Skip Navigation
Skip to contents

Ann Coloproctol : Annals of Coloproctology

OPEN ACCESS
SEARCH
Search
Advanced Search
mobile menu button

Articles

Page Path
HOME > Ann Coloproctol > Volume 42(3); 2026 > Article
Original Article
Others
 Decreased fluorescence latency and plateau intensity ratio as predictive markers of intraoperative hypoperfusion: quantitative indocyanine green perfusion analysis in colorectal cancer surgery
Paulina Daniluk-Marsy1orcid, Johannes Castelein2,3,4orcid, Karol Połom5,6orcid, Luigi Marano5orcid, Ronald J. H. Borra4,7orcid
Annals of Coloproctology 2026;42(3):324-332.
DOI: https://doi.org/10.3393/ac.2025.01298.0185
Published online: May 27, 2026

1Department of Surgical Oncology, Transplant and General Surgery, Faculty of Medicine, Medical University of Gdansk, Gdansk, Poland

2Department of Radiology, University Medical Center Groningen, Groningen, the Netherlands

3Department of Biomedical Sciences, Faculty of Health and Medical Sciences, University of Copenhagen, Copenhagen, Denmark

42nd Department of Radiology, Medical University of Gdańsk, Gdansk, Poland

5Department of Medicine, Academy of Applied Medical and Social Sciences (AMiSNS), Elblag, Poland

6Department of Gastrointestinal Surgical Oncology, Greater Poland Cancer Centre, Poznan, Poland

7Department of Nuclear Medicine and Molecular Imaging, University of Groningen, University Medical Center Groningen, Groningen, the Netherlands

Correspondence to: Paulina Daniluk-Marsy, MD Department of Surgical Oncology, Transplant Surgery and General Surgery, Faculty of Medicine, Medical University of Gdansk, Mariana Smoluchowskiego 17, Gdansk 80-214, Poland Email: p.daniluk@gumed.edu.pl
• Received: October 27, 2025   • Revised: January 10, 2026   • Accepted: March 22, 2026

© 2026 The Korean Society of Coloproctology

This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0/) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.

