Kaposi‚s sarcoma (KS) is the most common malignancy in AIDS patients. Cutaneous KS is often painful and can be a significant cosmetic problem impacting negatively on the patient‚s quality of life. Despite a wide variety of therapeutic approaches, effective palliative treatment is still difficult to achieve[2-4]. Photodynamic therapy (PDT) utilizes visible light activation of tissue-bound photosensitizers resulting in local necrosis of the lesion in a relatively selective manner. PDT with Photofrin as the photosensitizer has been approved for treatment of partially and completely obstructive esophageal tumors in the United States of America, and elsewhere for treatment of cancers of the bladder, lung, and early stage cancers of the stomach and cervix. In addition, there are numerous reports of its efficacy at other sites, e.g. head and neck, brain, and skin[6,7]. At Roswell Park Cancer Institute PDT with Photofrin has been applied to a variety of cutaneous tumors in >100 patients with >1000 lesions. Several of these studies included individuals with classic KS who responded to PDT, and as Photofrin has no known systemic effect, this preliminary experience suggested that PDT could offer an effective palliative treatment for epidemic cutaneous KS without further compromise of an already fragile immune system. Therefore, a Phase I/II study was performed to determine the feasibility of utilizing PDT in epidemic KS.
Materials and methods
All enrollees had HIV infection confirmed by accepted serologic methods as well as pathologically confirmed cutaneous Kaposi‚s sarcoma, Eastern Cooperative Oncology Group (ECOG) performance status of 0-3 with an expected survival of at least 3 months. Adequate bone marrow, renal and hepatic function was required for entry. This study was approved by the Institutional Review Board of Roswell Park Cancer Institute and all subjects provided written informed consent after the risks associated with the study had been explained to them.
Prior therapy, no less than 4 weeks before initiation of treatment with Photofrin, was permitted. Concurrent cytotoxic drugs or biologics were not allowed [however, growth factors such as granulocyte colony stimulating factor and erythropoietin were permitted]. All patients were cautioned to protect exposed body surfaces from bright light for 4 weeks after injection of Photofrin to avoid cutaneous phototoxicity induced by it. All patients were provided with sunglasses to be worn outdoors during the 4-week period.
Eligible patients with one or more lesions on the skin were injected with Photofrin and warned to use protective clothing and sunglasses when out of doors. Lesions were selected for treatment based on size (generally ≤3cm maximum diameter) and symptomatology. No more than three lesions were treated on the face and neck area at any one time as experience with patients with basal cell carcinoma predicted excessive edema. In the initial exploratory phase of the study, the light dose ranged from 100 to 250J/cm2 with most patients receiving multiple light doses. As these doses generally did not produce a high complete response rate, the study was modified to allow light doses ranging from 300 to 450J/cm2. In this group of patients, the dose was escalated based upon the following criteria: if at any light dose one-third or more of the patients experienced unacceptable toxicity (increase in ECOG status by 1 unit) or normal skin toxicity of necrosis and/or scaring within the treatment field, then one additional patient was treated under the same conditions. If the second patient demonstrated the same toxicity, then dose escalation was stopped. If this patient did not demonstrate same toxicity, dose escalation continued.
Multiple treatment sessions on subsequent days up to 72 h post-injection were allowed depending on number of lesions and patient agreement. Often it was not possible to treat all lesions due to the presence of a large number of lesions. In most patients this allowed for comparison between treated and untreated lesions; however, this was not a requirement of the study.
All patients received 1.0mg/kg body weight Photofrin (porfimer sodium; QLT PhotoTherapeutics, Vancouver, British Columbia, Canada), which was reconstituted in 5% glucose (D5W) to 2.5mg/ml immediately before bolus intravenous injection 2 days prior to light treatment. Light at 630m from a Coherent Innova argon laser pumping a Coherent 599 dye laser (Palo Alto, CA) was delivered through a 400μm or 600μm single quartz fiber (QLT PhotoTherapeutics) fitted with a micro lens to produce a homogenous circular pattern for surface treatment. Escalating surface treatment light doses of 100-400J/cm2 were delivered at 150mW/cm2 (requiring 11.1-44.4 min). In the latter part of the study special fiber holders were used which could be fixed directly to the patient‚s skin; the use of an eight-way splitter attached to the laser allowed eight individual lesions and/or fields to be treated simultaneously, thus substantially reducing the time required for treatment. In a few cases, especially for larger lesions, multiple treatments were given over a few months. In general, this did not improve outcome and was not continued.
