Introduction
The development of a real-time intraoperative detection system that is sensitive and specific for tumors will help ensure complete tumor resection, as a clear boundary between tumor and normal tissues is ideal for real-time surgical imaging, particularly for tiny (<5 mm) liver metastases and superficial liver tumors that cannot be observed before surgery (van der Vorst et al.
2012). Although sensitive optical sensors and probes have been developed, including nanoparticles, fluorescent proteins, or near-infrared (NIR) fluorescent molecules, time is necessitated for these entities to be used routinely in clinical practice (de Chermont et al.
2007; Adusumilli et al.
2006; Veiseh et al.
2007).
As a surgical navigation technology, an indocyanine green (ICG) fluorescence imaging (FI) system has had a positive effect on the surgical treatment of gastric cancer, breast cancer, and skin cancer, as well as other tumors (Miyashiro et al.
2008; Kitai et al.
2005; Tsujino et al.
2009; Kurihara et al.
2015; Morita et al.
2013). Further, ICG is used in liver function tests and accumulates in hepatocellular carcinoma (HCC) tissues postintravenous injection (Shibasaki et al.
2015). In 2009, Ishizawa et al. reported the first application an ICG–FI navigation system in the surgical treatment of liver tumors (Lim et al.
2014). Moreover, ICG is a relatively safe reagent that is clinically approved by the United States Food and Drug Administration (FDA) for examination of hepatic function, cardiac output, and retinal angiography and has been used for these purposes for 20 years (Lim et al.
2014; Kitai et al.
1999). Given that ICG is already used in clinical imaging examinations, the use of ICG in intraoperative imaging may help improve the accuracy of tumor and liver tissue demarcation and identification.
In the present study, we introduced an efficient ICG–FI system that allows for accurate navigation during liver resection using an intraoperative ICG delivery method in China.
Materials and methods
This study was approved by the Institutional Review Board of the Tianjin First Central Hospital.
Research objective
Fifty patients underwent liver resection between Sep 2014 and May 2015 at Tianjin First Central Hospital. The presence of focal liver lesions was confirmed with abdominal ultrasound, enhanced computed tomography (CT), or magnetic resonance imaging (MRI), and distant involvement and metastases were ruled out. Liver function was assessed to determine the Child-Pugh score, and an ICG skin test was performed before surgery.
According to the actual tumor status and operative goal, patients were divided into two groups: group I, determination of the range in size of superficial tumors and identification of a small tumor; and group II, determination of a liver resection margin and real-time navigation during a liver resection.
Patient inclusion criteria
To be enrolled in this study, patients had to meet the following criteria: confirmed lesion occupying the liver using abdominal ultrasound, abdominal CT scan, and MRI plus four-phase CT imaging before surgery; exclusion of distant involvement and metastasis; Child A liver function during a preoperative assessment; did not have embolization treatment before surgery; and negative ICG skin test before surgery.
Experimental reagents and equipment
In this study, we used an ICG injection (Dandong Medical and Pharmaceutical Co, Ltd). According to the manufacturer’s instructions for the ICG solution, it was used at a concentration of 2.5 mg/ml, and light exposure was avoided. Further, a photodynamic eye (PDE, Hamamatsu Photonics Trading Co, Ltd) was used.
Research methods
In this study, we had a group comprising patients who underwent the following: determination of the range in size of liver superficial tumors and identification of small tumors. For the ICG procedures, we used the following technique. First, an intravenous injection of ICG at a 0.25-mg/kg dose was administered during surgery through the following routes: (1) portal vein puncture; (2) right vein of the stomach; and (3) central venous catheter. Second, an infrared fluorescence observation camera was turned on, using the imaging mode on the camera. Third, we adjusted the brightness, contrast, and sharpness of the photon eye. Fourth, a surgeon used a satirized package of lenses and cables for the photon eye. Fifth, the surgeon irradiated the liver with the photon eye at a distance of 5–10 cm above the liver surface. Lastly, PDE equipment was used to directly examine the liver, so the liver cancer could be detected; subsequently, the cutting edge of the liver could be confirmed. After surgery, the specimens were sent to pathology, and the intraoperative observations were compared with the pathological findings of the liver.
In this study, we had a second group that underwent the following: determination of hepatic resection margin and real-time navigation during a hepatectomy. To do so, the following procedures were performed. (1) The precut half of the liver was marked using the following methods: Intraoperatively, the hepatic portal vein was dissected to free it, along with its left and right branches. The portal vein branches in the resected liver were ligated. ICG was infused at a dose of 0.25 mg/kg through a central venous catheter. PDE was used to clearly observe the line at which the liver would be cut in half. Then, the liver resection line was marked in the PDE approach guide.
We used the following methods to mark a standard cut segment of the liver. According to the preoperative imaging findings, the left and/or right pedicle of the liver was dissected and freed intraoperatively. Then, the liver parenchyma was removed using Cut-ultrasound aspiration (CUSA), and the Glisson’s sheath of the segments or subsegments was fully visualized. The portal vein was freed, and the distal portal vein on the precut side was clamped. After injection of 1–2 mg of ICG through the puncture site and release of the vessel clip, PDE allowed for the clear observation of the status of the liver segment and was used to determine the hepatectomy line.
