Introduction

Colorectal surgery is associated with 35% morbidity in some series, and anastomotic leak (AL) is one of the most feared complications, with reported rates between 2 and 24%^1,2. Not only does it carry an immediate risk due to high rate of reoperation, but it can also worsen the oncological prognosis³. Multiple factors have been associated with the development of AL, such as male sex, age, comorbidities, ASA III-IV, alcohol abuse, malnutrition, obesity, smoking, tumor stage, diverticulitis, low rectal anastomosis, prolonged operative time, poor anastomotic perfusion, postoperative transfusions, and perioperative sepsis⁴. Morbidity, mortality and costs resulting from AL can be minimized with prevention strategies.

Well-perfused, non-rotated, tension-free anastomoses prevent AL, but their assessment is based on the subjective view of the attending surgeon. Indocyanine green (ICG) fluorescent angiography (FA) allows the assessment of intestinal perfusion and anastomosis with low inter-observer variability⁵. We found no reports of the use of FA in the Argentine population.

The primary aim of this paper was to assess the intestinal perfusion with FA during colonic surgery, and to quantify its impact in the change of the surgical strategy on the site of resection and anastomosis. The secondary aim was recording clinical or radiological anastomotic leaks and their association with the utilization of previous FA.

Material and methods

We prospectively included All ASA I-III patients undergoing scheduled colonic resection with anastomosis confection performed with ICG-FA assistance between January 30, 2016 and June 30, 2018 in the Division of Gastroenterology Surgery, Coloproctology Section, of the Hospital de Clínicas “José de San Martín”, University of Buenos Aires. All participants consented to be included in the study, and the protocol was approved by the hospital’s Ethics Committee.

Surgeries were performed following the usual indications of the division, and the operating approach was not induced by the use of FA. Laparoscopy and conventional approaches were followed. Converted laparoscopy was also performed. The data recorded in each case were age, sex, comorbidities, initial diagnosis, definitive diagnosis, type of surgery, and type of anastomosis performed.

Fluorescent angiography was carried out using a special tower for fluorescence (Karl Storz, Tuttlingen, Germany) which has a xenon white light source that alternates with a near infrared light source of 860 nm, laparoscopic optics or VITOM® 3D system for open surgery, specially designed for fluorescence (Karl Storz, Tuttlingen, Germany). Once intestinal mobilization and mesocolon ligation were performed, and the site for intestinal resection and anastomosis were chosen, the site was marked according to standard surgical criteria for the vitality of the intestinal segment: dye, bleeding at the mucosal end, mobility) and the evaluation of the attending surgeon. This surgical moment, called ‘basal’, was documented with a digital image capture (AIDA™, Karl Storz, Tuttlingen, Germany). Next, an intravenous bolus injection of 3 ml of indocyanine green (5 mg/ml, Verdye, Diagnostic Green, Aschheim- Dornach, Germany) diluted in 7 ml of saline solution was administered, and then washed with additional 20 ml of saline solution. At the time of injection, the operating room lights were dimmed, and the light source was changed to infrared light. FA perfusion of the intestinal loop was assessed at the site of resection (basal); this second moment -assessing the same segment as in the basal moment, after indocyanine green injection- was called “perfusion of the intestinal tract” and was digitally captured. A second ICG injection was administered upon completion of the anastomosis to check its perfusion; this third moment was called “perfusion of the anastomosis” and was digitally documented. Regarding the use of FA, the impossibility to perform it, additional operating time to perform it, number of ICG injections, intestinal perfusion classified as unidentified, adequate, borderline, or deficient, change of operating strategy based on FA results (basal resection site different from the actual intestinal resection site) were recorded. The additional operating time was calculated by measuring the minutes elapsed after preparing the intestinal tract to be resected or anastomosed during white-light assessment, until completion of the infrared light use for perfusion assessment with fluorescence. The following categories for intestinal perfusion were defined: unidentified, it is not possible to detect fluorescence in the operative field; adequate, high intensity fluorescence signal and equivalent to the control in the chosen resection site (basal); low intensity fluorescence signal -lower than the control- in the basal site; and deficient, absence of fluorescence signal in the basal site. The intensity of the fluorescence signal was compared between the intestinal loops not involved in the resection (normal perfusion, control) and the signal at the basal site. Postoperative anastomotic leaks with clinical manifestations were detected⁶. The radiological study was performed in those patients clinically suspected of AL, using contrast-computed tomography for AL targeting.

