Near-Infrared intraoperative imaging with indocyanine green is feasible in segmentectomy for patients with chronic lung diseases: A retrospective single-center propensity-score matched analysis


 Objective

To investigate whether video-assisted thoracoscopic segmentectomy using near-infrared fluorescence imaging had better intersegmental plane visualization and peri-operative outcome in patients with chronic lung diseases.
Methods

Data were collected retrospectively from March 2014 and August 2019. A total of 92 patients with pulmonary nodules underwent near-infrared fluorescence guided uni-port thoracoscopic segmentectomy(NIF-VATS), 149 patients underwent thoracoscopic segmentectomy with inflation-deflation method(ID-VATS). After 1:1 propensity matching, perioperative outcomes between NIF-VATS and ID-VATS was compared.
Results

Incision size was 3 cm in both group.Mean operative time was 79 minutes in NIF-VATS group and 96 minutes in ID-VATS group. The intersegmental plane was not clear in 33 cases of ID-VATS group, and no clear boundary was found after prolonged waiting time. Emphysema or pulmonary bullae could be found in chest CT scan in these patients, they all were diagnosed as chronic obstructive pulmonary disease. In NIF-VATS group, the intersegmental plane was not clear in 8 cases. Under the guidance of three-dimensional reconstruction and preoperative positioning, the oncological margin length of both groups met the requirements of surgical quality control. The intraoperative blood loss, number of lymph node resection, showed no statistical difference between the two groups. Postoperative air leakage was more often observed in ID-VATS group. The postoperative drainage duration, postoperative hospitalization time was shorter in ID-VATS group.
Conclusions

Compared with inflation-deflation method, segmentectomy using NIF imaging is feasible for patients with chronic lung diseases with better intersegmental plane, shorter operation time, less complications, it might lead to faster recovery.


Introduction
Lung cancer is a malignant tumor with the highest incidence and mortality in the world (1). With the popularization of low dose computed tomography screening, the epidemiological characteristics of lung cancer have changed greatly (2). Segmentectomy is recommended for surgical treatment of early stage lung cancer, especially for patients with limited pulmonary function preserved (3). One of the most important factors for segmentectomy is to identify the accurate intersegmental plane, which is particularly important to ensure su cient distance between the cutting edges and lesion (4). In ationde ation method is the widely used, however, the intersegmental plane was usually hard to be identi ed clearly using in ation-de ation method in patients with chronic lung diseases due to decreased lung compliance. Near-Infrared Fluorescence (NIF) imaging with indocyanine green (ICG) is a recent advancement in segmentectomy with high success rates of intersegmental border visualization (5).
Few studies compared near-infrared uorescence (NIF) imaging with indocyanine green and in ationde ation method in identifying intersegmental border. The objective is to investigate whether nearinfrared uorescence imaging with indocyanine green had better intersegmental border visualization and peri-operative outcome in patients with chronic lung diseases underwent segmentectomy.

Patient Selection
A retrospective analysis was performed, a total of 92 patients with chronic lung diseases underwent thoracoscopic segemtectomy with near-infrared uorescence (NIF) imaging. The rst near-infrared uorescence guided thoracoscopic segmentectomy was performed in September 2017, the rst thoracoscopic segmentectomy using in ation-de ation method was performed in March 2014, and the initial 30 cases were not included due to the learning curve effect(6), a total of 149 patients with chronic lung diseases were included in control group. All surgeries were performed by highly-experienced experts in this specialized procedure. All surgeries were completed through one single incision. Consultants in our department all agreed that either technique was suitable for each patient. Cases with conversion to thoracotomy or lobectomy were excluded.
This study for the application of thoracoscopic segmentectomy and near-infrared uorescence imaging was approved by by the institutional review board at Nanjing Chest Hospital(number of the ethics approval: 2014-KL002-02 for uniportal thoracoscopic segmentectomy, 2017-KL-002-03 for near-infrared uorescence guided thoracoscopic segmentectomy), written informed consent about operative techniques and to the data-use agreement was obtained from all patients before surgery.
Based on preoperative CT ndings, patients with comorbid COPD(emphysema or chronic bronchitis), ILD(idiopathic pulmonary brosis (IPF) and interstitial pneumonia(IP), or CPFE(Combined pulmonary brosis and emphysema) were included. Other inclusion criteria for NIF-VATS or ID-VATS segmentectomy in patients with chronic lung diseases: Pulmonary nodules with a diameter of less than or equal to 2 cm in thin-slice chest CT and have at least one of the following characteristics: 1. adenocarcinoma in situ; 2.
composition of ground glass appearance is greater than or equal to 50%; 3. lesion doubling time is greater than or equal to 400 d.
Exclusion criteria were pulmonary nodules that cannot be removed by segmental pulmonary resection, multiple primary tumors require surgery other than pulmonary segmentectomy, wedge resection, patients unable to tolerate surgery (Fig. 1).
Segmentectomy was categorized into simple or complex from the standpoint of technical aspect: resection of superior segment of the lower lobe or the lingular segment was de ned as simple. Complex segmentectomy was resection of segment other than that to be resected in the simple segmentectomy, beside, complex segmentectomy was de ned as resection of a segment that had more than one intersegmental plane(7).