prev next
  • 910 Views
  • 55 Download
Full Article

Download PDF Download PDF

  • Purpose
    Intraoperative hypoperfusion is a major risk factor for postoperative anastomotic complications in colorectal surgery. Although most perfusion assessments emphasize arterial inflow, the roles of latency, sustained fluorescence, and venous outflow remain underexplored. This study aimed to identify quantitative indocyanine green (ICG) fluorescence parameters predictive of intraoperative hypoperfusion.
  • Methods
    Eighty patients who underwent colorectal cancer resection between 2018 and 2022 were analyzed. Intraoperative perfusion was evaluated qualitatively using near-infrared ICG angiography and quantitatively through post hoc analysis of fluorescence time-intensity curves. Associations with hypoperfusion were assessed using group comparisons and receiver operating characteristic (ROC) analysis.
  • Results
    Qualitative evaluation indicated hypoperfusion requiring intraoperative relocation of the anastomotic site in 12 patients (15%), with postoperative leakage in 1 patient (1.2%) despite revision of the anastomosis. Quantitative analysis showed that latency was significantly shorter in hypoperfused cases (5.5 seconds [interquartile range (IQR), 1.0–13.0 seconds] vs. 20.0 seconds [IQR, 6.8–27.0 seconds]; P=0.003), and the plateau intensity ratio (PIR; Fplateau/Fmax) was markedly decreased (0.6 [IQR, 0.6–0.7] vs. 0.8 [IQR, 0.6–0.9]; P=0.045). The ROC analysis demonstrated good diagnostic performance for latency (area under the curve [AUC], 0.76; sensitivity, 33.3% [95% confidence interval (CI), 13.8–60.9]; specificity, 92.8% [95% CI, 84.1–96.9]; positive likelihood ratio, 4.6) and moderate accuracy for PIR (AUC, 0.68; sensitivity, 66.7% [95% CI, 39.1–86.2]; specificity, 68.1% [95% CI, 56.4–77.9]; positive likelihood ratio, 2.1).
  • Conclusion
    Quantitative fluorescence analysis identified latency and PIR as complementary markers of perfusion dynamics. Whereas latency reflects early arterial inflow kinetics, PIR captures outflow characteristics related to microvascular and venous integrity. Combining these metrics may improve intraoperative detection of hypoperfusion and support data-driven decisions regarding anastomotic site selection.
  • Keywords: Perfusion; Colorectal surgery; Quantitative imaging; Fluorescence imaging; Indocyanine green
Colorectal cancer is among the most prevalent malignancies of the gastrointestinal tract, and surgical resection remains the cornerstone of curative treatment. Anastomotic leakage (AL) is a frequent and severe postoperative complication after colorectal surgery and is associated with increased morbidity, mortality, length of hospitalization, and healthcare costs. Impaired perfusion at the anastomotic site is widely recognized as a major contributing factor to AL, underscoring the need for reliable intraoperative methods to assess bowel perfusion and thereby minimize the risk of AL [1].
Near-infrared fluorescence (NIRF) imaging with intravenous indocyanine green (ICG) provides real-time visualization of tissue perfusion and has been increasingly adopted in surgical practice. Although numerous studies support its feasibility and safety, standardized protocols for ICG dose, injection timing, and interpretation of fluorescence signals remain undefined. This variability contributes to heterogeneity in reported outcomes and limits the reproducibility and generalizability of ICG-guided approaches [2].
Evidence from both prospective and retrospective studies suggests that ICG angiography may reduce AL rates, although the results have been mixed. The PILLAR II study, which focused on left-sided hemicolectomy and anterior resection, reported an AL rate of 1.4% in 139 patients evaluated with ICG perfusion assessment [3]. Randomized trials by De Nardi et al. [4] and Alekseev et al. [5] demonstrated benefit only in selected patient subgroups or in low rectal anastomoses. Meta-analyses have reinforced the potential benefit of ICG-NIRF, particularly in high-risk patients, but substantial heterogeneity across studies remains [6].
Beyond visual assessment, quantitative analysis of fluorescence kinetics has emerged as a promising approach for reducing interobserver variability and improving objectivity in perfusion assessment. Parameters such as latency to fluorescence onset, time to maximum intensity, slope of the ingress curve, and outflow plateau characteristics provide a multiparametric characterization of perfusion dynamics and may predict AL risk more accurately than visual judgment alone [7–9]. Nevertheless, the clinical applicability of these quantitative metrics has not yet been fully validated.
Given these limitations, the present study sought to determine whether intraoperative ICG-NIRF perfusion assessment, using both visual interpretation and quantitative fluorescence metrics, can guide surgical decision-making regarding anastomotic site selection and reduce AL risk. Specifically, we analyzed whether modifications to the anastomotic level based on perfusion signals improved outcomes and assessed the predictive value of quantitative indices such as Fplateau/Fmax, Tplateau/Tmax, Fplateau/Tplateau, and T½max/Tmax.
Ethics statement
The study protocol was approved by the Institutional Review Board of Scientific Research in Gdansk, Poland (No. NKBBN/94/2018 for the retrospective phase and No. NKBBN/944/2021 for the prospective phase). All patients included in the study were considered suitable candidates for surgery by a multidisciplinary oncologic board. Written informed consent for publication of the research details and clinical images was obtained from all patients. The study was conducted in accordance with the principles of the Declaration of Helsinki.
Study setting
This study combined retrospective and prospective data collection: retrospective data, including fluorescence video recordings and clinical information, were collected between April 2018 and December 2021; and prospective data collection began in January 2022. All procedures were performed at the former Department of Surgical Oncology, University Clinical Center in Gdansk, Poland.
