ORIGINAL ARTICLE

ANZJSurg.com

Does indocyanine green improve the evaluation of perfusion during

laparoscopic colorectal surgery with extracorporeal anastomosis?
Kenneth Buxey , Francis Lam, Mark Muhlmann and Shing Wong
Department of Colorectal Surgery, Prince of Wales Hospital, Sydney, New South Wales, Australia

Key words Abstract

anastomotic perfusion, colorectal surgery, general
surgery, indocyanine green.
Correspondence
Dr Kenneth Buxey, Department of Colorectal
Surgery, Prince of Wales Hospital, Barker Street,
Randwick, NSW 2031, Australia.
Email: knbuxey@gmail.com
K. Buxey MBBS (First Class Honours), FRACS;
F. Lam MBBS, PhD, FRACS; M. Muhlmann
MBBS, FRACS; S. Wong MBBS, MSc, FRACS.
Accepted for publication 12 May 2019.
doi: 10.1111/ans.15320
Background: There has been recent interest in indocyanine green (ICG) to assess anasto-
motic perfusion in colorectal surgery. We describe our experience using ICG when per-
forming laparoscopic segmental colorectal resections with extracorporeal anastomotic
technique and a highly standardized approach for clinically assessing blood flow.
Methods: We recruited 20 consecutive patients to undergo segmental laparoscopic re-
section and determined an appropriate point to transect mesentery proximally confirming
pulsatile arterial flow at this level. Once confirmed, we did a further perfusion study using
ICG to ascertain if this would change intraoperative decision-making.
Results: Twenty segmental colonic resections were assessed in nine female and 11 male
patients aged 26–91 years. ICG administration was safe with no adverse outcomes docu-
mented. ICG demonstrated anastomotic perfusion in all cases. We observed no cases where-
with pulsatile blood flow at the cut edge of the mesentery, ICG showed inadequate
perfusion at this level. We did find in 25% of cases ICG showed perfusion beyond the cut
edge of the mesentery to a distance of up to 2.5 cm.
Conclusion: ICG perfusion is safe and straightforward to carry out. However, when pulsa-
tile arterial bleeding is demonstrated clinically it does not add anything to assessment of per-
fusion in our study. Furthermore, in 25% of cases perfusion can be demonstrated beyond
the cut edge of the mesentery up to a distance of 2.5 cm. This raises the possibility that an
organ well perfused with ICG may have less than ideal blood flow when assessing for this
with a view to constructing an anastomosis.

perform this important intraoperative assessment. Existing publica-

Introduction
Indocyanine green (ICG) has recently gained interest within the
specialty of colorectal surgery as an adjunct to assist in various
aspects of resectional surgery. Initially described as a dye used in
photography it has had medical application since the 1960s, in the
assessment of hepatic function1 and subsequently in the measure-
ment of renal blood flow2 and in ophthalmology as a contrast agent
to perform fluorescent angiography of the choroidal circulation of
the eye.3 More recently interest has centred on its application as a
contrast agent used to enhance or augment minimally invasive sur-
gery with the application of inbuilt infrared imaging systems with
laparoscopic and robotic operating platforms. It is in this context
that there been interest in its application in colorectal surgery. There
have been publications describing the application of ICG to assess
anastomotic perfusion when performing colorectal surgery and
there is interest in the concept this may improve our ability to
tions, however, while reporting on their experience with ICG and
its comparison to standard assessment of anastomotic perfusion, do
not describe or define what a standard assessment actually is.4–6
This makes it difficult to draw conclusions about what additional
value an ICG assessment of an anastomosis offers in resectional
colorectal surgery. We describe our initial experience assessing
anastomotic perfusion with ICG in comparison to using our own
institutions highly standardized method of anastomotic perfusion
assessment with a view to determining if the ICG study is reproduc-
ible and importantly whether it is equivalent or indeed superior to
the standard assessment.
Methods
Institutional approval and local health district ethics approval was
gained for the intravenous use of Verdye (Diagnostic Green

