Citations

Clear

6273 citation(s) found for "To, Andrew".

In silico approaches to patient selection: Credentialing elraglusib as a novel treatment in metastatic melanoma resistant to checkpoint inhibitors.
McDermott, Joseph , Weiskittel, Taylor M. , Borden, Brittany A. , Carneiro, Benedito A. , Hu, Li , Mazar, Andrew P. , Cavalcante, Ludimila , Giles, Francis Joseph

Mayo Clinic

Journal of Clinical Oncology | June 4, 2023 | Volume: 41 | Issue: 16_suppl | Pages: 3079-3079

10.1200/jco.2023.41.16_suppl.3079 Get It

3079 Background: Elraglusib (9-ING-41), a novel inhibitor of GSK-3β, has been evaluated in >230 patients (pts) with advanced malignancies including metastatic melanoma in the 1801 phase 1/2 trial (NCT03678883). Pts with melanoma treated with single agent elraglusib demonstrated clinical benefit including a durable complete response despite multiple prior lines of therapy including checkpoint inhibitors (CPI). Methods: We collected data from numerous -omic modalities in the 1801 trial to characterize markers of elraglusib response with bioinformatics and machine learning (ML). Genomic panels were collected on 105 pts. The most frequent tumor types included colorectal cancer (n=35), pancreas (n=31), lung (n=15) and metastatic melanoma (n=12). Pts were treated with elraglusib as a single agent or in combination with chemotherapy. Pathway based features were generated by quantifying the percent of each Reactome pathway that was mutated. These pathway features were iteratively evaluated for their performance in shallow neural networks predicting patient RECIST with each model using 5-15 randomly selected features at a time. Publicly available genomic data from an anti-PD-1 trial in patients with metastatic melanoma (Hugo et al., 2017) was transformed into pathway-based features and input into the final model to predict elraglusib response in this population. Results: Our final pathway based genomic model for elraglusib response performed with a training accuracy of 93% and a test accuracy of 88% on the 1801 patients. We then tested the model on metastatic melanoma patients who were treated with anti-PD-1 therapy. We found that patients from Hugo et al. were predicted to respond to elraglusib (22/26). In addition, most patients with tumors resistant to anti-PD-1 therapy were also predicted to respond to elraglusib (10/12). Finally, we tested the model on metastatic melanoma patients treated with elraglusib in the 1801 study that had failed CPI and for whom -omic data was available. The model correctly identified 5/5 patients that demonstrated clinical benefit (1 CR, ongoing>1400 days; 4 prolonged OS of 107, 256, 357, and 556 days). Conclusions: We applied robust in-silico modeling developed using 1801 trial data to develop a rationale for the development of elraglusib in the treatment of metastatic melanoma including patients that failed or were refractory to anti-PD-1 treatment. This model correctly identified metastatic melanoma patients in 1801 with good clinical outcomes that had failed prior CPI treatment. Further, we present this in silico approach as a robust way to repurpose prior clinical study genomics and ML modeling for population enrichment during the clinical development of novel drugs. Clinical trial information: NCT03678883 .
Plasma cytokine profiles and survival outcomes in the 1801 phase 1/2 clinical trial of 9-ING-41 (elraglusib) in patients with advanced cancer.
Weiskittel, Taylor M. , Carneiro, Benedito A. , El‐Deiry, Wafik S. , Hu, Li , Mazar, Andrew P. , Huntington, Kelsey E. , Cavalcante, Ludimila , Mahalingam, Devalingam , Giles, Francis Joseph , Münster, Pamela N.