Quality control standard
All patients underwent contrast-enhanced computer tomography(CT), three-dimensional reconstruction was completed using EDDA-IQQA software or 3D-slicer(before November 2017). CT guided preoperative localization was performed (Fig. 2), except lesion located in apical, lingular or superior segment. Oncological margin length should be larger than 2 cm or maximal diameter of lesion.

Anaesthesia and incision
The anaesthesia procedures were the same in both groups. All patients received combined intravenous and inhalation anesthesia, with double-lumen endotracheal intubation. The patient was kept in lateral decubitus position, both cranial and caudal side was pushed down like a fold knife to make intercostal space wider, uni-portal VATS was applied. All patients were extubated after operation. Postoperative patient-controlled analgesia was provided.
A single incision, about 3 cm long, was performed at the fth intercostal space along the anterior axillary line (Fig. 3). Plastic wound protector was used. During operation, a 10-mm 30-degree thoracoscope (Pinpoint, Novadaq Technologies ULC, Stryker, for NIF-VATS segmentectomy; Karl Storz for ID-VATS segmentectomy) was placed at the superior side of the incision, several thoracoscopic instruments were simultaneously tted into the uni-port beneath the thoracoscope.
Segmentectomy using in ation-de ation method or NIF imaging Following ligation of the corresponding segmental artery, bronchus, and vein according to preoperative CT and three-dimensional reconstruction, intersegmental plane was identi ed by in ation-de ation method or near-infrared uorescence imaging.
Using in ation-de ation method, the collapsed pulmonary tissue was initiated to fully re-expand with controlled airway pressure under 20 cm H 2 O by pure oxygen, followed by single lung ventilation. After an interval of approximately 20 minutes, an irregularly curved demarcation was identi ed naturally between the de ated preserving segments and the in ated target segment.
We observed ICG uorescence using a near-infrared thoracoscope. ICG was reconstituted into distilled water to produce a 2.5 mg/ml solution, and a volume of 6-10 mL (dose of 15 mg to 20 mg) was injected in a peripheral intravenous catheter, followed by 10 ml of saline solution. The intersegmental plane of the targeted segment became clearly visible on infrared thoracoscopy about 15 seconds after injection, and could be marked using electrocautery.
A monopolar cautery hook was used to divide the pulmonary parenchyma of the intersegmental plane from the central side toward the peripheral side along the marked line, and anatomical segmentectomy was achieved using autosuture's cartridges.
When the ICG uorescence was too dim to enable any visualization of the intersegmental border, or only part of the intersegmental border could be found in NIF-VATS group; the target segment was not demarcated from the rest of the lung in ID-VATS group, successfully resection along the border could not be achieved, it was regarded as "unclear intersegmental boarder". Then intersegmental vein should be fully dissected, the target segment was excised along it.
Surgical margin, the distance between the cut end uncovered by visceral pleura and the tumor margin, was evaluated macroscopically to con rm that the surgical margin was not less than the maximum tumor diameter or 20 mm. If tumor free margin was not achieved, extended segmentectomy would be performed.
Lymph node sampling was performed when inraoperative pathology showed minimal invasive adenocarcinoma(MIA) or invasive adenocarcinoma(IAC).
To control air leakage after segmentectomy, partial closure of the visceral pleura, the use of brin glue, or reinforcement with a absorbable pad on the intersegmental plane were allowed.

