritional status significantly impact the recurrence risk. Oncologists should consider the role of inflammation-nutritional status when making decisions for patients with locally advanced ICC.

Abstract PO1-02-14: Breast cancer in older women - long-term prognosis by age and subtype in a large population-based cohort

Abstract Background Compared to middle-aged women, older women have a higher risk of breast cancer death. Whether the increased risk of breast cancer death is present in all subgroups of older adults with breast cancer, or negligible in the subgroups receiving adequate surgical and oncological treatment, has been debated. We studied long-term prognosis by age and subtype, adjusting for stage, treatment, socioeconomic status and comorbidity, clarifying which subgroups of older patients that are at increased risk of breast cancer death. Methods In a nation-wide cohort of 45,949 women aged ≥60 years at diagnosis of breast cancer 1997-2014, we estimated crude probabilities of breast cancer death and from other causes using competing risk methods. Cox proportional hazards models were used to estimate adjusted hazard ratios (HR) with 95% confidence intervals (CI. Results In this large population-based cohort study of older women with breast cancer, we found an increasing risk of breast cancer death with increasing age at diagnosis, both at 5- and 10-year follow up. The strong association between age and breast cancer death was found for all breast cancer subtypes, being most prominent in Her2-positive breast cancer. The age effect was also significant in Luminal A breast cancer where treatment opportunities are not necessarily affected by age. High-risk tumors were equally common in all age groups above 60 years. In multivariable analysis, stage, subtype, grade and treatment were important explanatory variables to the age-related differences in survival, while comorbidity and socioeconomic status contributed minorly. After adjustments, older age remained an independent risk factor for both overall and breast cancer survival. Older women with triple-negative (60-79 years) or Her2-positive (60-84 years) tumors had a 5-year risk of breast cancer death exceeding the risk of death by other causes. Conclusions Older women had a significantly higher risk of breast cancer death irrespective of breast cancer subtype, also in Luminal A disease where treatment opportunities are not necessarily affected by age. Table 3. Crude probabilities of breast cancer specific death and death by other cause (competing cause of death) at 5, 10 and 15 years. Table 4. Multivariable analysis of prognostic factors affecting 15-year overall (OS) and breast cancer- specific survival (BCSS) Table 5. Multivariable analysis of prognostic factors affecting 15-year overall (OS) and breast cancer- specific survival (BCSS) by age at diagnosis and subtype. Citation Format: Amelia Chiorescu, Anna L V Johansson, Antonis Valachis, Alexios Matikas, Jan Frisell, Lars Holmberg, Irma Fredriksson. Breast cancer in older women - long-term prognosis by age and subtype in a large population-based cohort [abstract]. In: Proceedings of the 2023 San Antonio Breast Cancer Symposium; 2023 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2024;84(9 Suppl):Abstract nr PO1-02-14.

Impact of Radiation Therapy Modalities on Loco-regional Control in Inflammatory Breast Cancer

In inflammatory breast cancer, radiation therapy intensification is considered a standard of care by some teams, although the level of evidence remains low. We sought to analyze the impact of radiation therapy modalities on the risk of loco-regional and distant relapse. This retrospective multicenter study included patients with localized inflammatory breast cancer treated between 2010 and 2017. Standard postmastectomy radiation therapy consisted of daily fractions to a total dose of 50 Gy equivalent without a boost or bolus, while intensified radiation therapy referred to the use of a boost or bolus. The cumulative incidence curves of locoregional and distant recurrence were displayed using the competing risk method. Of the 241 included patients, 165 were treated with standard and 76 with intensified radiation therapy. There was significantly more nodal involvement in the intensified group. With a median follow-up of 40 months postradiation therapy, there was no difference between standard versus intensified radiation therapy regarding the cumulative incidence of locoregional (P = .68) or distant recurrence (P = .29). At 5 years, the risks of locoregional and distant recurrence were 12.1% (95% CI, 7.5; 17.7) and 29.4% (95% CI, 21.8; 37.3) for patients treated with standard radiation therapy and 10.4% (95% CI, 4.4; 19.3) and 21.4% (95% CI, 12.6; 31.9) for those treated with intensified radiation therapy. In multivariate analyses, triple-negative subtype and absence of complete pathologic response were associated with a higher risk of loco-regional recurrence. Radiation therapy intensification had no significant impact on locoregional and distant recurrence. For patients with a non-complete pathologic response (n = 172, 71.7%), no significant differences were observed between the 2 groups for loco-regional (P = .80) and distant (P = .39) recurrence. Severe toxicity rates were similar in both groups. Contrary to other important series, this large retrospective multicentric study did not show a locoregional or distant control benefit of intensified radiation therapy. Pooled prospective studies and meta-analyses of intensified radiation therapy are warranted to endorse this approach.

