Historically, the use of a fluorophore for highlighting malignant glioma tissue was first described by Moore et al. [10] as a means of localizing brain tumors. Equipped only with topographical information from neurological examinations and vague morphological–anatomical data from ventriculographic studies, and intra-operatively with a brain needle for finding tumor-afflicted brain and biopsies of tissue, these authors employed a mercury vapor lamp and a Wood’s filter together with intravenous fluorescein to locate and resect tumors. They noted that in some cases, they were able to identify additional small regions of tumor, which escaped notice when viewed with ordinary illumination, thus first defining the two key elements of the use of fluorescence for surgery of brain tumors:

Regarding fluorescein, they acknowledged, however, that “the brain tissue immediately surrounding the tumor is edematous and edematous brain tissue is shown to retain a greater amount of dye than otherwise normal brain” and “edematous tissue surrounding the tumor does fluoresce, but to a lesser degree…” thus already pointing out possible limitations of dyes which reach the brain tumor via a breached blood–brain barrier in the tumor, i.e., of passive permeability markers. At that point of time, the concept of a blood–brain barrier was long accepted, but the lack of such in tumor was still in the realm of theory and put forward by Moore et al. [10] as a possible explanation for the differential uptake of dye when comparing tumor, edematous tissue, and normal brain. In summary however, the authors felt fluorescence to have “definitive clinical value.” On a by note, this paper already envisages a coupling of 131-iodine to fluorescein dyes for better diagnosis of malignant brain tumors.

After this first report little was heard about fluorescence in neurosurgery until the late 1990s when Kuoriwa and co-workers from Japan presented three publications on fluorescein for glioma surgery the last in 1999 [11–13].

In 1998, we published our first reports [14–16] on 5-aminolevulinic acid (5-ALA) for fluorescence-guided resections providing data on both experimental data and data on the first use of this compound in humans in a small phase I/II patient cohort. In contradistinction to fluorescein, 5-ALA is a prodrug, which is converted into a fluorescing fluorophore within tumor cells [17, 18] and thus is unique so far, compared to other fluorophores that have been investigated. 5-ALA has been studied extensively for fluorescence-guided resections since then.

Recently, with new filters integrated into operating microscopes, fluorescein is again receiving some, albeit controversial, interest [19–22].

Both fluorochromes will be discussed in detail in the course of this paper. Currently, new approaches with fluorophores targeted to surface markers of tumors cells are being investigated and are on the verge of clinical translation.

5-Aminolevulinic acid (ALA) is a natural metabolite, which is involved in heme-biosynthesis. When given in excess, this compound results in the accumulation and/or selective retention of fluorescing porphyrins, particularly protoporphyrin IX (PPIX) in epithelia and more predominantly in cancers arising from such epithelia and in many brain tumors [17, 18, 23]. Under physiological circumstances, PPIX is converted into non-fluorescent proto-heme by iron chelation and from there into hemoglobin. In experiments utilizing the C6 glioma cell line in vitro and in vitro we were able to first demonstrate selective retention in malignant glioma cells [15], an observation, which was safely extended to human malignant gliomas [16]. The reasons for this selectivity are not completely understood. Relationships have been postulated to decreased expression of ferrochelatase in tumors [24], an enzyme which converts PPIX to proto-heme by chelating iron into the porphyrin ring. Other possible explanations have been put forward, such as a decreased activity of the ATP-binding cassette transporter (ABCB6) [25] which eliminates porphyrins from cells, as well as tumor cell density, tumor cell proliferative activity and a disturbed blood–brain barrier permeability in the region of the tumor [26, 27].