© 2019 Royal Australasian College of Surgeons ANZ J Surg (2019)

2 Buxey et al.
Pharmaceuticals, Aschheim, Germany, ICG dye for intravenous
use). A Stryker (Kalamazoo, MI, USA) 1588 AIM laparoscopic
imaging system which contains a built-in near infra-red camera was
utilized to visualize ICG fluorescence. Twenty consecutive patients
were enrolled to assess the application of ICG for assessment of
anastomotic perfusion. The exclusion criteria were:
• Known allergy to ICG
• Receiving medications that interact with the ICG such as anti-
convulsants, bisulphite containing drugs, nitrofurantoin and
methadone
• Patients with significant hepatic dysfunction (MELD
score >10 or renal dysfunction eGFR <40).
All were elective colorectal resections with a restorative anasto-
mosis. Our institution uses a standard technique for these resec-
tions. This is as follows:
• Right hemicolectomy: laparoscopic mobilization of the right
colon and intracoporeal ligation of the ileocolic pedicle with
specimen exteriorization via a small midline wound. Division
of the ileal and colonic mesentery extracorporeally. The
colonic mesentery is divided perpendicular to the colon at least
5 cm from any tumour and the cut edge is assessed for pulsa-
tile arterial bleeding which is then controlled with sutures. In
the absence of pulsatile bleeding, the point of transection is
taken back until this is achieved. Intestinal continuity is
achieved by means of a stapled side to side anti-peristaltic
anastomosis reinforced with sutures.
• Left hemicolectomy: laparoscopic mobilization of the colon
based on the left colic artery, which is ligated intracorporeally.
Subsequently the colonic mesentery is divided proximally and
distally perpendicular to the colon and the cut edge is assessed
for pulsatile arterial bleeding proximally and distally. If no
pulsatile blood flow is demonstrated, further resection is
performed.
• Anterior resection: the inferior mesenteric artery is divided high
intracorporeally. The inferior mesenteric vein (IMV) is ligated
high as indicated and splenic flexure mobilization is performed
as needed to achieve a tension free anastomosis. For high ante-
rior resection, an end to end double stapled anastomosis is per-
formed. For low and ultralow anterior resection, a colonic J
pouch is performed. In all cases, distal division is performed
intracorporeally and the colonic conduit then exteriorized. The
mesentery is prepared as for the previously described resections
and only pulsatile arterial bleeding is accepted and in its absence
a more proximal point of transection is chosen. Diverting
ileostomies are performed in all cases of complete total mes-
orectal excision and in cases of extraperitoneal partial mes-
orectal excision where chemoradiation preoperative has been
administered.
ICG was applied as follows: following transection of the mesen-
tery and confirmation of pulsatile bleeding, intravenous ICG was
administered at a dose of 0.1–0.2 mg/kg up to a maximum of
10 mg by reconstituting 25 mg of ICG (Verdye) into a volume of
10 mL with 0.9% saline. An assessment of the point of angio-
graphic ischaemic demarcation was made and compared with the
point of mesenteric pulsatile bleeding. Early in our experience, we
attempted to simply assess perfusion post sectioning of the main
vascular pedicle but prior to division of the mesentery, so as to
allow the ICG to determine where to divide the mesentery. How-
ever, our experience of this was that it is difficult to determine an
exact point of ischaemic demarcation using this technique with typ-
ically no clear demarcation point and areas of relative ischaemia
still giving positive ICG signal. Furthermore, once the mesentery is
sectioned the tissues retain the ICG for extended periods of time
meaning the test cannot be reliably repeated to then determine
ischaemic demarcation precisely. We therefore abandoned this
approach for the above described technique. Upon returning the
bowel intracorporeally and laparoscopic assessment of perfusion
was made including an assessment of stump perfusion if an anterior
resection was performed. Following anastomosis, a further assess-
ment of anastomotic perfusion was made which for left sided resec-
tions involved administering a second dose of ICG.
We recorded the point of ischaemic demarcation demonstrated on
ICG testing and its relationship to the point of pulsatile bleeding. We
documented if the ICG testing affected the final operative strategy.
We also noted stump perfusion and whether the assessment of this
changed our intraoperative strategy. Post-operatively we recorded
whether the operation was complicated by anastomotic leak.
Results
We evaluated anastomotic perfusion using ICG in 20 consecutive
colectomies performed laparoscopically with an extracorporeal
anastomosis. Patients from 26 to 91 years of age with a mean age
of 66; there were 11 male patients and nine female patients. Resec-
tions performed were 10 right hemicolectomies including one sin-
gle incision laparoscopic resection, five high anterior resections,
two low anterior resection, two ultralow anterior resections and one
total colectomy with ileorectal anastomosis. Three patients received
protective loop ileostomies. A summary of the indications for the
resections and in the case of left sided resections whether the infe-
rior mesenteric vein was high ligated and whether the splenic flex-
ure was mobilized as shown in Table 1. In all cases, ICG was given
and this was able to achieve appropriate fluorescence of the bowel
to allow an assessment of perfusion of the bowel. An example of
this is shown in Figure 1. There were no recorded adverse reactions
to the administration of ICG. We had no cases where the bowel did
not perfuse to at least the level of the cut edge of the mesentery
where we had already demonstrated pulsatile arterial blood flow,
supporting the concept that when one has achieved this, one has
achieved adequate bowel perfusion. However, in five cases (25% of
the time) we noted demarcation of the ICG fluorescence beyond the
point of the mesenteric cut edge by distances ranging from 1 to
2.5 cm. This did not change our management intraoperatively as
we still transected the bowel at the level of the cut mesenteric edge
but in the setting of relying only on an ICG assessment this finding
needs careful consideration and potentially effects how one inter-
rupts the ICG study. In all cases, the distal bowel perfused normally
with ICG. There were no cases of proven anastomotic leak. One
patient who had a total colectomy and ileorectal anastomosis devel-
oped prolonged ileus post-operatively. All other patients had
uncomplicated recoveries and remained well on subsequent
follow up.
© 2019 Royal Australasian College of Surgeons
 Table 1 Assessment of perfusion with indocyanine green