Tandem High-Dose Chemotherapy Increases the Risk of Secondary Malignant Neoplasm in Pediatric Solid Tumors.

This study aimed to investigate the incidence and risk factors for secondary malignant neoplasms (SMN) in pediatric solid tumors, focusing on the effects of tandem high-dose chemotherapy (HDCT). Patients (aged < 19 years) diagnosed with or treated for pediatric solid tumors between 1994 and 2014 were retrospectively analyzed. The cumulative incidence of SMN was estimated using competing risk methods by considering death as a competing risk. A total of 1,435 patients (413 with brain tumors and 1,022 with extracranial solid tumors) were enrolled. Seventy-one patients developed 74 SMNs, with a 10-year and 20-year cumulative incidence of 2.680±0.002% and 10.193±0.024%, respectively. The types of SMN included carcinoma in 28 (37.8%), sarcoma in 24 (32.4%), and hematologic malignancy in 15 (20.3%) cases. Osteosarcoma and thyroid carcinoma were the most frequently diagnosed tumors. Multivariate analysis showed that radiotherapy (RT) > 2, 340 cGy, and tandem HDCT were significant risk factors for SMN development. The SMN types varied according to the primary tumor type; carcinoma was the most frequent SMN in brain tumors and neuroblastoma, whereas hematologic malignancy and sarcomas developed more frequently in patients with sarcoma and retinoblastoma, respectively. The cumulative incidence of SMN in pediatric patients with solid tumors was considerably high, especially in patients who underwent tandem HDCT or in those who received RT > 2,340 cGy. Therefore, the treatment intensity should be optimized based on individual risk assessment and the long-term follow-up of pediatric cancer survivors.

Less than whole uterus irradiation for patients with locally advanced cervical cancer

IntroductionCurrent consensus guidelines for definitive cervical cancer intensity modulated radiation therapy (IMRT) recommend inclusion of the entire uterus within the clinical target volume, however this is debated. We aimed to evaluate outcomes of patients with cervical cancer who were treated with less than whole uterus irradiation. MethodsWe identified 109 patients with FIGO Stage IB-IVA cervical cancer treated definitively with concurrent chemoradiation, including IMRT and brachytherapy, from 2010 to 2022 at a single institution where the practice was to include the gross cervix tumor with an internal target volume with differences in bladder filing accounted for, plus additional 5 mm planning target volume (PTV) margin. Local, regional, and distant recurrences were analyzed using competing risk methods, and a Wilcoxon rank sum test was performed to assess differences in dose to organs at risk based on the proportion of the uterus included in the PTV, with the median proportion of the uterus included (75 %) used as the cut-point. ResultsThe median follow-up time was 65 months (range 3–352 months). The 2-year cumulative incidence of LR for the entire cohort was 4.2 % (95 % confidence interval [CI] 1.3–9.7). Compared with patients who had ≥ 75 % of the uterus included in the PTV, patients who had < 75 % of the uterus included in the PTV had significantly lower bowel D200cc (p = 0.02). The cumulative incidence of local failure (LR) was not significantly different between the two groups. ConclusionsIncluding less than the whole uterus for definitive cervix cancer IMRT does not seem to compromise local control. Less than whole uterus irradiation could be considered for carefully selected cervix cancer patients to decrease bowel dose and possible treatment-related toxicity.

Clinical outcomes of simultaneous pancreas-kidney transplantation in elderly type II diabetic recipients

The effect of age on outcomes after simultaneous pancreas-kidney transplantation (SPK) among type II diabetes (T2DM) recipients remains inconclusive. This study aimed to analyze the relationship between the age at time of transplantation and mortality, graft loss and metabolic profiles of T2DM SPK recipients. A retrospective cohort consisting of T2MD SPK recipients in a single transplant center was established. The baseline clinical characteristics and outcomes were collected and analyzed based on the age groups divided by 55-year-old. Time-to-event data analysis was performed using Kaplan-Meier method, and competing risk method was adopted to calculate the cumulative incidence of graft loss. A mixed regression model was applied to compare metabolic outcomes including glycated hemoglobin (HbA1c), fasting blood glucose, triglyceride, cholesterol, low-density lipoprotein, and higher estimated glomerular filtration rate (eGFR). 103 T2DM SPK recipients were included, of which 35 were > = 55 years old and 68 were < 55 years old. Baseline characteristics were comparable between age groups. The results indicated that comparable 5-year survival outcomes between groups with functioning grafts perioperatively. Additionally, no relationship of age with graft loss, complications and metabolic outcomes was detected.