Study population
Eligible patients were aged ≥18 years, were of either sex, and had histologically confirmed colorectal cancer. Inclusion required adequate functional status, defined as a Karnofsky Performance Status of ≥80 and an Eastern Cooperative Oncology Group (ECOG) performance status grade of 0 or 1. Exclusion criteria were age <18 years, pregnancy, hepatic or renal failure, known hypersensitivity to ICG or iodine-containing compounds, and active hematologic disorders. A total of 80 patients met these criteria and were included in the final analysis.
Surgical approach and imaging protocol
All surgical procedures followed the Enhanced Recovery After Surgery (ERAS) protocol. Patients underwent standardized open colorectal resection with intraoperative assessment of bowel perfusion.
NIRF imaging was performed using the Quest Spectrum system (Quest Medical Imaging), which is optimized for both open and minimally invasive surgery. The device provides simultaneous visible-light and NIRF visualization, using an excitation wavelength of 807 nm and capturing emission at approximately 822 nm. To improve image quality, ambient lighting was minimized, and the surgical field was illuminated with the system’s integrated white-light source. The camera was mounted on a flexible arm 20 to 30 cm above the bowel, providing stable, hands-free, real-time imaging.
Fluorescence clips were recorded at 50 frames per second using the Quest Spectrum system. A standardized bolus of 5 mg ICG (5 mg/mL Verdye, Carl Roth) was administered via a central venous catheter after colon transection and before creation of the anastomosis, followed by a 10-mL saline flush. Perfusion dynamics, including latency and time to maximum fluorescence, were observed at the bowel margins designated for anastomosis. Surgeons had continuous access to both white-light and NIRF images during intraoperative decision-making.
Hypoperfusion status was assessed intraoperatively before anastomosis creation by the same experienced colorectal surgeon. The assessment was based on ICG inflow time, duration of fluorescence persistence, and comparative perfusion of the proximal and distal bowel segments intended for anastomosis. Cases with delayed inflow, reduced fluorescence persistence, and visibly impaired perfusion of either bowel end were classified as hypoperfusion-positive (HP+).
Venous drainage was managed according to standard oncologic principles, with inferior mesenteric vein ligation performed as required to achieve adequate mobilization. Quantitative fluorescence analysis was therefore intended to capture the functional consequences of venous outflow adequacy at the anastomotic site rather than to evaluate the anatomical level of venous ligation per se.
Region of interest
All fluorescence recordings were analyzed post hoc using Quest ResearchTool ver. 4.7.2 (Quest Software). Three circular regions of interest (ROIs; diameter, approximately 20 pixels) were defined per frame using the integrated ROI tracker: (1) ROI 1, ischemic zone; (2) ROI 2, transitional/watershed zone; and (3) ROI 3, well-perfused reference.
The automated tracker compensated for tissue motion, and manual correction was applied if drift exceeded 1 pixel. The software generated time-intensity curves (arbitrary unit, AU) for each ROI (Figs. 1, 2). The system allowed simultaneous visualization of white-light, grayscale NIRF, and NIRF overlay views, which facilitated ROI selection and verification.
Quantitative analysis of colonic perfusion
Time-intensity curves derived from each ROI were analyzed to obtain quantitative perfusion metrics in accordance with previously described models of fluorescence kinetics [10]. Arterial inflow was assessed using maximum fluorescence intensity (Fmax), latency to onset (latency), time to maximum intensity (Tmax), and the maximal slope of the ingress curve (Slopemax = Fmax/Tmax). Venous outflow and washout characteristics were characterized by plateau fluorescence intensity (Fplateau), plateau duration (Tplateau), the plateau time ratio (PTR; Tplateau/Tmax), which describes the relative duration of perfusion stabilization, and the plateau intensity ratio (PIR; Fplateau/Fmax), which quantifies the fraction of peak fluorescence sustained during the plateau phase. Additional derived parameters included the plateau slope (Slopeplateau = Fplateau/Tplateau), the time to half maximum intensity (T½max), and the relative time ratio (TR; T½max/Tmax). Together, these indices provided a multiparametric characterization of arterial inflow and venous washout dynamics at the intended anastomotic site.
Statistical analysis
Sample size estimation was performed using G*Power ver. 3.1 (Heinrich Heine University Düsseldorf). Assuming a moderate effect size (Cohen d=0.7), α=0.05, and 80% power, at least 68 patients (12 with significant hypoperfusion and 56 without hypoperfusion) were required. The final cohort of 80 patients, including 12 (15.0%) with intraoperative hypoperfusion prompting relocation of the anastomosis, exceeded this threshold and therefore provided sufficient statistical power.
Continuous variables were presented as mean±standard deviation for normally distributed data or as median and interquartile range (IQR) for nonnormally distributed data. Categorical variables were expressed as frequencies and percentages. Normality was assessed using the Kolmogorov-Smirnov test.
Group comparisons between patients with and without intraoperative hypoperfusion suspected on the basis of qualitative inspection of the ICG signal were conducted using the independent-samples t-test or the Mann-Whitney U-test, depending on data distribution. Categorical variables were analyzed using the chi-square test or Fisher exact test, as appropriate. A two-sided P-value of <0.05 was considered statistically significant, whereas results with P<0.1 were interpreted as trends.
