Outcomes of Anastomotic Evaluation Using Indocyanine Green Fluorescence during Minimally Invasive Esophagectomy

Abstract

Background

Limited evidence exists assessing whether anastomotic evaluation using indocyanine green fluorescence (IGF) during minimally invasive esophagectomy (MIE) predicts or improves outcomes. We hypothesized that IGF helps surgeons predict anastomotic complications and reduces anastomotic leaks after MIE.

Methods

In September 2019, our institution began routinely using IGF for intraoperative evaluation of anastomoses during MIE. Data were collected from patients undergoing MIE in the two years before and after this technology began being routinely used. Baseline characteristics and outcomes, including anastomotic leak, in patients who underwent indocyanine green fluorescence evaluation (ICG) and those who did not (nICG) were compared. Outcomes were also compared between ICG patients with normal versus abnormal fluorescence.

Results

Overall, 181 patients were included. Baseline demographic and clinical characteristics did not differ between the ICG and nICG groups. ICG patients experienced higher rates of anastomotic leak (10.2% vs. 1.6%, P = .015) and 90-day mortality (8.5% vs. 1.6%, P = .04) compared to nICG patients. Due to lack of equipment availability, 19 nICG patients underwent MIE after the use of IGF became routine, and none developed leaks. ICG patients with abnormal fluorescence had higher rates of anastomotic leak (71.4% vs 1.9%, P < .001) and 30-day mortality (28.6% vs 0%, P = .012) compared to those with normal fluorescence.

Discussion

Abnormal intraoperative IGF was associated with increased rate of anastomotic leak, suggesting predictive potential of IGF. However, its use was associated with an increased leak rate and higher mortality. Further studies are warranted to assess possible physiologic effects of indocyanine green on the esophageal anastomosis.

Keywords

esophagus/foregut thoracic surgery

Key Takeaways

• Use of intraoperative indocyanine green fluorescence (IGF) imaging during esophagectomy is an emerging application of IGF, and its effect on patient outcomes is still under investigation.

• IGF results that were judged by operating surgeons to be abnormal were associated with significantly increased risk of anastomotic leak, suggesting that IGF results could help predict complications.

• Patients who underwent IGF imaging had a higher rate of anastomotic leak, so future studies should continue to evaluate the safety and physiologic effects of IGF during esophagectomy.

Introduction

Despite improving survival rates over the past several decades due to treatment innovations, esophageal cancer continues to have one of the lowest survival rates for all cancers at all stages with an overall 5 year survival of 20%.^1,2 The cornerstone of curative treatment for patients with locally advanced esophageal cancer is resection, usually combined with chemoradiation depending on the clinical stage of presentation. Currently, minimally invasive esophagectomy (MIE) is a widely accepted surgical approach.³ While esophagectomy improves overall survival, it carries its own morbidity and mortality risks.⁴

Anastomotic leak is a feared complication of esophagectomy and occurs in approximately 2–25% of patients following esophageal resection and anastomosis.^5‐9 Mortality associated with esophageal anastomotic leaks may be as high as 12%.⁷ Risk factors for anastomotic leakage include advanced age, diabetes, heart failure, hypertension, smoking, and renal disease.^4,10 Healing of the anastomotic region is related to adequate arterial perfusion and optimal tension of the anastomosis, both of which are surgically modifiable and serve to prevent leakage.¹¹

The use of intraoperative indocyanine green fluorescence (IGF) imaging is emerging as a means to assess the blood supply after anastomosis.¹² Indocyanine green is injected intravenously and then an infrared camera is used to visualize the blood supply in the form of fluorescence around the anastomosis.¹¹ While the use of IGF has been described previously in colorectal surgical cases, its use in esophagectomy cases is relatively new, and current evidence on whether anastomotic evaluation using IGF during MIE can predict esophageal anastomotic leak occurrence or improve outcomes is mixed.^8,11,13,14 In this study, we aimed to assess the predictive value of IGF and whether its use improves patient outcomes. We hypothesized that its use reduces anastomotic leaks and effectively predicts anastomotic complications after MIE. We found that while IGF evaluation of the anastomosis may be helpful in predicting anastomotic complications, its use was associated with a higher rate of anastomotic leaks.