Mayo Clinic

Journal of Clinical Oncology | June 4, 2023 | Volume: 41 | Issue: 16_suppl | Pages: e14555-e14555

10.1200/jco.2023.41.16_suppl.e14555 Get It

e14555 Background: 9-ING-41 (elraglusib) is a novel inhibitor of GSK-3β that has shown early signs of clinical benefit in a phase 1/2 clinical trial (ACT1801; NCT03678883). Early data support a multi-modal mechanism of action for elraglusib that targets both the tumor and immune system. Thus, the discovery and validation of biomarkers of drug sensitivity to elraglusib is a high priority. Here, we take inspiration from the literature on checkpoint inhibitor therapy and explore the potential of cytokines to predict elraglusib clinical response. Methods: Peripheral plasma samples from 59 patients enrolled in the 1801 phase 1/2 trial of elraglusib were assayed for 40 circulating cytokine levels. Patients advanced solid tumors refractory to standard treatments and received either elraglusib (1.0-15.0 mg/kg) as a single agent or in combination with other chemotherapies. Samples were taken from time points prior to elraglusib administration (pre-dose), after cycle 1, or after cycle 2. Each cytokine time point combination was used to stratify patients into high and low categories using an optimal cut-point determined using maximally selected rank statistics. Kaplan Meier and Cox Proportional Hazards analysis was used to determine the efficacy of each cytokine as a biomarker. The change in cytokine values between pre-dose and cycle 1 or pre-dose and cycle 2 was evaluated in a similar manner. Results: A total of 142 samples spanning eight different tumor histologies were collected and analyzed. A total of 240 cytokine-timepoint combinations were evaluated, and 53 significantly stratified patients by overall survival (p-value < 0.05). Of those, 35 showed that higher levels of the given cytokine correlated with a survival benefit and 18 a survival decrease. Notable among the cytokines demonstrating survival benefits were Granzyme B at Cycle 1 Day 4 (HR: 9.54, p-value:0.015) and TGF β at Cycle 1 Day 4 (HR: 8.48, p-value:0.0063). Strong hazard ratios were also observed for cytokines that correlated with survival decreases including TRAIL-R2 at Cycle 1 Day 2 (HR: 9.56e-02, p-value:0.0001), and IL-10 at Cycle 1 Day 4 (HR: 8.77e-02, p-value:0.007). Delta between timepoints was also able to stratifying patients by overall survival. Conclusions: Plasma cytokine profiling is a promising approach for discovery and validation of biomarkers prognostic for elraglusib clinical benefit. Numerous cytokines measured pre-dose stratified patients into significant survival groups suggesting a potential strategy for patient enrichment in clinical trials of elraglusib. These initial findings are being expanded by 1) increasing the cohort to accommodate histology specific investigations; 2) examining which cytokines are elraglusib response-specific versus drug-agnostic; and 3) combining cytokines in multivariate and machine learning models for optimal predictive capacity. Clinical trial information: NCT03678883 .
The SPEARHEAD-1 trial of afamitresgene autoleucel (afami-cel [formerly ADP-A2M4]): Pooled analysis of cytokine release syndrome across cohorts in patients with advanced synovial sarcoma.
Ganjoo, Kristen N. , Araujo, Dejka M. , Razak, Albiruni Ryan Abdul, Cesne, Axel Le , Schuetze, Scott M. , Wagner, Michael J. , Attia, Steven , Forcade, Édouard , Druta, Mihaela , Bai, Jane P. F., Lunt, Colin, Sun, Amy , Yule, Murray, McAlpine, Cheryl, Williams, Dennis, Tine, Brian Andrew Van, D'Angelo, Sandra P , Blay, Jean‐Yves , Valverde, Claudia María , Pollack, Seth M.

Mayo Clinic

Journal of Clinical Oncology | June 4, 2023 | Volume: 41 | Issue: 16_suppl | Pages: e14531-e14531

