Automated NLP Extraction of Clinical Rationale for Treatment Discontinuation in Breast Cancer

2021 ◽  
pp. 550-560
Author(s):  
Matthew S. Alkaitis ◽  
Monica N. Agrawal ◽  
Gregory J. Riely ◽  
Pedram Razavi ◽  
David Sontag

PURPOSE Key oncology end points are not routinely encoded into electronic medical records (EMRs). We assessed whether natural language processing (NLP) can abstract treatment discontinuation rationale from unstructured EMR notes to estimate toxicity incidence and progression-free survival (PFS). METHODS We constructed a retrospective cohort of 6,115 patients with early-stage and 701 patients with metastatic breast cancer initiating care at Memorial Sloan Kettering Cancer Center from 2008 to 2019. Each cohort was divided into training (70%), validation (15%), and test (15%) subsets. Human abstractors identified the clinical rationale associated with treatment discontinuation events. Concatenated EMR notes were used to train high-dimensional logistic regression and convolutional neural network models. Kaplan-Meier analyses were used to compare toxicity incidence and PFS estimated by our NLP models to estimates generated by manual labeling and time-to-treatment discontinuation (TTD). RESULTS Our best high-dimensional logistic regression models identified toxicity events in early-stage patients with an area under the curve of the receiver-operator characteristic of 0.857 ± 0.014 (standard deviation) and progression events in metastatic patients with an area under the curve of 0.752 ± 0.027 (standard deviation). NLP-extracted toxicity incidence and PFS curves were not significantly different from manually extracted curves ( P = .95 and P = .67, respectively). By contrast, TTD overestimated toxicity in early-stage patients ( P < .001) and underestimated PFS in metastatic patients ( P < .001). Additionally, we tested an extrapolation approach in which 20% of the metastatic cohort were labeled manually, and NLP algorithms were used to abstract the remaining 80%. This extrapolated outcomes approach resolved PFS differences between receptor subtypes ( P < .001 for hormone receptor+/human epidermal growth factor receptor 2− v human epidermal growth factor receptor 2+ v triple-negative) that could not be resolved with TTD. CONCLUSION NLP models are capable of abstracting treatment discontinuation rationale with minimal manual labeling.

Cancers ◽  
2019 ◽  
Vol 11 (6) ◽  
pp. 737 ◽  
Author(s):  
Denis M. Collins ◽  
Neil T. Conlon ◽  
Srinivasaraghavan Kannan ◽  
Chandra S. Verma ◽  
Lisa D. Eli ◽  
...  

An estimated 15–20% of breast cancers overexpress human epidermal growth factor receptor 2 (HER2/ERBB2/neu). Two small-molecule tyrosine kinase inhibitors (TKIs), lapatinib and neratinib, have been approved for the treatment of HER2-positive (HER2+) breast cancer. Lapatinib, a reversible epidermal growth factor receptor (EGFR/ERBB1/HER1) and HER2 TKI, is used for the treatment of advanced HER2+ breast cancer in combination with capecitabine, in combination with trastuzumab in patients with hormone receptor-negative metastatic breast cancer, and in combination with an aromatase inhibitor for the first-line treatment of HER2+ breast cancer. Neratinib, a next-generation, irreversible pan-HER TKI, is used in the US for extended adjuvant treatment of adult patients with early-stage HER2+ breast cancer following 1 year of trastuzumab. In Europe, neratinib is used in the extended adjuvant treatment of adult patients with early-stage hormone receptor-positive HER2+ breast cancer who are less than 1 year from the completion of prior adjuvant trastuzumab-based therapy. Preclinical studies have shown that these agents have distinct properties that may impact their clinical activity. This review describes the preclinical characterization of lapatinib and neratinib, with a focus on the differences between these two agents that may have implications for patient management.


2016 ◽  
Vol 34 (10) ◽  
pp. 1134-1150 ◽  
Author(s):  
Lyndsay N. Harris ◽  
Nofisat Ismaila ◽  
Lisa M. McShane ◽  
Fabrice Andre ◽  
Deborah E. Collyar ◽  
...  

Purpose To provide recommendations on appropriate use of breast tumor biomarker assay results to guide decisions on adjuvant systemic therapy for women with early-stage invasive breast cancer. Methods A literature search and prospectively defined study selection sought systematic reviews, meta-analyses, randomized controlled trials, prospective-retrospective studies, and prospective comparative observational studies published from 2006 through 2014. Outcomes of interest included overall survival and disease-free or recurrence-free survival. Expert panel members used informal consensus to develop evidence-based guideline recommendations. Results The literature search identified 50 relevant studies. One randomized clinical trial and 18 prospective-retrospective studies were found to have evaluated the clinical utility, as defined by the guideline, of specific biomarkers for guiding decisions on the need for adjuvant systemic therapy. No studies that met guideline criteria for clinical utility were found to guide choice of specific treatments or regimens. Recommendations In addition to estrogen and progesterone receptors and human epidermal growth factor receptor 2, the panel found sufficient evidence of clinical utility for the biomarker assays Oncotype DX, EndoPredict, PAM50, Breast Cancer Index, and urokinase plasminogen activator and plasminogen activator inhibitor type 1 in specific subgroups of breast cancer. No biomarker except for estrogen receptor, progesterone receptor, and human epidermal growth factor receptor 2 was found to guide choices of specific treatment regimens. Treatment decisions should also consider disease stage, comorbidities, and patient preferences.


2015 ◽  
Vol 33 (10) ◽  
pp. 1136-1142 ◽  
Author(s):  
Ian E. Krop ◽  
Thomas M. Suter ◽  
Chau T. Dang ◽  
Luc Dirix ◽  
Gilles Romieu ◽  
...  

Purpose Trastuzumab emtansine (T-DM1), an antibody–drug conjugate comprising the cytotoxic agent DM1, a stable linker, and trastuzumab, has demonstrated substantial activity in human epidermal growth factor receptor 2 (HER2) –positive metastatic breast cancer, raising interest in evaluating the feasibility and cardiac safety of T-DM1 in early-stage breast cancer (EBC). Patients and Methods Patients (N = 153) with HER2-positive EBC and prechemotherapy left ventricular ejection fraction (LVEF) ≥ 55% received (neo)adjuvant doxorubicin plus cyclophosphamide or fluorouracil plus epirubicin plus cyclophosphamide followed by T-DM1 for four cycles. Patients could then receive three to four cycles of optional docetaxel with or without trastuzumab. T-DM1 was then resumed with optional radiotherapy (sequential or concurrent) for 1 year (planned) of HER2-directed therapy. The coprimary end points were rate of prespecified cardiac events and safety. Results Median follow-up was 24.6 months. No prespecified cardiac events or symptomatic congestive heart failures were reported. Four patients (2.7%) had asymptomatic LVEF declines (≥ 10 percentage points from baseline to LVEF < 50%), leading to T-DM1 discontinuation in one patient. Of 148 patients who received ≥ one cycle of T-DM1, 82.4% completed the planned 1-year duration of HER2-directed therapy. During T-DM1 treatment, 38.5% and 2.7% of patients experienced grade 3 and 4 adverse events, respectively. Approximately 95% of patients receiving T-DM1 plus radiotherapy completed ≥ 95% of the planned radiation dose with delay ≤ 5 days. Conclusion Use of T-DM1 for approximately 1 year after anthracycline-based chemotherapy was feasible and generally well tolerated by patients with HER2-positive EBC, providing support for phase III trials of T-DM1 in this setting.


Sign in / Sign up

Export Citation Format

Share Document