Methods

Study Population

The study was approved by the Kaiser Permanente Northern California (KPNC) Institutional Review Board with a waiver of the requirement for informed consent. We reviewed all adult patients undergoing MIE at Kaiser Permanente Oakland between September 1, 2017 and November 15, 2021. In September 2019, Kaiser Permanente Oakland began routine use of IGF for intraoperative evaluation of the anastomosis during MIE. Therefore, the study period included all cases from approximately two years before and after this change in protocol.

Study Design

Patients were identified via diagnosis and procedure codes from KPNC electronic medical records (EMR). Baseline demographic characteristics such as age, sex, and race/ethnicity as well as clinical characteristics such as body mass index (BMI), Charlson Comorbidity Index (CCI), smoking history, prior chemotherapy, prior radiotherapy, intraoperative anastomotic rating, operative duration, cancer histology, and stage were extracted from EMRs and confirmed by manual chart review. The primary outcome of interest was anastomotic leak. Secondary outcomes included anastomotic stricture, time until liquid oral tolerance, hospital length of stay, requirement for repeat interventions, readmission, 30-day mortality, and 90-day mortality.

Baseline characteristics and outcomes were compared between patients who underwent intraoperative indocyanine green fluorescence (ICG group) and patients who did not (nICG group). Among patients who underwent indocyanine green fluorescence, we also compared subgroups of patients based on whether the operating surgeon determined that they had normal or abnormal fluorescence of the anastomosis at the time of surgery. Of note, the abnormal fluorescence group included anastomoses that were considered to have borderline or abnormal fluorescence.

Technique

Each MIE was performed by one of five thoracic surgeons within our institution with minimal variability in technique. Each patient underwent a minimally invasive Ivor Lewis esophagectomy with a stapled anastomosis as we have previously described in detail.⁹ Upon stapling of the intrathoracic anastomosis, indocyanine green is injected peripherally and subsequent fluorescence was visualized along the anastomosis. Upon inspection, the operating surgeon graded the anastomotic perfusion and recorded results in the operative note (Figure 1).

Figure 1.

Example of borderline indocyanine green fluorescence of the gastroesophageal anastomosis. Short arrow represents the anastomosis. Long arrow represents the gastric conduit.

Statistical Analysis

Baseline characteristics between ICG and nICG groups were assessed in bivariate analyses using Chi-square tests and Fisher exact tests for categorical variables and t-tests for normally distributed continuous variables. Two sample Wilcoxon rank sum tests were used for non-normally distributed continuous variables. All analyses were performed using SAS v9.4 (Cary, NC, USA).

Results

Overall, 181 patients were included in the four-year period. Of these, 59 patients (32.6%) were in the ICG group and 122 (67.4%) patients were in the nICG group. In the time period after IGF began to be routinely used within our institution, 19 out of 78 patients did not receive indocyanine green due to lack of availability of the necessary equipment on the day of the operation. ICG and nICG groups did not differ significantly in age, sex, BMI, Charlson Comorbidity Index, race/ethnicity, smoking history, operative duration, and cancer stage/histology. Within the ICG group, 52 patients (88.1%) had normal fluorescence and 7 patients (11.9%) had abnormal fluorescence as assessed by the surgeon intraoperatively. Of note, all of the patients in the abnormal group were considered to have borderline rather than truly abnormal fluorescence. None of these patients underwent anastomotic revision after IGF evaluation. Patients with abnormal fluorescence had significantly longer operative durations (P = .001). Baseline demographic and clinical features can be reviewed in Table 1.

Table 1.

Overall Cohort Demographics and Clinical Characteristics.

