What is the future of indocyanine green and near‐infrared imaging in the surgical management of endometriosis?

We thank colleagues Turco et al for their interest in our recently published article about the use of indocyanine green (ICG) during the surgical management of endometriosis,1 and take the opportunity to respond to their comments.2 The discussion about the ideal ICG exposure time is certainly of importance. As ICG is bound to plasma proteins within seconds and 97% of the dye is excreted after 20 minutes, the optimal time to visualize tissue perfusion is 1–5 minutes as described in colorectal or liver surgeries.3 Our results indicate the necessity to wait for extravasation of the ICG dye into the tissue to clearly delineate endometriosis lesions.1 However, no evidence-based recommendations exist regarding ideal ICG exposure time. We do not agree that including data from a “sub-treated population” has caused bias in our study.2 To determine the ideal ICG exposure time was the aim of our study. In the optimally exposed group of 35 women, the ICG detection rate increased with increased exposure time even though the diagnostic value did not improve (five additional lesions were detected that had no histologically confirmed endometriosis).1 Turco et al doubt whether women with advanced-stage endometriosis did obtain an adequate exposure of the surgical field.2 Certainly, we performed an adequate adhesiolysis before inspection with near-infrared (NIR)-ICG, as the focus of our study was to detect subtle lesions. The identification of deep infiltrating endometriosis nodules, such as in the rectovaginal septum, does not imply difficulties. However, an extensive adhesiolysis leading to small areas of bleeding might interfere with the NIR fluorescence detection of endometriosis lesions.1 We defined the dose of ICG in accordance with the current literature analyzing visualization of tissue perfusion in the liver and in colorectal surgery (0.2-0.5 mg/kg),3 because the Gre-Endo Trial4 was not published at that time. We agree that a dose of 0.25 mg/kg of ICG might be sufficient, although we did not observe any interfering “background noise”. To date, no study has reported on the optimal dose of ICG to visualize endometriosis. Furthermore, our group has been using ICG since 2012 in sentinel lymph node biopsy for gynecological cancer and has published more than 25 articles on this topic, including one specifically looking at the impact of different doses of ICG on sentinel lymph node detection.5 We do agree that practical skills are necessary to identify endometriosis pathologically in fluorescent tissue. In our certified Endometriosis Center, we treat about 750 women annually and our team is experienced both in the histological diagnosis of endometriosis and in the examination of fluorescent tissue,5 applying immunohistochemistry if necessary. As a result of the high negative predictive value of white-light laparoscopy, we omitted control biopsies1 to reduce costs and minimize morbidity. We defined the true negatives for NIR-ICG as lesions positive with white-light laparoscopy, negative for NIR-ICG and with negative histology. True negative lesions that were negative with both NIR-ICG and white-light imaging could not be included. This might explain the poor negative predictive value of NIR-ICG imaging reported in our study. We agree that future studies are required to answer several questions. However, we believe that the future of ICG in endometriosis is in fluorescence-guided surgery for deep infiltrating nodules and not in its diagnostic potential.