A review on statistical and machine learning competing risks methods.

When modeling competing risks (CR) survival data, several techniques have been proposed in both the statistical and machine learning literature. State-of-the-art methods have extended classical approaches with more flexible assumptions that can improve predictive performance, allow high-dimensional data and missing values, among others. Despite this, modern approaches have not been widely employed in applied settings. This article aims to aid the uptake of such methods by providing a condensed compendium of CR survival methods with a unified notation and interpretation across approaches. We highlight available software and, when possible, demonstrate their usage via reproducible R vignettes. Moreover, we discuss two major concerns that can affect benchmark studies in this context: the choice of performance metrics andreproducibility.

Natural history of cancer-associated splanchnic vein thrombosis

BackgroundThere is uncertainty in the management of cancer-associated isolated splanchnic vein thrombosis (SpVT). ObjectivesTo describe the natural history of SpVT by cancer type and thrombus composition and to review anticoagulation (AC) practices and associated rates of usual-site venous thromboembolism (VTE), major and clinically relevant nonmajor bleeding (MB/CRNMB), recanalization/progression, and mortality. MethodsWe performed a retrospective cohort study in patients with SpVT at 2 cancer care centers in Houston, Texas. We estimated the incidence of usual-site VTE and MB/CRNMB at 6 months using competing risk methods and examined venous patency in a subset of patients with repeat imaging. We assessed associations with mortality using Cox regression. ResultsAmong 15 342 patients with an incident cancer diagnosis from 2011 to 2020, we identified 298 with isolated SpVT. Patients with hepatocellular carcinoma (HCC) and SpVT (n = 146) had the highest disease prevalence (20%), lowest rate of AC treatment (2%), and similar rate of usual-site VTE (4.2%) vs those without SpVT (5.2%) at 6 months, though tumor thrombus vs bland was associated with worse overall survival. In patients with non-HCC bland SpVT (n = 114), AC (n = 37) was more common in those with non-upper gastrointestinal cancers and fewer comorbidities. AC was associated with more recanalization (44% vs 15%, P = .041) but no differences in usual-site VTE, MB/CRNMB, or mortality at 6 months. ConclusionCancer-associated isolated SpVT is a common but heterogeneous thrombotic disease that is treated differently from usual-site VTE. Tumor thrombus is a negative prognostic factor. Initiation of AC in bland thrombi requires judicious consideration of thrombotic and bleeding risk.

The cumulative live birth rates of 18 593 women with progestin-primed ovarian stimulation-related protocols and frozen-thawed transfer cycles.

