Assessment of Non-Coherent Light Sources for Photodynamic Therapy

Roy H. Pottiert, Eva F. Dicksont, Pierre Nadeaut, Helmut Wielandt, Kasimir Sosint, and James C. Kennedyt

t Royal Military College of Canada, Dept. of Chemistry and Chemical Engineering, Kingston, Ontario, Canada
K7K 5L0 and t Queen's University, Dept. of Oncology, Kingston, Ontario, Canada K7L 3N6.

ABSTRACT

Several types of non-coherent light sources can be used for dermatological applications of photodynarnic therapy 
(PDT). Such sources are relatively inexpensive, stable and easy to operate, requiring very little maintenance. A 
further advantage is that large lesions can be irradiated in a single treatment. In order to assess the matching of the 
spectral output of the light source with the photochemo-therapeutic agent chromophore, a measurement of spectral 
irradiance with a monochromator must first be carried out. Also, in order to assure flatness of irradiation field over 
the entire bandwidths an irradiance measurement as a function of position in the beam must be perforrned. Data of 
this type obtained from different light sources is presented and discussed in terms of predicted PDT efficacy

Keywords: photodynamic therapy, protoporphyrin (IX), 5-aminolevulinic acid, light source characterization
1. INTRODUCTION

In the past, many of the applications of PDT have involved the use of a laser source for irradiation of the involved 
area, particularly for internal use.'~4 Currently, the use of noncoherent broad-band sources for irradiation of sma!! 
and/or large external areas is becoming more common.56 These sources differ fundamentally from laser sources in 
their output characteristics, important differences being the large beam size and broad spectral output. For 
illumination of large areas in particular, such sources must have reasonably uniform output properties over the 
entire treatment field in order to ensure uniform light dose delivery. We have developed an automated data 
acquisition system for measuring the magnitude and uniformity of several light output parameters of such sources 
in order to evaluate their suitability for treatment of derrnatological lesions of relatively large area.

2. LIGHT SOURCE CHARACTERIZATION PARAMETERS

The parameters which we have chosen to characterize for the non-coherent, broad-band light sources of interest 
are: (i) Irradiance: The magnitude of the total light output of the source measured at a particular location over the 
total spectral bandwidth of the light source, in units of mW-cm~2; (ii) Uniformity of irradiance: The relative and 
absolute variation of the irradiance over the entire illuminated area; (iii) Spectral irradiance: The light ouput of the 
source at a particular location in a particular 1 nm bandwidth, in units of mW-cm~2 nm~'; (iv) Uniformity of 
spectral irradiance: The relative and absolute variation of the spectral irradiance at a chosen wavelength over the 
entire illuminated area; also the relative variation of the ratio of the spectral irradiances at two chosen wavelengths 
over the entire illuminated area.

Once these parameters have been measured, a calculation is performed to determine the spectral matching 
between the light source and the in vivo absorption of the photosensitizer of choice, and the hence the relative 
effectiveness of each source. In this laboratory the main PDT method of interest is 5-aminolevulinic acid based 
PDT (ALA-PDT) in which the photosensitizes is endogenous protoporphyrin (IX) (PpIX).

3. LIGHT SOURCE CHARACTERIZATION SYSTEM

3.1. Description of system

In order to obtain the desired irradiance and spectral irradiance data in an automated manner we have constructed 
the computerized, automated acquisition system illustrated in Figure l. The apparatus can be set up for two 
different types of data acquisition, namely relative irradiance measurements, and spectral irradiance

measurenlelltsv from which are also derived irradiance values. Rapid relative irradiance measurements are 
performed using a calorimetric detector (Scientech 38/0101), or a suitably filtered radiometric detector 
(International Light SED033 silicon detector), to obtain relative irradiance in a narrow bandwidth, covered with a l 
mm pinhole and diffuser. Alternately, spectral irradiance measurements are performed with a spectral acquisition 
setup, using a fiber bundle (6 mm quartz fiber), whose output is directed into a motorized monochromator (JY 
H-20; 8 nm bandpass); the monochromator output is directed to a radiometric detector (International Light 
SED033 silicon detector) and the input end of the fiber bundle is covered with a pinhole (l mm diameter) and 
diffuser. The results of the spectral irradiance acquisition are used to calculate irradiance as a function of X,Y,Z 
position.