Overall cohort					Within ICG group
Characteristic, N (%)	Total(N = 181)	ICG(N = 59)	nICG(N = 122)	P value	Normal fluorescence (N = 52)	Abnormal fluorescence (N = 7)	P value
Age, years, mean±SD	65.2 ± 9.7	66.9 ± 8.6	64.4 ± 10.2	.105	67.2 ± 8.9	64.7 ± 6.2	.476
Sex
Male	144 (79.6)	46 (78.0)	98 (80.3)	.712	40 (76.9)	6 (85.7)	1.00
Female	37 (20.4)	13 (22.0)	24 (19.7)		12 (23.1)	1 (14.3)	1.00
Race/Ethnicity
White	126 (69.6)	41 (69.5)	85 (69.7)	.877	36 (69.2)	5 (71.4)	.482
African-american	6 (3.3)	3 (5.1)	3 (2.5)		2 (3.9)	1 (14.3)
Hispanic	17 (9.4)	5 (8.5)	12 (9.8)		4 (7.7)	1 (14.3)
Asian	24 (13.3)	7 (11.9)	17 (13.9)		7 (13.5)	0
Other	8 (4.4)	3 (5.1)	5 (4.1)		3 (5.8)	0
BMI, mean±SD	26.8 ± 5.1	25.9 ± 4.9	27.3 ± 5.2	.080	25.7 ± 5.0	27.2 ± 4.8	.300
Smoking history
Current	14 (7.7)	4 (6.8)	10 (8.2)		2 (3.8)	2 (28.6)	.054
Never	63 (34.8)	20 (33.9)	43 (35.2)	.915	19 (36.5)	1 (14.3)	.054
Former smoker	104 (57.5)	35 (59.3)	69 (56.6)		31 (59.6)	4 (57.1)	.972
CCI, mean±SD	6.8 (2.5)	6.5 (2.7)	6.9 (2.4)	.336	6.5 (2.7)	6.4 (2.8)	.972
Prior chemotherapy	166 (91.7)	52 (88.1)	114 (93.4)	.255	47 (90.4)	5 (71.4)	.190
Prior radiation	140 (77.3)	44 (74.6)	96 (78.7)	.536	11 (21.2)	4 (57.1)	.062
Operative duration, min, mean±SD	224.8 ± 70.4	214.7 ± 74.4	229.6 ± 68.1	.183	200.4 ± 59.9	321.1 ± 89.4	.001
Cancer stage
I	60 (33.1)	20 (33.9)	40 (32.8)		16 (30.8)	4 (57.1)	.535
II	42 (23.2)	12 (20.3)	30 (24.6)	.391	11 (21.2)	1 (14.3)
III	49 (27.1)	20 (33.9)	29 (23.8)		19 (36.5)	1 (14.3)
IV	30 (16.6)	7 (11.9)	23 (18.9)		6 (11.5)	1 (14.3)
Histology
Adenocarcinoma	156 (86.2)	49 (83.1)	107 (87.7)		45 (86.5)	4 (57.1)	.035
Squamous cell	23 (12.7)	9 (15.3)	14 (11.5)	.472	7 (13.5)	2 (28.6)
Other	2 (1.1)	1 (1.7)	1 (.8)		0 (0)	1 (14.3)

ICG, indocyanine green fluorescence evaluation; nICG, no indocyanine green fluorescence evaluation; SD, standard deviation; BMI, body mass index; CCI, Charlson Comorbidity Index; min, minutes.

ICG Fluorescence

Patients in the ICG group were found to have significantly higher rates of anastomotic leak (P = .015). There was no significant difference between ICG and nICG groups in rates of anastomotic stricture (P = .426). There were also no significant differences between ICG and nICG groups in time to oral liquid tolerance (P = .860), length of stay (P = .125), or repeat interventions during hospitalization (P = .781). Similarly, there was no significant difference in hospital readmission within 90 days (P = .286) or 30-day mortality (P = .105). However, patients in the ICG group had significantly higher rates of 90-day mortality than those in the nICG group (P = .038). These findings are presented in Table 2. None of the 19 patients in the nICG group who underwent MIE after the institution of IGF imaging developed anastomotic leaks.

Table 2.

Comparison of Minimally Invasive Esophagectomy Patient Outcomes with and without Indocyanine Green Fluorescence Evaluation.

