Laser-induced fluorescent visualization and photodynamic therapy in surgical treatment of glial brain tumors

: High-grade gliomas have a diffuse and infiltrative nature of the growth of tumor cells, due to which the achievement of radical resection is difficult. Surgical resection completeness of brain tumors is an important factor in prolonging the life of patients. An accurate definition of tumor boundaries and residual fluorescent regions is impossible due to imperfections of the equipment used for fluorescent imaging. 5-aminolevulinic acid (5-ALA) is a precursor of protoporphyrin IX (PpIX) in humans and is clinically used to detect and treat tumors. Currently, fluorescence-guided surgery with PpIX used a surgical microscope with an excitation wavelength in the blue spectrum range. Because of its low ability to penetrate into biological tissue, blue light is ineffective for providing high-quality fluorescent navigation. Also, when performing an operation using radiation in the blue spectrum range, the photosensitizer’s surface layer (PS) often bleaches out, which leads to frequent errors. The use of red light emission makes it possible to slow down the PS bleaches out due to the absorption properties of PpIX, but this task is technically more complicated and requires highly sensitive cameras and specialized optical filters. The new two-channel video system for fluorescent navigation has a radiation source in the red range of the spectrum, the penetration depth of which is greater than the blue light, which makes it possible to increase the depth of probing into biological tissues. The study’s clinical part involved 5 patients with high grade glioma and 1 patient with low grade glioma: grade III oligodendrogliomas (2), grade IV glioblastomas (3), and grade II diffusion astrocytoma (1).

In the case of HGG, the degree of surgical resection (DSR) is an important prognostic factor for further recurrence [6,7]. Surgical resection of a recurrent tumor with DSR > 98% favorably affects the patient's life expectancy. In patients with initially diagnosed HGG, resection with DSR ≥ 78% can prolong overall survival [8].
After the studies had demonstrated clinical advantages in terms of completeness of tumor resection (65% of complete resections with 5-ALA versus 36% under white light) and survival without tumor progression, fluorescence control with 5-ALA has been successfully applied for resection of malignant gliomas in Europe [9]. This ratio is not observed in glioblastoma cases, where the difference in percentage values is either minimal or less significant.
Achieving the maximum degree of resection during surgery is often difficult due to inaccurate determination of the tumor invasion during direct imaging. Comparing the obtained data of fluorescent navigation with 5-ALA and intraoperative navigation under white light, in more than half of the patients, the tumor boundaries did not coincide [10][11][12]. To improve the quality of monitoring and PDT treatment of the postoperative tumor bed, a clinical study was carried out using a two-channel video system for fluorescence diagnostics and photodynamic therapy instead of the commonly used surgical microscope.
Usually, to perform a fluorescence-guided resection, a surgical microscope with 405 nm excitation is used. Since the depth of light penetration into the tissue for this wavelength is about 2-3 mm, and the depth of fluorescence probing is 0.5 mm, this choice of the excitation wavelength is not optimal [13]. In addition, when performing an operation using the radiation in the blue spectrum range, the surface layer of the photosensitizer burns out faster than when excited with a red light. The wavelength of the exciting radiation used in the video system is 635 nm, which falls into the optical transparency window of biological tissue (625 -950 nm) [14]. It has a greater depth of probing compared to the one used in microscope. Based on [15,16], the penetration depth of laser radiation with a wavelength of 635 nm into the biological tissues of the human head is 7-13.5 mm.
A malignant tissue biopsy is performed before surgery to determine the type of tumor with histological analysis. When conducting a biopsy, it is necessary to minimize damage to healthy tissues. Therefore, the study is carried out use a stereotactic approach. It is necessary to ensure the exact location of the needle in the tumor for correct analysis. This study presents the using of the developed stereotactic needle for biopsy with simultaneous spectroscopic control [17] in the scope of applying fluorescence diagnostics (FD) and photodynamic therapy. The stereotactic approach was performed in a case when the tumor was being deep-seated. A stereotactic needle was used to determine the concentration of the PS accumulated in the tissues and to establish the actual tumor boundaries by using recorded fluorescence spectra.

