OBJECTIVEA large body of experimental and clinical observations indicates that disturbances in cerebral blood flow (CBF) and impaired cerebrovascular autoregulation are important in the pathogenesis of germinal matrix-intraventricular hemorrhage (GMH-IVH) and periventricular leukomalacia (PVL), the 2 most important forms of brain injury in pretmature infants. Near-infrared spectroscopy (NIRS) has been used recently to estimate CBF in human newborns. The objectives of this study were to evaluate the correlation of NIRS estimations and cerebral blood flow in newborn piglets, which in turn may help provide the ideal NIRS estimation reflecting the changes of cerebral blood flow and cerebrovascular autoregulation.

METHODSTen newborn piglets, aged 1 - 3 days, were randomly assigned to one of the following groups: normal control group (n = 6) and hypotension group (n = 4). Hypotension was induced by withdrawing blood from an arterial catheter. We NIRS was used to determine quantitative changes in cerebral concentrations of oxygenated hemoglobin (DeltaHbO(2)) and deoxygenated hemoglobin (DeltaHHb), then calculated NIRS estimations DeltatHb (DeltaHbO(2)+DeltaHHb) and DeltaHbD (DeltaHbO(2)-DeltaHHb). Cerebral blood flow (CBF) was determined by colored microspheres, and mean artery blood pressure (MABP) measured by arterial catheter pressure transducer was recorded simultaneously. Linear regression methods were used to analyze the relationships between NIRS estimations, CBF measured by micropheres, and MABP.

RESULTSThe correlation of NIRS estimations and CBF was quantitated by calculating coherence scores. A coherence of 1.0 indicates perfect correlation, a coherence of 0 indicates a complete lack of correlation. In the norm group, the experimental study showed strong correlations beween DeltaHbD, DeltatHb and changes in global CBF (GCBF), cerebral cortex CBF (CBFc), coherence scores r(1a) = 0.409, r(1b) = 0.440, r(2a) = 0.394 and r(2b) = 0.400, respectively, P < 0.05. In the hypotension group, the decrease of CBF was significant when the MABP dropped to 35 mm Hg (P < 0.05). With the decreasing MABP, there was a notable increase of DeltaHHb (P < 0.01), a modest increase (P < 0.05) at the beginning and then a marked fall (P < 0.01) of DeltaHbO(2) and DeltatHb was noted when the MABP dropped to 35 mm Hg. DeltaHbD decreased in parallel with the decline in CBF determined by colored microspheres, DeltaHbD varied with CBF during hypotensive episodes. Notably, there was a very strong correlations between DeltaHbD and changes in CBF (coherence scores GCBF r(3a) = 0.890, CBFc r(3b) = 0.887, P < 0.01); Importantly, decreases in DeltatHb did not correlate significantly with decreases in CBF during hypotension (coherence scores GCBF r(4a) = 0.395, CBFc r(4b) = 0.375, P > 0.05). Concordant changes (correlation coefficient > 0.5) in DeltaHbD, CBF and MABP, consistent with impaired cerebrovascular autoregulation, were observed in newborn piglets when MABP was less than 35 mm Hg. When MABP was more than 35 mm Hg, newborn piglets with intact cerebrovascular autoregulation in which CBF are maintained constant despite alternations in MABP have shown inconsistent changes in DeltaHbD, CBF and MABP (correlation coefficient < 0.5).

CONCLUSIONDeltaHbD signal is more sensitive to changes in CBF than DeltatHb signal, in terms of cerebral hemodynamic changes both in normal and hypotensive conditions, while DeltatHb in normal condition. The lower limit of CBF autoregulation in newborn piglets aged 1 - 3 days was 35 mm Hg, and correlation between NIRS estimation (DeltaHbD) and MABP could be used to identify cerebrovascular autoregulation in newborn piglets.