ICG Spots More Positive Sarcoma Margins, Surgeons Are More Specific

An intraoperative indocyanine green (ICG) fluorescence imaging protocol developed specifically for sarcoma surgery identified 14 of 15 pathology-confirmed positive margins, compared with 4 of 15 identified by surgeons, according to an observational study published in the Journal of Surgical Oncology.¹ However, the ICG approach had a specificity of 41.7% vs 88.9% for surgeon impression, and the authors recommended that future studies combine the 2 methods rather than use imaging alone.

Sarcomas are rare cancers that arise from connective tissues such as muscle, bone, nerve, fat, and cartilage. The American Cancer Society estimates that about 13,910 soft tissue sarcomas will be diagnosed in the US in 2026, with about 5400 deaths.² Surgical removal with negative histopathological margins is the foundation of treatment, and positive margins—which previous studies have reported in 10% to 30% of patients—are associated with local recurrence, the study authors noted.¹ They explained that intraoperative margin assessment relies largely on surgeon experience and judgment and that frozen section analysis is expensive, time-consuming, and limited to the areas a surgeon selects.

“Sarcoma surgeons are guided almost entirely by the subjective parameters of experience, training, and clinical judgment,” the authors wrote. “More precise, objective measurements are warranted to potentially decrease the incidence of local recurrence.”

The University of Pittsburgh team previously described ICG use in an earlier cohort of 26 patients with soft tissue sarcoma, using imaging parameters developed in breast cancer surgery.³ The current analysis expanded the cohort to develop sarcoma-specific parameters, which the authors described as the first such protocol in the literature.¹

In the prospective, observational, nonrandomized study (NCT04719156), 67 patients were enrolled between April 2021 and April 2024, and 51 were included in the analysis after 16 were excluded. Patients received intravenous ICG (Spy Agent Green; Stryker Endoscopy) on the day of surgery at 2.5 mg/kg for the first 15 patients and 2.0 mg/kg thereafter. After each resection, the surgeon left the operating room to remain blinded while a trained resident or research staff member used a near-infrared camera to measure fluorescence in the specimen and at 5 locations in the tumor bed. A signal-to-background ratio (SBR) was calculated by dividing tumor bed signal fluorescence by average background fluorescence, surgeons recorded their margin assessments on a questionnaire, and no clinical decisions were based on ICG results. Final pathology served as the definitive measure of margin status.

Indocyanine Green Sensitivity vs Surgeon Specificity for Sarcoma Margins

Most patients had soft tissue sarcoma (40 of 51; 78.4%), the mean (SD) age at surgery was 58.8 (18.4) years, and 35 of 51 tumors (68.6%) were in the lower extremity. Final pathology identified positive margins in 15 patients (29.4%), while surgeons predicted 8 (15.7%). Surgeons correctly identified 4 of 15 positive margins (sensitivity, 26.7%) and 32 of 36 negative margins (specificity, 88.9%), with an accuracy of 70.6%. ICG classified with a threshold adapted from breast cancer surgery performed worse, with a sensitivity of 6.7% and specificity of 94.4%.

Receiver operating characteristic curve analysis of SBR produced an area under the curve of 0.67 and an optimal ICG-SBR cut point of 1.67. Classifying margins as positive at an SBR of 1.67 or greater yielded a sensitivity of 93.3%, specificity of 41.7%, positive predictive value of 40%, negative predictive value of 93.8%, and accuracy of 56.9%. The difference in SBR between pathology-positive and pathology-negative margins was not statistically significant (P = .06).

The authors recommended that future studies use an intraoperative strategy that combines the high sensitivity of ICG with the high specificity of surgeon impression.

“The ICG detector can ‘see’ things that the surgeon cannot. Surgeons, by virtue of their experience, possess clinical knowledge that the detector does not,” the authors wrote.

Local Recurrence, Survival, and Study Limitations

Over a median follow-up of 25.2 months (95% CI, 20.9-30.5), 7 patients (13.7%) experienced local recurrence and 17 (33.3%) died. Among the 7 patients with recurrence, surgeons had predicted 3 positive margins, pathology reported 5, and ICG detected 6. Only pathology-determined margin status was associated with recurrence-free survival (log-rank P = .01; HR adjusted for age and sex, 5.8; 95% CI, 1.1-30.4). Neither the surgeon’s impression nor ICG at the 1.67 cut point were associated with recurrence-free survival, and none of the 3 methods were associated with overall survival.

The authors explained that the close link between final pathology and local recurrence is concerning if the goal is a tool to lower recurrence, and they cited the small sample size and limited follow-up as possible explanations. The study was also limited by its observational, nonrandomized design, a cohort with only 15 positive margins, the heterogeneity of sarcoma subtypes, and a surgeon impression form that was not a standardized metric. They called for a prospective, randomized, multi-institutional trial to determine whether sarcoma-specific ICG protocols reduce local recurrence.

“Perhaps the greatest lesson we have learned through this study is that ICG is very sensitive and surgeons are very specific,” the authors concluded. “ICG measurements ought not be acted upon without incorporating the surgeon’s clinical judgement.”

References

1. Bhogal S, Gong M, Li W, et al. Intraoperative evaluation of sarcoma surgical margins with indocyanine green fluorescence imaging: results from an observational study. J Surg Oncol. Published online September 17, 2026. doi:10.1002/jso.70383

2. Key statistics for soft tissue sarcomas. American Cancer Society. Updated February 9, 2026. Accessed October 9, 2026. https://www.cancer.org/cancer/types/soft-tissue-sarcoma/about/key-statistics.html

3. Gong MF, Li WT, Bhogal S, et al. Intraoperative evaluation of soft tissue sarcoma surgical margins with indocyanine green fluorescence imaging. Cancers (Basel). 2023;15(3):582. doi:10.3390/cancers15030582