5-aminolevulinic acid
5-ALA is a precursor of PpIX in humans and is clinically used for the detection and treatment of tumors as an FDA-approved drug [18]. Once inside the cell, 5-ALA takes part in the heme biosynthesis, where it undergoes the conversion into porphobilinogen, hydroxymethylbilane, uroporphyrinogen IX, PpIX, and, finally, heme. In tumor cells, at the stage of converting 5-ALA to porphobilinogen, the activity of porphobilinogen deaminase is high, which leads to a rapid conversion of porphobilinogen to uroporphyrinogen III [19]. The conversion of PpIX to heme is, however, slow due to low ferrochelatase activity. This regulation results in selective accumulation of PpIX in tumor cells, which results in fluorescence when exposed to laser radiation.
The most intense absorption of PpIX is 405 nm and 632 nm. Peaks of PpIX fluorescence are observed at 635 nm and 705 nm. According to the results of scientific studies, the maximum fluorescence in a brain tumor is achieved 6 hours after the introduction of PpIX into the human body [18]. Low-intensity fluorescence begins to appear after 3 hours, while from 9 hours on, the fluorescence decreases.

Clinical part
The clinical part of the study involved 5 patients who underwent surgery after a glioma diagnosis based on the results of histology and preoperative diagnostics MRI at the Moscow State University of Medicine Dentistry (Table 1). Preoperative diagnoses in 2 patients -grade 3 oligodendroglioma, in 3 patients -grade 4 glioblastoma recurrence, and in one patient -grade II diffusion astrocytoma. The patients in the clinical part of the study had different tumor sizes. All tumors were localized in the cerebral cortex. The preoperative treatment received for each patient varied depending on the presence of a relapse. The state of the patients was assessed according to the Karnofsky Performance Scale (KPS).

Two-channel video-system
The two-channel video system for fluorescence diagnostics (UFPh-630-01, BIOSPEC) was used for laser-induced fluorescent monitoring in a clinical study. Figure 1 shows the scheme of a two-channel video-system. The setup consisted of a laser source, a white light source, a system of optical filters, and an optical endoscope, using which the surgeon could analyze the regions of interest of the brain surface.The laser light source had a radiation wavelength of 635 nm, which corresponds to the wavelength of the PpIX absorption peak. The PpIX fluorescent signal was measured at a wavelength of 705 nm.
The two-channel video system allows obtaining images in black-and-white and color modes. These images can then be superimposed, with the fluorescent zones colored in green for visual assessment of the tumor foci borders. The system also allows for the evaluation of the fluorescence index in the investigated tumor region. A rigid endoscope with a 45 • tilt at the distal end was used to detect residual fluorescence.

Stereotactic device
The stereotactic needle for biopsy with simultaneous spectroscopic control with optical fibers fixed inside was connected to a laser radiation source and a fiber optic spectrometer (LESA-01-Biospec), respectively. Figure 2(a) shows the image of stereotactic device. The stereotactic needle was inserted into an outer stereotactic needle with a receiving window located on the lateral side closer to the distal end. The receiving window was provided for that the surgeon could easily take a biopsy of tumor tissue during the FD for further histological analysis.The stereotactic needle consisted of two light fibers, one for emission and one for detection, and a reflective 45• surface. Laser radiation was reflected through a light guide from an inclined surface and fell on a biological object with an accumulated PS.  (1), an optical fiber (2) connected to it, a Luer tee (3), a stereotactic biopsy needle (4) connected to the receiving fiber (5), and an optical fiber, a spectrometer (6) with a polychromator (8) with fiber-optic radiation input comprising a filter (7), a photodiode array (9), a registration unit (10), and a personal computer (11).
The fluorescence signal was recorded with a fiber optic spectrometer. Figure 2(b) shows the scheme of experimental setup. The calculated fluorescence index has carried the information about the concentration of the PS in the tumor tissue. The length of the stereotactic needle was 295 mm, and the diameter was 1.50 mm. The diameter of each optical fiber was 250 microns. The length of the receiving window -11 mm. The device's sensitivity made it possible to detect the PS concentration of 0.25 mg/kg in the optical phantom of a brain tumor.
The detection window in the first version of the device allows taking a tissue biopsy during stereotactic examination. But sticking of blood / tissue on the detection window of the stereotactic device can affect the received signal. Second version of the stereotactic device has a transparent shell in the detection window, which prevents biological fluid got into the detected area.

