Decrease in arterial oxygen partial pressure within the first 24 h of rhGM-CSF administration in AML patients

GM-CSF may induce pulmonary complications, such as dyspnea with temporary decreases in oxygen saturation described as first dose effect for higher dosages of intravenous rhGM-CSF. This study investigated possible pulmonary disturbances in adult de novo AML patients receiving yeast rhGM-CSF 24 h prior to chemotherapy under phase II/III conditions. Eighteen patients were monitored for 22 treatment episodes. GM-CSF was given s.c. 1 q.d., 2 q.d. or continuously i.v. at 250 micrograms/m2/d 24 h prior to induction chemotherapy (TAD9, n = 18) and consolidation (TAD9, n = 4). Spirometry, bodyplethysmography, single breath-diffusion capacity (DLCO) and arterial blood gas analyses were obtained prior to GM-CSF, and repeated after 24 h. Pulse oxymetric oxygen saturation (saO2) was registered continuously for the first 16 h within day 1 of rhGM-CSF treatment. Patients were aged 21-75 years. The saO2 monitoring did not reveal any first dose effect. PaO2 values decreased from 78.9 mmHg before GM-CSF to 72.8 mmHg after 24 h (p < 0.01, maximum shift 15 mmHg). PaO2 shifts occurred mainly with pre-existing lowered paO2, but otherwise were independent of age, the route of GM-CSF administration, leukocyte levels, or increase of leukocytes with GM-CSF. Increases in AaDO2 reflected the paO2 shifts (p < 0.05). No dyspnea corresponded to these changes. DLCO values did not decrease significantly after 24 h. Summarily, contemporary dosage of yeast rhGM-CSF avoids short-term oxygen desaturations, but leads to clinically benign impairment in oxygen tension, based on ventilation/perfusion mismatches. This should be taken into account for patients starting at subnormal paO2.