Study population
A total of 80 patients were included in the final analysis. Of these, 12 patients (15%) were included in the HP(+) group due to intraoperative hypoperfusion based on qualitative visual interpretation of the ICG signal, prompting relocation of the anastomotic site. Among them, 1 patient (1.2%) developed postoperative AL within 30 days despite intraoperative revision of the anastomosis. The remaining 68 patients (85%) showed adequate perfusion on qualitative visual inspection of the ICG signal, required no modification of the anastomotic level, and were classified as the hypoperfusion-negative (HP–) group.
The mean age was comparable between groups, and the distribution of patients aged <70 and ≥70 years was also similar (P=0.999). Likewise, sex was not associated with intraoperative hypoperfusion; men comprised 75% of the HP+ group and 58.8% of the HP– group (P=0.356). Body mass index (BMI) was significantly associated with hypoperfusion. Patients with a BMI of <25 kg/m2 were more frequently hypoperfused than those with a BMI of ≥25 kg/m2 (66.7% vs. 33.3%, P=0.017). The prevalence of diabetes did not differ significantly between groups (P=0.322). No patient with type 1 diabetes developed hypoperfusion, whereas type 2 diabetes was more common in the HP+ group (33.3% vs. 19.1%). Smoking status was comparable between groups (P=0.446). Regarding surgical procedures, most patients underwent anterior resection, and the distribution of procedure types did not differ significantly between groups (P=0.375). No hypoperfusion was observed after right hemicolectomy, whereas left hemicolectomy and anterior resection showed similar hypoperfusion rates (Table 1).
Perfusion time–related factors
Among the time-dependent fluorescence parameters, latency was significantly shorter in the HP+ group than in the HP– group (5.5 seconds [IQR, 1.0–13.0 seconds] vs. 20.0 seconds [IQR, 6.8–27.0 seconds]; P=0.003), indicating earlier onset of the fluorescence signal in cases with intraoperative hypoperfusion (Fig. 3A). Other time-based variables, including Tmax (16.5 seconds [IQR, 8.0–25.8 seconds] vs. 18.0 seconds [IQR, 9.0–31.5 seconds]; P=0.821), T½max (7.0 seconds [IQR, 2.3–15.8 seconds] vs. 10.0 seconds [IQR, 3.5–13.5 seconds]; P=0.543), and Tplateau (17.0 seconds [IQR, 7.0–39.3 seconds] vs. 17.0 seconds [IQR, 11.0–26.0 seconds]; P=0.770), did not differ significantly between groups.
Similarly, the perfusion ratios TR (0.6 [IQR, 0.2–0.7] vs. 0.5 [IQR, 0.4–0.7]; P=0.648) and PTR (1.2 [IQR, 0.7–1.6] vs. 0.9 [IQR, 0.5–2.0]; P=0.639) did not differ significantly between groups. Overall, these findings suggest comparable inflow dynamics between groups despite the presence of localized hypoperfusion identified intraoperatively (Table 2).
Fluorescence intensity–related factors
Absolute intensity parameters, such as Fmax (94.0 AU [IQR, 58.5–115.0 AU] vs. 100.0 AU [IQR, 57.5–100.0 AU]; P=0.859) and Slopemax (4.4 AU/sec [IQR, 3.7–6.6 AU/sec] vs. 4.4 AU/sec [IQR, 2.2–8.7 AU/sec]; P=0.593), were similar between HP+ and HP– groups. However, perfusion indices describing sustained fluorescence, reflecting capillary filling and venous drainage, showed clearer separation between groups. Median Fplateau was lower in the HP+ group (51.5 AU [IQR, 24.5–87.5 AU] vs. 70.0 AU [IQR, 38.5–90.0 AU]; P=0.390), and Slopeplateau was also lower (2.8 AU/sec [IQR, 1.4–5.2 AU/sec] vs. 3.3 AU/sec [IQR, 2.1–5.7 AU/sec]; P=0.500), although neither difference reached statistical significance.
PIR showed the strongest discriminatory potential and was significantly lower in the HP+ group (0.6 [IQR, 0.6–0.7] vs. 0.8 [IQR, 0.6–0.9]; P=0.045). This finding indicates that hypoperfused segments, despite reaching near-normal peak fluorescence, failed to maintain signal stability during the plateau phase, suggesting impaired venous clearance or reduced microvascular reserve (Table 2, Fig. 3B).
Analysis of perfusion markers
Among all quantitative fluorescence parameters, latency showed the best discriminatory performance for identifying intraoperative hypoperfusion, with an area under the receiver operating characteristic curve (AUC) of 0.76, indicating good diagnostic accuracy. The optimal threshold for latency (approximately 1.5 seconds) yielded a sensitivity of 33.3% (95% confidence interval [CI], 13.8–60.9) and a specificity of 92.8% (95% CI, 84.1–96.9), corresponding to a positive likelihood ratio of approximately 4.6 (Table 3, Fig. 4).
PIR showed moderate discriminatory ability for distinguishing hypoperfused (HP+) from normally perfused (HP–) segments, with an AUC of 0.68. Using an optimal cutoff of approximately 0.68, PIR yielded a sensitivity of 66.7% (95% CI, 39.1–86.2) and a specificity of 68.1% (95% CI, 56.4–77.9), resulting in a positive likelihood ratio of approximately 2.1 (Table 3, Fig. 4).
This study evaluated both qualitative and quantitative intraoperative fluorescence perfusion assessment in colorectal cancer surgery and demonstrated that both early inflow dynamics and plateau-phase stability play distinct but complementary roles in identifying intraoperative hypoperfusion. Among all analyzed parameters, latency, representing the time from ICG injection to fluorescence onset, showed the strongest discriminatory performance, whereas the PIR provided additional insight into the maintenance of microvascular perfusion over time. Together, these findings underscore the clinical relevance of dynamic perfusion analysis beyond simple visual interpretation.
The shorter latency observed in the HP+ group may appear counterintuitive, because delayed inflow has traditionally been associated with ischemia. However, this phenomenon may reflect compensatory redistribution of perfusion or technical factors associated with superficial capillary hyperfluorescence in hypoperfused segments, resulting in earlier but less sustained fluorescence. Thus, early fluorescence onset does not necessarily indicate adequate tissue perfusion and may instead reflect superficial or transient inflow without sustained microvascular integrity, emphasizing the importance of interpreting latency in conjunction with plateau-based parameters such as PIR. Importantly, although inflow parameters (latency, Tmax, and Slopemax) describe arterial perfusion kinetics, their relationship to tissue viability is limited without corresponding assessment of venous outflow.
