Introduction

Over the last few years, frameless stereotactic biopsies have progressively replaced frame-based stereotactic procedures as a first line diagnostic approach to various brain lesions. The combined use of intraoperative magnetic resonance imaging (MRI) further increases the spatial resolution of the probe target correlation in these procedures [7]. Yet, despite the increasing accuracy of frameless stereotactic procedures, failure to correctly establish tumor histology and dignity occurs in approximately 25% of cases, mainly due to tissue heterogeneity and sampling bias [3, 7]. In cases where the frameless stereotactic biopsy is based solely on preoperative computertomographic (CT) or MRI images, the results are even less favorable with 37% [1] to 49% [5] of incorrectly diagnosed gliomas. In order to avoid sampling bias, obtaining multiple specimens has been advocated [12]. Since stereotactic procedures are the diagnostic method of choice for deep-seated lesions, within the brain stem, or within the so-called eloquent areas, taking additional samples during the same procedure to improve the diagnostic accuracy can be risky or even impossible. A precise histological diagnosis, however, is crucial for appropriate treatment guidance and prognosis. To decrease the morbidity and mortality risks associated with multiple sampling, we introduced fluorescence-controlled stereotaxy in our department [4]; additional cases were published in support of the concept by others [8]. We here present a series of 17 consecutive patients where we combined 5-aminolevulinic acid (5-ALA)-induced fluorescence with frameless stereotaxy for rapid biopsy accuracy verification in cases of suspected malignant or high-grade brain tumors.

Methods

Four hours prior to surgery the patient is given an oral dose of a 20-mg/kg body weight aqueous solution of 5-aminolevulinic acid hydrochloride (Gliolan® medac GmbH, Hamburg, Germany) [11]. The MRI data used for biopsy planning and guidance are obtained either preoperatively using a standardized high-resolution imaging protocol or, if available, intraoperatively after the patient has been positioned on the operating table. After general anesthesia, burr hole placement and dural opening at the determined entry point are performed in the usual fashion. Using frameless stereotaxy navigation (StealthStation®, Medtronic, Louisville, KY, USA), the biopsy needle is then advanced into the targeted lesion, guided by the MRI data, and a first sample is taken. After removal of the needle it is placed under a 405–440-nm blue light source (either a dedicated light source or the light source of an operating microscope fitted for 5-ALA-induced PpIX fluorescence) with the specimen still in place to look for fluorescence. In case of strong positive fluorescence, no frozen section is obtained, and no further samples are taken.

Illustrative case

A 71-year-old male (Table 1, case 01) with multiple intracerebral lesions of unknown origin underwent a frameless stereotactic biopsy in the periventricular area of the right frontal horn. As a proof of concept, the accuracy of the calculations and needle placement was verified by obtaining an intraoperative MRI scan with the titanium biopsy needle (Pajunk Medizintechnologie, Geisingen, Germany) still in place. After removal of the needle and blue light illumination of the tissue sample, a positive red fluorescence could be observed. As the opening of the biopsy needle (10 mm) was larger than the target lesion (5 mm), and the needle itself has a 2-mm forerun (Fig. 1, top), only the distal part of the actual specimen was expected to hold pathological tissue, which was confirmed by visible 5-ALA-induced protoporphyrin IX (PpIX) fluorescence only within the tumor tissue (Fig. 1, bottom). Histology and immunohistochemistry revealed a malignant lymphoma (B-cell lymphoma), and showed the border between the target lesion and the surrounding brain (Fig. 2) as predicted by fluorescence and MRI.

