1. Introduction
This paper analyzes prior studies to show that a new treatment for glioblastoma (GB) that uses ultra low intensity transcranial light photodynamic treatment with drug augmented photodynamic treatment (LIT-PDT) can be an effective treatment. In PDT treatment, oxygen is transformed to its reactive singlet state by light, catalyzed by a 5-aminolevulinic acid (5-ALA) metabolite, PpIX. This singlet state is a reactive oxygen species (ROS) that oxidizes, and thereby destroys or damages, vital GB cell structures leading to cell death [
1,
2,
3]. After oral administration 5-ALA is preferentially taken up by GB cells and converted to PpIX to a greater degree than the surrounding brain tissue. This has two important consequences. 1) PpIX fluoresces red after illumination with 415 nm light, enabling more complete intraoperative identification of tumor margins and hence more complete resection. 2) PpIX will transfer 635 nm light energy to ground state oxygen to generate ROS, preferentially killing GB cells [
1,
2,
3].
Survival at 2 years after a diagnosis of GB remains <20% despite multimodal treatment. This paper presents a data analysis and synthesis of two data sets that can be combined to make a new treatment for GB. The first data set,
Section 2 below, analyzes peer-reviewed PDT treatment of GB using orally administered 5-ALA as the photosensitizer. Our analysis concludes that ultra low light fluences delivered noninvasively, transcranially, over 24 hours is both feasible, safe, and predicted to be more effective than short duration, intraoperative high fluence 5-ALA PDT as currently used in treating GB.
The second data set, presented in
Section 3 below, analyzes peer-reviewed data on four drugs from general medical practice - the antibiotic ciprofloxacin (cipro), the iron chelator deferiprone, and either the hypertension treatment drug telmisartan or the antipsychotic drug ziprasidone to increase effect of 630 nm light PDT in treating GB. They all have good evidence of increasing 5-ALA PDT effectiveness.
However, although many variables will influence survival, in the summer of 2024, standard treatment of GB with maximal safe resection, followed by irradiation and temozolomide, in a general GB population, still results a median overall survival under 2 years due to tumor recurrence [
4,
5].
Dozens of clinical trials of new medicines, and different forms and schedules of ionizing irradiation tried over the last 20 years have failed to greatly prolong survival [
6,
7]. The addition of the wearable Optune® device (Novocure Ltd., Haifa, Israel) that delivers 200 KHz electromagnetic radiation (radio frequency, non-ionizing) to the brain and tumor area has demonstrated good tolerability, some GB mitotic slowing and other potentially advantageous metabolic changes but only slight prolongation of median overall survival [
8]. Using intraoperative PpIX fluorescence leads to better tumor demarcation, resection and slightly longer time to progression [
9]. Ten years ago it was said that GB's fatal recurrence rate is almost 100% [
10]. Despite some improvements in the standard of care, the same can be said today.
The usual first GB recurrence is within the first 20 mm of the resection cavity wall, and most of these are within the original irradiation field [
11]. Extending resection, so-called extra marginal resection may prolong time to recurrence but not prevent it. At the time of diagnosis GB has already spread throughout the entire brain.
Based on our analysis of the transcranial light brain penetration studies of Tedford et al and Mathews et al, vide infra, we determined that i) noninvasive transcranial 630 to 660 nm light can deliver an ultra low fluence rate (17 μW / cm
2) to a GB, to the post-resection peritumoral area, and to the entire brain and, ii) prolonged and repetitive delivery of such fluence is effective in mediating 5-ALA PDT killing of GB cells [
12,
13].
Thus the LIT-PDT Regimen includes two components: 1) ultra low light levels delivered continuously noninvasively over 24 hours with 5-ALA PDT, and 2) three orally administered drugs from general, non oncology medicine - ciprofloxacin, deferiprone, and telmisartan or ziprasidone - will augment capture of 630 nm photon energy, increasing ROS generation after 5-ALA. Details follow.
2. Multiple array, Ultra Low Fluence, Transcranial, Repetitive, 5-ALA PDT
In PDT of GB as currently constituted, oral 5-ALA is preferentially taken up by GB cells and metabolized to PpIX. PpIX transfers 630 nm light energy to ground state O
2 (
3O
2), thereby generating singlet O
2 (
1O
2), one of several ROS [
14,
15,
16,
17]. See
Table 1. for oxygen related ROS definitions.