What are the odds of achieving pregnancy when adopting progestin-primed ovarian stimulation (PPOS)-related protocols combined with repetitive frozen-thawed transfer (FET) cycles in patients with different clinical characteristics? The cumulative live birth rates (CLBRs) of women undergoing different PPOS-related protocols can be significantly and consistently enhanced within six FET cycles when the female age is <40 years (or even <45 years) and when >5 oocytes are retrieved, regardless of antral follicle count (AFC). There have been numerous studies on the live birth rate of the first FET cycle in patients with PPOS-related protocols. These studies have focused mainly on comparing pregnancy outcomes with those of other stimulation protocols. However, owing to the unique features of the PPOS-related strategy, such as its flexible timing of oocyte retrieval and repeated transfer of frozen embryos, studies using the CLBR as an overall indicator of success and investigating which types of patients would benefit from this protocol are lacking. This retrospective cohort study included 18 593 women who underwent PPOS-related protocols (dydrogesterone + hMG, medroxyprogesterone acetate + hMG, micronized progesterone + hMG treatment, and luteal-phase ovarian stimulation protocol) from 1 March 2011 to 31 September 2022 in our centre. The population was categorized by female age, number of oocytes retrieved, and AFC in the analysis of CLBR within six FET cycles. The age groups (Groups 1-5, respectively) were <30, 30-34, 35-39, 40-44, and ≥45 years. The number of oocytes retrieved was grouped as 1-5, 6-10, 11-15, 16-20, and >20. AFC was grouped as <5, 5-10, 11-15, and >15. The Kaplan-Meier analysis (optimistic method), which hypothesized that patients who did not continue treatment had the same chance of achieving a live birth as those who continued, and the competing risk method (conservative method) which hypothesized they had no chance of achieving a live birth, were applied. In further analyses, the Cox model and Fine-Gray model were adopted: the former corresponds to the optimistic scenario, and the latter corresponds to the pessimistic scenario. CLBR had a declining trend with female age over six FET cycles (Groups 1-5, respectively: optimistic: 96.9%, 96.6%, 91.4%, 67.3%, and 11.7%; conservative: 87.3%, 85.0%, 74.0%, 41.3%, and 7.5%), requiring more FET cycles to achieve a success rate of at least 50% (Groups 1-5, respectively: optimistic: 2, 2, 2, 4, and >6 cycles; conservative: 2, 2, 2, >,6 and >6 cycles). CLBR showed an increasing trend with the number of oocytes retrieved (Groups 1-5, respectively: optimistic: 93.8%, 94.3%, 95.8%, 96.0%, and 95.6%; conservative: 66.2%, 78.3%, 85.6%, 88.9%, and 91.0%). All groups needed the same number of FET cycles to achieve a success rate of at least 50% (Groups 1-5, respectively: optimistic: 2, 2, 2, 2, and 2 cycles; conservative: 2, 2, 2, 2, and 2 cycles). Furthermore, the CLBR within six FET cycles had an increasing trend with AFC number (Groups 1-4, respectively: optimistic: 89.2%, 94.8%, 95.9%, and 96.3%; conservative: 67.4%, 78.2%, 83.9%, and 88.1%), with all four groups achieving a success rate of at least 50% by the second FET cycle. The current research is limited by its retrospective design and single-centre nature, which may restrict the generalizability of our findings. This work describes two models (the Kaplan-Meier analysis and the competing risk method) to evaluate the clinical outcome of patients using PPOS-related protocols, which are especially useful for patients of advanced age or those with diminished ovarian reserve. Our findings encourage patients below 45 years old, especially younger than 40 years, and patients with lower AFCs and fewer retrieved oocytes to try this new protocol. Moreover, this study demonstrates the degree of improvement in the CLBR within six FET cycles for patients with different clinical characteristics, providing a valuable point of reference to determine whether to continue ART after a transfer failure. The study was supported by grants from the National Natural Science Foundation of China (82071603 to L.W., 82001502 to Y.L.). There are no conflicts of interest to declare. N/A.

Prospective Clinical and Biomarker Validation of the ASTCT Consensus Definition for Transplant-Associated Thrombotic Microangiopathy (TA-TMA)