10.1200/jco.2023.41.16_suppl.e14531 Get It

e14531 Background: Afami-cel is an autologous, T-cell receptor T-cell therapy engineered to target melanoma-associated antigen A4 (MAGE-A4) in human leukocyte antigen (HLA) A*02–eligible patients (pts) with advanced solid tumors. SPEARHEAD-1 (NCT04044768) is a Phase 2, two cohort, open-label trial to evaluate the efficacy and safety of afami-cel in pretreated pts with advanced/metastatic synovial sarcoma or myxoid/round cell liposarcoma. Here we report data on cytokine release syndrome (CRS) in the pts with advanced synovial sarcoma treated with afami-cel across Cohort 1 and Cohort 2. Methods: Eligible pts had advanced synovial sarcoma and were HLA-A*02 positive with MAGE-A4–positive tumors. Pts received afami-cel after lymphodepleting chemotherapy containing fludarabine and cyclophosphamide. Tocilizumab administration was permitted for Grade 1 CRS if symptoms persisted for ≥ 24 hours or if the pt had co-morbidities, and for Grade ≥2 CRS. Results: As of August 29, 2022, 72 pts with synovial sarcoma received afami-cel (n = 44 Cohort 1; n = 28 Cohort 2). Afami-cel dose range was 1.92–10.00×109 transduced T-cells. The median age was 39.5 years (range: 16–73 years), 47.2% of pts were female, 88.3% of pts were white, and tumor MAGE-A4 expression H-score was 244.5 (range: 66–300). CRS occurred in 72.2% of the pts with synovial sarcoma receiving afami-cel, including Grade 1 in 50% (36/72), Grade 2 in 20.8% (15/72), and Grade 3 in 1.4% (1/72) of pts as per American Society for Transplantation and Cellular Therapy grading. The median time to onset was 2.5 days (range: 1–5 days), and the median time to resolution was 3 days (range: 1 to 14 days). Twenty-seven of 72 (37.5%) pts received at least one dose of tocilizumab. Consistent with expectations, higher serum IL-6 levels were associated with CRS events and the indicated use of tocilizumab for management. RECIST v1.1 responses were observed in pts with and without CRS. Correlation of clinical factors and CRS with response will be presented. Conclusions: CRS was common after afami-cel administration but was mostly low grade. Administration of tocilizumab in SPEARHEAD-1 was effective in the management of CRS. Clinical trial information: NCT04044768 .
Characteristics and predictive modeling of HER2 mutation in patients with non-small cell lung cancer.
Varolgüneş, Uras , Mittal, Garima , Parikh, Kaushal , Salas, Maribel , Arunachalam, Meena, Chen, Haoqian , Chen, Wenjin , Yu, Dantong, Setoguchi, Soko , Ip, Andrew , Malhotra, Jyoti Dhar , Bates, Benjamin

Mayo Clinic

Journal of Clinical Oncology | June 4, 2023 | Volume: 41 | Issue: 16_suppl | Pages: e13559-e13559

10.1200/jco.2023.41.16_suppl.e13559 Get It

e13559 Background: Recent advances in targeted therapies have increased the survival and quality of life for non-small cell lung cancer (NSCLC) patients. Understanding clinical characteristics and predicting HER2 status may facilitate efficient use of next-generation sequencing (NGS) testing and promote precision medicine/targeted therapies. Methods: NSCLC patients ≤65 years old, diagnosed between 01/2011 and 06/2021 who have NGS test results were identified in from two National Cancer Institute designated cancer centers in the Northeastern region of the United States. Demographics, behavioral, and clinical characteristics were extracted from electronic health record (EHR) data and EHR-linked cancer registry. Prevalence prediction modelling of HER2/ERBB2 mutation was performed using machine learning (ML) methods including logistic regression (LR), random forest (RF), naïve Bayes (NB), extreme gradient boosting (XGBoost) and multi-layer perceptron (MLP). 10- fold cross validation and k-nearest neighbor method were used to impute missing values. Under-sampling, oversampling and feature selection methods were tried. Results: Among 761 NSCLC patients with NGS reports available, 353 (46%) were ≤65 years old. Among 353 (mean age 57, 58% female, 88% non-Hispanic, 69% white, 57% had metastasis, 58% prescribed chemotherapy) 25 (7%) had a HER2 mutation. Based on univariate significance tests with LR on the complete data: cerebrovascular disease, BMI between 27.5-29.9, KRAS, EGFR and MET were significant with odds ratios (ORs) ranging from 0.2-3.9. In our prediction models, BMI, cerebrovascular disease, weight loss, hepatic metastases, diabetes, congestive heart failure, and KRAS, EGFR, MET and RET alteration status were significant predictors in multivariate analyses, suggesting the potential usefulness of these features in predicting HER2 mutation among patients with NSCLC. Area under the ROC curve (AUC) varied across different models from 70.9% (RF), 70% (XGB and MLP), 69.5% (LR) to 68.2% (Naïve Bayes). The best performing model with RF yielding an AUC ~71% can provide relative high specificity and acceptable sensitivity depending on the cut off. Conclusions: Our prediction models yielded AUC around 70% with the best performing model’s AUC being 71% in a subgroup aged ≤65. After further validation, our prediction model may be used to identify patients appropriate for repeat NGS testing who did not have HER2 testing in first molecular profiling. Future work is needed to predict status in repeat testing with previously test negative patients who relapse or have progression of disease. Random forest (RF) Model performance across different cutoff for the predicted probability. [Table: see text]
Effect of DUAL pharmacological blockade of AURKA and PD-L1 pathways on plasticity and metastasis for triple negative breast cancer.
Takchi, Andrew , Haddad, Tufia C.