The light source to be chsuractenzed is located on an angle adjustable stage with motorized vertical adjustment. 
Either the calorimetric detector head, or the input end of the fiber bundle, is mounted on a motorized XYZ 
translation stage (Techno), whose overall displacement in any direction is as much as 50 cm. The X and Y 
displacements correspond to the honzontal and vertical positions of the detector in the beam, respectively, while 
the Z displacement corresponds to the distance of the detector from the face of the light source. A reference 
radiometric detector (Ophir PD2-A) filtered at a chosen wavelength (638 nm interference filter, 1() nm bandwidth) 
observes a small fraction of the scattered light output of the source in order to determine the overall stability of the 
output with time and to permit correction for small fluctuations. Acquisition of data from the detectors and 
movement of the monochromator drive and the motorized stage is computer controlled via a personal computer.

3.2. DescriDtion of measurements

For determination of relative irradiance as a function of X,Y,Z position, a series of measurements are performed at 
the chosen Z distance from the sourcF, for various X,Y coordinates at even increments (usually 5 to 10 mm apart), 
over a rectangular or circular grid, as appropriate to the shape of the output beam.

For a measurement of spectral irradiance uniformity, a similar series of measurements are performed using the 
spectral acquisition system, usually at larger X,Y increments due to the increased

acquisition time required. The full output spectrum of the source, in increments of 5 nm, is scanned at each X,Y 
location. These measurements initially yield uncorrected spectra, which must be corrected for the wavelength 
dependency of the system response, to give the corrected spectra in terms of relative spectral irradiance; then the 
data must be further corrected to obtain the absolute spectral irradiance curves at each X,Y,Z location. Finally 
these curves are integrated to obtain irradiance as a function of position. The calibration and correction procedures 
are described in the following section. Acquisition times as a function of X,Y position for a given Z distance are in 
the range of minutes to hours depending on the type of measurement, resolution and illumination area required.

3.3. Calibration of sYstem and correction of data

The relative irradiance data include, as needed, corrections to take into account fluctuations in overall source 
output with time during acquisition.

The data obtained with the spectral acquisition system must be corrected to yield both spectral irradiance curves, 
and irradiance, as a function of X,Y,Z position. The first correction procedure is performed using as light

source a standard lamp (quartz-halogen NBS-calibrated source) whose spectral shape is known over the region of 
300 nm to 800 nm. From this measurement correction factors for wavelength sensitivity are generated to yield 
relative speckal irradiance curves. The second correction procedure involves the acquisition of spectra from a 
suitable light source using the spectral acquisition system, while under the same conditions in the same X,Y,Z 
location performing a concomitant measurement of irradiance using the calorimeter detector (appropriately 
shielded to observe only a 1 cm2 area of the source). The integrated relative spectral irradiance curves and the 
irradiance values obtained from the calorimeter are ratioed and a second correction factor is obtained, which 
converts the relative spectral irradiance curves to corrected spectral irradiance curves in units of mW cm~2 nm '. 
Several measurements are generally performed in order to assure that a constant correction factor is obtained.

4. RESULTS

Measurements were performed initially on a modified slide projector (Kodak, with 500 W bulb and Hoya R-60 
600 nm long-pass filter) which had been used extensively for PDT treatment in the past, due to its inexpensive 
nature and ease of operation.5 Data were obtained on an illuminated area of approximately 7.5 x 9 cm at a Z 
distance of 27.5 cm from the face of the lens, with the lens pulled fully out. The relative irradiance data as a 
function of X,Y position in increments of 5 mm are illustrated in Figure 2 in a three-dimensional representation, 
showing a peak in the irradiance at the centre of the spot, and a rapid drop-off at the edges.

Table 1 illustrates a summary of the data obtained on irradiance and its uniformity. In order to compare with 
conditions actually used for treatment. in which the

affected area to be treated can be masked off to a smaller area than the actual illuminated area, calculations have 
been performed using extracted irradiance values falling within a radius of either 1.5 or 3 cm from the centre of 
the spot. Uniformity is expressed in terms of the relative standard deviation, as a percent of the average value, and 
the total absolute deviation, which is the maximum minus the minimum values.

