Decision Support Tool Identifies Ulcerative Colitis Patients Most Likely to Achieve Remission With Vedolizumab vs Adalimumab

Background & Aims We have previously validated a clinical decision support tool (CDST) (vedolizumab CDST [VDZ-CDST]) for clinical and endoscopic remission with VDZ in ulcerative colitis (UC). We aim to expand the validation for predicting histoendoscopic mucosal improvement (HEMI) with VDZ vs adalimumab (ADA). Methods In a post hoc analysis of a clinical trial for VDZ vs ADA in moderate to severe UC (VARSITY trial; NCT02497469), comparative accuracy was evaluated for the VDZ-CDST among an external validation cohort of VDZ- and ADA-treated patients for week 52 HEMI (Mayo endoscopic subscore 0-1 and Geboes score <3.2). Comparative effectiveness of VDZ and ADA was assessed after stratifying the cohort by baseline probability of response to VDZ using the VDZ-CDST. Results A total of 419 patients were included. The majority of patients enrolled in the VARSITY trial had a high (61%) or intermediate (29%) baseline predicted probability of response to VDZ. The baseline VDZ-CDST score was significantly more likely to predict week 52 HEMI for VDZ (area under the curve , 0.712; 95% confidence interval, 0.636-0.787) relative to ADA-treated patients (area under the curve, 0.538; 95% confidence interval, 0.377-0.700; P < .001 for AUC comparison). A significant (P < .001) association was observed between the VDZ-CDST and measured VDZ drug exposure over 52 weeks. Superiority of VDZ to ADA was only observed in patients with a high baseline predicted probability of response to VDZ. Conclusions Superiority of VDZ to ADA is dependent on baseline probability of response, and a VDZ-CDST is capable of identifying UC patients most appropriate for VDZ vs ADA.

396. Disparities in SARS-CoV-2 Antibody Prevalence: Findings from a Citywide Serosurvey in Holyoke, Massachusetts, November 2020–January 2021

Background Seroprevalence studies are important tools to estimate the prevalence of prior or recent SARS-CoV-2 infections. This information is critical for identifying hotspots and high-risk groups and informing public health responses to the COVID-19 pandemic. We conducted a city-level seroprevalence study in Holyoke, Massachusetts to estimate the seroprevalence of SARS-CoV-2 antibodies and risk factors for seropositivity. Methods We invited inhabitants of 2,000 randomly sampled addresses to participate between November 5 and December 31, 2020. Participants completed questionnaires measuring sociodemographic and health characteristics, and COVID-19 exposure history, and provided dried blood spots for measurement of SARS-CoV-2 IgG and IgM antibodies. To calculate total and group seroprevalence estimates, inverse probability of response weights were constructed based on age, gender, race/ethnicity and census tract to ensure estimates represented the city’s population. Results We enrolled 280 households including 472 individuals. 328 underwent antibody testing. The citywide weighted seroprevalence of SARS-CoV-2 IgG or IgM was 13.9% (95%CI 7.8 - 21.8) compared to 9.8% based on publicly reported case counts. Seroprevalence was 16.8% (95%CI 5.7 – 28.0) among individuals identifying as Hispanic compared to 8.9% (95%CI 3.0 - 14.7) among those identifying as White. Seroprevalence was 20.7% (95%CI 2.2 – 39.2) for ages 0-19; 13.8% (95%CI 5.6 – 22) for ages 20 – 44; 9.6% (95%CI 0 – 20.5) for ages 45 – 59; 4.8% (95%CI 0 – 10.2) for ages 60 – 84; and 42.9% (95%CI 0 – 100) for ages >85. Table 1. Seroprevalence by antibody positivity profile Table 2. Unweighted and weighted seroprevalence by sociodemographic characteristics Figure 1. Seroprevalence by Medical, Symptom, Testing and Exposure History. Conclusion The measured SARS-CoV-2 seroprevalence in Holyoke was only 13.9% during the second surge of SARS-CoV-2 in this region, far from accepted thresholds for “herd immunity” and highlighting the need for expanding vaccination. Individuals identifying as Hispanic were at high risk of prior infection. Subsequent community-level serosurveys are necessary to guide local responses to the SARS-CoV-2 pandemic. Disclosures All Authors: No reported disclosures

Gender Differences or Gender Bias?