Mayo Clinic

Journal of Clinical Oncology | June 4, 2023 | Volume: 41 | Issue: 16_suppl | Pages: e13100-e13100

10.1200/jco.2023.41.16_suppl.e13100 Get It

e13100 Background: Aurora-A mitotic kinase (AURKA) plays a pivotal role in Triple Negative Breast Cancer (TNBC) progression through activation of epithelial to mesenchymal transition (EMT)-mediated cancer cell plasticity. Cancer cells that undergo EMT acquire a CD44high/CD24low and/or ALDHhigh cancer stem-like phenotype (Cancer Stem Cells) characterized by high self-renewal capacity and intrinsic drug resistance. Cancer Stem Cells (CSCs) also shows high PD-L1 expression that plays a critical role in inducing tumor immune escape through CD8+ T-cells exhaustion. For these reasons, molecular characterization of oncogenic pathways that promote stemness and intra-tumoral PD-L1 expression is essential to evaluate whether novel combinatorial therapeutic strategies aimed to inhibit cancer cell plasticity, immune evasion capacity and organ metastasis may impact cancer progression. Methods: (1) TNBC Models: MDA-MB 231 and variant MDA-MB 231/LM cells, and unique TNBC-M40 cells established from metastatic PDXs; (2) In Vitro Studies: TNBC cells were treated with the AURKA inhibitor alisertib (50nM). PD-L1 expression was evaluated by immunoblotting assay. Apoptosis was measured by immunofluorescence using Cleaved-PARP antibodies. To assess stemness capacity, TNBC cells were cultured under non-adherent conditions to form mammospheres. ALDH activity (functional stemness marker) was measured using the Aldefluor Kit. (3) In Vivo Studies: 1x106 MDA-MB 231/LM cells were injected into the mammary fat pad of humanized NSG-CD34+ female mice. Animals were randomized into four groups (5 mice per group) and treated with anti-PD-L1 ICI atezolizumab (20 mg/Kg) and alisertib (25 mg/Kg) as monotherapy or in combination. Control groups were treated with saline solution placebo. PD-L1, CD44, Vimentin expression and tumor infiltration of CD8+ T-cells were assessed by immunofluorescence. Results: Treatment of MDA-MB 231/LM xenografts with alisertib decreased the levels of CD44 and PD-L1 expression and increased CD8+ T-cells tumor infiltration. Because PD-L1 genetic targeting enhanced in vitro alisertib-induced apoptosis, MDA-MB 231/LM xenografts were treated with alisertib and the anti-PD-L1 ICI atezolizumab. Combination of atezolizumab and alisertib resulted in the reduced expression of the EMT marker vimentin that was linked to lack of organ metastasis. Remarkably, combination of atezolizumab and alisertib also induced a significant reduction of ALDHHigh CSC in ex-vivo 3D-Organoids established from TNBC-M40 cells. Conclusions: This study provides innovative pre-clinical rationale for combining AURKA inhibitors with ICIs to impair cancer cell plasticity, immune evasion capacity and halt the progression of metastatic TNBCs that currently has limited effective therapeutic options.
Pirtobrutinib in covalent BTK-inhibitor (cBTKi) pre-treated mantle cell lymphoma (MCL): Updated results and subgroup analysis from the phase 1/2 BRUIN study with >3 years follow-up from start of enrollment.
Shah, Nirav N. , Eyre, Toby A. , Ujjani, Chaitra S. , Alencar, Alvaro J. , Lewis, David , Maddocks, Kami J. , Munir, Talha, Balbas, Minna, Tsai, Donald E. , Wang, Chunxiao , Wang, Michael , Jurczak, Wojciech , Zinzani, Pier Luigi , Cheah, Chan Y. , Izutsu, Koji , Ma, Shuo , Flinn, Ian W. , Patel, Krish , Wang, Yucai , Zelenetz, Andrew D.