The spectrum obtained of the filtered slide projector output and its uniformity are illustrated in Figure 3. Figure 3a 
illustrates a typical spectral irradiance curve. Figure 3b illustrates the spectral uniformity as a function of radial 
distance r from the centre of the illuminated spot, expressed in terms of the ratio of the spectral irradiances at 630 
and 670 nm; these wavelengths have been chosen due to their potential relevance in ALA-PDT as the last 
absorption band of PpIX is in the region of 630 to 640 nm in vivo, while its primary photoproduct

absorbs in the region of 670 nm. Some spectral nonuniformity is observed closer to the edges of the spot where the 
longer wavelengths contnbute relatively less to the spectrum.

Another example of a relative irradiance uniformity measurement is illustrated in Figure 4, a three-dimensional 
representation of the irradiance uniformity of a prototype commercial illuminator for PDT (called the PDTI) 
currently being evaluated in our laboratories. The greatly increased uniformity over a much larger area compared 
to the slide projector can be easily visualized, with the standard deviation reduced to 7% over a 9 cm spot.

The relative efficiency of these two light sources for ALA-PDT based applications has been compared by 
calculating the relative light absorbed at each wavelength by PpIX assuming an in vivo spectnum comparable to 
the in vivo surfacedetected fluorescence excitation spectrum measured in our laboratory for ALA-induced PpIX in 
mouse skin. Because of the fact that this spectnum has been obtained in vivo by exciting and

. . _
observing from the slain surface, it takes into account, to a certain extent, effects of tissue penetration on the 
exciting light. For this calculation, the in vivo excitation spectrum is assumed to have the same shape as the 
effective absorption spectrum; from the is? vivo excitation spectnum, an absorption spectnum is generated for an 
arbitrary low concentration of PpIX in vivo. In order to calculate the number of photons absorbed at each 
wavelength in a linear manner, the absorption spectrum is converted to a transmittance spectnum, T(X). Then the 
spectral irradiance of each light source as a function of wavelength, I(X), is multiplied by (I - T(x)) to give the 
light absorbed at each wavelength, I(x) x (1 - T(X)). Examples are given in Figure 5 for these calculations

performed for both the slide projector and the PDTI. Each curve is then integrated and divided by the integrated 
value of I(1) to give the fraction of emitted light absorbed by the PpIX under the assumed conditions, as

illustrated in Table 2. Thus assuming equal overall irradiance for e?,~%.h enilrrv tea DNTr

would be 2.7 X as efficient in exciting the assumed PplX in vivo, due to its better spectral matching.

S. CONCLUSIONS

We have developed an automated data acquisition system in order to permit precise characterization o broadband 
sources suitable for PDT applications. Data on irradiance, spectral irradiance, and source stability can be obtained 
in an automated manner, and uniformity of irradiance and spectral irradiance as a function of the position in the 
light beam can be calculated. Results of the characterization of a PDT-adapted slide projector and a prototype 
PDT illuminator for external use are presented, and their output characteristics compared for suitability of use 
with ALA-PDT. Such an acquisition system should permit routine, quantitative-and detailed comparison of PDT 
sources currently in use and under development.

6. ACKNOWLEDGEMENTS

This work was supported in part by a Canadian Department of National Defence Grant (FUHFV) and by DUSA 
Pharmaceuticals Inc.

7: REFERENCES

1. D.R. Doiron, "Photophysics of and instrumentation for porphyrin detection and activation," Porphrvin 
Localization and Treatment of Tumors, D.D. Doiron and C.J. Gomer, eds., pp. 41-73, Alan R. Liss Inc., NY, 1984. 
2. J.A. Parrish and B.C. Wilson, "Current and future trends in laser medicine," Photochem. Photobiol. 53, 731-738 
(1991).
3. T.J. Dougherty and S.L. Marcus, "Photodynamic Therapy", Eur. J Cancer 28A(10), 1734-42 (1992).
4. Photodynamic Therapy and Biomedical Lasers, P. Spinelli, M. Dal Fante and R. Marchesini, eds., Excerpta 
Medica, N. Y., 1992.
5. C. Whitehurst, K. Byrne and J.V. Moore, "Development of an alternative light source to lasers for
photodynamic therapy: 1. Comparative in X itro dose response characteristics," Lasers in Medical Science 8, 259
267 (1993)
6. J.C. Kennedy, R. H. Pottier and D.C. Pross, "Photodynam ic therapy with endogenous protoporphyrin IX: Basic 
principles and present clinical experience", J. Photochem. Photobiol. B 6, 143-148 (1990).