An outline of the history of PDT

J. Moan and Q. Peng

Table of Contents

1.1 Introduction 3
1.2 The 'photodynamic action' 3
1.3 Phototherapy 4
1.4 Photochemotherapy 5
1.5 The early days of photodynamic therapy 5
1.5.1 Hematoporphyrin and hematoporphyrin derivative 6
1.5.2 Other photosensitizers introduced for PDT 8
1.6 Why do some photosensitizers localize selectively in tumors? 8
1.7 The action mechanisms of PDT at the cellular level 12
1.8 The action mechanisms of PDT at the tissue level 12
1.9 Photochemical internalization 13
References 13

1.1 Introduction

Photodynamic therapy, PDT, has now reached the level of being an accepted treatment for a number of diseases, among which are several forms of cancer. Many countries have approved its use. The number of articles on PDT published in a year, both clinical and basic, seems to be steadily increasing. It has, however, been observed that many of the investigators are obviously unaware of the early work done in this field and hence, repeat many of the experiments reported earlier (before the internet and the modern database were established). Therefore, in the present historical review, the early work is weighted more heavily than the recent work that is more easily accessible to the readers.

1.2 The 'photodynamic action'

The term 'photodynamic action' ('photodynamische Wirkung') was introduced in 1904 by one of the pioneers of photobiology: Professor Hermann von Tappeiner, director of the Pharmacological Institute of the Ludwig-Maximilians University in Munich. It is not clear why he called the process 'dynamic'; it might have been to distinguish this biological phenomenon from the reactions taking place in the photographic process that had been discovered a few years earlier. Actually, von Tappeiner was not completely happy with his term, as he says in the foreword of his book Die Sensibilisierende Wirkung Fluorescierender Substanzen: 'Whether or not the name is to be used further or dropped, must be left to the discretion of my colleagues'. Also, Blum in his textbook on photosensitization expresses objections to the term: 'The choice of 'photodynamic action' is not altogether a happy one, but has advantage of priority and usage'. Feeling the need for a more correct and descriptive term we have tried to replace 'photodynamic therapy', PDT, by 'photochemotherapy', PCT without overwhelming success. There may be several reasons for this. First, PDT has the advantage of priority. Secondly, photochemotherapy has a wider definition (see later) including processes not requiring oxygen. Thirdly, PCT is an abbreviation for porphyria cutanea tarda, a disease that involves photosensitivity. However, the name 'photochemotherapy' would have paved the way for clinical applications better than the name 'photodynamic therapy' since all oncologists are familiar with chemotherapy, while 'photodynamic' may give associations to 'biodynamic' which, at the best, is regarded as a quasi-scientific term. According to the original definition, as well as to Blum's later recommendation, the term 'photodynamic action' should be used only for photosensitized reactions requiring oxygen. It should be remarked that oxygen is not only involved in photosensitization of Type II, but also usually involved in photosensitization of Type I. A Type I reaction is a radical or redox reaction in which a photosensitizer, excited to the triplet state (3S), interacts with a neighbouring molecule (A) by exchange of an electron or a hydrogen atom:

3S + A [right arrow] S·- + A·+ (1)

followed by

A·+ + 3O2 [right arrow] Aox (2)

S·- + O2 [right arrow] S0 + O2·- (3)

O2·- + A [right arrow] Aox, (4)

in both cases giving an oxidized biomolecule Aox.

An alternative type I reaction pathway might be:

3S + 3O2 [right arrow] S+ + O2·-, (4)

followed by

S+ + A [right arrow] S0 + A·+ (6)

Reaction (4) may follow after reaction (5), and reaction (2) may follow after reaction (6). Type I reactions may also be independent of oxygen, as is the case for the reactions of psoralens with DNA.

A photosensitized process of Type II is by definition an energy transfer process. The most common Type II process is oxidation via singlet oxygen (1O2) formation:

3S + 3O2 [right arrow] S0 + 1O2

1O2 + A [right arrow] Aox

A further description of 1O2 reactions in biological systems can be found in ref. 5.

Many authors use the expression 'photodynamic action' and 'PDT' synonymously with '1O2 reactions' and even with: 'Type II reactions', but according to the definitions given above this is not strictly correct. However, in most practical cases it may be acceptable since most PDT sensitizers act via 1O2 which is formed in an electron exchange process.