Ferrochelatase mediates iron incorporation into PpIX, transforming PpIX to non-ROS generating heme. PpIX accumulates in large quantities in GB cells after 5-ALA administration due to increased uptake of 5-ALA compared to non-GB brain tissue, and reduced ferrochelatase activity and increased heme demand in GB cells [
15,
18,
19,
20,
21].
Intraoperative 5-ALA PDT heretofore has been by using ~200 J / cm
2 total, of 635 nm light delivered to the resection cavity wall in five fractions of 12 minutes with 2 minute pause between the four periods. The light exposure is performed with the direct placement of up to 4 diffusers or through an Intralipid ™ filled Foley catheter intracavitary diffuser, for a delivery total of 720,000 mW.s = 720 J / cm
2 total, fractionated over 1 hr [
22].
This, delivered intraoperatively as a single session before closure immediately after fluorescence guided resection [
19,
21,
22,
23]. Exploration began twenty years ago of low fluence, mW 5-ALA PDT, delivered over extended periods of time to increase selective tumor cell kill through apoptosis. [
24,
25,
26]. See
Table 2 for definitions of light measurement terms.
The limitations of 5-ALA PDT as currently practiced are restricted to a single intraoperative PDT session. This means that high fluence light is required. As a consequence necrosis predominates as mode of death, as opposed to the preferable apoptosis that would predominate if ultra low fluence could be used. Approximately 4 to 13 mm deep to the surface of the resection cavity receives enough photon energy to generate cytotoxic ROS during intraoperative PDT [
23,
27]. Also blood clots and surgical debris may interfere in the efficacy of the intracavitary light delivery. Also, in current practice intraoperative high light flux PDT quickly exhausts available O
2 supply even when ventilating the patient with 100% oxygen during the treatment. PDT requires O
2 to be effective.
Prolonged or repetitive PDT leaving the resection cavity open with the attendant risks and procedure complications [
28], or fully implantable light and power sources, the Globus Lucidus [
29], have been in development to allow long term, repeated PDT. However we know that PpIX positive GB cells reside within the first 20 mm of the resection cavity wall [
30,
31]. Those cells will not receive enough short term light to be killed by intraoperative PDT, even by high flux delivery of 200 J / cm
2 [
12,
13].
In vitro work by Mathews et al however has opened up another avenue to overcome the current limitations of one shot intraoperative PDT. Mathews et al [
13] and others [
32,
33] have shown that ultra low fluence (<50 μW / cm
2] delivered continuously over 24 hours can be repeated with resulting deep blockage of GB growth. Cytotoxicity was primarily by apoptosis, not necrosis. Mathews et al even showed no growth using four 24 hr periods of 17 μW / cm
2 flux, 1.5 J / cm
2 total delivered, 635 nm light, separated by 3 day intervals [
13]. Note that 1.5 J / cm
2 per 24 illumination day is less than 1% of the fluence used in today’s clinical intraoperative 5-ALA PDT:
If given once at a single session, Mathews reported that 12 J / cm
2 was the required light dose to achieve similar growth suppression, an 8 times higher total energy delivery compared to the repeated 24 hour low flux treatment [
13].
This small fluence can be delivered non-invasively, transcranially, to GB tissue from the scalp surface by using fully external multi LED 630 nm light array, through intact human skin and skull. The calculation behind this statement:
Lapchak measured light attenuation across the human calvaria and found that 4.2% of the incident power was transmitted with minor corrections for variations in depth and hydration [
34,
35,
36]. Tedford et al measured attenuation versus depth in human brain parenchyma to be 2.4 / cm
2 depending slightly on wavelength and geometry [
12]. Assuming the head is broadly illuminated so that scattering may be neglected, we can reasonably estimate light power at the tumor in terms of power incident on the scalp reduced by these factors.
where
p = delivered power density at the tumor in W / cm2
P = incident power density at the scalp in W / cm2
d = depth of the tumor beneath the cortex in cm
Matthews found in vitro that a protocol of long-term exposure at low power was effective on glioma spheroids at 17 µW / cm
2. Inverting the formula, we can solve for the incident power density P sufficient to deliver a desired power density p at the tumor as a function of the tumor depth d. See
Table 3 for the resultant predicted external 630 nm flux requirement noninvasively applied to the head to achieve 17 µW / cm
2 at a GB post-resection area at the listed depth from the external light source.