Mayo Clinic

Journal of Clinical Oncology | June 4, 2023 | Volume: 41 | Issue: 16_suppl | Pages: 7514-7514

10.1200/jco.2023.41.16_suppl.7514 Get It

7514 Background: Pirtobrutinib is a highly selective, non-covalent (reversible) BTKi. Here, we report updated results of pirtobrutinib in patients (pts) with cBTKi pre-treated relapsed/refractory (R/R) MCL and more than 3 years follow-up from start of enrollment. Methods: Pts with cBTKi pre-treated R/R MCL received pirtobrutinib monotherapy in a multicenter phase 1/2 BRUIN trial (NCT03740529). Efficacy was assessed in the prespecified primary efficacy cohort that comprised the first 90 enrolled pts who had measurable disease, had received a prior cBTKi, and had no known central nervous system involvement. The primary endpoint was overall response rate (ORR) as assessed by independent review committee. Secondary endpoints included duration of response (DOR) and safety. A data cut of 29 July 2022 was utilized. Results: Among MCL pts who received a prior cBTKi (n=90), median age was 70 years (range, 46-87), median prior lines of therapy were 3 (range, 1-8), 82% discontinued a prior cBTKi due to disease progression, and 78% had intermediate/high risk sMIPI score. Of samples available, 17/36 (47%) had TP53 mutations and 25/34 (74%) had Ki67 ≥30%. The ORR was 57% (95% CI, 46-67), including 19% complete responses (n=17) and 38% partial responses (n=34). At a median follow-up time of 13 months, the median DOR among the 51 responding pts was 17.6 months (95% CI, 7.3-27.2). The 12- and 18-month estimated DOR rates were 58% (95% CI, 41-72) and 45% (95% CI, 27-61), respectively. ORR and DOR by subgroups are shown in the Table. The median progression-free survival was 7.4 months (95% CI, 5.3–13.3). The median overall survival was 23.5 months (95% CI, 15.9-NE). In the MCL safety cohort (n=166), the most frequent treatment-emergent adverse events (TEAE) were fatigue (31%), diarrhea (22%), and anemia (17%). The most common Grade ≥3 TEAE was neutropenia (15%). Grade ≥3 TEAE of hemorrhage (3%) and atrial fibrillation/flutter (2%) were infrequent. Only 5 (3%) pts discontinued due to a treatment-related AE. Conclusions: Pirtobrutinib continues to show durable efficacy and a favorable safety profile in heavily pre-treated R/R MCL pts with prior cBTKi therapy. Responses were observed in pts with high-risk disease features including pts with blastoid/pleomorphic variants, elevated Ki67 index, and TP53 mutations. Clinical trial information: NCT03740529 . [Table: see text]
RANO 2.0: Proposal for an update to the Response Assessment in Neuro-Oncology (RANO) criteria for high- and low-grade gliomas in adults.
Wen, Patrick Y. , Groot, John Frederick De, Lassman, Andrew B. , Molinaro, Annette M. , Shankar, Lalitha , Reardon, David A. , Vogelbaum, Michael A. , Chang, Susan M. , Wen, Patrick Yung , Bent, Martin J. van den , Youssef, Gilbert , Cloughesy, Timothy F. , Ellingson, Benjamin M. , Weller, Michael , Galanis, Evanthia , Barboriak, Daniel P. , Gilbert, Mark R. , Huang, Raymond Y. , Mehta, Minesh P. , Molinaro, Annette M. , Preusser, Matthias , Wick, Wolfgang