1.3 Phototherapy

Phototherapy can be defined as the use of light alone for therapeutic purposes. However, endogenous sensitizers are usually involved, so phototherapy often relies on photodynamic processes. Solar light has been used to treat a number of disorders such as vitiligo, psoriasis, rickets, skin cancer and even psychosis. Heating, as well as psychological effects mediated by vision, may have played roles in these therapies, but the effect of visible light and ultraviolet radiation on the skin was probably more important. Phototherapy has been applied by humans for 3000 years and was known by the Egyptians, the Indians and the Chinese. In Greece, Herodotes called it 'heliotherapy' and recommended it for 'restoration of health' in the 2nd century BC. In the 18th century the effect of sunlight on rickets was known. In 1815, Carvin wrote that sunlight had a curing effect on 'scrofula', rickets, rheumatism, scurvy, paralysis and muscle weakness. The Polish physician Sniadecki documented in 1822 the importance of sun exposure for the prevention of rickets. Later, in 1903, the Dane, Niels Finsen, was awarded the Nobel Prize for his work on the use of light from the carbon arc in the treatment of lupus vulgaris (skin tuberculosis) and was acknowledged as the founder of modern phototherapy. He also treated smallpox with red light and found that this treatment prevented suppuration of pustules.

In the 1950s Richard Cremer in Essex, England, after listening to an observant nurse, introduced phototherapy as a treatment of jaundice in newborn babies. This is the most widely used form of phototherapy today. However, light as a therapeutic agent for depression and for maintenance of biological rhythms may in the future become equally important. It seems that, in the latter context, light may act not only via vision but also via absorption in skin.

Since several of the above-mentioned forms of phototherapy may belong to photochemotherapy or even PDT, contact between PDT scientists and scientists in the field of phototherapy might be mutually beneficial.

1.4 Photochemotherapy

In one of India's sacred books Atharava-veda (1400 BC) it is described how seeds of the plant Psoralea corylifolia can be used for the treatment of vitiligo. Psoralens are the photoactive components of these seeds, just as in the extracts of the plant, Ammi majus, which grows on the banks of the Nile, was used by the Egyptians to treat vitiligo. For centuries photochemotherapy made no further progress until 1974 when PUVA (i.e. treatment with psoralens and UVA radiation) was reported to be an efficient treatment of psoriasis. Photochemotherapy can also be tailored to act on the immune system such as in extracorporeal photophoresis of mucoses fungoides, cutaneous T-cell lymphoma.

1.5 The early days of photodynamic therapy

The very first attempts to apply PDT to treatment of tumors and other skin diseases, such as lupus of the skin and chondylomata of the female genitalia, were performed by the group of von Tappeiner in 1903–1905. They tried a number of dyes: eosin, fluorescein, sodium dichloroanthracene disulfonate and 'Grubler's magdalene red'. The dyes were in most cases topically applied, but intratumoral injections were also attempted. Favorable results were reported, but there was no long-term follow-up, and PDT was soon forgotten, probably because of the advent of ionizing radiation in cancer therapy.

The story of how this German group hit upon the idea of using dyes as biological sensitizers for light is fascinating: one of von Tappeiner's students, Oscar Raab, was investigating the toxic effects of acridine on paramecia. In one experiment the paramecia survived incubation with a given acridine concentration for about 1.5 h, while in another experiment they survived for about 15 h under identical conditions, except, as recorded by the observant student, during one of the experiments there was a heavy thunderstorm. So he started to wonder whether light might have played a role — this resulted in the discovery of photodynamic action. Characteristic of that time: the supervisor published the finding before the student. The group performed a large amount of work on photosensitization, discovered that oxygen was required for the photodynamic effect and summarized their work in a book. At about the same time as Raab made his discovery, a French neurologist administered eosin orally for epilepsy and found that the patient got dermatitis on light-exposed areas of the body.

1.5.1 Hematoporphyrin and hematoporphyrin derivative

No history of PDT can be told without mentioning hematoporphyrin. This compound was first produced (in an impure form) by Scherer who removed iron from dried blood in 1841 by treatment with sulfuric acid. The spectrum of this red substance, as well as its fluorescence, was described by Thudichum in 1867, and the name 'hematoporphyrin' was dedicated to it by Hoppe-Seyler in 1871. In the period 1908–1913, a number of photobiological experiments were carried out with hematoporphyrin, demonstrating how it sensitized paramecia, erythrocytes, mice, guinea pigs and humans to light. The German doctor Friedrich Meyer Betz became extremely photosensitive for more than two months after injecting 200 mg of hematoporphyrin into himself. This dose is similar to that used of HPD and Photofrin for PDT today, and one may wonder whether the hematoporphyrin used by Meyer Betz was HPD rather than pure hematoporphyrin which turns out to be a rather poor photosensitizer due to its high water solubility. Hans Fischer, who was awarded the Nobel Prize for his work on porphyrins, reported that uroporphyrin was almost as phototoxic as hematoporphyrin. This is surprising in view of the fact that the water solubility of uroporphyrin is even greater than that of hematoporphyrin.