Discussion
In the present study, using an intraoperative administration of ICG with an FI navigation system and a PDE, we found 12 small tumors in 8 patients in whom a preoperative imaging examination did not indicate that these tumors existed. Of these eight patients, five patients had hepatocellular carcinoma; we also found the following conditions: recurrent nodular cirrhosis (two patients), liver macrovesicular steatosis (two patient), hemangioma (two patients), and hepatic focal hyperplasia (two patients). With our intraoperative ICG–FI navigation system, the smallest detected tumors were approximately 2 mm in diameter.
ICG is an FDA-approved reagent that is safe for examination of hepatic function, cardiac output, and retinal angiography (Lim et al.
2014; Kitai et al.
1999). ICG is a near-infrared fluorescent dye that is stimulated by light wavelengths of 750–810 nm that produce near-infrared light with a wavelength of 850 nm (Porcu et al.
2016). Near-infrared light can be observed using an ICG–FI system with a PDE and is displayed with the help of a developing device (van der Vorst et al.
2012; Porcu et al.
2016; Abo et al.
2015).
ICG fluorescence can penetrate living tissue deeply and thus can be advantageous for visualizing these tissues (Gotoh et al.
2009). In a preliminary study, Kitai et al. reported that fluorescence was observed from an ICG solution embedded 1 cm deep in material with optical properties compatible with human tissues; after an intravenous injection of ICG, it was rapidly ingested by the liver, as indicated by the stimulation of emission light (Miyashiro et al.
2008). To that end, after several hours, the liver can completely excrete ICG, and the fluorescence gradually dissipates (de Graaf et al.
2011). However, in cases of liver cirrhosis, liver regeneration nodules, liver cancer, and other types of liver dysfunction, the secretion and excretion function of liver cells is impaired, and ICG remains in lesions; thus, the dissipation of fluorescence is delayed (Gotoh et al.
2009; de Graaf et al.
2011; van der Vorst et al.
2013).
ICG–FI is a safe and minimally invasive technique (Gotoh et al.
2009). Several studies have reported that an intravenous ICG administration of 0.25–0.5 mg/kg from 12–24 h to 14 days before surgery helps to identify tumors with clear boundaries with intraoperative FI (Tsujino et al.
2009; Morita et al.
2013; Lim et al.
2014; Kawaguchi et al.
2013). In a report by Kawaguchi et al., an 81-year-old man with recurrent HCC secondary to hepatitis C was administered an ICG injection 4 days prior to surgery that revealed clear fluorescence of the tumor on the liver surface during surgery (Kawaguchi et al.
2013). In the present study, fluorescence contrast between normal liver and tumor tissues was obvious in 32 of 35 patients with intraoperative ICG administration. A boundary for half the liver or specific liver segments was determined in nine patients by examining the portal vein anatomy after ICG injection.
At present, the mechanisms that mediate the accumulation of ICG in HCC nodules relative to normal liver tissue remain unknown. Normal liver tissue can rapidly uptake ICG, and ICG is usually eliminated in bile; however, severely cirrhotic liver tissue may not be able to eliminate ICG (Sear
1990; Verbeek et al.
2012). In contrast, it has been shown that ICG passively accumulates in HCC if ICG is administered 1–8 days prior to surgery; therefore, borders of superficial liver tumors can be visualized clearly and accurately with a PDE (Verbeek et al.
2012). Using this principle, a boundary of a tumor can be determined that can aid in the determination of the liver resection line.
Three intraoperative routes of administration of ICG were clinically validated in this study. Administering ICG into the portal vein or right vein of the stomach, fluorescence of normal liver tissue developed within 1–2 min, while fluorescence of the cirrhotic tissue developed slowly and non-uniformly. However, administering ICG through a central venous catheter, fluorescence of normal liver tissue developed uniformly in 5–10 min, but the fluorescence pattern observed with this route of administration did not differ from that observed with administration through the portal vein or right vein of the stomach. Moreover, central venous ICG administration resulted in slightly slower and weaker fluorescence, but with an increased dose, the results significantly improved. We observed that central venous administration avoided the anatomy of the portal hepatic and right gastroepiploic veins, shortening the operative time and reducing injuries.
An ICG fluorescent system has the ability to detect small lesions that preoperative imaging examinations cannot help identify (van der Vorst et al.
2012). Preoperative imaging examinations such as CT and ultrasound have limited detection rates for small tumors in the liver (van der Vorst et al.
2012). Even with intraoperative ultrasounds, tiny lesions on the liver surface can be easily missed, leading to the occurrence of liver metastasis in the short term. With preoperative ICG administration, it can be difficult to detect a demarcation line in the liver parenchyma, as cirrhotic liver nodules have the same appearance as tumor tissues (Kurihara et al.
2015). At this time, one negative is that malignant tumors cannot be distinguished from benign tumors, thus leading to a high false positive rate (Morita et al.
2013). In this study combining ICG imaging with the use of a PDE, eight patients had small tumors that were not identified during a preoperative imaging examination, including five patients who had malignant lesions.
Because of more accurate preoperative imaging assessments and intraoperative real-time three-dimensional navigation instructions, liver resection has become more accurate (Morise et al.
2014). It can also be used repeatedly, avoiding the disadvantages of repetitive dyeing of the liver with methylene blue (Ishizawa et al.
2012; Ishizawa and Kokudo
2013; Lim and Vibert
2013).
This study had some limitations. First, this study had a small sample size. Additional investigation with a larger sample size is necessitated to demonstrate these results. Second, this study did not compare the results demonstrated in the two groups with those of a comparator group. Third, this study did not compare the current results with those of prior studies. Therefore, future study is needed to compare preoperative ICG injection timing and various routes of ICG administration and to further apply this tool in clinical practice.