Chest Tube Remove
One 22F chest drain was inserted to thoracic cavity at the end of the operation, it was placed at posterior part of the uniportal incision. Drain removal criteria were as follows: no observed air leak and total drainage less than 200 ml in 24 hours; normal chest roentgenograph; normal vital signs; good overall medical status. No patient was discharged with a chest tube in situ.

Postoperative Treatment and Follow-Up
Chest radiography was performed at the rst day postoperatively and every 3 days until discharge. Drinking and meal intake were resumed after bowel sounds returned with no nausea or vomiting in both groups.
Postoperative wound pain was monitored using the Visual Analogue Scales(VAS). The scale is an integer scale of 0-10, where 0 is no pain and 10 is the worst pain imaginable. VAS was evaluated on postoperative days 1, 3, 7, 14, 30, 60.
Statistical Analysis SPSS 16.0 for Windows (IBM, Armonk, NY) was used for analysis. To minimize the impact of potential confounders and selection bias, propensity score analysis was used to compensate for the differences in baseline patient characteristics between the two groups of patients. Patients in the two groups were 1:1 matched using the nearest propensity score on the logit scale. Variables that could in uence the outcomes of treatment were matched, including age, gender, body mass index (BMI), ASA status class, and maximal lesion size. After PSM, differences in continuous and categoric clinical characteristics were compared.
Continuous data are presented as mean and SD and were analyzed with two-sample Student's t tests for independent data. Categorical variables are given as a count and percentage of patients and analyzed with the χ 2 or Fisher's exact test. All tests were two-sided, P-values < 0.05 were considered statistically signi cant. SPSS Statistics 16.0 (IBM Corp, Armonk, NY) was used for statistical evaluations.

Results
Between September 2017 and August 2019, a total of 122 patients with chronic lung diseases underwent uni-portal thoracoscopic segemtectomy with near-infrared uorescence (NIF) imaging, 30 cases were not included due to the learning curve effect. A total of 149 patients with chronic lung diseases underwent uni-portal thoracoscopic segemtectomy using in ation-de ation method was included in control group. Before matching, there were no signi cant differences between the two groups in terms of gender, age, smoking history, pulmonary function(FEV1, %FEV1, FEV1/FVC, %DLCO),preoperative CT ndings, BMI, ASA score, comorbidity, maximal tumor size on CT scan, the lung lobe where lesion located, type of segmentectomy. After 1:1 matching, 184 patients remained for analysis, baseline demographic and clinical variables were comparable between the two groups ( Table 1). Position of lesion and extent of surgical area included RS 1 , RS 2 , RS 3 , RS 2 b + S 3 a, RS 1 a + S 2 a, RS 3 a + S 1 b, RS 6 , RS 7 + 8 , RS 9 + 10 , RS 7-10 , LS 1 + 2+3 , LS 1 + 2 , LS 1 + 2 b + c, LS 4 + S 1 + 2 c, LS 3 , LS 4 + 5 , LS 6 , LS 8 , LS 9 + 10 , LS 7-10 .  Under the guidance of three-dimensional reconstruction and preoperative positioning, the oncological margin length of both groups met the requirements of surgical quality control. There was no case of conversion to thoracotomy in either group.
The intraoperative blood loss, number of lymph node resection showed no statistical difference between the two groups. Histopathologic examination of patients with a pulmonary nodule included AAH, AIS, MIA, IAC and other benign lesion( Table 2).
Postoperative prolonged air leakage (indicating the time of chest tube leakage ≥ 5 days) was more often observed in ID-VATS group(7.6% in NIF-VATS group and 14.1% in ID-VATS group). Postoperative drainage duration and postoperative hospitalization time was shorter in NIF-VATS group, mean chest tube duration was 4.6 days, mean postoperative hospital stay was 5.6 days. Rate of other complications was not different, including postoperative hemoptysis, atrial brillation, pleural effusion, irritable cough (Table 2). There were no serious postoperative complications or death within 30 days after surgery in either group. There was no signi cant difference of VAS on postoperative day 1, 3, 7, 30, 60.
The median (range) follow-up period was 16(3-31) months for the NIF-VATS group, and 38(8-70) months for the ID-VATS group. All patients were alive. No tumor recurrence or metastasis was noticed during follow-up in both groups.

