We are grateful for the comments offered by Drs. Lam and Warner regarding the effect of morphine and fentanyl on intracranial pressure (ICP) in severe head injury patients. They express concern that our methods in autoregulation testing need further interpretation and refer to the possibility that an insufficient mean arterial pressure (MAP) change observed by the doses of opioids administered could have unmasked our results. 1
The first issue they address is the use of transcranial Doppler ultrasonography (TCD) for the study of autoregulation. Although flow velocity measurements performed with use of TCD would have provided an additional estimation of cerebral blood flow (CBF) during autoregulation studies, it is worth reemphasizing that TCD velocities do not always mirror blood flow. The ability of TCD to assess cerebral vasoreactivity assumes that changes in the diameter of the insolated vessels, usually the middle cerebral artery, are negligible. However, it has been shown that mid-sized arteries, even the internal carotid artery, may be implicated in maintaining a constant CBF when cerebral perfusion pressure is modified. 2Thus, errors in flow assessment may occur because of changes in middle cerebral artery or internal carotid artery diameter during systemic blood pressure changes. Most of our patients had a diffuse brain injury, and, therefore, global autoregulation measurements such as arteriojugular venous oxygen content difference (AVDo2) were likely to be more representative of the autoregulatory status than evaluation of bilateral changes in hemispheric middle cerebral artery flow velocities. Furthermore, corrections for changes in arterial carbon dioxide tension (Paco2) are easier to perform with use of this test than with use of TCD.
Focusing on the second issue, we completely agree with Lam and Warner that autoregulation capacity may be influenced by Paco2. However, hypocapnia and hypercapnia do not affect autoregulation per se but may influence the autoregulatory response of cerebral vessels that are already vasoconstricted and vasodilated, respectively. To avoid this known artifact, before testing autoregulation, we always manipulate the ventilator settings to obtain a baseline Paco2in the normoventilation range. The actual Paco2values during autoregulation testing in our study were (before and after inducing hypertension) 38 ± 4 and 39 ± 3 mmHg in the morphine group and 37 ± 3 and 38 ± 4 mmHg in the fentanyl group (mean ± SD). As shown in table 1, Paco2values remained unchanged after opiates were administered. The percentage change of 1/AVDo2relative to the resting values was corrected for spontaneous changes in Paco2with use of the absolute CO2reactivity (CO2Rabs), which was calculated as the change in AVDo2divided by the measured change in Paco2(Δ AVDo2/Δ Paco2).
The third issue addressed is our “arbitrary classification of autoregulatory pattern.” Our method of interpreting autoregulation is based on experimental models, and we agree that it is arbitrary, as any other classification used before in clinical and experimental studies. However, contrary to Lam and Warner’s opinion, we believe that taking ICP into consideration when testing autoregulation may help to clarify our results. If autoregulation is tested only through changes in CBF, some paradoxical observations, such as the false autoregulation phenomenon, can be observed, and patients may be classified wrongly as “preserved autoregulation.” In the study referenced by Lam and Warner, Enevoldsen and Jensen 3described false autoregulation (pseudoautoregulation) as an alteration of autoregulation in which the apparent maintenance on a constant CBF when increasing cerebral perfusion pressure is caused by an increase in brain tissue pressure. In these patients with impaired or abolished vasoconstrictory response to increased cerebral perfusion pressure, increasing MAP induces parallel changes in water filtration through the blood–brain barrier. Because of the compression of the cerebral microcirculation, these changes always induce an ICP increase with an unpredictable change in CBF. This fact makes interpretation of the results with only CBF measurements as a basis difficult, and, consequently, we believe that changes in both CBF and ICP must be taken into consideration when characterizing autoregulatory status. It is puzzling to observe some studies that classify patients with a constant CBF but with 10–15 mmHg increases in ICP (a relatively common phenomenon in the acute phase of severe head injuries) as “intact autoregulation.” We agree that impaired autoregulation is not the same as abolished autoregulation, but it would be necessary to test autoregulation in a wide range of MAP (including hypotension) to distinguish one from the other, which we believe is not ethically possible in severe head injury patients.
The last issue refers to the magnitude of MAP change caused by the doses of opioids administered in our study. We agree with Lam and Warner that the small decrease in MAP (3–4 mmHg) could not be sufficient to stimulate the vasodilatory cascade. However, the ICP increase seen in patients with preserved and impaired autoregulation suggests that cerebrovascular autoregulation may not be the only probable mechanism responsible for morphine- and fentanyl-induced increases in ICP. As we stated in the discussion, opioids may interact directly with receptors located on brain blood vessels, and μ receptor activation has been proven to cause direct cerebral vasodilation in some experimental models. 4Although we cannot rule out the hypothesis that greater doses of opioids may increase ICP through an autoregulatory response, we believe that further studies are needed to elucidate the role of opioids on cerebral circulation during brain injury.
We appreciate Lam and Warner’s insights on our work and the chance to clarify some controversial issues about the classification of autoregulation in severe head injury patients.