Multiphoton microscopy in the diagnostic assessment of pediatric solid tissue in comparison to conventional histopathology: results of the first international online interobserver trial

Abstract

Purpose: Clear resection margins are paramount for good outcome in children undergoing solid tumor resections. Multiphoton microscopy (MPM) can provide high-resolution, real-time, intraoperative microscopic images of tumor tissue. Objective: This prospective international multicenter study evaluates the diagnostic accuracy, feasibility, and interobserver congruence of MPM in diagnosing solid pediatric tissue and tumors for the first time. Material and methods: Representative fresh sections from six different neonatal solid tissues (liver, lung, kidney, adrenal gland, heart muscle, testicle) and two types of typical pediatric solid tumors (neuroblastoma, rhabdomyosarcoma) with adjacent nonneoplastic tissue were imaged with MPM and then presented online with corresponding H&E stained slides of the exact same tissue region. Both image sets of each tissue type were interpreted by 38 randomly selected international attending pediatric pathologists via an online evaluation software. Results: The quality of MPM was sufficient to make the diagnosis of all normal tissue types except cardiac muscle in >94% of assessors with high interobserver congruence and 95% sensitivity. Heart muscle was interpreted as skeletal muscle in 55% of cases. Based on MPM imaging, participating pathologists diagnosed the presented pediatric neoplasms with 100% specificity, although the sensitivity reached only about 50%. Conclusion: Even without prior training, pathologists are able to diagnose normal pediatric tissues with valuable accuracy using MPM. While current MPM imaging protocols are not yet sensitive enough to reliably rule out neuroblastoma or rhabdomyosarcoma, they seem to be specific and therefore useful to confirm a diagnosis intraoperatively. We are confident that improved algorithms, specific training, and more experience with the method will make MPM a valuable future alternative to frozen section analysis. Registration: The trial was registered at www.researchregistry.com, registration number 2967.