Representative biopsies were carried out in all patients before treatment to confirm the diagnosis of KS. Additional biopsies were obtained at selected times post-treatment (30 min to several weeks) to determine response to treatment as well as depth of effect induced by PDT. Those taken from 30 min to 5 days were defined as short-term; those obtained after 5 days were defined as long-term. Biopsies were examined by two pathologists blinded to the clinical history of each sample. Biopsies were processed in a routine fashion, embedded in paraffin blocks and stained with hematoxylin and eosin. Tissue sections were cut at 5μm. In addition, biopsies were stained with antibodies to identify immune system effector cells UCHL-1 and CD3 (T cells), L26 (B cells), CD68 (histiocytes), as well as antibodies against Factor XIIIA to identify dermal dendrocytes and CD34 to identify KS cells. The superficial and deep areas were scored separately using a scale as follows: 0, no staining; 1+, 1-4 per medium power field; 2+, 5-25 per medium power field; 3+, 25-50 per medium power field; 4+, 50-100 per medium power field. All primary antibodies were visualized by the DAB Detection Kit (Ventana Medical System, Tucson, Arizona, USA).
Analysis of pathologic response
The superficial (0-2mm) and deep areas of the punch biopsies were evaluated separately and designated according to the presence or absence of KS: no tumor response (KS present and intact); partial pathologic response (intact KS detected in small areas with evidence of disruption); complete pathologic response (no KS found).
Analysis of clinical response
A complete clinical response was defined as visual total disappearance of the lesion or complete flattening of the lesion without regrowth for at least 4 weeks. A partial response was a ≥50% reduction in the sum of the product of the two longest perpendicular diameters of measurable lesions for at least 4 weeks. Clinical failure was <50% reduction in the sum of the product of the longest perpendicular diameter of measurable lesions or no flattening. Progressive disease was a >25% increase in the sum of the product of the two longest perpendicular diameters of the lesion or appearance of new lesions within the treatment field.
Systemic toxicity was evaluated on a weekly basis and graded by common toxicity criteria. Normal tissue toxicity within the treatment field was graded by a scale established by previous studies: 0, no effect; 1.0, erythema; 2.0, edema; 2.5, erythema plus edema; 3.0, eschar; 4.0, necrosis, 5.0, scarring.
Twenty-five male patients-18 Caucasian, four Hispanic and three African-American-were enrolled in this study; 348 lesions were treated. Seventeen of the patients were receiving antiretroviral therapy consisting of zidovudine, which reflected current practice during the study period; 16 of these patients had been treated previously for their KS: cytotoxic agents (seven patients), radiation therapy (10 patients), and biological response modifiers (seven patients). Twenty-two out of the 25 patients were taking reverse transcriptase inhibitors. Of the 25 patients, one had documented visceral disease. CD4 cell counts were in the range 1-590×106/l (median of 141×106/l); only six of the 25 patients had CD4 cell counts ≥200×106/l.
As shown in Table 1, for the 25 patients enrolled in this study, tumor response correlated with light dose as the only predictive parameter with a response rate of 100% achieved at doses of 250J/cm2 or higher (54% with a complete clinical response). Because the number of lesions treated at the highest light doses of 350-400J/cm2 was small, for statistical analysis the lesions were grouped by light dose into three groups: 100-199J/cm2, 200-299J/cm2 and 300-399J/cm2. Analysis using the χ2 test showed a highly significant trend in the increase in the rate of complete clinical response, with increasing light dose (P<0.0001) (Table 2).
