In vivo assessment of the window of barrier opening after osmotic blood-brain barrier disruption in humans

Object. Osmotic blood-brain barrier (BBB) disruption induced by intraarteri al infusion of mannitol is used in conjunction with chemotherapy to treat h uman brain tumors. The time course to barrier closure, or the so-called the rapeutic window, has been examined in animals but little information is ava ilable in humans. The authors, therefore assessed the time course to barrie r closure after osmotic BBB disruption in humans. Methods. Disruption of the BBB was demonstrated using Tc-99m-glucoheptonate (TcGH) single-photon emission computerized tomography (SPECT) scanning in 12 patients who were treated monthly with combination chemotherapy in conju nction with BBB disruption. The primary diagnosis was primary central nervo us system lymphoma in seven patients and primitive neuroectodermal tumors i n five. The TcGH (20 mCi) was injected at 1- to 480-minute intervals after osmotic BBB disruption, and patients underwent SPECT scanning after 4 hours . A total of 38 studies was performed. Good-to-excellent BBB disruption was obtained in 29 procedures and poor-to-moderate disruption was seen in the other nine studies. The TcGH indices correlated with the degree of BBB disruption as measured p ostprocedure on contrast-enhanced CT scans (r = 0.852). Mean baseline TcGH indices were 1.02 +/- 0.07. For the group of patients with good-to-excel le nt disruptions the mean indices at 1 minute postdisruption measured 2.19 +/ - 0.18. After 40 minutes no significant change was noted (mean index 2.13 /- 0.2). Then the indices declined more steeply and at 120 minutes after th e disruption the index was 1.36 +/- 0.02. A very slow decline was noted bet ween 120 and 240 minutes after mannitol infusion. At 240 minutes the barrie r was still open for all good-to-excellent disruptions (index 1.33 +/- 0.08 ) but at 480 minutes the mean indices had returned to the baseline level. Conclusions. Results of these in vivo human studies indicate that the time course to closure of the disrupted BBB for low-molecular-weight complexes i s longer than previously estimated. The barrier is widely open during the f irst 40 minutes after osmotic BBB disruption and returns to baseline levels only after 6 to 8 hours following the induction of good or excellent disru ption. These findings have important clinical implications for the design o f therapeutic protocols.