Mayo Clinic

Journal of Clinical Oncology | June 4, 2023 | Volume: 41 | Issue: 16_suppl | Pages: 2017-2017

10.1200/jco.2023.41.16_suppl.2017 Get It

2017 Background: Response Assessment in Neuro-Oncology (RANO) criteria for high-grade and low-grade glioma (HGG and LGG) were developed to improve reliability of response assessment in glioma trials. Several limitations of the original RANO criteria have been reported. Methods: To address limitations of the original RANO criteria, a large cohort of patients with newly diagnosed and recurrent glioblastoma were evaluated comparing RANO-HGG with modified RANO (mRANO) and immunotherapy RANO (iRANO) criteria to inform the following proposed updates (RANO 2.0). Results: Based on the 2021 WHO classification of glioma, we recommend a single common set of criteria for all gliomas regardless of WHO grade and IDH mutational status. These criteria will be used for all trials regardless of the treatment modalities being evaluated. In newly diagnosed glioblastoma, the post-radiotherapy MRI, usually obtained approximately 4 weeks after completion of radiotherapy, rather than the post-surgical MRI as proposed in original RANO criteria, will be used as the baseline for comparison of subsequent scans. Since the incidence of pseudoprogression is high in the first 3 months following radiotherapy, confirmation of progression during this period with a repeat MRI (in 4-8 weeks) will be required for determining radiographic progression of disease (PD) if the patient is clinically stable. However, confirmation scans are not mandatory after this period nor for recurrent tumors since these scans do not appear to improve reliability in determining progression. For agents with a high likelihood of producing pseudoprogression such as viral therapies and some other immunotherapies, mandatory confirmation of progression with a repeat MRI is an option. The primary measurement will remain the 2-dimensional maximum cross-sectional area of tumor as the simplest method, but where resources are available, volumetric measurements are an option. For IDH wild-type tumors with contrast enhancement, non-enhancing tumor will no longer be evaluated with the possible exception of patients receiving antiangiogenic therapies. However, in tumors with a significant non-enhancing component, both the contrast enhancing and non-enhancing components will be evaluated as target lesions. Conclusions: We hope that these and other proposed changes by the RANO working group will improve response assessment in glioma clinical trials and help the development of more effective therapy for patients.
Management and Outcomes of Wilms Tumor With Suprarenal Intravascular Extension
Naik‐Mathuria, Bindi , Dasgupta, Roshni , Short, Scott S. , Lovvorn, Harold N. , Newman, Erica, Rich, Barrie S. , Piché, Nelson , Glick, Richard D. , Petroze, Robin T. , Polites, Stephanie F. , Alore, Elizabeth , Sutthatarn, Pattamon , Wong‐Michalak, Shannon , Romao, Rodrigo , Al‐Hadidi, Ameer , Marquart, John , Gainer, Hailey, Johnson, Mike, Boehmer, Chloé , Rinehardt, Hannah , Seemann, Natashia M. , Davidson, Jacob , Polcz, Valerie E., Lund, Sarah , McKay, Katlyn G. , Correa, Hernán , Lund, Sarah B, Dasgupta, Roshni , Utria, Alan Fremder , Ehrlich, Peter F. , Aldrink, Jennifer H. , Murphy, Andrew J. , Lautz, Timothy B. , Kim, Eugene S. , Lal, Dave R. , Kastenberg, Zachary J. , Malek, Marcus M. , Whitlock, Richard S. , Chen, Stephanie , Rubalcava, Nathan S. , Gainer, Hailey, Rothstein, David H.

