Robotic MRI/CT Guided Multiregional ‘smart’ Biopsy for Characterization of Tumor Heterogeneity: A Prospective Development Study

Intratumoral heterogeneity means single site tumor biopsy might not be representative. Here we develop and optimize an MRI-informed robotic multiregional ‘smart’ biopsy technique in retroperitoneal and pelvic sarcomas (RPS). We also explore the relationship between imaging and histological biomarkers. Participants with suspected RPS underwent multiparametric (mp)MRI within a week prior to biopsy. Three target regions with differing MRI characteristics were contoured. Robotic or freehand multiregional biopsy was performed, collecting samples from each target region in separate specimen pots. CT/MRI fusion extracted quantitative imaging biomarkers for correlation with histological biomarkers at precise biopsy sites. The primary endpoint was feasibility and safety. Spearman's correlation explored the relationship between imaging and histological biomarkers. Twelve participants (7 women), median age 58.6 years interquartile range [IQR]: 52 – 75 years underwent biopsy. All procedures were technically successful with same-day discharge. The within-tumor range of Apparent Diffusion Coefficeint correlated very strongly with the within-tumor range of Ki-67 proliferation index; Spearman’s ρ = 0.91 (95% CI 0.68 to 0.98). Ranges represent intratumoral heterogeneity uniquely obtained by multiregional biopsy. Multiregional robotic MRI-informed, CT-guided biopsy is feasible and safe in RPS. Sampling three distinct regions within tumors provides a more comprehensive and accurate representation of tumor biology than standard biopsy. The close relationship between imaging and histological heterogeneity biomarkers has broader implications for pancancer biopsy techniques, imaging characterization, and personalized treatment selection.