Introduction: Transplantation-associated thrombotic microangiography (TA-TMA) is a disorder that causes severe complications post-hematopoietic cell transplantation (HCT). Diagnosing TA-TMA is challenging due to non-standardized criteria. The American Society for Transplantation and Cellular Therapy (ASTCT) has proposed new consensus diagnostic criteria for standard risk (ASTCT-SR) and high risk (ASTCT-HR) but their correlation with older criteria based on microangiopathic hemolytic anemia (MAHA) or diagnostic complement biomarkers used in previous TA-TMA studies, such as levels of the terminal complement activation product sC5b-9 and the alternative complement pathway activation product Factor Ba, is unknown. Using data and samples from an ongoing prospective pediatric cohort study, we compared the two definitions and their associations with biomarkers. Methods: Patients aged 0 to 18 years who underwent first allogeneic HCT 2021-2023 at Texas Children's Hospital were included in the clinical and biomarker study. Those who did not have at least 100-day post-HCT event-free follow-up were excluded. Two different definitions for TA-TMA were applied. Clinical TMA (cTMA) was diagnosed if a patient 1) met 4/4 concurrent criteria of MAHA (schistocytosis, thrombocytopenia, anemia, and elevated lactate dehydrogenase (LDH)) at least twice in 14 days and 2) clinically adjudicated to exclude known mimics or alternative causes of TMA. In contrast, ASTCT-HR TMA was diagnosed if a patient 1) met 4/7 non-concurrent criteria (schistocytosis, thrombocytopenia, anemia, elevated LDH, elevated blood pressure, elevated urinary protein excretion, and elevated sC5b9) at least twice in 14 days and 2) had evidence of LDH &amp;gt;2x normal, proteinuria ≥1 mg/mg, elevated sC5b9, acute graft-versus-host disease, active bacterial/viral infection, or end-organ damage. Cumulative incidence of TA-TMA was assessed by the competing risk method. Overall survival was compared by the Kaplan Meier estimator and log-rank test. Biomarker levels for samples collected at day 15 (+/- 14 days) were compared using the t-test. Results: A total of 32 patients met the inclusion criteria. The median age was 5.7 years (range 1.1-18.8) and most were White (68.8%) and Hispanic (56.2%). Malignant disease was the indication for HCT in 50% of patients. While the classic MAHA-based cTMA definition had an incidence of 12.7%, the newly proposed ASTCT-HR definition doubled it to 28.5% by day-100 (Figure 1). In contrast to patients with concordant TA-TMA diagnosis (+/+) who had significantly worse post-transplant survival (p=0.0047) compared with those with concordant non-TMA (-/-), those reclassified as TA-TMA by the new definition only (-/+) had an unexpectedly good prognosis (100% survival at day 100) despite the lack of TMA-directed therapy. Longitudinal biomarker analysis was conducted on 29 patients. The highest average plasma levels of sC5b-9 and Ba during the first 100 days were observed in patients meeting both cTMA and ASTCT-HR criteria (+/+). When focused on early biomarkers measured around day 15, sC5b-9 (P=0.041) and Ba (P=0.004) were significantly elevated in the doubly positive TMA group (+/+) vs. non-TMA group (-/-) (Figure 2). We did not observe significant differences for ASTCT-HR singly positive (-/+) vs. non-TMA group (-/-). Conclusions: In the current prospective clinical and biomarker study, the recently proposed ASTCT-HR definition was more sensitive but less prognostic than the traditional cTMA definition for TA-TMA. Patients who solely met the ASTCT-HR diagnostic criteria but not the cTMA criteria had benign complement biomarker profiles and excellent survival prognosis without TMA-directed therapy. While clinicians need to be vigilant for this infrequent yet potentially severe complications, it is crucial that diagnostic criteria accurately identify affected patients given the expense and potential complications of treatment. We recommend ongoing validations of the new consensus diagnostic criteria from other prospective cohort studies with larger sample sizes and extended follow-up.

Investigation of Age-Specific Differences in Survival and Cause of Death Among Diffuse Large B-Cell Lymphoma Patients with and without HIV in the Modern Era of Antiretroviral Therapy: A Population-Based Analysis from 2010 to 2017

