Predicting likelihood of in vivo chemotherapy response in canine lymphoma using ex vivo drug sensitivity and immunophenotyping data in a machine learning model

Bohannan, Zach; Pudupakam, Raghavendra Sumanth; Koo, Joonhoi; Horwitz, Harrison; Tsang, Josephine; Polley, Amanda; Han, Enyang James; Fernández, E. Esteban; Park, Stanley; Swartzfager, Deanna; Qi, Nicholas Seah Xi; Tu, Chantal; Rankin, Wendi Velando; Thamm, Douglas H.; Lee, Hye Ryeon; Lim, Se-won

doi:10.1111/vco.12656

Cited by 8 publications

(11 citation statements)

References 49 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…It would be anticipated with more sophisticated in vitro modeling algorithms, chemotherapy response predictions could become more accurate and clinically actionable for optimizing treatment of canine uroepithelial carcinomas. Such is the paradigm currently being tested for canine lymphoma, whereby large data sets inclusive of clinical, molecular, and biologic data can be combined and analyzed with machine learning models to theoretically design optimal chemotherapy protocols [ 47 ].…”

Section: Discussionmentioning

confidence: 99%

Investigating PSMA differential expression in canine uroepithelial carcinomas to aid disease-based stratification and guide therapeutic selection

et al. 2022

View full text Add to dashboard Cite

Background In male dogs, uroepithelial cancers include invasive urothelial carcinoma (iUC) and prostate carcinoma (PCA). The inability to distinguish iUC involving the prostate from PCA results in indiscriminate clinical management strategies that could be suboptimal as first-line chemotherapy for iUC (cisplatin) and PCA (docetaxel) differ in people. Prostate specific membrane antigen (PSMA) is a transmembrane protein, and its overexpression has been identified in human prostate carcinoma and neovasculature associated with solid tumor growth. This study investigates whether differential PSMA expression exists between presumptive canine iUC and PCA among cell lines and archived patient samples, which might allow for improved accuracy in disease-based stratification and optimal chemotherapy selection. Additionally, in vitro sensitivities of reported canine iUC and PCA cell lines to uroepithelial directed chemotherapeutic agents were characterized. Results Normalized PSMA gene and protein expressions were not significantly different between 5 iUC and 4 PCA cell lines. PSMA protein expression was uniformly observed in uroepithelial cancers regardless of anatomic origin from archived patient samples, further confirming that PSMA cannot differentiate iUC from PCA. In vitro sensitivity of cell lines to uroepithelial directed chemotherapeutics revealed that vinblastine exerted the broadest cytotoxic activity. Conclusions Differential expression of PSMA was not identified between canine iUC and PCA cell lines or archived patient samples, and PSMA alone cannot be used for disease stratification. Nonetheless given its conserved overexpression, PSMA may be a targetable surface marker for both canine iUC and PCA. Lastly, in uroepithelial carcinomas, vinblastine might exert the broadest anticancer activity regardless of cellular origin.

show abstract

Section: Discussionmentioning

confidence: 99%

Investigating PSMA differential expression in canine uroepithelial carcinomas to aid disease-based stratification and guide therapeutic selection

et al. 2022

View full text Add to dashboard Cite

show abstract

“…In the previous study, we successfully developed the ML model for predicting in vivo response of single chemotherapeutic drugs using DS and FC data [10]. We first tried to emulate the previous work by training ML models with the drug sensitivity of only five drugs that constitute the (L-)CHOP chemotherapy and flow cytometry readouts.…”

Section: Discussionmentioning

confidence: 99%

“…Towards this aim, we previously reported the development of the ML approach for predicting in vivo response to a single chemotherapeutic drug [10]. Random forest (RF) models were trained using ex vivo drug sensitivity analysis and flow cytometry results to estimate the probability of positive response to a given drug.…”

Section: Introductionmentioning

confidence: 99%

Predicting Dynamic Clinical Outcomes of the Chemotherapy for Canine Lymphoma Patients Using a Machine Learning Model

Koo

Choi²,

Lee³

et al. 2021

Veterinary Sciences

Self Cite

View full text Add to dashboard Cite

First-line treatments of cancer do not always work, and even when they do, they cure the disease at unequal rates mostly owing to biological and clinical heterogeneity across patients. Accurate prediction of clinical outcome and survival following the treatment can support and expedite the process of comparing alternative treatments. We describe the methodology to dynamically determine remission probabilities for individual patients, as well as their prospects of progression-free survival (PFS). The proposed methodology utilizes the ex vivo drug sensitivity of cancer cells, their immunophenotyping results, and patient information, such as age and breed, in training machine learning (ML) models, as well as the Cox hazards model to predict the probability of clinical remission (CR) or relapse across time for a given patient. We applied the methodology using the three types of data obtained from 242 canine lymphoma patients treated by (L)-CHOP chemotherapy. The results demonstrate substantial enhancement in the predictive accuracy of the ML models by utilizing features from all the three types of data. They also highlight superior performance and utility in predicting survival compared to the conventional stratification method. We believe that the proposed methodology can contribute to improving and personalizing the care of cancer patients.

show abstract

“…The sensitivity of live tumor cells to 13 different drugs was quantified using a high-throughput ex vivo assay as previously described ( 3 ). Tumor cells were profiled at the single-cell level using multicolor flow cytometry and a panel of 9 primary antibodies as previously described ( 3 ).…”

Section: Methodsmentioning

confidence: 99%

“…In both species, the tumors are typically highly responsive to first-line combination therapies that include cyclophosphamide, doxorubicin, vincristine, and prednisone (CHOP). There is not yet a standard of care for either dogs or humans when patients relapse after first-line therapy ( 2 , 3 ). Patients may be reinduced with first-line therapy or treated with one of several different rescue therapies (salvage therapies).…”

Section: Introductionmentioning

confidence: 99%

Multimodal machine learning models identify chemotherapy drugs with prospective clinical efficacy in dogs with relapsed B-cell lymphoma

Callegari,

Tsang,

Park

et al. 2024

Front. Oncol.

Self Cite

View full text Add to dashboard Cite

Dogs with B-cell lymphoma typically respond well to first-line CHOP-based chemotherapy, but there is no standard of care for relapsed patients. To help veterinary oncologists select effective drugs for dogs with lymphoid malignancies such as B-cell lymphoma, we have developed multimodal machine learning models that integrate data from multiple tumor profiling modalities and predict the likelihood of a positive clinical response for 10 commonly used chemotherapy drugs. Here we report on clinical outcomes that occurred after oncologists received a prediction report generated by our models. Remarkably, we found that dogs that received drugs predicted to be effective by the models experienced better clinical outcomes by every metric we analyzed (overall response rate, complete response rate, duration of complete response, patient survival times) relative to other dogs in the study and relative to historical controls.

show abstract

Predicting likelihood of in vivo chemotherapy response in canine lymphoma using ex vivo drug sensitivity and immunophenotyping data in a machine learning model

Cited by 8 publications

References 49 publications

Investigating PSMA differential expression in canine uroepithelial carcinomas to aid disease-based stratification and guide therapeutic selection

Investigating PSMA differential expression in canine uroepithelial carcinomas to aid disease-based stratification and guide therapeutic selection

Predicting Dynamic Clinical Outcomes of the Chemotherapy for Canine Lymphoma Patients Using a Machine Learning Model

Multimodal machine learning models identify chemotherapy drugs with prospective clinical efficacy in dogs with relapsed B-cell lymphoma

Contact Info

Product

Resources

About