We report a case of non-cardiogenic pulmonary edema following granulocyte-colony stimulating factor (G-CSF) administration in a bone marrow donor. During leukapheresis the patient was asymptomatic, but the next day she developed dyspnea, hypoxia, and bilateral pleural effusions. The patient clinically improved with diuretic and steroid therapy. We propose the significant leukocytosis triggered an inflammatory response leading to acute lung injury.
A 21 year old white female at 28 weeks gestation had G-CSF administration for bone marrow donation for her sister. She was healthy and her pregnancy had been unremarkable. Her only medications were prenatal vitamins and iron. She received G-CSF 900 μg (10 μg/kg/day) intravenous daily for three days via a #12 French catheter in the internal jugular vein. On day four, the WBC peaked at 40.0 10e9/L and the patient underwent leukapheresis. The next morning, the catheter was removed taking care to avoid air embolism. Minutes later while walking to her car, the patient developed shortness of breath and cough. She denied any associated chest pain. She had a presyncopal episode and was taken to the emergency room. She was hypoxic with arterial oxygen tension of 88 mmHg on 100% fractional inspired oxygen via nonrebreather mask. On examination she was normotensive, alert and oriented, with diffuse bilateral rales. Laboratory data showed normal kidney and liver function. WBC had decreased to 33.0 10e9/L. Hemoglobin was 12.6 g/dL. Platelets were at a nadir of 83 10e9/L following apheresis. A CT angiogram performed immediately on arrival to emergency room showed bilateral pleural effusions without evidence of clot or air embolism. Echocardiogram showed normal systolic function. Serial cardiac enzymes and electrocardiogram were normal. The patient received lasix and methylprednisone 40 mg intravenous every 12 hours. After several days she had normal oxygen saturation on room air. Methylprednisone was stopped and the patient was discharged home.
Pulmonary edema has been reported in patients receiving G-CSF. Capillary leak sydrome has been described in bone marrow recipients pretreated with G-CSF and is thought to be due to leukocyte activation with recruitment of inflammatory mediators leading to systemic inflammation and increased capillary permeability. Acute lung injury in bone marrow donors pretreated with G-CSF has been associated with increased levels of interleukin-1 beta as in the case described by Arimura et al. We are not aware of any reported cases of non-cardiogenic pulmonary edema after G-CSF in pregnant donors. The differential diagnosis also included embolic phenomenon. Air embolism was ruled out by CT, but remains a diagnositc consideration in the appropriate clinical setting. Air embolism causes the formation of platelet-fibrin aggregates and microthrombi leading to pulmonary hypertension and ultimately capillary leak. Although G-CSF and air embolism can both result in capillary leak, the treament is very different for each of these conditions depending on the underlying mechanism.
In this case, marked leukocytosis following G-CSF mobilization of peripheral blood progenitor cells may have triggered the activation of inflammatory mediators in the lung resulting in non-cardiogenic pulmonary edema. Our patient clinically improved with diuretic as well as with the anti-inflammatory effect of steroids. Thus we propose the diagnosis of pulmonary edema due to G-CSF induced capillary leak.
Courtney Lucado, None.