PPIX is strongly fluorescent with an excitation maximum at approximately 405 nm (the so-called Soret band) with two emission peaks at 635 and 704 nm (Fig. 6.2) [28]. In vivo, the spectrum detected from tissue has an additional peak at between 450 and 500 nm, i.e., in the green range, resulting from underlying brain tissue autofluorescence from NADH, riboflavins and carotenoids and other autofluorophores in normal tissue. This autofluorescence is equally stimulated by blue excitation light [29]. Surgical microscopes adapted for protoporphyrin IX detection currently use xenon excitation light filtered for violet-blue with a wavelength peak at 375–440 nm. For observing fluorescence they do not simply use a long-pass or barrier filter for visualizing the 635 nm peak of protoporphyrin IX, that is, cutting away all wavelengths below for instance 600 nm. This would result in PPIX being visible on a black background, without allowing any background discrimination. We considered the possibility of background discrimination to be important for enabling resection and tissue manipulation by the surgeon using the fluorescence mode. Rather, microscopes by the major manufactures use barrier filters for fluorescence detection, which open at slightly below 440 nm. This allows a small fraction of the excitation light to be included in the observation pathway of the microscopes. In addition, tissue autofluorescence, which is located in the blue-green range, is also visualized. Remitted excitation light and tissue autofluorescence together enable background brain tissue to become visible with a green-blue tone [14].

5-ALA is currently approved for resections of malignant gliomas in many countries of the world after initial approval by the European Medicines Agency (EMA). Approval was granted based on a study [30], which randomized patients with malignant gliomas to be operated on either with or without 5-ALA for fluorescence-guided resections. Groups were compared for extent of resection and safety as well as for progression-free survival. Extents of resection were greater in the 5-ALA group, resulting in a prolongation of progression-free survival without significantly more frequent neurological deficits. Rates of complete resection of contrast-enhancing tumor were 65% for the 5-ALA versus 35% for the control arm with conventional white light illumination.

On a by note, the rate of 65% is frequently cited as the limit to what can be achieved using 5-ALA. This may not be the case however, since patients operated on using 5-ALA in the context of the study were the first dozen or so patients ever to be operated on by surgeons with relatively little experience with the novel technology. Newer studies put forward by experienced surgeons indicate a much higher potential of 5-ALA for facilitating gross total resections of enhancing tumors in up to 96% of cases (Table 6.1), when used in conjunction with neurophysiological mapping and monitoring [31–35]. In one recent study, intra-operative MRI in combination with neurophysiological methods resulted in 82% “complete” resections [35]. When augmented with 5-ALA-induced fluorescence, gross total tumor resection rates were stated to have been reached in 100%. Overall, a meta-analysis by Eljamel [36] indicated a rate of gross total resection, i.e., resection of more than 98% of enhancing tumor in over 75% of a total of 565 patients.

In retrospect, utilizing resection rates as a primary study aim for a randomized study may not have been a well-chosen endpoint for testing the usefulness of 5-ALA. Resection depends on many factors other than visualizing the tumor, foremost the worry about maintaining function, which depends strongly on the surgeon, the surgeon’s convictions regarding the value of resection and the use of mapping/monitoring for safely extending resections. Particularly, the concepts and views regarding these factors have changed over time. Today the value of safe resection is more generally accepted in the neurosurgical community and mapping/monitoring has become a common adjunct for surgery of malignant gliomas. Thus, it may have been more appropriate to test 5-ALA as an intra-operative diagnostic agent or contrast medium. This would have involved testing the compound for toxicological safety as well as for measures of diagnostic accuracy only, e.g., looking at the positive predictive value of fluorescence for demonstrating tumor rather than attempting to prove a therapeutic value of the use of 5-ALA. This potential therapeutic value is the sum of patient selection and intra-operative tumor detection but mostly of the surgeon performing the surgery and utilizing any available technology, the surgeon’s role being highly uncontrollable and variable.

5-ALA is presently being marketed as “Gliolan^®” in the EU and other parts of the world, and surgical microscopes modified for fluorescence-guided resections are available from the leading microscope manufacturers, for which fluorescence modules can be purchased as accessories. These microscopes uniformly feature xenon light sources generating conventional white light and are switchable to blue light illumination and the appropriate observation filters, usually by operating a switch or foot pedal, thus enabling free shuttling between both illumination modes.