The listed external flux can be easily delivered by many of the current commercially available photobiomodulation helmets for home use, for which no prescription is needed. If we take an average scalp area to be 700 cm2 then total light energy delivered to the entire scalp area would be ~12 W to reach the deepest GB.
So Part 1 of the LIT-PDT program is to deliver post resection 17 μW / cm2 630 nm illumination to the resection area / 24 hrs after oral preoperative 5-ALA. Light will be delivered by an external multi LED array cap, giving broad head illumination such that light scattering may be neglected.
Prolonged transcranial 630 nm light delivery from 10 to 20 W light sources has a well established safety history of use in humans, a procedure called photobiomodulation [
37,
38,
39,
40,
41]. Photobiomodulation refers to use of external low level (<20 mW / cm
2) LED arrays to deliver transcranial light at ~630 nm (red) light to the brain with the aim of improving brain function deficits.
Dozens of clinical studies in humans have explored 630 nm photobiomodulation treatment to reduce impairments of Alzheimer’s disease [
42,
43], cognitive or executive function [
44,
45], Parkinson’s disease [
46], traumatic brain injury [
47,
48], autism [
49], and others [
50,
51,
52]. While the results on brain function have not shown to have unequivocal benefit, all the studies have shown good safety with minimal or no side effects from this light delivery.
A wide variety of these photobiomodulation helmets delivering broad head illumination at 630 nm are currently commercially available for home use without prescription. Most of these can easily deliver 17 µW / cm
2 to any part of the brain.
Figure 1 shows a schematic of LIT-PDT illumination of a GB tumor area
5. Discussion and Conclusions
Cipro, deferiprone, and febuxostat are well tolerated drugs, usually without side effects. Although surprises cannot be excluded, there is no a priori reason to suspect drug-drug interaction that would change that low side effect risk when these three are used together.
Low fluence light delivery in 5-ALA PDT has several advantages over the current use of high fluence intraoperative PDT. Lower fluence over a longer time results predominantly in apoptosis and less inflammation. Higher fluence 5-ALA PDT over a shorter time results in predominantly necrosis with relatively more inflammation and edema [
25,
135]. A second advantage of low fluence light is that we can repeatedly and noninvasively deliver it.
The safety of the human brain’s exposure to low wattage extracranial 630 nm has been well established by the large body of research on photobiomodulation [
136,
137,
138,
139,
140,
141].
Preclinical study of Mathews et al explored 17 μW / cm
2 flux over considerably longer exposure times (24 hours, 1.5 J / cm
2 total) but repeating this every fourth day [
13]. A further advantage of low flux delivery is reduced immunosuppression elements [
142].
A pilot study is planned. 36 hours prior to primary surgery, one newly diagnosed GB will receive 5-ALA by mouth, then 32 hours of continuous noninvasive 630 nm transcranial illumination of fluence calculated to deliver 17 μW / cm2 / 24 hours. Calculated as follows:
24h x 60 min x 60 sec = 86,400 sec in 1 day.
17 μW / cm2 x 86,400 sec =
1,468,800 μW.sec / cm
2 / day = 1.5 J / cm
2 / d, the in vitro dose that Mathews et al showed stopped all GB cell growth [
13].
Thus the predicted total 630 nm light energy received by the tumor resection area will be 2.3 J / cm2 delivered over a continuous illumination time of 32 hours. Depending on results the next step would be the same illumination schedule, 32 hours on, off for 88 hours (1⅓ day on, 3⅔ days off) , for 4 cycles prior to surgery. Further schedules would be determined by these initial histology results. Evaluation of effect by necrosis, apoptosis, and K67 markers, and standard H&E pathology report.
As a consequence of this paper’s analysis of the data, a phased pilot study of LIT-PDT warranted and planned.