In this context, PIR, which quantifies the fraction of maximum fluorescence sustained during the plateau phase, emerged as the most physiologically meaningful indicator of adequate microcirculation. Despite preserved arterial inflow, hypoperfused bowel segments exhibited reduced PIR values, consistent with microvascular dysfunction that limited fluorescence maintenance through inadequate capillary perfusion or venous outflow [11]. Although vascular anatomy and ligation strategy may influence perfusion dynamics, intraoperative ICG fluorescence provides a direct tissue-level assessment that integrates arterial inflow, collateralization, and microcirculatory function, potentially reducing the impact of anatomical heterogeneity.
These findings are consistent with previous reports indicating that venous outflow and the capillary equilibrium phase, rather than initial arterial inflow, are critical determinants of anastomotic perfusion stability. Iwamoto et al. [7] identified delayed T0 as a marker of increased leakage risk but noted that maximum-intensity parameters alone were not predictive, supporting our observation that static inflow metrics (such as T½) provide limited prognostic value. Likewise, Larsen et al. [12] highlighted substantial interobserver variability when perfusion assessment relied solely on visual inflow impressions, emphasizing the need for objective quantitative evaluation of the plateau phase. The present findings extend this understanding by demonstrating that even when peak inflow appears normal, reduced fluorescence stability, reflected by a lower PIR, may reveal subclinical hypoperfusion at the microvascular level.
From a physiological perspective, these results support the concept that anastomotic failure is not solely a consequence of impaired arterial supply but also reflects insufficient venous clearance and capillary homogeneity. Effective microvascular outflow is required to sustain oxygen exchange and prevent tissue congestion, both of which may be compromised when the plateau phase decays prematurely. Similar perfusion kinetics have been described in esophageal surgery, where well-perfused conduit regions show a balanced wash-in/washout pattern, whereas ischemic zones exhibit unstable fluorescence plateaus [13]. Applied to colorectal surgery, monitoring both the onset and stability of fluorescence may allow surgeons to detect subtle perfusion deficits before constructing the anastomosis.
Systemic hemodynamic variables such as mean arterial pressure, vasopressor use, and hemoglobin concentration may influence fluorescence signal characteristics; however, these data were not uniformly available for multivariable adjustment. Future studies integrating synchronized anesthetic and perfusion data are warranted to clarify further the independent contribution of quantitative fluorescence metrics. Nevertheless, because ICG fluorescence reflects the net functional interaction between systemic hemodynamics and local microcirculatory regulation, it remains a clinically meaningful intraoperative endpoint.
From a practical standpoint, integrating quantitative fluorescence metrics such as latency and PIR could substantially enhance intraoperative decision-making. Real-time ICG analysis software can now automatically generate time-intensity curves, allowing surgeons to evaluate both inflow and plateau characteristics within seconds. The clinical value of quantitative ICG perfusion lies primarily in its capacity to guide real-time intraoperative decision-making. By identifying hypoperfusion before anastomosis construction, these metrics may allow surgeons to modify the anastomotic level or surgical strategy, potentially preventing postoperative complications. Accordingly, low leakage rates after such interventions should be interpreted as evidence of effective risk mitigation rather than as a sign of limited clinical relevance.
If future multicenter studies validate standardized threshold values, such as latency of <1.5 seconds and PIR of <0.68 as observed in our cohort, these metrics could serve as actionable indicators prompting revision of resection margins or consideration of a diverting stoma. Unlike qualitative fluorescence impressions, such data-driven thresholds offer reproducibility and objectivity suitable for integration into enhanced recovery pathways and emerging digital surgical platforms.
Limitations
This study has several limitations. First, although the sample size was adequately powered for the primary comparisons, it remains modest. Larger multicenter studies are needed to validate the generalizability of plateau-based indices. Second, technical factors such as camera positioning, movement, ambient light, and ICG dosing, although standardized in our protocol, may have influenced fluorescence kinetics. Third, although our findings strongly suggest that venous outflow is a key determinant of AL, perfusion is multifactorial and may also be affected by patient comorbidities, surgical technique, and anastomotic tension.
Conclusions
This study demonstrates that intraoperative quantitative fluorescence analysis can objectively characterize both arterial inflow and venous outflow dynamics at the anastomotic site. Although latency provided the highest diagnostic accuracy for detecting intraoperative hypoperfusion, PIR emerged as the most physiologically relevant marker of sustained microvascular integrity. These findings support a comprehensive, multiphasic assessment of fluorescence perfusion that integrates both inflow and plateau parameters to improve intraoperative decision-making and reduce the risk of postoperative anastomotic failure.