Case Sex Age Resection Indication IMV Splenic Colonic demarcation Stump/ileal Anastomotic Case Leak DLIO
number flexure perfusion perfusion influenced
by ICG

© 2019 Royal Australasian College of Surgeons

Indocyanine green use in colorectal surgery

1 M 40 HAR Diverticulitis No No At mesenteric division Normal Normal No No No

2 M 57 HAR Sigmoid cancer No No At mesenteric division Normal Normal No No No
3 F 49 LAR Rectal cancer Yes Yes At mesenteric division Normal Normal No No Yes
4 M 26 RHC Ceacal cancer NA NA 1 cm beyond Normal Normal No† No NA
5 M 86 RHC Ascending colon cancer NA NA At mesenteric division Normal Normal No No NA
6 F 63 TAC/IRA Synchronous cancer Yes Yes NA Normal Normal No † No
and polyposis
7 M 67 RHC Ascending colon cancer NA NA 1.5 cm beyond Normal Normal No† No NA
8 M 70 LAR Rectal cancer Yes Yes 2 cm beyond Normal Normal No† No No
9 M 78 RHC Hepatic flexure cancer NA NA At mesenteric division Normal Normal No No NA
10 M 63 ULAR Low rectal cancer post Yes Yes 2 cm beyond Normal Normal No† No Yes
chemoRxTx
11 F 50 HAR Sigmoid cancer No No At mesenteric division Normal Normal No No No
12 M 90 RHC Cecal cancer NA NA At mesenteric division Normal Normal No No NA
13 F 91 RHC Hepatic flexure cancer NA NA 1.5 cm beyond Normal Normal No† No NA
14 F 68 SILS RHC Cecal polyp NA NA At mesenteric division Normal Normal No No NA
15 F 73 RHC Ileal cancer NA NA At mesenteric division Normal Normal No No NA
16 M 80 ULAR Low rectal cancer post Yes Yes 2 cm beyond Normal Normal No No Yes
chemoRxTx
17 M 65 HAR Sigmoid cancer No No At mesenteric division Normal Normal No No No
18 F 77 RHC Cecal cancer NA NA At mesenteric division Normal Normal No No NA
19 F 61 RHC Cecal cancer NA NA At mesenteric division Normal Normal No No NA
20 F 57 HAR Sigmoid cancer Yes Yes 2.5 cm beyond Normal Normal No† No No

