The clinical use of sodium fluorescein as an aid in the diagnosis and localization of brain tumors has been previously documented (6, 11, 14). Briefly, the technic used at the present time is as follows: All patients subjected to craniotomy are injected immediately preceding surgery with 1.0 gram of sodium fluorescein. At operation, the cortex of the brain is examined under ultraviolet light. If a superficial tumor is present, it can be easily distinguished and its limits outlined by the presence of the yellow-green fluorescence typical of fluorescein. If the tumor is below the cortex and gives no indication of its presence on examination and palpation, brain needles are commonly used to probe for it and to obtain biospy material. With singular consistency, the presence or absence of fluorescence of the tissue fragments procured has been indicative of the presence or absence of abnormal tissue. Edematous brain tissue fluoresces slightly, and therefore, even if the surgeon does not strike the tumor directly, its proximity can be predicted by the presence of a slight fluorescence of the biopsy material. If a second biopsy taken from an area adjacent to the first reveals no fluorescence, the surgeon knows that the tumor is in the opposite direction. Among the last 104 patients suspected of having brain tumors in whom the fluorescein method was used, there were 96 with tumors which fluoresced, 8 whose tumors did not fluoresce, 6 who had no tumor and no fluorescent tissue, and only one in whom fluorescence was described and no tumor found. Figure 1 is an operative specimen photographed with ordinary illumination. The tumor, a recurrent astrocytoma, is surrounded by a rim of normal brain tissue. Figure 2 is a photograph of the same specimen exposed to ultraviolet light; the centrally situated tumor tissue is fluorescent in contrast to the peripherally located normal brain tissue. The patient was given one gram of sodium fluorescein (5 per cent solution) one hour before operative removal of the specimen. The 8 tumors which did not fluoresce were 3 diffuse astrocytomas in patients with long histories, 1 angioblastoma which was almost completely necrotic, 1 medulloblastoma, and 3 diagnosed as cholesteatomas. These errors represent the limitations of the procedure. These tumors, with a single exception, were benign and slow-growing. The affinity of fluorescein for tumor tissue is not specific but can be demonstrated in any lesion of the central nervous system which breaks down the so-called “blood-brain-barrier mechanism” (BBB). Thus one would expect, as mentioned previously, that areas of brain edema would fluoresce slightly, as would the capsules of brain abscesses, granulomas, and traumatized tissue. The membranes of subdural hematomas also fluoresce. For the most part, these can be easily distinguished from the more vividly fluorescent neoplastic lesions. Metastatic tumors, regardless of type, fluoresce brilliantly.
Read full abstract