Conflict of interest

Luigi Marano is an editorial board member of this journal, but was not involved in the peer reviewer selection, evaluation, or decision process of this article. Ronald J. H. Borra is a medical consultant for Perfusion Tech. No other potential conflict of interest relevant to this article was reported.

Funding

None.

Author contributions

Conceptualization: PDM; Data curation: PDM; Formal analysis: PDM; Investigation: PDM; Methodology: PDM, JC, LM; Supervision: PDM, KP, RJHB; Validation: PDM, KP, RJHB; Visualization: PDM, JC; Writing–original draft: PDM, JC, LM; Writing–review & editing: all authors. All authors read and approved the final manuscript.

Fig. 1.
Example of 3 circular regions of interest (ROIs) superimposed on every frame with the integrated ROI‑tracker on the bowel loop. Red dot indicates ROI 1, the ischemic region; green dot indicates ROI 2, the transitional/watershed zone; and blue dot indicates ROI 3, the well-perfused reference.
ac-2025-01298-0185f1.jpg
Fig. 2.
Example of time-intensity curves for each region of interest (ROI). AU, arbitrary unit.
ac-2025-01298-0185f2.jpg
Fig. 3.
Comparison of key perfusion parameters between hypoperfusion-negative (HP–) and hypoperfusion-positive (HP+) groups. (A) Latency was significantly shorter in the HP+ group (P=0.003), indicating faster fluorescence onset. (B) The plateau intensity ratio (PIR) was significantly lower in the HP+ group (P=0.045), suggesting impaired maintenance of perfusion despite initial inflow. Bars represent median values, and error bars represent interquartile ranges. *P<0.05, **P<0.01.
ac-2025-01298-0185f3.jpg
Fig. 4.
Receiver operating characteristic curves for quantitative fluorescence perfusion parameters predicting intraoperative hypoperfusion. The latency parameter (green line) demonstrated the best diagnostic performance with an area under the curve (AUC) of 0.76, indicating good discriminatory ability between hypoperfused (hypoperfusion-positive) and normally perfused (hypoperfusion-negative) bowel segments. The plateau intensity ratio (PIR) showed moderate performance (AUC, 0.68), reflecting differences in sustained fluorescence intensity during the plateau phase. The diagonal line represents the line of no discrimination (AUC, 0.5).
ac-2025-01298-0185f4.jpg
ac-2025-01298-0185f5.jpg
Table 1.
Baseline demographic and clinical characteristics of patients stratified by hypoperfusion status (n=80)
Characteristic No. of patients (%) P-value
HP– group (n=68) HP+ group (n=12)
Age (yr) 0.999
 <70 29 (42.6) 5 (41.7)
 ≥70 39 (57.4) 7 (58.3)
Sex 0.356
 Male 40 (58.8) 9 (75.0)
 Female 28 (41.2) 3 (25.0)
Body mass index (kg/m2) 0.017
 <25 19 (27.9) 8 (66.7)
 ≥25 49 (72.1) 4 (33.3)
Diabetes 0.322
 No 48 (70.6) 8 (66.7)
 Type 1 7 (10.3) 0 (0)
 Type 2 13 (19.1) 4 (33.3)
Smoking status 0.446
 Never 41 (60.3) 8 (66.7)
 Former 11 (16.2) 3 (25.0)
 Current 16 (23.5) 1 (8.3)
Surgery type 0.375
 Anterior resection 26 (38.2) 5 (41.7)
 Anterior resection with ileostomy 11 (16.2) 2 (16.7)
 Right hemicolectomy 18 (26.5) 0 (0)
 Left hemicolectomy 13 (19.1) 5 (41.7)

Hypoperfusion status was assessed intraoperatively using indocyanine green fluorescence before performing an anastomosis. P-values were calculated using the chi-square test or the Fisher exact test, as appropriate.

HP–, hypoperfusion-negative; HP+, hypoperfusion-positive.

Table 2.
Comparison of quantitative indocyanine green fluorescence parameters between HP– and HP+ groups (n=80)
Parameter HP– group (n=68) HP+ group (n=12) P-value
Perfusion time–related
 Latency (sec) 20.0 (6.8–27.0) 5.5 (1.0–13.0) 0.003
 Tmax (sec) 18.0 (9.0–31.5) 16.5 (8.0–25.8) 0.821
 T½max (sec) 10.0 (3.5–13.5) 7.0 (2.3–15.8) 0.543
 Tplateau (sec) 17.0 (11.0–26.0) 17.0 (7.0–39.3) 0.770
 TR (T½max/Tmax) 0.5 (0.4–0.7) 0.6 (0.2–0.7) 0.648
 PTR (Tplateau/Tmax) 0.9 (0.5–2.0) 1.2 (0.7–1.6) 0.639
Fluorescence intensity–related
 Fmax (AU) 100.0 (57.5–100.0) 94.0 (58.5–115.0) 0.859
 Fplateau (AU) 70.0 (38.5–90.0) 51.5 (24.5–87.5) 0.390
 Slopemax (Fmax/Tmax) (AU/sec) 4.4 (2.2–8.7) 4.4 (3.7–6.6) 0.593
 Slopeplateau (Fplateau/Tplateau) (AU/sec) 3.3 (2.1–5.7) 2.8 (1.4–5.2) 0.500
 PIR (Fplateau/Fmax) 0.8 (0.6–0.9) 0.6 (0.6–0.7) 0.045

Values are presented as median (interquartile range). Comparisons between groups were performed using the Mann-Whitney U-test.

HP–, hypoperfusion-negative; HP+, hypoperfusion-positive; Tmax, time to maximum fluorescence intensity; T½max, time to half maximum intensity; Tplateau, plateau duration; TR, time ratio; PTR, plateau time ratio; Fmax, maximum fluorescence intensity; AU, arbitrary unit; PIR, plateau intensity ratio.

Table 3.
ROC curve analysis of perfusion parameters
Parameter Optimal cutoffa Sensitivity (%) (95% CI) Specificity (%) (95% CI) Positive LR AUC P-value
Latency (sec) 1.5 33.3 (13.8–60.9) 92.8 (84.1–96.9) 4.6 0.76 0.003
PIR 0.68 66.7 (39.1–86.2) 68.1 (56.4–77.9) 2.1 0.68 0.045

ROC, receiver operating characteristic; CI, confidence interval; LR, likelihood ratio; AUC, area under the curve; PIR, plateau intensity ratio.

aDetermined using the Youden index.