Sadism, defined by the infliction of pain and suffering on others for pleasure or subjugation, has recently garnered substantial attention in the psychological research literature. The Assessment of Sadistic Personality (ASP) was developed to measure levels of everyday sadism and has been shown to possess excellent reliability and validity using classical test theory methods. However, it is not known how well ASP items discriminate between respondents of different trait levels, or which Likert categories are endorsed by persons of various trait levels. Additionally, individual items should be evaluated to ensure that men and women of similar levels of sadism have an equal probability of response endorsement. The purpose of this research was to apply item response theory (IRT) and differential item functioning (DIF) to investigate item properties of the ASP across its three translations: English, Polish, and Italian. Overall, the results of the IRT analysis showed that with the exception of Item 9, the ASP demonstrated sound item properties. The DIF rate analyses identified two items from each questionnaire that were of practical significance across gender. Implications of these results are discussed.

Exploring Heterogeneity in Histology-Independent Technologies and the Implications for Cost-Effectiveness

Background The National Institute for Health and Care Excellence and a number of international health technology assessment agencies have recently undertaken appraisals of histology-independent technologies (HITs). A strong and untested assumption inherent in the submissions included identical clinical response across all tumour histologies, including new histologies unrepresented in the trial. Challenging this assumption and exploring the potential for heterogeneity has the potential to impact upon cost-effectiveness. Method Using published response data for a HIT, a Bayesian hierarchical model (BHM) was used to identify heterogeneity in response and to estimate the probability of response for each histology included in single-arm studies, which informed the submission for the HIT, larotrectinib. The probability of response for a new histology was estimated. Results were inputted into a simplified response-based economic model using hypothetical parameters. Histology-independent and histology-specific incremental cost-effectiveness ratios accounting for heterogeneity were generated. Results The results of the BHM show considerable heterogeneity in response rates across histologies. The predicted probability of response estimated by the BHM is 60.9% (95% credible interval 16.0; 91.8%), lower than the naively pooled probability of 74.5%. A mean response probability of 56.9% (0.2; 99.9%) is predicted for an unrepresented histology. Based on the economic analysis, the probability of the hypothetical HIT being cost-effective under the assumption of identical response is 78%. Allowing for heterogeneity, the probability of various approval decisions being cost-effective ranges from 93% to 11%. Conclusions Central to the challenge of reimbursement of HITs is the potential for heterogeneity. This study illustrates how heterogeneity in clinical effectiveness can result in highly variable and uncertain estimates of cost-effectiveness. This analysis can help improve understanding of the consequences of histology-independent versus histology-specific decisions.

Bayesian biomarker-driven outcome-adaptive randomization with an imperfect biomarker assay

Objective: We investigate the impact of biomarker assay’s accuracy on the operating characteristics of a Bayesian biomarker-driven outcome-adaptive randomization design. Methods: In a simulation study, we assume a trial with two treatments, two biomarker-based strata, and a binary clinical outcome (response). P bt denotes the probability of response for treatment t ( t = 0 or 1) in biomarker stratum ( b = 0 or 1). Four different scenarios in terms of true underlying response probabilities are considered: a null ( P00 = P01 = 0.25, P10 = P11= 0.25) and consistent ( P00 = P10 = 0.25, P01 = 0.5) treatment effect scenario, as well as a quantitative ( P00 = P01 = P10 = 0.25, P11 = 0.5) and a qualitative ( P00 = P11 = 0.5, P01 = P10 = 0.25) stratum-treatment interaction. For each scenario, we compare the case of a perfect with the case of an imperfect biomarker assay with sensitivity and specificity of 0.8 and 0.7, respectively. In addition, biomarker-positive prevalence values P( B = 1) = 0.2 and 0.5 are investigated. Results: Results show that the use of an imperfect assay affects the operational characteristics of the Bayesian biomarker-based outcome-adaptive randomization design. In particular, the misclassification causes a substantial reduction in power accompanied by a considerable increase in the type-I error probability. The magnitude of these effects depends on the sensitivity and specificity of the assay, as well as on the distribution of the biomarker in the patient population. Conclusion: With an imperfect biomarker assay, the decision to apply a biomarker-based outcome-adaptive randomization design may require careful reflection.