Variable, N (%)	Total (N = 181)	ICG (N = 59)	nICG (N = 122)	p value
Anastomotic leak	8 (4.4)	6 (10.2)	2 (1.6)	.015
Anastomotic stricture	34 (18.8)	13 (22.0)	21 (17.2)	.426
Time to oral liquid tolerance, days, mean±SD	1.3 ± 1.3	1.4 ± 1.8	1.3 ± 0.9	.860
Length of index hospital stay, days, mean ±SD	3.8 ± 4.3	4.6 ± 5.3	3.5 ± 3.7	.125
Repeat intervention during hospital stay	16 (8.8)	6 (10.2)	10 (8.2)	.781
Hospital readmission within 90 Days	29 (16.0)	12 (20.3)	17 (13.9)	.286
30-Day mortality	2 (1.1)	2 (3.4)	0 (0)	.105
90-Day mortality	7 (3.9)	5 (8.5)	2 (1.6)	.038

ICG, indocyanine green fluorescence evaluation; nICG, no indocyanine green fluorescence evaluation; SD, standard deviation.

Of the seven patients who died within 90 days of surgery, three had developed anastomotic leaks. Two of the patients with leaks had developed bronchoesophageal fistulas. The other four patients died from aspiration pneumonia, congestive heart failure exacerbation, diabetic ketoacidosis, and complications from metastatic disease.

Intraoperative Anastomosis Classification

Patients with abnormal fluorescence had significantly higher rates of anastomotic leak (P < .001), repeat interventions during hospitalization (P = .018), and 30-day mortality (P = .012). Patients with abnormal fluorescence also had significantly longer times to oral tolerance (P = .002) and length of stay (P = .012). There was no significant difference in rates of anastomotic stricture (P = .330), hospital readmission within 90 days (P = .141) or 90-day mortality (P = .102) between patients with normal fluorescence vs. abnormal fluorescence. These findings are illustrated in Table 3.

Table 3.

Normal ICG Fluorescence vs. Abnormal ICG Fluorescence Outcomes.

Variable, N (%)	ICG group total(N = 59)	Normal fluorescence(N = 52)	Abnormal fluorescence (N = 7)	P value
Anastomotic leak	6 (10.2)	1 (1.9)	5 (71.4)	<.001
Anastomotic stricture	13 (22.0)	13 (25.0)	0 (0)	.330
Time to oral liquid tolerance, days, mean SD	1.4 ± 1.8	1.1 ± 0.6	3.8 ± 5.1	.002
Length of index hospital stay, days, mean SD	4.6 ± 5.3	3.9 ± 4.3	9.7 ± 9.2	.012
Repeat intervention during hospital stay	6 (10.2)	3 (5.8)	3 (28.6)	.018
Hospital readmission within 90 Days	12 (20.3)	9 (17.3)	3 (42.9)	.141
30-Day mortality	2 (3.4)	0 (0)	2 (28.6)	.012
90-Day mortality	5 (8.5)	3 (5.8)	2 (28.6)	.102

ICG, indocyanine green fluorescence evaluation; SD, standard deviation.

Discussion

Although patients who underwent IGF evaluation of the anastomosis during MIE had higher rates of anastomotic leak, the degree of fluorescence in patients who received IGF was predictive of outcomes. Though somewhat paradoxical, these results highlighted that IGF may be helpful in predicting which esophagectomy cases may result in an anastomotic leak. This study adds to the limited literature on the use of IGF in esophagectomy procedures, which is a relatively new and emerging application of indocyanine green.

Those who were deemed to have abnormal IGF by the operating surgeon had significantly increased rates of anastomotic leak, repeat interventions during hospitalization, and 30-day mortality. These patients with abnormal IGF also had longer operative times, though this correlation likely results from the fact that cases with abnormal fluorescence may be more difficult and may carry higher risks of complications, including anastomotic leak and conduit necrosis.^15,16 Length of stay was not noted to be statistically different between the two groups, though, the mean length of stay in the nICG group of 3.5 days was closer to our prior data (3.6 days).⁹ As previously described, our patients’ length of stay is largely facilitated by our standardized post-operative protocol.⁹ This involves early oral nutrition and ambulation with eligibility for discharge on post-operative day three assuming the patient meets several discharge criteria such as absence of leak on esophagram, lack of tachycardia, absence of fever, and ability to sustain hydration and nutrition with over two liters of fluid orally per day.⁹

The correlation between abnormal IGF and significantly increased rates of anastomotic leak suggested that abnormal IGF may be helpful in predicting which cases may be complicated by anastomotic leak and could potentially be managed in a different fashion than a standard esophagectomy case with normal IGF intraoperatively. Previous studies have shown that IGF results can predict increased risk of leakage in colorectal surgical cases; however, the use of IGF and its impact on outcomes after esophagectomy has not been as well described and has yielded mixed results.^{8,11,13,14,17} While some studies suggest that IGF can be used to predict which areas have adequate perfusion, others have found that IGF evaluation did not lead to reduced anastomotic leakage rates.^11,14 Therefore, the utility of IGF in esophagectomy remains unclear.