Discussion
Segmentectomy is fundamentally a more limited parenchymal resection and is being advocated for earlystage lung cancer. Several clinical studies have shown that the long-term e cacy of segmentectomy for early NSCLC with a diameter of less than 2 cm is similar to that of lobectomy, with more lung function preserved (3). The advantages of segmentectomy could be preservation of lung function with low frequency of surgical morbidity and mortality, especially for patients with compromised lung function.
Adequate tumor free margin is one of the most important factors in segmentectomy (8,9). Precise identi cation of the intersegmental border is critical for successful anatomical pulmonary segmentectomy. In ation-de ation method is the widely used, however, it usually takes about 15-20 minutes to show a intersegmental plane after in ation (10,11). For patients with impaired pulmonary function, chronic obstructive pulmonary disease or extensive pleural adhesion, the intersegmental plane was hard to be identi ed clearly using in ation-de ation method due to decreased lung compliance (12). In this report, patients with COPD or other factors related with impaired pulmonary function, intersegmental plane was not clear in segmentectomy using in ation-de ation method.
Near-Infrared Fluorescence (NIF) imaging with indocyanine green (ICG) is a recent advancement in minimally invasive segmentectomy (13). Following ligation of the corresponding segmental artery, vein, and bronchus within the target segment, ICG is then injected into a peripheral vein (14). The surgical eld is visualized with the uorescence imaging. The target segment, which was now isolated from the pulmonary vasculature and therefore not perfused with ICG, exhibited no uorescence and remained dark. The remainder of the lung, which was perfused with ICG, became uorescent green. a clear delineation of the intersegmental plane was then identi ed as the line separating the dark lung parenchyma from the green lung parenchyma. The operation time was signi cantly shortened. Besides, the brightness of ICG uorescence mainly depends on both the ICG concentration in blood vessels and the blood ow in the pulmonary artery (15), it is also in uenced by the condition of the pulmonary parenchyma, including anthoracosis and emphysema (16). It may be di cult to clearly visualize the entire intersegmental plane using ICG in patients with severe emphysema, as the blood ow in emphysematous regions is quite low.
When intersegmental plane was not clear, segmental hilum should be opened wider by electrocautery or other energy divice to identify intersegmental vein (17,18), which would cause more damage to lung tissue. For patients with impaired lung function, postoperative air leakage was more often observed in ID-VATS group.
Whether the target segmental arteries were cut precisely also in uence the indenti cation of the intersegment plane. Recognition of anatomy based on three-dimensional reconstruction is essential for safe and precise anatomical segmentectomy (19). It allows the surgeon to be guided in classi cation of individual anatomical patterns, identi cation of anomalies, and intraoperative navigation for veins, arteries, and bronchi.
NIF imaging could last for about 30 seconds in clear condition, it seemed to be a little short, but usually enough to mark the plane using cautery by experienced surgeon.
The median duration of postoperative hospital stay was relatively long, hospitalization in China is relatively long because patients usually do not have to pay high costs for prolonged hospitalization. Patients prefer to stay in the hospital for an additional 1 or 2 days after an operation despite having met the criteria for discharge in our hospital.
There were several limitations to this present study. For retrospective nature of the study, the randomization was absent, and selection bias cannot be eliminated. Although geriatric patients underwent segmentectomy seemed to have better peri-operative result, prospective research was needed to further con rm the conclusion. Small sample size and short follow-up time were also the main limitations, long-term and subjective patient outcomes should be established in future studies to assess both peri-operative outcome and oncologic e cacy.

Conclusion
Compared with in ation-de ation method, segmentectomy using NIF imaging is feasible for patients with chronic lung diseases with better intersegmental plane, shorter operation time, less complications, it might lead to faster recovery. Ethics approval and consent to participate: This study for the application of thoracoscopic segmentectomy and near-infrared uorescence imaging was approved by by the institutional review board at Nanjing Chest Hospital(number of the ethics approval: 2014-KL002-02 for uniportal thoracoscopic segmentectomy, 2017-KL-002-03 for near-infrared uorescence guided thoracoscopic segmentectomy), written informed consent about operative techniques and to the data-use agreement was obtained from all patients before surgery. Consent for publication: Consent form from Nanjing Chest Hospital was obtained from patients included in the research.
Availability of supporting data: The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request. Figure 1 The owchart of the study.  Three-dimensional reconstruction was completed before surgery.

Figure 4
Page 20/20 The intersegmental plane visualisation during thoracoscopic segmentectomy. A: The intersegmental plane was not clear using in ation-de ation method, B: near-infrared intraoperative imaging with indocyanine green is associated with better intersegmental plane visualization.