Point-of-Care Diagnosis of Endometrial Cancer Using the Surgical Intelligent Knife (iKnife)—A Prospective Pilot Study of Diagnostic Accuracy

SIMPLE SUMMARY: The iKnife is an emerging tool which uses standard electrosurgical methods to generate surgical aerosols that are then interrogated by a mass spectrometer to provide real-time tissue signatures. It has been shown to correctly identify different tissue types including lung, colon and liver. The aim of this study was to ascertain whether the iKnife can correctly identify endometrial cancer from endometrial Pipelle biopsy samples, thereby facilitating point-of-care diagnosis. One hundred and fifty endometrial Pipelle samples were analysed in this study. The iKnife reliably diagnosed endometrial cancer in seconds, with a diagnostic accuracy of 89%, minimising the current delays for women whilst awaiting a histopathological diagnosis. The findings presented in this study can pave the way for new diagnostic pathways. ABSTRACT: Introduction: Delays in the diagnosis and treatment of endometrial cancer negatively impact patient survival. The aim of this study was to establish whether rapid evaporative ionisation mass spectrometry using the iKnife can accurately distinguish between normal and malignant endometrial biopsy tissue samples in real time, enabling point-of-care (POC) diagnoses. Methods: Pipelle biopsy samples were obtained from consecutive women needing biopsies for clinical reasons. A Waters G2-XS Xevo Q-Tof mass spectrometer was used in conjunction with a modified handheld diathermy (collectively called the ‘iKnife’). Each tissue sample was processed with diathermy, and the resultant surgical aerosol containing ionic lipid species was then analysed, producing spectra. Principal component analyses and linear discriminant analyses were performed to determine variance in spectral signatures. Leave-one-patient-out cross-validation was used to test the diagnostic accuracy. Results: One hundred and fifty patients provided Pipelle biopsy samples (85 normal, 59 malignant, 4 hyperplasia and 2 insufficient), yielding 453 spectra. The iKnife differentiated between normal and malignant endometrial tissues on the basis of differential phospholipid spectra. Cross-validation revealed a diagnostic accuracy of 89% with sensitivity, specificity, positive predictive value and negative predictive value of 85%, 93%, 94% and 85%, respectively. Conclusions: This study is the first to use the iKnife to identify cancer in endometrial Pipelle biopsy samples. These results are highly encouraging and suggest that the iKnife could be used in the clinic to provide a POC diagnosis.