The correlation between abnormal IGF and increased complications suggests that surgeon interpretation of IGF results may effectively assess whether anastomotic regions have “adequate” or “inadequate” perfusion. Therefore, future studies should continue to evaluate whether changes in management in response to abnormal IGF improve patient outcomes.

Pertinently, the intraoperative course for all patients in this study was unchanged in response to abnormal IGF results, and no revisions to the anastomotic site were performed in response to abnormal IGF results. This was based on surgeon discretion, as the severity of the abnormality in these patients was not deemed significant enough to cause a change in the operative course. In each of these cases, the fluorescence was considered to be borderline rather than frankly abnormal. These cases were also likely technically more difficult with higher inherent risk of anastomotic leak as evidenced by their longer durations. However, given the higher anastomotic leak rate among patients with even borderline fluorescence, these cases may be candidates for intraoperative revision if feasible.

Whether or not the anastomosis is revised, understanding which patients are at particularly high risk of anastomotic leak may still prove beneficial as it can heighten a surgeon’s degree of suspicion for anastomotic leak in the post-operative period. This can potentially lead to faster identification and intervention for patients who do develop leaks.

The higher rates of anastomotic leak among ICG patients compared to nICG patients were somewhat surprising and raise some concerns. In general, the ICG group in our study is a more recent cohort that was compared to a historical cohort, so it is possible that an unidentified confounding factor may have changed at the same time of implementation of indocyanine green use. However, 19 patients in this later period did not undergo IGF imaging simply due to an unexpected lack of equipment availability at the time of surgery. None of the patients in this unintentional subgroup of the later temporal cohort developed leaks, which raises concern regarding the use of indocyanine green. It is also worth emphasizing that age, BMI, Charlson Comorbidity Index, smoking history, operative duration, stage, and histology were not different between the ICG and nICG groups. While prior studies on IGF imaging in esophagectomy have suggested that it is generally safe, studies assessing its physiology specifically related to esophageal tissue are lacking.⁸ Given the retrospective nature of our study, we are unable to definitively identify a cause for an increase in anastomotic leaks among the ICG group. Overall, the anastomotic leak rate of 10.2% in the ICG group is similar to the NSQIP data, and part of the reason we may have noticed an increase in anastomotic leak rate in our ICG group is that, historically, our leak rate has been lower at 2.1%.⁹ A prospective, randomized trial may be able to shed further light on any possible detrimental effect of indocyanine green on the gastroesophageal anastomosis. Additionally, further assessment of indocyanine green in terms of vasoactivity or toxicity within esophageal tissue may help identify potential concerns with its use. Further investigation is warranted because indocyanine green could be a helpful adjunct in identifying high risk anastomoses but only if its use does not cause harm.

Our study results should be interpreted in the context of its limitations. First, the lack of randomization of our patient groups affected our ability to limit confounders. Additionally, the evaluation of IGF results was inherently subjective. However, the success among our surgeons in interpreting normal vs. abnormal fluorescence was evident in the adverse outcomes associated with patients deemed to have abnormal fluorescence at the anastomosis.

Overall, while our study showed the value of IGF imaging in terms of predicting anastomotic complications, our findings of increased risk of anastomotic leak upon its use were concerning. Further studies are warranted to further elucidate its safety and its physiologic effects in the setting of esophageal surgery.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Kaiser Permanente Northern California Graduate Medical Education Program and Kaiser Foundation Hospitals.