  • 1. Daniluk P, Mazur N, Swierblewski M, Chand M, Diana M, Polom K. Fluorescence imaging in colorectal surgery: an updated review and future trends. Surg Innov 2022;29:479–87. ArticlePubMedPDF
  • 2. Dip F, Boni L, Bouvet M, Carus T, Diana M, Falco J, et al. Consensus conference statement on the general use of near-infrared fluorescence imaging and indocyanine green guided surgery: results of a modified Delphi study. Ann Surg 2022;275:685–91. ArticlePubMedPMC
  • 3. Jafari MD, Wexner SD, Martz JE, McLemore EC, Margolin DA, Sherwinter DA, et al. Perfusion assessment in laparoscopic left-sided/anterior resection (PILLAR II): a multi-institutional study. J Am Coll Surg 2015;220:82–92.e1. ArticlePubMed
  • 4. De Nardi P, Elmore U, Maggi G, Maggiore R, Boni L, Cassinotti E, et al. Intraoperative angiography with indocyanine green to assess anastomosis perfusion in patients undergoing laparoscopic colorectal resection: results of a multicenter randomized controlled trial. Surg Endosc 2020;34:53–60. ArticlePubMedPDF
  • 5. Alekseev M, Rybakov E, Shelygin Y, Chernyshov S, Zarodnyuk I. A study investigating the perfusion of colorectal anastomoses using fluorescence angiography: results of the FLAG randomized trial. Colorectal Dis 2020;22:1147–53. ArticlePubMedPDF
  • 6. Arezzo A, Bonino MA, Ris F, Boni L, Cassinotti E, Foo DC, et al. Intraoperative use of fluorescence with indocyanine green reduces anastomotic leak rates in rectal cancer surgery: an individual participant data analysis. Surg Endosc 2020;34:4281–90. ArticlePubMedPDF
  • 7. Iwamoto H, Matsuda K, Hayami S, Tamura K, Mitani Y, Mizumoto Y, et al. Quantitative indocyanine green fluorescence imaging used to predict anastomotic leakage focused on rectal stump during laparoscopic anterior resection. J Laparoendosc Adv Surg Tech A 2020;30:542–6. ArticlePubMedPDF
  • 8. Frank AH, Heiliger C, Andrade D, Werner J, Karcz K. Intra-arterial versus negative-staining of embryonal resection borders with indocyanine green (ICG) fluorescence for total mesorectal excision in colorectal cancer: an experimental feasibility study in a porcine model. Minim Invasive Ther Allied Technol 2022;31:107–11. ArticlePubMed
  • 9. Rentea RM, Halleran DR, Ahmad H, Sanchez AV, Gasior AC, McCracken K, et al. Preliminary use of indocyanine green fluorescence angiography and value in predicting the vascular supply of tissues needed to perform cloacal, anorectal malformation, and Hirschsprung reconstructions. Eur J Pediatr Surg 2020;30:505–11. ArticlePubMed
  • 10. Han SR, Lee CS, Bae JH, Lee HJ, Yoon MR, Al-Sawat A, et al. Quantitative evaluation of colon perfusion after high versus low ligation in rectal surgery by indocyanine green: a pilot study. Surg Endosc 2022;36:3511–9. ArticlePubMedPDF
  • 11. Son GM, Kwon MS, Kim Y, Kim J, Kim SH, Lee JW. Quantitative analysis of colon perfusion pattern using indocyanine green (ICG) angiography in laparoscopic colorectal surgery. Surg Endosc 2019;33:1640–9. ArticlePubMedPDF
  • 12. Larsen PO, Nerup N, Andersen J, Dohrn N, Klein MF, Brisling S, et al. Anastomotic perfusion assessment with indocyanine green in robot-assisted low-anterior resection, a multicenter study of interobserver variation. Surg Endosc 2023;37:3602–9. ArticlePubMedPMCPDF
  • 13. de Groot EM, Kuiper GM, van der Veen A, Fourie L, Goense L, van der Horst S, et al. Indocyanine green fluorescence in robot-assisted minimally invasive esophagectomy with intrathoracic anastomosis: a prospective study. Updates Surg 2023;75:409–18. ArticlePubMedPDF

Figure & Data

References

    Citations

    Citations to this article as recorded by  

      • Cite this Article
        Cite this Article
        export Copy Download
        Close
        Download Citation
        Download a citation file in RIS format that can be imported by all major citation management software, including EndNote, ProCite, RefWorks, and Reference Manager.
        Format:
        • RIS — For EndNote, ProCite, RefWorks, and most other reference management software
        • BibTeX — For JabRef, BibDesk, and other BibTeX-specific software
        Include:
        • Citation for the content below
        Decreased fluorescence latency and plateau intensity ratio as predictive markers of intraoperative hypoperfusion: quantitative indocyanine green perfusion analysis in colorectal cancer surgery
        Ann Coloproctol. 2026;42(3):324-332.   Published online May 27, 2026
        DOI: https://doi.org/10.3393/ac.2025.01298.0185
        Close
      • XML DownloadXML Download
      Figure
      • 0
      • 1
      • 2
      • 3
      • 4
      Decreased fluorescence latency and plateau intensity ratio as predictive markers of intraoperative hypoperfusion: quantitative indocyanine green perfusion analysis in colorectal cancer surgery
      Image Image Image Image Image
      Fig. 1. Example of 3 circular regions of interest (ROIs) superimposed on every frame with the integrated ROI‑tracker on the bowel loop. Red dot indicates ROI 1, the ischemic region; green dot indicates ROI 2, the transitional/watershed zone; and blue dot indicates ROI 3, the well-perfused reference.
      Fig. 2. Example of time-intensity curves for each region of interest (ROI). AU, arbitrary unit.
      Fig. 3. Comparison of key perfusion parameters between hypoperfusion-negative (HP–) and hypoperfusion-positive (HP+) groups. (A) Latency was significantly shorter in the HP+ group (P=0.003), indicating faster fluorescence onset. (B) The plateau intensity ratio (PIR) was significantly lower in the HP+ group (P=0.045), suggesting impaired maintenance of perfusion despite initial inflow. Bars represent median values, and error bars represent interquartile ranges. *P<0.05, **P<0.01.
      Fig. 4. Receiver operating characteristic curves for quantitative fluorescence perfusion parameters predicting intraoperative hypoperfusion. The latency parameter (green line) demonstrated the best diagnostic performance with an area under the curve (AUC) of 0.76, indicating good discriminatory ability between hypoperfused (hypoperfusion-positive) and normally perfused (hypoperfusion-negative) bowel segments. The plateau intensity ratio (PIR) showed moderate performance (AUC, 0.68), reflecting differences in sustained fluorescence intensity during the plateau phase. The diagonal line represents the line of no discrimination (AUC, 0.5).
      Graphical abstract

      Fig. 1.