A vast majority (91%) of the lesions were 0.5-3.0cm in diameter. However, there were 12 lesions each at <0.5cm and >3.0cm. In all cases a margin of 0.5cm was used for the treatment field. In some cases larger treatment fields (up to 7cm) were used to treat multiple lesions within a larger mass of tumor. However, none of these were evaluable at follow-up due to the continued progression of surrounding untreated tumor.
Median duration of response for complete responders was 26 weeks (range, 4-44 weeks) and for partial responders, 8 weeks (range, 4-56 weeks).
Twenty-three pre-treatment biopsies were evaluated for cell content (Table 3). All 23 showed essentially the same pattern: many lesional KS cells (CD34, 4+), significant numbers of histiocytes (CD68 and FX111a, 3+), some T lymphocytes (CD3 and UCHL-1, 2+) and rare B lymphocytes (L26, 1+). Post-therapy biopsies were not necessarily from the same sites.
Forty-three biopsies taken between 0.5 h and 4 months (Table 3) post-treatment were obtained from 15 patients (Fig. 1). The average thickness of the biopsies was 3.9mm (range, 1.1-6mm), and the average thickness of the KS lesions was 3.3mm (range, 1.0-6mm). The majority of short-term biopsies (defined as 30 min to 5 days post-treatment) demonstrated disruption of the superficial KS (i.e., up to 2mm). There was a trend towards greater depth of necrosis at the 300-400J/cm2 light doses (2.4mm average) as opposed to the lower dose range of 100-300J/cm2 (1.8mm average). Necrotic depth ranged from 0mm in a patient receiving 100J/cm2 to 4.8mm in a patient receiving 300J/cm2 light dose. It was noted that when necrosis did not appear throughout the entire depth of the biopsy, the superficial dermis demonstrated histological abnormalities consisting of cellular debris, congested capillaries, hemorrhage and perivascular edema. The overlying epidermis and the underlying deep dermis were found to be relatively unchanged (Fig. 2). Complete clinical responses correlated with destruction of KS in the superficial area of biopsies, but not necessarily with pathological destruction of KS cells at the full depth of biopsy. In 14 biopsy samples from lesions judged to show complete clinical response, only two showed complete absence of the KS cells throughout the biopsy samples; the remainder had residual KS in the deep dermis (Fig. 3).
The number of KS cells in short-term biopsies detected by immunohistochemical staining was generally lower in the superficial dermis compared with the relatively untouched deep dermis. These changes were evident even at 30 min post-treatment. No influx of immune cells was seen in these short-term biopsies. The lack of increase in inflammatory cells detected using immunohistochemical markers in the short term biopsies (up to 5 days) indicates there was no influx of such cells in the immediate post-therapy period (Table 3, Fig. 4). Neither were any increases noted in the 24-48 h period (data not shown). In longer term biopsies (1-10 months), 10 out of 12 demonstrated replacement of KS cells by fibrosis in the superficial dermis (considered complete pathologic response superficially). In these cases, there were very few inflammatory cells in the scarred area (Table 3). Fig. 5 shows a complete pathologic response that correlated with the complete clinical response illustrated in Fig. 6. As in the short-term biopsies, with the exception of one biopsy sample, there was no evidence of influx of immune cells in the areas affected by treatment compared to those that did not appear to be affected by treatment. Fig. 7.
Cutaneous phototoxicity reactions were seen in seven of the 25 patients (27%), all occurring within the 4 week period during which patients were cautioned to avoid bright lights and sunlight without protection. This is similar to the rate of non-compliance in other Photofrin trials. The sunburn-type reaction of slight erythema to slight edema occurred only on sunlight-exposed surfaces, most frequently on the face and exposed parts of the neck. The onset of reaction is rapid, appearing within a few minutes of exposure and generally accompanied by a ‚prickling‚ sensation. The reactions all subsided within 72 h without sequelae. No other phototoxicity was noted.