Mayo Clinic

Annals of Surgery | June 2, 2023 | Volume: 279 | Issue: 3 | Pages: 528-535

10.1097/sla.0000000000005921 | PMID 37264925 Get It

Objective: The purpose of this study was to describe management and outcomes from a contemporary cohort of children with Wilms tumor complicated by inferior vena caval thrombus. Background: The largest series of these patients was published almost 2 decades ago. Since then, neoadjuvant chemotherapy has been commonly used to manage these patients, and outcomes have not been reported. Methods: Retrospective review of 19 North American centers between 2009 and 2019. Patient and disease characteristics, management, and outcomes were investigated and analyzed. Results: Of 124 patients, 81% had favorable histology (FH), and 52% were stage IV. IVC thrombus level was infrahepatic in 53 (43%), intrahepatic in 32 (26%), suprahepatic in 14 (11%), and cardiac in 24 (19%). Neoadjuvant chemotherapy using a 3-drug regimen was administered in 82% and postresection radiation in 90%. Thrombus level regression was 45% overall, with suprahepatic level showing the best response (62%). Cardiopulmonary bypass (CPB) was potentially avoided in 67%. The perioperative complication rate was significantly lower after neoadjuvant chemotherapy [(25%) vs upfront surgery (55%); P=0.005]. CPB was not associated with higher complications [CPB (50%) vs no CPB (27%); P=0.08]. Two-year event-free survival was 93% and overall survival was 96%, higher in FH cases (FH 98% vs unfavorable histology/anaplastic 82%; P=0.73). Neither incomplete resection nor viable thrombus cells affected event-free survival or overall survival. Conclusions: Multimodal therapy resulted in excellent outcomes, even with advanced-stage disease and cardiac extension. Neoadjuvant chemotherapy decreased the need for CPB to facilitate resection. Complete thrombectomy may not always be necessary.
Developing nonlinear k-nearest neighbors classification algorithms to identify patients at high risk of increased length of hospital stay following spine surgery
Chan, Andrew K. , Bisson, Erica F. , Bydon, Mohamad , Glassman, Steven D. , Foley, Kevin T. , Shaffrey, Christopher I. , Potts, Eric A. , Shaffrey, Mark E., Coric, Domagoj , Knightly, John J. , Park, Paul , Wang, Michael Y. , Fu, Kai-Ming, Slotkin, Jonathan R., Asher, Anthony L. , Virk, Michael S. , Michalopoulos, Giorgos D. , Haid, Regis W., Chou, Dean , Mummaneni, Praveen V. , Michalopoulos, Giorgos , Shahrestani, Shane , Guan, Jian , Agarwal, Nitin

Mayo Clinic

Neurosurgical Focus | June 1, 2023 | Volume: 54 | Issue: 6 | Pages: E7

10.3171/2023.3.focus22651 | PMID 37283368 Get It

OBJECTIVE Spondylolisthesis is a common operative disease in the United States, but robust predictive models for patient outcomes remain limited. The development of models that accurately predict postoperative outcomes would be useful to help identify patients at risk of complicated postoperative courses and determine appropriate healthcare and resource utilization for patients. As such, the purpose of this study was to develop k-nearest neighbors (KNN) classification algorithms to identify patients at increased risk for extended hospital length of stay (LOS) following neurosurgical intervention for spondylolisthesis. METHODS The Quality Outcomes Database (QOD) spondylolisthesis data set was queried for patients receiving either decompression alone or decompression plus fusion for degenerative spondylolisthesis. Preoperative and perioperative variables were queried, and Mann-Whitney U-tests were performed to identify which variables would be included in the machine learning models. Two KNN models were implemented (k = 25) with a standard training set of 60%, validation set of 20%, and testing set of 20%, one with arthrodesis status (model 1) and the other without (model 2). Feature scaling was implemented during the preprocessing stage to standardize the independent features. RESULTS Of 608 enrolled patients, 544 met prespecified inclusion criteria. The mean age of all patients was 61.9 ± 12.1 years (± SD), and 309 (56.8%) patients were female. The model 1 KNN had an overall accuracy of 98.1%, sensitivity of 100%, specificity of 84.6%, positive predictive value (PPV) of 97.9%, and negative predictive value (NPV) of 100%. Additionally, a receiver operating characteristic (ROC) curve was plotted for model 1, showing an overall area under the curve (AUC) of 0.998. Model 2 had an overall accuracy of 99.1%, sensitivity of 100%, specificity of 92.3%, PPV of 99.0%, and NPV of 100%, with the same ROC AUC of 0.998. CONCLUSIONS Overall, these findings demonstrate that nonlinear KNN machine learning models have incredibly high predictive value for LOS. Important predictor variables include diabetes, osteoporosis, socioeconomic quartile, duration of surgery, estimated blood loss during surgery, patient educational status, American Society of Anesthesiologists grade, BMI, insurance status, smoking status, sex, and age. These models may be considered for external validation by spine surgeons to aid in patient selection and management, resource utilization, and preoperative surgical planning.
Which supervised machine learning algorithm can best predict achievement of minimum clinically important difference in neck pain after surgery in patients with cervical myelopathy? A QOD study
Park, Christine , Mummaneni, Praveen V. , Gottfried, Oren N., Shaffrey, Christopher I. , T'Ang, A , Bisson, Erica F. , Asher, Anthony L. , Coric, Domagoj , Potts, Eric A. , Foley, Kevin T. , Wang, Michael Y. , Fu, Kai-Ming, Virk, Michael S. , Knightly, John J. , Meyer, Scott A., Park, Paul , Upadhyaya, Cheerag D., Shaffrey, Mark E., Buchholz, Avery L., Tumialán, Luis M. , Turner, Jay D. , Sherrod, Brandon A. , Chou, Dean , Haid, Regis W., Bydon, Mohamad , Chan, Andrew K. , Agarwal, Nitin