      Fig. 2.

      Fig. 3.

      Fig. 4.

      Graphical abstract

      Decreased fluorescence latency and plateau intensity ratio as predictive markers of intraoperative hypoperfusion: quantitative indocyanine green perfusion analysis in colorectal cancer surgery
      Characteristic No. of patients (%) P-value
      HP– group (n=68) HP+ group (n=12)
      Age (yr) 0.999
       <70 29 (42.6) 5 (41.7)
       ≥70 39 (57.4) 7 (58.3)
      Sex 0.356
       Male 40 (58.8) 9 (75.0)
       Female 28 (41.2) 3 (25.0)
      Body mass index (kg/m2) 0.017
       <25 19 (27.9) 8 (66.7)
       ≥25 49 (72.1) 4 (33.3)
      Diabetes 0.322
       No 48 (70.6) 8 (66.7)
       Type 1 7 (10.3) 0 (0)
       Type 2 13 (19.1) 4 (33.3)
      Smoking status 0.446
       Never 41 (60.3) 8 (66.7)
       Former 11 (16.2) 3 (25.0)
       Current 16 (23.5) 1 (8.3)
      Surgery type 0.375
       Anterior resection 26 (38.2) 5 (41.7)
       Anterior resection with ileostomy 11 (16.2) 2 (16.7)
       Right hemicolectomy 18 (26.5) 0 (0)
       Left hemicolectomy 13 (19.1) 5 (41.7)
      Parameter HP– group (n=68) HP+ group (n=12) P-value
      Perfusion time–related
       Latency (sec) 20.0 (6.8–27.0) 5.5 (1.0–13.0) 0.003
       Tmax (sec) 18.0 (9.0–31.5) 16.5 (8.0–25.8) 0.821
       T½max (sec) 10.0 (3.5–13.5) 7.0 (2.3–15.8) 0.543
       Tplateau (sec) 17.0 (11.0–26.0) 17.0 (7.0–39.3) 0.770
       TR (T½max/Tmax) 0.5 (0.4–0.7) 0.6 (0.2–0.7) 0.648
       PTR (Tplateau/Tmax) 0.9 (0.5–2.0) 1.2 (0.7–1.6) 0.639
      Fluorescence intensity–related
       Fmax (AU) 100.0 (57.5–100.0) 94.0 (58.5–115.0) 0.859
       Fplateau (AU) 70.0 (38.5–90.0) 51.5 (24.5–87.5) 0.390
       Slopemax (Fmax/Tmax) (AU/sec) 4.4 (2.2–8.7) 4.4 (3.7–6.6) 0.593
       Slopeplateau (Fplateau/Tplateau) (AU/sec) 3.3 (2.1–5.7) 2.8 (1.4–5.2) 0.500
       PIR (Fplateau/Fmax) 0.8 (0.6–0.9) 0.6 (0.6–0.7) 0.045
      Parameter Optimal cutoffa Sensitivity (%) (95% CI) Specificity (%) (95% CI) Positive LR AUC P-value
      Latency (sec) 1.5 33.3 (13.8–60.9) 92.8 (84.1–96.9) 4.6 0.76 0.003
      PIR 0.68 66.7 (39.1–86.2) 68.1 (56.4–77.9) 2.1 0.68 0.045
      Table 1. Baseline demographic and clinical characteristics of patients stratified by hypoperfusion status (n=80)

      Hypoperfusion status was assessed intraoperatively using indocyanine green fluorescence before performing an anastomosis. P-values were calculated using the chi-square test or the Fisher exact test, as appropriate.

      HP–, hypoperfusion-negative; HP+, hypoperfusion-positive.

      Table 2. Comparison of quantitative indocyanine green fluorescence parameters between HP– and HP+ groups (n=80)

      Values are presented as median (interquartile range). Comparisons between groups were performed using the Mann-Whitney U-test.

      HP–, hypoperfusion-negative; HP+, hypoperfusion-positive; Tmax, time to maximum fluorescence intensity; T½max, time to half maximum intensity; Tplateau, plateau duration; TR, time ratio; PTR, plateau time ratio; Fmax, maximum fluorescence intensity; AU, arbitrary unit; PIR, plateau intensity ratio.

      Table 3. ROC curve analysis of perfusion parameters

      ROC, receiver operating characteristic; CI, confidence interval; LR, likelihood ratio; AUC, area under the curve; PIR, plateau intensity ratio.

      Determined using the Youden index.

      Table 1.

      Table 2.

      Table 3.

      Ann Coloproctol : Annals of Coloproctology Twitter Facebook
      © Korean Society of Coloproctology.
      TOP