Results

The results of 37 patients were presented; mean age was 68.7 years (SD 12), and mean BMI was 28.7 (SD 11.7); 17 (45%) were male, 11 patients (64%) showed associated comorbidities, 16 (43.2%) were ASA I and 21 (56.76%) ASA II. Thirty-five (94.6%) patients underwent colorectal cancer surgery. Nine anterior resections, 2 subtotal, 16 left and 9 right colectomies, and 1 transit reconstruction were performed. Real-time assessment of colonic perfusion was achieved in all patients (37/37, 100%). The additional operating time needed to perform FA was 3-9 minutes (average, 5.8 minutes). Perfusion site after the anastomoses was adequate in all cases (100%) (Figure 1). The anastomosis site was changed in 4 patients (10.8%) due to poor perfusion of that site or to adequate perfusion in a site that appeared ischemic under white-light assessment (Figures 2 & 3) (Table 1). In one right colectomy, an additional 5 cm of the transverse colon had to be resected; in a left colectomy, a rectal stump was preserved, which appeared hypoperfused under the assessment with white light and presented optimal perfusion with FA; while in 2 cases with anterior resection, it was necessary to change to proximal in 2 and 3 cm the resection site in the left colon. Postoperative complications included anastomotic leaks in 2 patients (5.41%). One was a 45-year-old male patient with a BMI of 31 and a synchronous colon tumor, who underwent subtotal colectomy with ileorectal mechanical anastomosis; the second one was a 72-year-old female patient with a BMI of 26 and a manual ileostransverse anastomosis, who underwent right colectomy for cancer. None of the cases with AL showed poor perfusion of the anastomotic tracts, or of the anastomosis once it was performed.

Figure 1 Determining the colonic resection site during a laparoscopic sigmoidectomy. A. White-light assessment during laparoscopy, after mesocolon ligation; arrow shows the proximal perfused area. B. Infrared light assessment, and indocyanine-green intravenous injection. The perfused colonic wall emits a bluish-violet fluorescence (outlined by the white solid line), while the nonperfused ischemic segment emits no light and is not observed in the image (black area, outlined by dotted line). C. Black and white image where fluorescent emission is identified as white tones with an overlapping pattern of the perfused (striped) and nonperfused (grid) areas. 

Figure 2 Transverse colon during conventional right colectomy. A. White-light assessment with adequate dye, and bleeding from the edge of the tract; the arrow indicates the proposed anastomotic site. B. During fluorescent angiography, distribution of indocyanine green in patches was tar geted at the level of the mesocolon and the colonic wall, showing deficient perfusion. The resection site was then changed distally by 5 cm to a homogeneous perfusion area. Perfused areas were outlined with solid lines, and ischemic areas with dotted lines. C. Black and white image where fluorescent emission is identified as white tones with an overlapping pattern of the perfused (striped) and nonperfused (grid) area. 

Figure 3 Rectal end during conventional left colectomy. A. White-light assessment with inadequate dye to determine vitality; the arrow indicates the proposed resection site. B. During fluorescent angiography, intramural vessels (arrow) and green light emission of the wall in the rectal stump can be observed, indicating adequate stump perfusion. Perfused areas were outlined with solid lines, and ischemic areas with dotted lines. C. Black and white image where fluorescent emission is identified as white tones with an overlapping pattern of the perfused (striped) and nonperfused (grid) area. 

Table 1 Resultados 

Discussion

As experience was gained and the use of FA in surgery spread, it was proposed as a technique to assess intestinal perfusion during colorectal surgery with less subjectivity when performing a well-perfused anastomosis. A history of diabetes mellitus, anticoagulation therapy, preoperative chemotherapy and operating time were significantly associated with poor intestinal perfusion⁷.

Perfusion assessment with FA was technically possible for us (100% of cases), and added a few minutes of extra operating time (6 minutes) to the total procedure time. Therefore, we believe that using FA is feasible and does not cause a substantial impact on the organization of the operating room. With proper training, operating the fluorescence system is intuitive for the surgeon. Another FA advantage is getting information on intestinal perfusion in real time; the assessment takes place at the same time as the study is performed, and allows intraoperative decisions to be made.

