Physiologic approach to dialysis-induced hypoxemia. Effects of dialyzer material and dialysate composition

This study was undertaken to evaluate the effects of membrane-related complement activation and dialysate composition on dialysis-associated hypoxemia. Seven chronic hemodialysis patients were hemodialyzed 3 times sequentially with the following three combinations; Cuprophan membrane with acetate dialysate, polymethylmethacrylate (PMMA) membrane with acetate dialysate, and PMMA membrane with bicarbonate dialysate. During dialysis with acetate dialysate, the pulmonary diffusing capacity (DLco) at 30 min after the start of dialysis was decreased to 88% (p less than 0.01) of the predialysis value with PMMA and to 79% (p less than 0.01) with Cuprophan, and the degree of DLco on PMMA membrane was different from that on Cuprophan (p less than 0.01). The degree of leukopenia with PMMA was less than that with Cuprophan. However, the fall in DLco did not alter the alveolar-arterial O2 tension gradient. Although the changes in transcutaneous PO2 (tcPO2) were not constant in all three combinations, a distinct fall in tcPO2 was observed in the first half of dialysis with acetate dialysate. During dialysis with acetate dialysate but not with bicarbonate, the extracorporeal dialyzer removed an average of 60 ml/min of CO2, and the respiratory quotient dropped from a mean predialysis value of 0.86 to 0.59 (p less than 0.001). The arterial CO2 tension was not significantly changed throughout dialysis, but the alveolar ventilation decreased significantly in proportion to the fall in carbon dioxide output. The arterial tension fell from a control level of 91 +/- 6 to 77 +/- 8 mm Hg (p less than 0.01) in 30 min. It is concluded that, in spite of a fall in DLco, dialysis-induced hypoxemia in this group of patients on maintenance dialysis is caused by CO2 loss via the dialyzer, resulting in reflex hypoventilation.