Discussion
A preliminary study by Vidal et al. reported transcutaneous NIRS measurements in 24 children after KT [7]. Renal rSrO2 measured by NIRS significantly increased over time during the first three days after surgery and a significant correlation was found with serum creatinine, estimated glomerular filtration rate (eGFR) and with the decrease of post-operative urine-neutrophilgelatinase-associated lipocaine (u-NGAL), which is a marker of tubular injury reflecting the ischemic and reperfusion damage of the kidney. In this series, DUS did not identify any abnormality in renal vascularisation and no vascular complications were reported. To date, four patients experienced a delayed graft function (DGF) without any peculiar modifications of rSO2.
Despite these promising results, Skowno et al. raised some criticisms [8]. First, DUS, the gold standard for the assessment of allograft perfusion, was not related to rSrO2. Recently, Malakasiotiet al . identified a significant correlation between renal rSO2 measured by NIRS and resistive index derived from DUS in a series of 29 paediatric patients [9]. Once again, no complications were reported in the cohort.
Second, the region of interest of NIRS may reach up to 4 cm below the skin, according to the manufacturers, and its effectiveness in monitoring somatic perfusion was validated in infants weighing less than 10 kg [10]. This body weight was sensibly lower than those of the patients undergone KT. However, Skowno et al. confirmed the effect of body size on transcutaneous NIRS only in a porcine model, suggesting its in vivo application up to 1 cm of depth [11].
Nevertheless, Pérez Civantos et al. assessed renal rSrO2 in 61 adults undergone KT, using probes with a maximum depth of 2.5 cm [12]. NIRS readings were significantly correlated to the decreasing values of serum lactate at eight hours and 24 hours, which is another marker of ischemic injury. Furthermore, renal rSrO2 was related to initial diuresis at three hours and to mixed central venous oxygen saturation. On the other hand, no correlation with DUS was found. In addition, in this series arterial thrombosis and bleedings were encountered, showing a maintained decrease in rSrO2 which might anticipate the vascular events. Moreover, the same study was replicated in liver transplantations. Even in this scenario, NIRS decreased early during bleeding or in case of thrombosis [13].
The last concern about NIRS regarded the placement of the probes. The vascularisation over the surgical wounds may be altered by the healing process, Skowno et al. speculated that a placement directly over that area might influence NIRS readings [8]. Therefore, experimental studies on reliable animal models, such as rats or piglets, should better clarify the real impact of abdominal wall thickness and of the healing process on transcutaneous NIRS monitoring.
Furthermore, the normal range of renal rSrO2 values need to be established by investigating a larger number of patients. Only after this process, NIRS might help in the prompt identification of acute adverse events, which is the final aim of real-time NIRS monitoring. At the moment, NIRS contributed to the early diagnosis of vascular complications in two children undergone liver transplantation [15]. However, even if renal rSrO2 was not altered among the patients affected by DGF in the series by Vidal et al. , the Authors suggested that NIRS might help in discriminating the causes of oligo-anuria in the early post-operative time after KT, helping in managing the postoperative fluid balance [7].
In conclusion, preliminary studies showed the correlation between NIRS, graft function and graft perfusion in paediatric KTs. Further investigation should also relate rSrO2 to perfusion renal scintigraphy, which is the most objective modality to assess renal allograft vascularisation and function [4]. Before large-scale in vivoutilisation, experimentation of NIRS on animal models might investigate factors influencing the measurements of renal rSrO2, such as the depth of the region of interest and the correct placement of the probes. Furthermore, animal models or selected cohorts of patients might help in defining the normal values of renal rSrO2.