GP2d - Limitations of Virtual Surgical Planning in Osteoradionecrosis: The Role of Imaging in Margin Determination

Abstract: Introduction: Osteoradionecrosis (ORN) of the craniomaxillofacial skeleton is a severe complication of head and neck radiation therapy that patients are at a lifelong risk for. Surgical resection is often required for advanced disease. Accurate preoperative determination of disease margins is critical for surgical planning, particularly when virtual surgical planning (VSP) is employed for microvascular reconstruction as VSP cannot compensate for the lack of accuracy of surgical margins. However, no validated imaging-based criteria exist for defining adequate bony resection margins. This study aims to evaluate the accuracy of imaging modalities in defining ORN margins and their implications for VSP reliability.

Methods: A systematic literature search was conducted in MEDLINE. Studies evaluating computed tomography (CT), magnetic resonance imaging (MRI), cone-beam computed tomography (CBCT), panoramic radiography, bone scintigraphy, single photon emission computed tomography (SPECT/CT), and photon emission computed tomography (PET/CT) for ORN margin determination were included. Sensitivity, specificity, and diagnostic accuracy were extracted where available. Findings were analyzed in the context of their impact on surgical planning and VSP accuracy.

Results: The search identified 150 studies of which 17 studies met the inclusion criteria for systematic review. SPECT/CT demonstrated the highest sensitivity (95-100%) and specificity (91%) for determining ORN extent and margins. MRI showed high sensitivity for detecting bone marrow changes that may precede CT-visible destruction, though limited specificity. Panoramic radiography was useful for screening, but underestimated disease extent compared to CT. However, paired CT and panoramic imaging significantly improved diagnostic performance (P < .001). PET/CT remains limited in distinguishing ORN from tumor recurrence due to overlapping standard uptake values; CT-based radiomics models have achieved 88% accuracy in differentiating ORN from healthy tissue. CBCT detected more ORN lesions than SPECT/CT (100% vs 92%) though its role in margin determination requires further validation. Furthermore, the analysis revealed a lack of correlation between histologic margin status and disease progression. No imaging modality reliably predicted bone viability or correlated with histologic margins. This contributes to frequent intraoperative modification and reduced VSP accuracy. ORN is associated with increased intercondylar/intergonial error in VSP and poorer reconstructive outcomes, with traditional VSP demonstrating inferior bony healing compared to dosimetry-guided approaches.

Conclusion: The primary limitation with VSP in ORN is the inability to accurately determine disease margins preoperatively. Intraoperative bleeding is currently the gold standard for margin determination which creates an inherent conflict with VSP, which requires margin determination before surgery. Despite advances in imaging, no single modality provides sufficiently reliable preoperative margin determination to fully support accurate VSP. SPECT/CT appears to be the most promising modality for preoperative margin determination given its demonstrated correlation with pathological findings and high sensitivity/specificity. The combination of CT with functional imaging (SPECT or DCE-MRI) may be the best approach for virtual surgical planning. The literature reveals a critical gap in validated imaging criteria for preoperative margin determination in ORN. While multiple modalities can detect and characterize ORN, the correlation between radiologic findings and histologic margins remains poorly established, thus limiting the precision of preoperative planning and leading to higher complication rats and intraoperative adjustments. These findings highlight the need for further clinical investigation of multimodal imaging techniques to enhance accuracy and clinical utility of VSP in ORN.


References:

1. Johal M, Ma JNB, Parthasarathi K, et al. Institutional-Based and Commercial Virtual Surgical Planning in Maxillomandibular Reconstruction - Comparing the Digital Plan and Postoperative Scan. Journal of Plastic, Reconstructive & Aesthetic Surgery : JPRAS. 2022.


2. Comparative Evaluation of SPECT/-CT and CBCT in Patients With Mandibular Osteomyelitis and Osteonecrosis.

Clinical Oral Investigations. 2019. Malina-Altzinger J, Klaeser B, Suter VGA, et al.