Plasma exchange morbidity in Guillain-Barré syndrome: results from the French prospective, randomized, multicenter study. The French Cooperative Group.

OBJECTIVES

To describe all adverse events occurring during plasma exchange sessions in adult patients with the Guillain-Barré syndrome. To analyze these events with regard to the technical modalities and biological changes induced by sessions, and to try to identify a population at high risk for adverse events.

DESIGN

Double-blind, randomized, prospective, multicenter trial.

SETTING

A total of 28 French and Swiss intensive care units.

PATIENTS

The study is based on 220 patients allocated either to plasma exchange (n = 109) or not (n = 111). This study focused on 105 patients who received at least one plasma exchange, with replacement fluid secondly allocated by randomization to albumin, or fresh frozen plasma. A total of 105 patients underwent 390 plasma exchanges. Fifty-five patients received albumin (208 sessions) as replacement fluid, and 50 patients received fresh frozen plasma (182 sessions).

INTERVENTIONS

Prospective monitoring of patients for each session including technical modalities, adverse effects, and biological parameters.

MEASUREMENTS AND MAIN RESULTS

A total of 253 adverse incidents were recorded. At least one adverse incident occurred in 39% of plasma exchange sessions among 80 (76%) patients. In 15 patients, plasma exchange treatment had to be discontinued because of severe intolerance (six patients, including three patients with severe bradycardias), intercurrent complications, mainly infections (four patients), and technical difficulties. One patient with pneumococcal septicemia and pneumonia died during the second plasma exchange session. Fresh frozen plasma was associated with more adverse incidents than albumin (135 vs. 118, p = .008). The occurrence of adverse events was also related to the preplasma exchange hemoglobin level assessed before the session (p = .04). Otherwise, the frequency of adverse effects did not depend on technical modalities (type of equipment, anticoagulation). Age, sex, previous history, neurologic severity, and the need for mechanical ventilation, as assessed on inclusion in the study, did not modify the risk of adverse effects. Finally, occurrence of bradycardia did not rely on initial neurologic severity.

CONCLUSIONS

These results confirm that fresh frozen plasma should be abandoned as replacement fluid in plasma exchanges of Guillain-Barré syndrome patients. They also underline the need for close monitoring of patients during sessions and, especially, the respect of treatment contraindications. Some adverse incidents could be attributed to the underlying disease rather than to the plasma exchange session.