The first observation of porphyrin fluorescence from tumors was published by Policard in 1924. The red fluorescence from endogenously produced porphyrins in experimental rat sarcomas was attributed to bacteria infecting the tumors. A few years later Kördler detected a similar fluorescence in breast carcinomas and some other superficial tumors but found no evidence for the involvement of bacteria. Several investigators reported preferential accumulation of porphyrins and porphyrin precursors in neoplastic tissue. Of particular interest is 5-aminolevulinic acid (ALA), a precursor to heme synthesis; Rubino and Rasetti stated that the preferential tumor accumulation of porphyrins was not related to an elevated activity of the enzymes of heme synthesis. Later work showed that this may not be a universal observation since the activity of porphobilinogen deaminase seems to be high in some tumors while the activity of ferrochelatase as well as the concentration of iron seems to be low. Auler and Banzer were probably the first to study the accumulation of injected porphyrins in tumors. They reported that hematoporphyrin injected in rats accumulated in primary and metastatic tumors as well as in lymph nodes. Animal tumors were treated with light after injection of this dye, and promising results were obtained. Figge et al. showed that a number of porphyrins have a selective affinity, not only for neoplastic tissue, but also for embryonic and regenerating tissues in rodents. Surprisingly, it was observed that tumor-bearing mice tolerated larger doses of 65Zn hematoporphyrin than mice without tumors. This observation probably indicated that the tumors accumulated a significant fraction of the dye, thus protecting sensitive organs from radiation damage. This group was probably the first to demonstrate that hematoporphyrin also had a tumor-localizing ability in a variety of human malignancies. A large amount of data on the tumor-localizing ability of porphyrins was published by Lipson et al. at the Mayo Clinic during 1960–1967. Their work was inspired by that of Dr. Samuel Schwartz, who was interested in porphyrins from the point of view of radiosensitization. Some of the results of Schwartz and co-workers are remarkable and deserve to be mentioned even though they concern ionizing radiation. Some 153 mice-bearing rhabdomyosarcoma tumors were injected with different doses of crude hematoporphyrin, ranging from 10 to 1250 µg. Three hours later the tumors were exposed to ionizing radiation. Twenty seven of the mice that had got 50 µg were all cured, while among the remaining 126 animals no cures were observed. The conclusion was that low or intermediate concentrations of hematoporphyrin have a stronger sensitizing effect than either very low (10 µg) or very high (> 250 µg) doses. We once looked for the radiation modifying effect of hematoporphyrin, but found no such effect, possibly because we used too large concentrations.

Schwartz realized that commercial samples of hematoporphyrin were impure and tried to purify them. Surprisingly, he found that pure hematoporphyrin was a poor tumor-localizer. Treatment with acetic-sulfuric acid mixture gave some components which had better properties with respect to tumor-localization. These components came to be known as 'hematoporphyrin derivative', HPD. They were later used by Lipson and co-workers and by a large number of clinical investigators, both for diagnostic and therapeutic purposes. HPD contains several porphyrin monomers as well as dimers and oligomers. A debate about its composition lasted for several years. Among others we studied HPD by means of HPLC and fluorescence methods and found three different groups of components, monomers (hematoporphyrin stereoisomers, hematoporphyrin vinyl deuteroporphyrin isomers and protoporphyrin) with a high fluorescence quantum yield but with a poor tumor uptake, dimers with a lower fluorescence quantum yield but with a higher tumor uptake and non-fluorescent aggregates with the best tumor-localizing properties. The cellular uptake of these components, and of other related porphyrin compounds, increased with increasing lipophilicity, and so did the quantum yield of cell inactivation. The non-fluorescent fraction, the aggregates, had a low photosensitizing ability although they seemed to be slightly more photoactive than one might expect on the basis of their weak or non-existent fluorescence. The aggregates seemed to be the best tumor-localizers. Several investigators proposed that the aggregates might decompose to fluorescent and photoactive components after being taken up by the tumor, but this was never proven. Kessel and others continued the investigation to find the chemical identity of HPD and concluded that the dimers and oligomers were coupled by ether as well as ester linkages. Dougherty and his co-workers partly purified HPD by removing the monomers. The resulting product was called Photofrin and is still the most widely used sensitizer for clinical PDT.

The modern era of PDT was founded in the 1970s with the pioneering work of Dougherty and co-workers at the Roswell Park Memorial Cancer Institute in Buffalo who used HPD/Photofrin. This history is well known and has already been elegantly described in several articles.