In a multicenter study of 139 patients undergoing colectomy, FA achieved perfusion assessment in all cases with a change in the surgical plan in 7.9% of the surgeries⁸. In a specialized center, Boni et al reported 107 FA colectomies during laparoscopic surgery with a 4% change in the strategy⁹. Kawada reported the use of FA in laparoscopy with colorectal anastomosis. FA resulted in a proximal change of the transection line by 5 mm in 26.5 % of the patients, and by 50 mm in 4.4 % of the patients⁷. In our series, perfusion assessment with fluorescence was possible in all cases, in line with the authors above. Moreover, FA was not only possible, but it also influenced the surgical strategy in a significant number of cases, modifying the approach to the intestine anastomosis in 10.8% of the patients.

Another element to discuss is whether use and change of resection site result in decreased incidence of AL. In a series conducted in a single center with recent and comparable historical control, Boni found no significant difference with FA in 42 patients; it is noteworthy, however, that there were 2 patients with AL in the control group, but the resection site was changed in the FA group and there were no cases of AL¹⁰. In this group, 4.7% of the resection site changes resulted in the absence of leaks. Kudszus reported a 3.5% AL risk reduction in the FA group, as opposed to 7.5% in the control group without FA¹¹. A systematic review on the value of the analysis of anastomotic perfusion with fluorescence highlighted the difference between FA groups and the control groups, the average AL being 3.83% and 7.58%, respectively¹². A subject-matched study showed a 6% reduction in the incidence of AL with the use of FA, being the rate of AL in the FA group 8.8%, and 14.7% in the control group¹³. A systematic review and meta-analysis concluded that the use of FA reduces the risk of AL with an odds ratio of 0.51¹⁴. It follows from this analysis that the value of FA increases in oncological resection, and left colon -particularly rectum- where AL reduction would be 81% when ICG-FA is used (OR 0.19). Another meta-analysis that only included colorectal cancer surgeries also showed reduced risk for AL in patients undergoing ICG-FA, changing the anastomotic site in 4.7-19% of the cases, with an incidence of AL being lower, statistically significant, with no bias¹⁵. This meta-analysis concludes that ICG-FA is associated with a lower AL rate after colorectal resection. Those results were validated by a third meta-analysis, which confirmed the results of the previous analyses with a risk reduction of 12% for AL. [15] A recent randomized trial evaluating the reduction of AL rate in colorectal surgery showed no difference between the treatment groups¹⁶. The trial involved 252 patients randomized 1:1 to white-light perfusion assessment or to ICG-FA. The primary aim was to assess whether ICG-FA could lead to a reduction in AL rate. Secondary outcomes were possible changes in the surgical strategy. A change in surgical strategy was reported in 13 patients (11%): 11 (9%) AL in the white-light group and 6 (5%) in the FA group; this difference was not statistically significant.

Our series showed an AL rate of 5.41% (2 cases), a value that compares to those from other series, such as the ones mentioned above. Considering that the aim of this study was to assess the technical possibility and its impact on the change of surgical strategy, no control group was defined in terms of the incidence of AL without using FA. However, we believe it is important to discuss this point because it is the major impact factor of FA.

While FA offers us the possibility of real-time assessment of adequate intestinal perfusion, it presents a degree of subjectivity when assessing the intensity of the fluorescence signal on the video screen. On the other hand, there are variations in the colors and the intensity of the video signal depending on the manufacturers; there are also technical factors such as the ICG dose, ICG injection or perfusion speed, distance from the source and camera to the intestine, quality of the fiber optic cable, intensity of the infrared light and characteristics of the patients -such as blood pressure, arteriosclerosis and BMI- that may influence the intensity of the signal. To solve this issue, systems for quantifying fluorescence are under development. Fluorescence emission curves at the level of the intestinal wall have been described¹⁷. These curves would classify patients into risk groups for anastomosis and eliminate the subjectivity of the video assessment by replacing it with standardized numerical values.

Some disadvantages in our report include the number of participants, the heterogeneity of surgeries performed, and the lack of a control group to compare the incidence of AL. Strengths include being a prospective report and the first in the national literature.

Conclusions

Indocyanine-green fluorescent angiography in colorectal surgery was feasible, adding minimal complexity and time to the traditional procedure and resulting in change of strategy for the intestinal resection in 11% of cases, with a low rate of anastomotic leak.