In Reply:--The authors thank Heinrich and Kao for clarifying the meaning of hepatic perfusion improvement after nitric oxide inhalation in the reported case. We kept this point relatively ambiguous because of space constraints, although we agree this aspect must be clarified. Hepatic arterial flow is proportional to the pressure gradient between mean arterial and hepatic venous pressures, whereas portal venous flow is dependent on the pressure gradient between portal venous and hepatic venous pressures. Considering these mechanical determinants of liver perfusion, we agree with Heinrich and Kao that the patient benefitted from an increased perfusion pressure in the portal circulation. Accordingly, a Starling resistance have been described in the portal circulation, and the high hepatic venous pressure observed before nitric oxide inhalation was much likely to be higher than the critical closing pressure described in the portal venous circulation for normal livers, leading to an impaired flow through this circuit.
In light of the deleterious effects of a hepatic congestion for the portal circulation emphasized by Heinrich and Kao, manipulation of the hepatic arterial flow might be of more limited interest in the reported case. First, the arterial buffer response, which has been confirmed to exist in humans, might have led to a relatively high hepatic arterial flow in this patient with a presumably highly reduced portal venous flow. Second, trying to increase portal inflow to the liver would imply acting on mesenteric blood flow and thus on systemic blood flow, which can be limited in the described clinical situation.
Concerning the atrial fibrillation, we agree on the comment raising the necessity to control the ventricular rate. However, cardioversion was judged dangerous at the acute phase of the decompensation and was considered as a consequence rather than the cause of the right heart failure. Accordingly, normal sinus rhythm spontaneously recovered after the hemodynamic improvement, suggesting that the atrial fibrillation resulted from rather than in hemodynamic dysfunction. Two years after this acute episode, the patient is still in good hemodynamic condition and has been removed from the heart transplantation list.
In conclusion, we believe that the observed beneficial effect of inhaled nitric oxide in this patient was the consequence of a significant decrease in hepatic venous pressure while mean arterial pressure was maintained. For the hospitals that "do not have the proper permission or setup for nitric oxide administration," we could advocate the use of a pharmacologic combination therapy to achieve these goals, although such a combination might be less effective or more difficult to use.
Claire Gatecel, M.D., Instructor.
Alexandre Mebazaa, M.D., Ph.D., Assistant Professor.
Robert Kong, M.B., B.S., F.R.C.A., Instructor.
Nathalie Guinard, M.D., Instructor.
Nathalie Kermarrec, M.D., Resident.
Joaquim Mateo, M.D., Instructor.
Didier Payen, M.D., Ph.D., Professor, Department of Anesthesiology and Critical Care, Hopital Lariboisiere, 2, rue, Ambroise Pare, 75010 Paris, France.
(Accepted for publication April 13, 1995.)