Metrological characteristics of the equipment
Before conducting the clinical study, the device was used to experiment with optical phantoms of a brain tumor with PpIX. PpIX in various concentrations (0.25, 0.5, 1, 2, 5, 10, 20 mg/kg) was mixed with a scattering medium (1% MLT/LCT Intralipid) and poured into test tubes. These optical phantoms were irradiated with a diode laser with a wavelength of 635 nm. The fluorescence signal with a wavelength of 700 nm was observed to determine the PS concentration in each optical phantom. The fluorescence index was calculated as the ratio of the PpIX fluorescence intensity in the 690-730 nm range to the intensity of the backscattered laser signal.
As a result of an experiment on freshly prepared optical phantoms of a brain tumor with PpIX using a stereotaxic needle for spectral diagnostics, the phantoms fluorescence spectra were recorded. The signal-to-noise ratio made it possible to detect the concentration of photosensitizers in each of the phantoms by the intensity of the emitted fluorescence ( Fig. 3(a)). By using the twochannel video system, the fluorescence index was recorded for each concentration of phantoms. The index was read from a marker placed in the fluorescent zone center ( Fig. 3(b)).

Fig. 3.
Fluorescence spectra of optical phantoms with PS PpIX obtained with a fiber optic spectrometer (a). Each color of the spectrum corresponds to a PpIX concentration of 0, 0.5, 1, 2, 5 mg/kg. The markers indicate the 635 nm laser peak to which the normalization and the PpIX fluorescence peak at a wavelength of 705 nm were performed. An increase of the fluorescence intensity in the spectra corresponds to an increase in the PpIX concentration in optical phantoms. Phantom frames with corresponding concentrations obtained using a two-channel video system (b). The emitted PpIX fluorescence in the optical phantom was highlighted in green and matched with the actual image.

Intraoperative video-fluorescent study plan
5-ALA was orally administered to the patients at a 1500 mg (20 mg/kg body weight) concentration before 6-8 hours before surgery. Open brain operations were performed under standard white light. A Carl Zeiss microscope equipped with a fluorescent mode with ultraviolet (UV) illumination at a wavelength of 400 nm and optical filters was used to determine the zones of maximum accumulation of the PS in the tumor. An optical system of surgical navigation (Medtronic StealthStation EM Navigation) with combined positioning based on magnetic resonance imaging (MRI) data obtained before surgery was used to determine the exact localization of the anatomical structures.
When conducting fluorescence diagnostics using a two-channel video system, the localization and size of regions with an increased fluorescence index were determined relative to the cerebral cortex tissue assumed to be healthy. Based on the zones with the highest fluorescence intensity and the duration of open brain surgery, the optimal laser power density for each patient was calculated before the PDT. In all cases, the applied dose of light to the tumor bed's entire surface was 30 J/cm 2 .
Data was recorded in the format of video files and pictures for further processing of the results. To assess the accumulation of the PS in tumor tissues, a comparative analysis of the fluorescence indexes from tumor tissues and optical phantoms of brain tissue with PpIX with different concentrations (0, 0.25, 0.5, 1, 2, 5, 10, 20 mg/kg) with a single calibration of the video system was performed.
Patients were awake during surgery. Tests to assess the cognitive ability to understand information and continuous intraoperative neuromonitoring were conducted by using somatosensory evoked potentials with motor response registration during direct electrical stimulation. Resection was interrupted when the residual part of the tumor occupied functionally significant regions of the brain, or no visible fluorescence was observed. Table 2 shows patient data after clinical operations. The value of PS concentration was obtained during general calibration by comparing the fluorescence intensity of optical phantoms and the fluorescence zones with most intensity of the tumor bed. The time when operation begins was chosen by considering the specific limitations. Immediately before the FD and PDT, PS accumulation was in the range from 5 to 7 hours. Holding the time before FD is necessary for precise diagnostics because the peak of PpIX accumulation in a brain tumor corresponds to 6 hours after the solution's injection. After 9 hours, the fluorescence intensity decreases due to PS removal from tissues [20].