Table 1 Demographic data, detected fluorescence and final pathology of 17 patients who underwent a frameless stereotactic brain biopsy after oral administration of Gliolan® (5-ALA). Fluorescence was assessed under a 440-nm blue light source
Full size table
Fig. 1
figure 1

The opening of the biopsy needle (10 mm) is larger than the target lesion (5 mm), and the needle itself has a 2-mm fore-run (Fig. 1, top). Only the distal part of the actual specimen was expected to hold pathological tissue, which was confirmed by visible 5-ALA-induced protoporphyrin IX (PpIX) fluorescence only within the tumor tissue (Fig. 1, bottom)

Full size image
Fig. 2
figure 2

Immunohistochemistry revealed a malignant lymphoma (B-cell lymphoma), and showed the border between the target lesion and the surrounding brain (CD 20)

Full size image

Results

Seventeen biopsies using a combination of frameless stereotaxy and 5-ALA-induced fluorescence were performed (Table 1). In 13 cases a positive fluorescence was detected and histological examination confirmed a malignant or high-grade brain neoplasm in all cases. The specimens of the remaining four biopsies were fluorescence-negative and required frozen section confirmation and/or further samples. In these cases histology showed either nonspecific gliotic changes or low-grade tumors, both of which are known not to induce any fluorescence after 5-ALA application.

In the fluorescence-positive cases the procedure time could be reduced by 30–45 min, since no frozen sections needed to be obtained and no resampling was necessary.

In this small series of patients there were no complications related either to the use of 5-ALA or to the surgical procedure itself.

Discussion

Stereotactic brain biopsies carry a morbidity and mortality risk of 4% and 0.9%, respectively, and lead to an inaccurate or imprecise diagnosis in one third to one half of glioma cases [5]. Further improvement can be achieved by obtaining multiple specimens and/or the additional use of intraoperative imaging [7]. Since patients with malignant gliomas fare better with gross total resection [6], bioptic procedures are usually limited to inoperable or small lesions that are deeply located (e.g., brain stem) or situated in eloquent cortical areas. Unfortunately, these typical areas for biopsy are the least suitable for multiple tissue sampling. Intraoperative frozen sections can aid the neurosurgeon in verifying the adequate biopsy sampling location, but they are time consuming and also prone to errors.

5-ALA is a non-fluorescent “pro-drug” that, due to a metabolism disturbance, is transformed into fluorescent PpIX in the tumor cells of high-grade gliomas [10] and other highly proliferative neoplasms. This property is used to achieve radical tumor resection and thereby extend progression-free survival in high-grade glioma patients [11]. In the case of stereotactic brain biopsy, the use of protoporphyrin fluorescence does not aim at optimal resection, but allows the procedure to be time-efficient and as minimally invasive as possible by reducing the risk of multiple tumor samplings. The sensitivity and specificity of malignant tissue detection have been reported to be 85% and 100%, respectively [2]. The PPV of observable intraoperative PpIX fluorescence for non-normal tissue is found to be as high as 98.9% [9]. Therefore, a sample showing strong positive fluorescence most likely represents the targeted lesion. Although we used intraoperative MR imaging (PoleStar N20®, Medtronic, Yokneam Elit, Israel) for navigation to assess the accuracy of the method described and to rule-out an acute hemorrhagic complication, the same methodology is applicable to frameless stereotactic procedures based on preoperative CT or MR images, the main purpose being that no further samplings are required in case of a positive fluorescence.

Negative fluorescence can either be due to a sampling error (small lesion, preoperative imaging not accounting for brain shift) or to the histological nature of the pathology (non-fluorescent tumor or lesion). In these cases frozen sections are required to assess the adequacy of the sampling location; if available, intraoperative imaging can also be used to correct the needle position relative to the intended target and can be of invaluable help in very small lesions.

Conclusion

The addition of 5-ALA-induced PpIX fluorescence proved useful in frameless stereotactic biopsies of malignant or high-grade brain tumors where immediate strong intraoperative fluorescence was observed. The procedure time could be reduced and its safety increased by eliminating the need for multiple samples. In these cases frozen section verification may not be mandatory, since the final pathology was conclusive in all instances of positive fluorescence. The method is easy to implement and requires no modification to the biopsy procedure itself. Fluorescence-negative cases represent either a missed target or a non-fluorescent pathology, and therefore do require frozen sections and/or intraoperative imaging to verify the adequacy of the sampled target.