CPX-351 Exposure-Response Analyses for Efficacy and Safety in Pediatric Patients with Relapsed or Refractory Acute Myeloid Leukemia (AML)

CPX-351 (Vyxeos®; daunorubicin and cytarabine liposome for injection), a dual-drug liposomal encapsulation of daunorubicin and cytarabine at a synergistic 1:5 molar ratio, is approved by the US FDA and EMA for the treatment of adults with newly diagnosed therapy-related AML or AML with myelodysplasia-related changes. Separate studies at Cincinnati Children's Hospital (CCH) and by the Children's Oncology Group (COG) evaluated the efficacy and safety of CPX-351 in pediatric patients (pts). CPX-351 was administered by IV infusion on Days 1, 3, and 5 for the first induction only. The CCH study included 2 dose levels, 100 and 134 units/m2, with 100 units/m2 identified as the recommended phase 2 dose (RP2D); pts with multiple relapsed or refractory AML were included in the exposure-response (E-R) analyses. The COG study established an RP2D of 135 units/m2; the majority of pts in the E-R analyses had first-relapsed AML. E-R analyses were conducted to evaluate the relationship between plasma exposures to CPX-351 and efficacy or safety endpoints. Plasma PK parameters of cytarabine and daunorubicin, including area under the curve up to 48 h post dose on Day 5 Cycle 1 (AUC48), maximum concentration on Day 5 Cycle 1 (Cmax), and concentration at 48 h post dose on Day 5 Cycle 1 (C48), were derived from a previously developed population PK model for pediatric and adult pts. The efficacy endpoints were complete remission (CR), CR+CRp (partial platelet count recovery), or CR+CRp+CRi (incomplete hematologic recovery). Age, sex, bone marrow blast, white blood cell, and platelet counts at baseline were included in the covariate analysis. The safety endpoints included grade ≥3 TEAEs. From the studies, both pediatric and adult pts were included in the E-R analyses for efficacy. Overall, the E-R efficacy population included 43 (53%) pediatric pts (1-17 y) and 38 (47%) adults (≥18 y); the majority were male (57%) and of white origin (63%). Of the 81 pts included in the analysis, 28 (31%), 16 (25%), and 1 (2%) pt achieved CR, CRp, and CRi, respectively; 39 (48%) pts had progressive disease or no response. Various models were used to describe the relationship between probability of response (CR, CRi, or CRp) and exposure parameters (AUC48, Cmax, and C48). C48 of cytarabine was associated with the best statistical goodness-of-fit in the logistic regression model. The effect of cytarabine C48 on the probability of response was statistically significant (P = 0.0144; odds ratio = 1.057). These results suggest the odds of response increased by ~6% for each 1-unit increment of cytarabine C48 (ie, 1 μg/mL). The estimated probability of response was separated into exposure quartiles: 33% for Q1, 49% for Q2, 59% for Q3, and 83% for Q4. The cytarabine C48 was mainly observed in Q1 and Q2 for the 100 units/m2 dose and in Q3 and Q4 for the 134-135 units/m2 doses. In a covariate analysis, none of the covariates tested explained the variability in response. AUC48 of cytarabine and daunorubicin and C48 of daunorubicin were also significantly correlated with probability of response to CPX-351. This was expected, as the PK of daunorubicin and cytarabine are highly correlated when administered as CPX-351. Logistic models were also developed to describe the relationship between probability of response and exposure parameters. Similarly, cytarabine C48 was a significant predictor of the probability of CR or CRi. However, no logistic models were found to describe the relationship of probability of CR+CRp (or CR only) and exposure parameters. Although pt populations of the CCH and COG studies were different, a similar correlation was shown for exposures and efficacy in the CCH study, where higher exposures were associated with higher response rates. Grade ≥3 TEAEs observed in pediatric pts were separated into exposure quartiles for Cmax of cytarabine or daunorubicin. The probabilities of grade ≥3 TEAEs for quartiles of cytarabine Cmax were 77% for Q1, 85% for Q2, 85% for Q3, and 71% for Q4. Similar results were observed with daunorubicin. In these post hoc E-R analyses, higher CPX-351 doses were associated with higher plasma exposures, which led to a higher probability of response (CR+CRp+CRi) in pediatric pts with relapsed or refractory AML. There was no correlation of plasma exposures to grade ≥3 TEAEs. These analyses should be interpreted with caution due to the small sample size; however, they support use of the higher dose of 135 units/m2. Disclosures Wang: Jazz Pharmaceuticals: Current Employment, Current equity holder in publicly-traded company. Absalon:Jazz Pharmaceuticals: Research Funding. Faderl:Jazz Pharmaceuticals: Current Employment, Current equity holder in publicly-traded company. Cooper:Celgene: Other: Spouse was an employee of Celgene (through August 2019). OffLabel Disclosure: Yes, in these studies, CPX-351 was evaluated in pediatric AML