Applying the device for FD and PDT with minimal invasiveness
Patient C. Preoperative diagnosis: Oligodendroglioma (III degree of malignancy) of left frontal lobe located in the motor zone (Fig. 4). Size of the main part of the tumor: 25.2 × 13 × 14.6 mm. PS was administered orally. The time of PS accumulation was 5.5 hours. The operation was started 4.5 hours after the PS administration. The tumor was not removed. A laser therapy device with a wavelength of 635 nm was used as a radiation source. The fluorescence spectra were recorded using the LESA-01-Biospec fiber optic spectrometer with an optical filter. Fluorescence diagnostics was carried out from the deepest part of the tumor towards the surface of the head. The needle step between the measurement points was 5 mm (Fig. 5). Each spectrum was recorded 5 times to exclude the possible signal loss due to a change in the distance between the receiving window and the tumor tissue. For photodynamic therapy, the tumor was split into 3 zones: on both borders of tumor and center. Each zone was irradiated by 4 fractions in position with the needle's rotate 90 degree. The irradiation time for one fraction was 6.5 min. For each zone irradiation time was 26 minutes. During the study, the fluorescence spectra were recorded before and after PDT. After processing the spectra, a graphical dependence of the fluorescence index on the needle's position was compiled (Fig. 5). The pre-resection fluorescence index was 325, the maximum value was 350, which corresponds to the maximum accumulation 6 hours after the PS administration. After several PDT sessions, the fluorescence index was decreased and was equal to 75. A decrease in the fluorescence index indicated the photobleaching of PS as a result of a photodynamic reaction.
According to the stereotaxic examination results, deviations of the registered fluorescence signal from the expected one were revealed. As the detector was moved relative to the tumor, the expected fluorescence index outside the tumor boundaries should correspond to the minimum accumulation due to the specifics of PS metabolism. Due to the selectivity of PS accumulation, the maximum accumulation occurs in tumor cells, while PS accumulation in healthy tissues is minimal. According to fluorescence diagnostics, the fluorescence index remained high outside the main part of the tumor. The possible explanation of the change in the fluorescence index by two causes: edema around the tumor was also accumulated PS; as we were moved to the outer border of the tumor, the signal reception window was obscured by a "biological substance" that accumulated the PS, which contributed to the recorded signal. The receiving window was modified to exclude the second reason (Fig. 2).

Craniotomy procedure cases
Patient D. Preoperative diagnosis: Glioblastoma (IV degree of malignancy) (Fig. 6). Size of the main part of the tumor: 30×30×20 mm. PS was administered orally. Accumulation time of PS: 6 hours. The operation was started 5 hours after the PS administration. The tumor was surgically removed. After the resection, the bed of the removed tumor was monitored to assess the result of surgical removal and the possibility of a PDT session on fluorescent regions of the tumor (Fig. 7  a, b).  After the FD, 3 consecutive PDT sessions were performed with the same radiation dose of 30 J/cm 2 . The first PDT session was performed on the entire tumor bed with a zone of 7 cm 2 for 10 min.The second and third sessions -on the residual fluorescent regions of the bed for 1.5 min each. During the PDT, the distal end of the optical fiber was fixed at a distance from the tumor bed surface. The entire fluorescent portion of the biological tissue was illuminated. After each irradiation, the change in the intensity of residual fluorescence was recorded (Fig. 7(c, d)).
The fluorescence indexes before PDT were 175 and 173. After the PDT sessions, the fluorescence indexes were changed to 90 and 98. After PDT, the high fluorescence index could be associated with the start of brain tissue necrosis and inaccurate camera focus. Further irradiation was not performed due to the time limit for open brain surgery.
Patient TAV. Preoperative diagnosis: Oligodendroglioma (III grade of malignancy) of the left frontal lobe (Fig. 8). Size of the main tumor part: 50 × 40 × 30 mm. PS was administered orally. PS accumulation time: 7 hours. Operation was started 6.5 hours after the PS administration. The tumor was surgically removed. After the resection, video fluorescence diagnostics of the tumor bed were carried out to determine the localization and the size of the residual regions with tumor cells. As a result of diagnostics, two regions of residual fluorescence with diameters of 50 mm and 40 mm were found ( Fig. 9(a, b)).PDT was performed in two regions of the tumor bed with radiation doses of 30 J/cm 2 . The irradiation time of the first region was 20 min, the second region -6 min. After the PDT sessions, FD was performed to assess the fluorescence intensity of certain zones of the tumor bed ( Fig. 9(c, d)). The fluorescence indexes before PDT were 33 and 24. After the PDT sessions, the fluorescence indexes changed to 03 and 11.  Figure 10 shows a fluorescence map of the first region of residual fluorescence of brain tissue before PDT, where the dark color corresponds to the highest fluorescence index in the studied region. A region of research is highlighted around as a black circle. The accumulation of the photosensitizer was recorded in the region of the brain, which functions linked to speech production. Thus, the residual fluorescent zones were not resected.
Patient PAV. Preoperative diagnosis: Glial tumor of left frontal lobe (Fig. 11). Size of main tumor part: 50 × 50 × 30 mm. PS was administered orally. PS accumulation time: 7.5 hours. Operation was started 6.5 hours after the PS administration. Surgical removal was performed.
After surgical resection, video-fluorescence diagnostics of the tumor bed were carried out to determine the localization and size of the residual regions with tumor cells. The diameter of the  tumor bed was 50 mm. After identifying the fluorescent regions ( Fig. 12(a, b)), a PDT session was performed for 18 minutes with an irradiation dose of 30 J/cm 2 . To control the regions of residual fluorescence and to assess the result of exposure to laser radiation after the PDT, FD was performed (Fig. 12(c, d)). The fluorescence indexes changed from 35 and 25 to 11 and 09, respectively. Patient NKA. Preoperative diagnosis: recurrence of glioblastoma (IV degree of malignancy) of the left temporal lobe (Fig. 13). Size of the main tumor part: 65 × 54 × 40 mm. PS was administered orally. PS accumulation time: 6.5 hours. The operation was started 4.5 hours after the PS administration. Surgical removal was performed.
Removal of the tumor was performed under fluorescence control using FD to assess the residual regions with tumor cells. Two regions with an increased fluorescence index were identified with diameters of 65 and 20 mm (Fig. 14(a, b)). One PDT session was performed on each region with the exposure time to laser radiation of 21 min and 2 min, respectively.  The irradiation dose for each region was 30 J/cm 2 . To assess the result of the PDT, secondary FD was performed (Fig. 14(c, d)). Due to the PDT, the fluorescence indexes changed from 51 and 74 to 30 and 10, respectively. Patient DLY. Preoperative diagnosis: relapse of glioblastoma (IV degree of malignancy) of the right frontal lobe (Fig. 15). Size of the main tumor part: 17 × 14 × 11 mm.PS was administered orally. PS accumulation time: 6 hours. The operation was started 5 hours after the PS administration. Surgical removal was performed. Resection of a part of the tumor that did not affect the functionally significant regions was performed under the control of FD. The localization and size of the residual fluorescent regions were determined ( Fig. 16(a, b)). Photodynamic therapy was performed on two regions 30 and 10 mm in diameter. An irradiation dose of 30 J/cm 2 was applied to each zone with an irradiation time of 5 min and 1 min, respectively. The repeated diagnostics was performed to assess PDT effect on fluorescent regions (Fig. 16(c, d)). The fluorescence indexes before PDT were 81 and 41. After the PDT sessions, the fluorescence indexes changed to 28 and 23. Pictures of fluorescent regions of brain tissue after PDT (c) and (d). The white square has marked a zone of most intense fluorescence.

