PET/MRI Machine Learning Model May Eliminate Sentinel Lymph Node Biopsy in Majority of Breast Cancer Patients

Patients with newly diagnosed breast cancer receive a PET/MRI scan to investigate involvement of the axillary lymph nodes. The radiologist then assesses whether lymph node involvement is present (node ​​positive vs. node negative) based on easily assessable lymph node morphological and metabolic criteria. Based on this data, a random forest model is trained. Thus, the most important lymph node criteria relevant for the assessment of the condition of the lymph nodes are determined. By adjusting the threshold, the sensitivity can now be increased by means of random forest such that 68.2% of patients can be spared an axillary biopsy. Credit: Journal of Nuclear Medicine (2022). DOI: 10.2967/jnumed.122.264138

Almost 70% of breast cancer patients could know if their cancer has spread to their lymph nodes without having to undergo an invasive sentinel node biopsy. New research published ahead of print in the Journal of Nuclear Medicine shows that with the help of machine learning (a type of artificial intelligence), axillary lymph node metastases can be reliably excluded based on PET/MRI imaging.

The presence of lymph node metastases plays a crucial role in treatment planning, especially regarding the extent of surgery and radiation therapy. Therefore, it is of great clinical importance to distinguish patients with lymph node metastases from patients without lymph node metastases.

“Sixty percent of patients do not have lymph node metastases at the initial diagnosis of breast cancer,” said study author Janna Morawitz, MD, a radiology resident at the Institute of Diagnostic and Interventional Radiology in Düsseldorf University Hospital in Germany.

“As such, it would be desirable to be able to prove negative lymph node status by imaging with a high degree of certainty to spare these patients the invasive procedure of biopsy or surgery.”

In the study, the researchers sought to determine whether machine learning prediction models could determine the status of lymph nodes in PET/MRI scans as accurately as an experienced radiologist could. A total of 303 patients with primary breast cancer from three medical centers were recruited for the study and were divided into a training group sample and a test group sample.

All patients underwent MRI and whole-body PET/MRI dedicated to 18F-FDG. Imaging datasets were assessed for axillary lymph node metastases based on structural and functional features. Machine learning models were developed based on the sample from the MRI and PET/MRI training group and were then applied to the sample from the test group.

The diagnostic accuracy of MRI was 87.5% for radiologists and the machine learning algorithm. For PET/MRI, accuracy was 89.3% for radiologists and 91.2% for machine learning. After fitting the machine learning model for PET/MRI, a sensitivity of 96.2% and a specificity of 68.2% were obtained.

“Based on the information gleaned from MRI and PET/MRI scans, decision trees can be developed to help radiologists, especially junior radiologists, determine if a sentinel node biopsy is warranted,” Morawitz noted. . “Integration of this model into daily practice could potentially replace sentinel lymph node biopsy in the future.”