Predictive Model of Response to Blinatumomab Therapy in Children and Adults with Relapsed/Refractory B-ALL

Background .Blinatumomab is a monoclonal BITe antibody recently introduced for the treatment of relapsed and refractory (r\r) B-ALL. Response rate reaches 65-85% in different subgroups of patients but currently there is no reliable methods to predict response. Using results of the analysis of the efficacy of blinatumomab therapy in heterogenic cohort r/r B-ALL patients, including adults and children, we built a mathematical model to predict response to this antibody. Patients and methods.A total of 128 pts with r\r B-ALL were included in the analysis, 52 children and 76 adults. Median age was 19 years (1-52). Hematological relapse (hr/r) and persistent measurable residual disease - molecular relapse (mr/r) were in 54% and 46% patients, respectively. Significant predictors of response in the multivariate analysis were used in the modeling for assessment of the probability of response to blinatumomab therapy. For validation of adjusted logistic regression parameters a k-fold cross-validation was performed. Confidence intervals for the model efficiency scores were calculated. Results. The response rate to blinatumomab in the study cohort was 75%. In the univariate and multivariate analysis the following parameters had the highest significance for response: patient age (OR=12,06, 95% CI 3,6-40, p<0,001), short mean time (less 18 months) to next line of treatment (OR=0,17, 95% CI 0,05-0,5, p=0,002) and bone marrow blast percentage (OR=0,06, 95% CI 0,01-0,2, p<0,001). The resulting ROC-curve for the regression model is given in figure 1A. The sensitivity/specificity trade-off depending on a decision threshold level in given in figure 1B. For the final classification model for predicting response to blinatumomab therapy in pts r/r B-ALL decision-making threshold was adjusted for obtaining maximum specificity- 0,85, sensitivity -0,5 (i.e. select the patients who will very likely respond). Using this model, the following guiding maps were calculated (figure 1C). The following confusion matrix was obtained when applying the resulting model to the full database: the specificity was 90% (ratio 30/3) and sensitivity was 50% (ratio 44/45). From the confusion matrix the following confidential internals for sensitivity and specificity were estimated: sensitivity - point estimation= 0,51 (95% CI 0,4-0,61), specificity - point estimation= 0.91 (95% CI 0,76-0,98). Conclusions.The possibility of response to blinatumomab therapy depends on patient's age, the bone marrow blast percentage and early relapse. The model predicts the response to blinatumomab therapy in patients with r/r B-ALL with high specificity and reasonable sensitivity and will be able to optimize the use of this very expensive therapy. Based on the model high probability of response in case of short time to next line of therapy is observed only for adult patients with low-to-moderate blast count, while in patients with long time to next line of therapy broader patient population has a high probability of response, including young children and patients with high blast count. Verification of this model in a larger multicenter study is required to confirm this hypothesis. Figure Disclosures No relevant conflicts of interest to declare.