Conclusion
Laser-induced fluorescence monitoring using a two-channel video system for fluorescence diagnostics improves the quality of surgical resection of malignant tumors. The percentage of resected volume from the visible fluorescence with visible fluorescence was 80-100% in the clinical study.
After conducting the comparative analysis of the fluorescence indexes from tumor tissues and optical phantoms of brain tissue with PpIX at various concentrations recorded during one calibration of the video system, the accumulation of the photosensitizer in tumor tissues with the corresponding concentrations was determined: Patient C -17.6 mg/kg, Patient D -17 mg/kg, Patient TAV -3 mg/kg, Patient PAV -3 mg/kg, Patient NKA -25 mg/kg, Patient DLY -15 mg/kg. The two-channel video system allows determining the concentration of the photosensitizer in the tumor tissue.
The use of a two-channel video-system for fluorescence diagnostics significantly increases the completeness of resection of glial brain tumors compared to resection performed under a surgical microscope with a laser wavelength in the blue range. The fluorescence indexes of the regions of interest in the brain were decreased several times after PDT. The video system can visualize the fluorescent zones that are invisible in a surgical microscope.
Fluorescence diagnostics and photodynamic therapy were performed using a stereotactic needle for biopsy with simultaneous spectroscopic control in a patient with a hard-to-reach and deep-seated tumor. According to the registered fluorescence spectra during diagnostics, a graphical distribution of the fluorescence index along the tumor depth was compiled. The fluorescence index before PDT was 325, with a maximum value of 350. After PDT, the value of the fluorescence index changed to 75. A decrease in the fluorescence index after PDT indicates the achievement of photobleaching.
The results of a clinical study can be used to improve the equipment used for diagnostics and in the development of new methods of treating neuro oncological diseases.
Disclosures. The authors declare no conflicts of interest.