Introduction Despite improved prognosis with combined antiretroviral therapy (cART) among patients with human immunodeficiency virus (PWHIV) diagnosed with diffuse large B-cell lymphoma (DLBCL), population-based studies suggest inferior survival rates compared to patients without HIV. Uncertainty remains about the persistence of this survival difference across age groups. Examining outcomes across ages may help identify higher-risk groups in the current era of cART, providing insights for designing and conducting age-specific interventions. Methods A retrospective population-based cohort was conducted using US population-based cancer registry data among patients aged 15-79 years with newly diagnosed DLBCL from 2010 to 2017, with follow-up through 2019. The 2010-2017 period reflects the availability of integrase inhibitors and the introduction of distinct lymphoma subtypes to the World Health Organization classification system in cancer registries. HIV status was reported at lymphoma diagnosis. Patients were classified into three age groups: adolescents and young adults (AYAs, 15-39 years), adults (40-59 years), and older adults (60-79 years). Overall survival (OS) was defined from diagnosis to death of any cause. Probabilities were estimated using the Kaplan-Meier method and compared with the log-rank test. Multivariable Cox regression models were fitted to assess all-cause mortality, adjusting for cancer stage, race/ethnicity, sex, year of diagnosis, and age. We estimated the cumulative incidence of infections and HIV-related deaths using the competing risk method. Results Patient characteristics: A total of 26,905 DLBCL patients were identified (1,030 PWHIV and 25,758 patients without HIV). Compared to patients without HIV, PWHIV were younger (median age at diagnosis 48 years vs. 62 years), predominantly male (85% vs. 56%), more likely to be non-Hispanic Black (38% vs. 8%) and Hispanic (24% vs. 18%), and more likely to have advanced-stage disease at diagnosis (71% vs. 56%) (Table 1). OS in DLBCL PWHIV vs. patients without HIV:Within each age group, OS was significantly lower among PWHIV with DLBCL compared to patients without HIV(Table 1 and Figure 1). With a median follow-up of 67 months, 5-year OS was particularly low among AYAs (53% vs. 87%, p&amp;lt;0.0001) and adults (48% vs. 74%, p&amp;lt;0.0001). Similarly, older adults experienced lower 5-year OS rates (45% vs. 57%, p&amp;lt;0.0001). Multivariable Cox models suggest that PWHIV has an increased risk of all-cause mortality within each age group (Figure 1). OS in different age groups by HIV status: For PWHIV, age was not a significant factor for increased all-cause mortality, with 5-year OS rates ranging from 45% in adults to 53% in AYAs (p=0.170). Conversely, in DLBCL patients without HIV, AYAs had superior 5-year OS rates (87%, p&amp;lt;0.001) compared to adults (74%) and older adults (57%). Cause of death by HIV status: Among both PWHIV and without HIV, lymphoma/leukemia-related mortality was the most common cause of death reported in death certificates (47% and 73% respectively). However, the 5-year cumulative incidence of infection mortality was higher among PWHIV (range 3-4% across age groups) compared to patients without HIV (1% across age groups). For PWHIV, other HIV-related mortality had a 5-year cumulative incidence ranging from 9% to 14% across age groups. Results stratified by cancer stage (early- or advanced-stage disease) were consistent with those presented above. Conclusion This study highlights worse outcomes among PWHIV diagnosed with DLBCL compared to patients without HIV within each age group at the population level in the modern era of cART, notably among AYAs. Unlike the findings in DLBCL patients without HIV, age was not significantly associated with mortality among PWHIV with DLBCL. Future studies should explore age-specific demographic, biological, or healthcare system factors mediating these outcome disparities. The considerable HIV-related mortality suggests refining HIV surveillance approaches among PWHIV with DLBCL in prospective studies or clinical trials. Identifying the optimal integration of chemotherapy and cART may improve outcomes in this vulnerable population.

Risk of Transformation By Frontline Management in Follicular Lymphoma and Marginal Zone Lymphoma: A US Population-Based Analysis

