Estimation of microvascular perfusion after esophagectomy: a quantitative model of dynamic fluorescence imaging

Abstract

Most common complications of esophagectomy stem from a perfusion deficiency of the gastric conduit at the anastomosis. Fluorescent tracer imaging allows intraoperative visualization of tissue perfusion. Quantitative assessment of fluorescence dynamics has the potential to identify perfusion deficiency. We developed a perfusion model to analyze the relation between fluorescence dynamics and perfusion deficiency. The model divides the gastric conduit into two well-perfused and two anastomosed sites. Hemodynamics and tracer transport were modeled. We analyzed the value of relative time-to-threshold (RTT) as a predictor of the relative remaining flow (RRF). Intensity thresholds for RTT of 20% to 50% of the maximum fluorescence intensity of the well-perfused site were tested. The relation between RTT and RRF at the anastomosed sites was evaluated over large variations of vascular conductance and volume. The ability of RTT to distinguish between sufficient and impaired perfusion was analyzed using c-statistics. We found that RTT was a valuable estimate for low RRF. The threshold of 20% of the maximum fluorescence intensity provided the best prediction of impaired perfusion on the two anastomosed sites (AUC = 0.89 and 0.86). The presented model showed that for low flows, relative time-to-threshold may be used to estimate perfusion deficiency.

Keywords: Esophagectomy; Fluorescence imaging; Gastric conduit model; Indocyanine green; Perfusion.

Fig. 1

An image of constructed gastric conduit (a), obtained intraoperatively during fluorescence imaging post-esophagectomy, demonstrated perfusion with ICG and reduced fluorescence in the collateral-dependent upper part. The corresponding fluorescence enhancement curves of each ROI were measured at the gastric conduit (b). The model (c): from bottom to top: the sites 1–4, which correspond, respectively, to the ROIs. A site consists of an arterial and a venous compartment directly connected through a capillary bed (horizontal line between compartments). Collaterals, as depicted by vertical lines between compartments, connect the adjacent sites. In this figure, sites 1 and 2 are well perfused, whereas sites 3 and 4 are anastomosed. Arrows indicate direction of flow after the intervention

Fig. 2

Example of contrast dynamics at the four sites with derivation of the respective times to threshold. Sites 1 and 2 have native perfusion, while sites 3 and 4 depend on collateral flow, and accordingly have slower contrast dynamics

Fig. 3

Result of simulation 1: relation between RTT and RRF over collateral conductance space, for sites 3 (left) and 4 (right). The plots show RTT for thresholds of contrast arrival at 20% (a), 30% (b), 40% (c), and 50% (d) of the maximum contrast signal at site 1 as a function of RRF. The corresponded fitting functions and the respective goodness-of-fit parameters were also displayed

Fig. 4

Result of simulation 2: variation in large vessels conductance changes the relation between true RRF and its estimate (RTT)

Fig. 5

Result of simulation 3: RRF and RTT relation as a function of the ratio of arterial volume (AV) and venous volume (VV). Note that not every symbol/color is visible due to the overlay

Fig. 6

The sample space of the model from the complete (pink) and physiological (green) parameter space of collateral conductances (G_ca, G_cv), large vessel conductance (G_LV), and vascular volume (AV, VV) was calculated from I₂₀ to I₅₀ (top to bottom)

Fig. 7

ROC curve of four threshold for site 3 (top) and 4 (bottom). The cutoff value was determined hypothetically as 0.5 and 0.4 for sites 3 and 4, respectively

Fig. 8

The effects of the balance of arterial and venous collateral conductance (a) and of the total collateral conductance (b) on the relative remaining flow (RRF) in the anastomotic site 4, demonstrating the log-linear dependence of RRF on total collateral conductance without an effect of arteriovenous balance. RRF for all levels of total collateral conductance and arteriovenous balance (c), demonstrating absence of interactive effects of both parameters