The products and lots of intravenous immunoglobulin preparations we used for patients 1 and 4 Table 1 of our article [1]) were 3 monthly infusions of the same lot of Gamimune 5% (Miles, Inc., West Haven, Connecticut); these patients developed aseptic meningitis after each infusion. Patient 1 received 14 additional twice monthly infusions, 6 with different lots of Gamimune and 8 with Gammagard (Baxter Healthcare, Glendale, California). He again developed aseptic meningitis after each infusion except with the last two. Patient 2 developed aseptic meningitis after receiving 1 infusion of Gamimune 5%; he subsequently received 13 monthly infusions of Gamimune 10% from different lots without recurrence of the meningitis. Patient 3 developed aseptic meningitis after the first infusion of Gamimune; he subsequently received 2 monthly infusions of Gamimune from a different lot without a recurrence. Patient 5 developed aseptic meningitis after each of 4 monthly infusions of Gamimune, including different lots. She subsequently received 5 monthly infusions of Gammar (Armour Pharmaceutical, Collegeville, Pennsylvania); aseptic meningitis developed after each infusion. One treatment with Sandoglobulin (Sandoz Pharmaceuticals, East Hanover, New Jersey) resulted in an even more severe episode of aseptic meningitis. A sixth patient, not included in our published series [1], received 24 monthly infusions of Gamimune but did not have symptoms of aseptic meningitis. After receiving Sandoglobulin, she developed the first episode of aseptic meningitis.

These observations indicate that 1) aseptic meningitis can occur with all products, even in the same person; 2) a particular product is not more "meningogenic" than others, but some lots of the same product can be; and 3) aseptic meningitis is unrelated to the duration of treatment (2 days compared with 5 days) or to the total dose because it can develop even during the first day of infusion [2]. The dose used for the treatment of neurologic autoimmune diseases is not higher, as previously reported [3], but rather is the same as that used for idiopathic thrombocytopenic purpura and Kawasaki disease. The development of aseptic meningitis with all the products but only with certain lots (derived from various pools of donors) suggests that the triggering factor may not be the stabilizing agents within the intravenous immunoglobulin but the IgG itself, its autoantibody specificity [4], or the concentration of soluble molecules, such as human leukocyte antigens and cytokines^[5].

Dr. Peetermans and colleagues recommend an empirical treatment with antibiotics until the spinal fluid cultures are negative for bacterial meningitis. I have a different view. We described a large series of patients to alert the physicians that the meningitis was aseptic—not bacterial—and that there were no signs of subarachnoid hemorrhage. This has been the experience of others, including several cases communicated to me. The temporal profile of the aseptic meningitis associated with intravenous immunoglobulin is predictable and characteristic [1]. The patients become sick but their reaction is short-lived, and they always recover within 2 to 3 days. Not one case of intravenous immunoglobulin-induced bacterial meningitis has been reported. My recommendation, therefore, is to clinically monitor the patients and to treat them symptomatically with narcotic analgesics and antiemetics. Treatment with intravenous immunoglobulin requires repeated infusions; should we do a lumbar puncture in the same patient each time aseptic meningitis develops? The clinical picture is also identical to subarachnoid hemorrhage; should we obtain a computed tomographic scan of the brain each time? I do not believe so, except if exceptional circumstances shake confidence in an expert clinical judgment.