INTRODUCTION Follicular lymphoma (FL) and marginal zone lymphoma (MZL) are commonly characterized by an indolent nature and favorable overall survival (OS). However, patients (pts) may experience high-grade transformation (HGT) into DLBCL, with an estimated annual risk of 1%-3% and cumulative incidence of HGT are 8% for FL and 10% for MZL at 10 and 12 years, respectively. The impact of treatment and surveillance strategies on the incidence of HGT remains uncertain, with some studies indicating reduced risk in treated pts without significant OS difference post-HGT. This study uses population-based data to assess the cumulative incidence and risk factors associated with HGT, post-HGT OS and lymphoma-specific survival (LSS) across lymphoma subtypes, with either treatment or surveillance as initial strategies. METHODS We extracted data for pts diagnosed with grades 1-2 FL, splenic MZL (SMZL), nodal MZL (NMZL), and extranodal MZL (MALT) from SEER database (period 2000-2015 with follow-up through 2018). We only included pts with histologic confirmed lymphoma as first malignancy, complete survival and treatment data, and a minimum of 3 months from FL and MZL diagnosis to HGT. Competing risk methods were used to estimate HGT cumulative incidence rates, with time to first event (HGT or death) defined from lymphoma diagnosis to date of first event; event-free pts were censored at date of last contact. Gray's test and Fine-Gray (FG) proportional subdistribution hazard modelling was employed to identify variables associated with risk of HGT. Post-HGT OS and LSS analyses utilized Cox and FG regression models, respectively. RESULTS 26827 pts met inclusion criteria. FL comprised 14229 pts (53%) while MZL types were SMZL 1182 (4.4%), NMZL 3916 (14.6 %) and MALT 7500 (28%). Treatment was given at diagnoses in 60.2%, 33.7%, 53.8% and 49.8% of pts with FL, SMZL, NMZL, MALT, respectively. 864 (3.2%) pts developed HGT to DLBCL, 553 (3.9%) in FL and 311 (2.5%) in MZL pts. After a median follow-up of 84 months (range 1-227), the 10-year cumulative incidence of HGT were for FL 4.24% (95%CI 3.88%-4.62%), SMZL 6.46% (95%CI 4.98%-8.20%), NMZL 3.28% (95%CI 2.70%-3.95%) and MALT 1.73% (95%CI 1.42%-2.09%). In FL, the 10-year cumulative incidence of HGT was 3.83% (95%CI 3.40-4.31%) in treated pts and 4.88% (95%CI 4.26%-5.54%) for those under surveillance (SHR=0.80, 95%CI 0.68-0.94; p=0.009) (Figure 1). This finding was confirmed in a multivariable analysis (p=0.001) that conversely identified female gender (p&amp;lt;0.001) and advanced stage at diagnosis (p&amp;lt;0.001) with higher risk of HGT (Figure 1). In MZL, the 10-year cumulative incidence rate of HGT in treated pts versus those managed with surveillance depicted by subtypes were as follows: SMZL 6.93% (95%CI 4.40%-10.2%) vs 6.29% (95%CI 4.53%-8.43%); NMZL 3.71% (95%CI 2.87%-4.70%) vs 2.78% (95%CI 2.03%-3.71%); MALT 2.01% (95%CI 1.54%-2.57%) vs 1.45% (95%CI 1.07%-1.93%). No significant difference between treatment versus surveillance were detected in univariable and multivariable analyses for SMZL (p=0.72 and 0.54) and NMZL (p=0.07 and 0.07), whereas surveillance pts had lower risk of HGT in MALT (p=0.02 and 0.01) (Figure 2). There were 429 (49.7%) deaths, including 337 (39%) lymphoma-specific deaths recorded in 864 pts with HGT in the entire cohort; and the median OS after HGT was 3.67 years (95%CI 2.5-6.25). Frontline therapy for FL was associated with higher risk of overall and lymphoma-specific deaths post-HGT (HR=1.32, 95%CI 1.03-1.69, p=0.02; and SHR=1.33, 95%CI 1.02-1.75, p=0.03); however, there was no significant difference in OS between frontline therapy and surveillance approach in SMZL (p=0.11), NMZL (p=0.73) and MALT (p=0.70). Interestingly, in SMZL worse LSS post-HGT was observed in those managed with frontline therapy (SHR=2.68, 95%CI 1.32-5.45, p=0.006). CONCLUSIONS Frontline treatment was associated to reduced HGT risk in FL, with advanced stage and female gender identified as risk factors for HGT. Contrarily, for SMZL, NMZL, and MALT, initial treatment neither diminished HGT risk nor improved overall survival post-HGT. Interestingly, in MALT initial treatment exhibits a higher HGT risk, likely due to selection bias, even though HGT cumulative incidence in MALT was lower compared to other subtypes. SMZL displayed a particularly high cumulative incidence of HGT and poorer LSS.

Competing risks modeling of length of hospital stay enhances risk-stratification of patient care: application to under-five children hospitalized in Malawi.

Length of hospital stay (LOS), defined as the time from inpatient admission to discharge, death, referral, or abscondment, is one of the key indicators of quality in patient care. Reduced LOS lowers health care expenditure and minimizes the chance of in-hospital acquired infections. Conventional methods for estimating LOS such as the Kaplan-Meier survival curve and the Cox proportional hazards regression for time to discharge cannot account for competing risks such as death, referral, and abscondment. This study applied competing risk methods to investigate factors important for risk-stratifying patients based on LOS in order to enhance patient care. This study analyzed data from ongoing safety surveillance of the malaria vaccine implementation program in Malawi's four district hospitals of Balaka, Machinga, Mchinji, and Ntchisi. Children aged 1-59 months who were hospitalized (spending at least one night in hospital) with a medical illness were consecutively enrolled between 1 November 2019 and 31 July 2021. Sub-distribution-hazard (SDH) ratios for the cumulative incidence of discharge were estimated using the Fine-Gray competing risk model. Among the 15,463 children hospitalized, 8,607 (55.7%) were male and 6,856 (44.3%) were female. The median age was 22 months [interquartile range (IQR): 12-33 months]. The cumulative incidence of discharge was 40% lower among HIV-positive children compared to HIV-negative (sub-distribution-hazard ratio [SDHR]: 0.60; [95% CI: 0.46-0.76]; P < 0.001); lower among children with severe and cerebral malaria [SDHR: 0.94; (95% CI: 0.86-0.97); P = 0.04], sepsis or septicemia [SDHR: 0.90; (95% CI: 0.82-0.98); P = 0.027], severe anemia related to malaria [SDHR: 0.54; (95% CI: 0.48-0.61); P < 0.001], and meningitis [SDHR: 0.18; (95% CI: 0.09-0.37); P < 0.001] when compared to non-severe malaria; and also 39% lower among malnourished children compared to those that were well-nourished [SDHR: 0.61; (95% CI: 0.55-0.68); P < 0.001]. This study applied the Fine-Gray competing risk approach to more accurately model LOS as the time to discharge when there were significant rates of in-hospital mortality, referrals, and abscondment. Patient care can be enhanced by risk-stratifying by LOS based on children's age, HIV status, diagnosis, and nutritional status.