Mayo Clinic

Neurosurgical Focus | June 1, 2023 | Volume: 54 | Issue: 6 | Pages: E5

10.3171/2023.3.focus2372 | PMID 37283449 Get It

OBJECTIVE The purpose of this study was to evaluate the performance of different supervised machine learning algorithms to predict achievement of minimum clinically important difference (MCID) in neck pain after surgery in patients with cervical spondylotic myelopathy (CSM). METHODS This was a retrospective analysis of the prospective Quality Outcomes Database CSM cohort. The data set was divided into an 80% training and a 20% test set. Various supervised learning algorithms (including logistic regression, support vector machine, decision tree, random forest, extra trees, gaussian naïve Bayes, k–nearest neighbors, multilayer perceptron, and extreme gradient boosted trees) were evaluated on their performance to predict achievement of MCID in neck pain at 3 and 24 months after surgery, given a set of predicting baseline features. Model performance was assessed with accuracy, F1 score, area under the receiver operating characteristic curve, precision, recall/sensitivity, and specificity. RESULTS In total, 535 patients (46.9%) achieved MCID for neck pain at 3 months and 569 patients (49.9%) achieved it at 24 months. In each follow-up cohort, 501 patients (93.6%) were satisfied at 3 months after surgery and 569 patients (100%) were satisfied at 24 months after surgery. Of the supervised machine learning algorithms tested, logistic regression demonstrated the best accuracy (3 months: 0.76 ± 0.031, 24 months: 0.773 ± 0.044), followed by F1 score (3 months: 0.759 ± 0.019, 24 months: 0.777 ± 0.039) and area under the receiver operating characteristic curve (3 months: 0.762 ± 0.027, 24 months: 0.773 ± 0.043) at predicting achievement of MCID for neck pain at both follow-up time points, with fair performance. The best precision was also demonstrated by logistic regression at 3 (0.724 ± 0.058) and 24 (0.780 ± 0.097) months. The best recall/sensitivity was demonstrated by multilayer perceptron at 3 months (0.841 ± 0.094) and by extra trees at 24 months (0.817 ± 0.115). Highest specificity was shown by support vector machine at 3 months (0.952 ± 0.013) and by logistic regression at 24 months (0.747 ± 0.18). CONCLUSIONS Appropriate selection of models for studies should be based on the strengths of each model and the aims of the studies. For maximally predicting true achievement of MCID in neck pain, of all the predictions in this balanced data set the appropriate metric for the authors’ study was precision. For both short- and long-term follow-ups, logistic regression demonstrated the highest precision of all models tested. Logistic regression performed consistently the best of all models tested and remains a powerful model for clinical classification tasks.