†Cases where perfusion was demonstrated past point of mesenteric transection. DLIO, diverting loop ileostomy; HAR, high anterior resection; ICG, indocyanine green; IMV, inferior mesenteric vein; IRA, ileorectal
anastomosis; LAR, low anterior resection; NA, not applicable; RHC, right hemicolectomy; SILS, single incision laparoscopic surgery; TAC, total abdominal colectomy; ULAR, ultralow anterior resection.
3
4 Buxey et al.
Fig. 1. Indocyanine green perfusion assessment.
Discussion
We have demonstrated that using intravenous ICG to assess bowel
perfusion when performing resectional colorectal surgery is
straightforward and appears very safe. What is less clear is firstly
whether it provides any additional information when performing
extracorporeal anastomosis, and secondly how to interpret the find-
ing when solely relying on it to assess a completely intracorporeal
anastomosis. We did not have any cases where the bowel failed to
perfuse with ICG to at least the level of the cut edge of the mesen-
tery. This seems to contradict some of the other published studies
looking at ICG where the bowel perfused not to the point of the
planned mesenteric transection as determined clinically. However,
in those studies the method used to determine this point is not well
described and it is unclear if pulsatile arterial bleeding was sought.
The importance of observing pulsatile flow has been emphasized
since the time of Lockhart-Mummery7 and it has been previously
described that performing an anastomosis and determining pulsatile
marginal blood flow does reduce leak rate and the absence of this
phenomenon increases leak rate.8 Khoury and Waxman, in a major
review of the subject of colonic anastomosis have previously
emphasized that only 5 mm of colon should be devoid of mesentery
to optimize marginal flow.9 If one simply applies an understanding
of the basic underlying anatomy and cardiovascular physiology of
the colon, it is hard to understand how the bowel would not be per-
fused if there was pulsatile arterial inflow at the point where this
was being assessed. This is analogous to demonstrating normal
peripheral pulses in the assessment of a limb – in this setting angi-
ography will not demonstrate ischaemia. We do not believe ICG
has much to add if pulsatile mesenteric bleeding has been demon-
strated, which we regard as the gold standard. What is more con-
cerning is that in some cases the ICG perfused the bowel beyond
this point by significantly more than the 5 mm emphasis by Koury
and Waxman. This makes anatomical and physiological sense, as
presumably blood reaches distally via the plexus of submucosal
blood vessels. However, such blood flow and therefore perfusion,
is likely to be poorer than at the cut edge of the mesentery. It is
unclear whether the blood flow would be adequate to reliably
prevent an anastomotic leak. Given we have shown ICG perfusion
can be demonstrated at least 2.5 cm beyond where the demonstrated
blood flow appears optimal, if one were using ICG solely to assess
the perfusion, one approach might be to resect the bowel at least
2.5 cm proximal to the point of ICG demarcation. It is possible that
any ICG perfusion represents adequate arterial perfusion, but this is
unknown and it seems from our observations that small amounts of
perfusion can be demonstrated in areas devoid of mesentery where
only the submucosal vasculature is perfusing the bowel, which we
would regard as inadequate. It is of course important to have a
means to assess the bowel perfusion if one is performing intra-
corporeal anastomosis. Whether this anastomotic technique if of
any value to patients is currently not established. Certainly in the