Blinatumomab Therapy Is Associated with Favorable Outcomes after Allogeneic Hematopoietic Cell Transplantation in Pediatric Patients with B Cell Acute Lymphoblastic Leukemia

Blinatumomab, a bispecific T cell engager that binds CD19 in leukemic cells and CD3 in cytotoxic T cells and leads to leukemic blast lysis, is often used in pediatric patients with relapsed/refractory (R/R) B cell acute lymphoblastic leukemia (B-ALL) prior to allogeneic hematopoietic cell transplantation (allo-HCT). Concerns about the potential risk of blinatumomab-related immune-mediated toxicities after allo-HCT have not been adequately addressed. These include graft-versus-host disease (GVHD), delayed engraftment, and graft failure or rejection. Pediatric-specific data reporting post-HCT outcomes of patients treated with blinatumomab are scarce and limited to small cohorts. We sought to investigate the clinical outcomes of pediatric patients with R/R B-ALL who received blinatumomab therapy pre-HCT, focusing on overall survival (OS), leukemia-free survival (LFS), cumulative incidence of relapse (CIR), and nonrelapse mortality (NRM), as well as the incidence of immune-mediated post-HCT complications including GVHD, delayed neutrophil or platelet engraftment, graft failure, and graft rejection. We also investigated blinatumomab's effects on B cell reconstitution based on achievement of i.v. immunoglobulin (IVIG) independence post-HCT. This single-center, retrospective study included patients with B-ALL receiving blinatumomab therapy before undergoing allo-HCT, with transplantation performed between 2016 and 2021 at our institution. Patients receiving blinatumomab for relapse after allo-HCT were excluded. Patients receiving chemotherapy alone before allo-HCT during the same period composed the control group. Seventy-two patients were included, 31 of whom received blinatumomab before allo-HCT. Survival estimates were obtained using the Kaplan-Meier method, and the log-rank test was used to analyze differences between groups. Categorical variables were compared between groups using the chi-square test or Fisher exact test, and continuous variables were compared using the Wilcoxon rank-sum test. Cumulative incidences were estimated using the competing risks method, and Gray's test was used to analyze differences between groups. A Cox proportional hazards regression model was used for univariate and multivariable analyses for OS. Landmark analysis was performed at the set time points of 30 days and 100 days post-allo-HCT. Most patients in the study cohort had high-risk relapsed B-ALL. Blinatumomab therapy induced minimal residual disease (MRD)-negative remissions in all patients, whereas 5 patients (12.2%) receiving chemotherapy alone had persistent MRD pre-allo-HCT. Time from the start of therapy to the date of allo-HCT was shorter for patients who received blinatumomab compared with those who received chemotherapy (P < .0001). Blinatumomab therapy was associated with greater LFS compared to chemotherapy alone (P = .049), but when limited to 1 year, LFS was not significantly different from control (P = .066). There appeared to be higher OS, lower CIR, and lower NRM in patients receiving blinatumomab compared to the control group; however, the differences were not significant. None of the variables assessed in multivariable analysis was associated with differences in OS. When compared to the controls, blinatumomab therapy did not result in a higher incidence of acute or chronic GVHD, delayed neutrophil or platelet engraftment, or graft failure or rejection. The time to IVIG infusion independence post-allo-HCT was similar in the 2 groups. This study supports the use of blinatumomab salvage therapy for R/R B-ALL before allo-HCT given its efficacy in inducing MRD-negative remissions and optimizing LFS, as well as its lack of association with an increased incidence of post-allo-HCT adverse immune-mediated toxicities. Larger, prospective studies are needed to confirm these findings and to investigate blinatumomab's effects in long-term post-allo-HCT events.