ORCID iD

Kian Banks

References

Cancer of the Esophagus - Cancer Stat Facts . SEER, 2022. https://seer.cancer.gov/statfacts/html/esoph.html

Siegel

Miller

Fuchs

Jemal

. Cancer statistics, 2022. CA A Cancer J Clin. 2022;72(1):7-33. doi:10.3322/caac.21708

Kang

. Current issues in minimally invasive esophagectomy. Korean J Thorac Cardiovasc Surg. 2020;53(4):152-159. doi:10.5090/kjtcs.2020.53.4.152

Kassis

Kosinski

Ross

Koppes

Donahue

Daniel

. Predictors of anastomotic leak after esophagectomy: an analysis of the society of thoracic surgeons general thoracic database. Ann Thorac Surg. 2013;96(6):1919-1926. doi:10.1016/j.athoracsur.2013.07.119

Urschel

. Esophagogastrostomy anastomotic leaks complicating esophagectomy: A review. Am J Surg. 1995;169(6):634-640. doi:10.1016/S0002-9610(99)80238-4

Alanezi

Urschel

. Mortality secondary to esophageal anastomotic leak. Ann Thorac Cardiovasc Surg Off J Assoc Thorac Cardiovasc Surg Asia. 2004;10(2):71-75.

Biere

SSAY

Maas

Cuesta

van der Peet

. Cervical or thoracic anastomosis after esophagectomy for cancer: a systematic review and meta-analysis. Dig Surg. 2011;28(1):29-35. doi:10.1159/000322014

Slooter

Eshuis

Cuesta

Gisbertz

van Berge Henegouwen

. Fluorescent imaging using indocyanine green during esophagectomy to prevent surgical morbidity: a systematic review and meta-analysis. J Thorac Dis. 2019;11(suppl 5):S755-S765. doi:10.21037/jtd.2019.01.30

Ashiku

Patel

Horton

Velotta

Ely

Avins

. A refined procedure for esophageal resection using a full minimally invasive approach. J Cardiothorac Surg. 2022;17(1):29. doi:10.1186/s13019-022-01765-2

10.

Manghelli

Ceppa

Greenberg

, et al.

Management of anastomotic leaks following esophagectomy: when to intervene?

J Thorac Dis. 2019;11(1):131-137. doi:10.21037/jtd.2018.12.13

11.

Karampinis

Ronellenfitsch

Mertens

, et al. Indocyanine green tissue angiography affects anastomotic leakage after esophagectomy. A retrospective, case-control study. Int J Surg. 2017;48:210-214. doi:10.1016/j.ijsu.2017.11.001

12.

Unno

Suzuki

Yamamoto

, et al. Indocyanine green fluorescence angiography for intraoperative assessment of blood flow: A feasibility study. Eur J Vasc Endovasc Surg Off J Eur Soc Vasc Surg. 2008;35(2):205-207. doi:10.1016/j.ejvs.2007.09.001

13.

Ladak

Dang

Switzer

, et al. Indocyanine green for the prevention of anastomotic leaks following esophagectomy: a meta-analysis. Surg Endosc. 2019;33(2):384-394. doi:10.1007/s00464-018-6503-7

14.

Casas

Angeramo

Bras Harriott

Dreifuss

Schlottmann

. Indocyanine green (ICG) fluorescence imaging for prevention of anastomotic leak in totally minimally invasive Ivor Lewis esophagectomy: a systematic review and meta-analysis. Dis Esophagus Off J Int Soc Dis Esophagus. 2022;35(4):doab056. doi:10.1093/dote/doab056

15.

Slooter

de Bruin

Eshuis

, et al. Quantitative fluorescence-guided perfusion assessment of the gastric conduit to predict anastomotic complications after esophagectomy. Dis Esophagus. 2020;34(5):. doi:10.1093/dote/doaa100

16.

Grimminger

Goense

Gockel

, et al. Diagnosis, assessment, and management of surgical complications following esophagectomy. Ann N Y Acad Sci. 2018;1434(1):254-273. doi:10.1111/nyas.13920

17.

Kawada

Hasegawa

Wada

, et al. Evaluation of intestinal perfusion by ICG fluorescence imaging in laparoscopic colorectal surgery with DST anastomosis. Surg Endosc. 2017;31(3):1061-1069. doi:10.1007/s00464-016-5064-x