authors view it makes sense to consider it in obese patients where it
allows for smaller, lower incisions and avoids traction on bulky and
short mesenteries. Beyond this, there are no clear established
patient benefits in laparoscopic cases. When using a robotic plat-
form with an intra-corporeal anastomosis we still caution the find-
ing that ICG perfusion seems to be demonstrable in the setting of
blood flow only being derived from submucosal vessels.
Conclusion
Although easy to perform and seemingly very safe to administer,
our study suggest that ICG perfusion may still reach areas of rela-
tive ischaemia as evidenced by perfusion being demonstrated
beyond the cut edge of the mesentery where the colon is ‘bare’.
The significance of this is uncertain but it would not be current
practice when doing a clinical assessment of colonic vascularity to
regard this as acceptable. We did not note this phenomenon to
occur beyond 2.5 cm and believe the observation is in keeping with
the known underlying anatomy and cardiovascular physiology of
colonic blood flow. Discarding the last 2.5 cm of any colon
assessed as well perfused by ICG alone would seem prudent in our
view. Further quantification of ICG perfusion with a ‘gold standard’
reference point may prove very useful. However, presently ICG
assessment is limited to a visual reference that is non-quantified on
the current laparoscopic and robotic platforms that have in-built
near infra-red imaging. We believe that in most cases extracorpo-
real anastomosis remains appropriate for laparoscopic and intra-
corporeal techniques may be considered on a select basis, mainly in
obese patients. With a robotic platform which makes intra-corporeal
anastomosis appealing it is still important to utilize a technique to
prevent anastomotic complications and in this current study we
have identified the potential for ICG-based assessments to result in
possible ischaemia if the results are not carefully interpreted.
Conflicts of interest
None declared.
References
1. Leevy CM, Mendenhall CL, Lesko W, Howard MM. Estimation of
hepatic blood flow with indocyanine green. J. Clin. Invest. 1962; 41:
1169–79.
© 2019 Royal Australasian College of Surgeons
Indocyanine green use in colorectal surgery
2. Reubi FC, Vorburger C, Tuckman J. Renal distribution volumes of
Indocyanine green, [51Cr]EDTA, and 24Na in man during acute renal
failure after shock. Implications for the pathogenesis of anuria. J. Clin.
Invest. 1973; 52: 223–35.
3. Kogure K, David NJ, Yamanouchi U, Choromokos E. Infrared absorption
angiography of the fundus circulation. Arch. Ophthalmol. 1970; 83: 209–14.
4. Ris F, Hompes R, Cunningham C et al. Near-infrared (NIR) perfusion
angiography in minimally invasive colorectal surgery. Surg. Endosc.
2014; 28: 2221–6.
5. Jafari MD, Wexner SD, Martz JE et al. perfusion assessment in laparo-
scopic left-sided/anterior resection (PILLAR II): a multi-institutional
study. J. Am. Coll. Surg. 2015; 220: 82–92.e1.
© 2019 Royal Australasian College of Surgeons
5
6. Blanco-Colino R, Espin-Basany E. Intraoperative use of ICG fluores-
cence imaging to reduce the risk of anastomotic leakage in colorectal sur-
gery: a systematic review and meta-analysis. Tech. Coloproctol. 2018;
22: 15–23.
7. Grace R. Peroperative observation of marginal artery bleeding; a predic-
tor of anastomotic leakage. Br. J. Surg. 1990; 77: 714.
8. Novell JR, Lewis AAM. Peroperative observation of marginal artery
bleeding: a predictor of anastomotic leakage. Br. J. Surg. 1990;
77: 137–8.
9. Khoury GA, Waxman BP. Large bowel anastomoses. I. The healing
process and sutured anastomoses. A review. Br. J. Surg. 1983;
70: 61–3.