The most common adverse reaction in the treatment field was mild pain following treatment (24-48 h) which was adequately controlled with oral analgesics.
Toxicity of normal tissue within the treatment field was scored on a scale of 0 to 5 (Table 4). Areas treated at 250J/cm2 or below showed only mild normal skin reactions of erythema and edema (scores ≤2.5) and healed without scabbing or necrosis. Light doses of 300-400J/cm2 tended to demonstrate more severe normal skin reaction (score ≥3).
The responses indicated in Table 4 are consistent with other studies that have shown that the effect of PDT is light dose-dependent; the depth of the biologic effect is consistent with findings of previous clinical trials of PDT for cutaneous lesions[10,11]. The overall response rate for the most therapeutically effective doses of 200-300J/cm2 was 84% with a complete clinical response of 33%. While duration of follow-up was limited as patients elected to enter other studies to control the systemic manifestations of KS or other aspects of their HIV infection, the results compare favorably with other therapeutic interventions. There are numerous approaches to the palliative treatment of AIDS-related KS. Although all of them have some degree of efficacy, they all have disadvantages as well. Among local treatments radiation therapy is effective and safe, but only a few lesions can be treated at one time and may further compromise an already fragile immune system. Cryosurgery is effective, especially for smaller lesions, but larger lesions may require multiple treatments over a few months, and this therapy may result in scarring and/or hypopigmentation. Intralesional chemotherapy with vinblastine can be effective but may be a painful procedure that is unresponsive to local anesthesia and is accompanied by subsequent hyperpigmentation. Among systemic therapies bleomycin/vincristine, adriamycin bleomycin/vincristine (ABV)[13,14], or liposomal doxorubicin (LD) have been shown to result in 50-80% responses, but few patients have complete responses and relapse is common after 4-6 months, even with combined treatment. The ABV and LD regimens are similar in efficacy; however, LD is associated with less toxicity. There has been considerable experience with alpha interferon: this biologic response modifier appears to be especially effective in individuals with CD4 counts >200×106 cells/l[17,18]. More recently, retinoids have been used with some benefit but their role in Kaposi‚s sarcoma remains to be defined.
Photodynamic therapy used as described in this report allows for treatment of as many as 40-50 lesions in a single setting, is generally not associated with pain during treatment (however, moderate pain is frequently seen following treatment) and can be used when patients no longer respond to the treatments noted above. Thus, PDT provides one more weapon in the arsenal against this condition. However, the photosensitizer used here (Photofrin®) induces cutaneous photosensitivity for 4-6 weeks during which time patients must protect themselves from bright light, especially sunlight. Total cost (including light source) is relatively high and is comparable to radiation therapy. The results reported here have been achieved in most patients with a single treatment session and are effective in those with all stages of HIV disease.
Cutaneous photosensitivity was of low frequency (27%) and was minor (slight sunburn on hand and/or face). Treatment on the face and neck often resulted in facial edema which resolved spontaneously within 72 h and is consistent with our previous experience utilizing PDT for other cutaneous malignancies on the face and neck.
Nineteen treatment areas resulted in scarring, all at light doses of 300-400J/cm2. However, 10 of those areas were in a single patient treated at 300J/cm2 who had previously received adriamycin which may have contributed to this adverse outcome. In this institution, severe reaction to PDT has also been observed in patients being treated for breast cancer metastases if adriamycin had been given previously. Because of normal tissue damage (in the treatment field) at light doses of greater than 350J/cm2 the maximum tolerated dose was thought to be 300J/cm2.
There was no correlation between peripheral blood CD4 cell counts and treatment response (data not shown), consistent with the lack of inflammatory reaction observed in these biopsy samples. The presence of immediate microscopic cellular debris suggests direct physical destruction of tumor, rather than destruction due to an inflammatory response induced by this therapy. An inflammatory component to the action of PDT had been suggested, both in pre-clinical and clinical studies[20-22].
Hebeda et al.  reported on the treatment of AIDS-related cutaneous KS lesions with 2.0mg/kg Photofrin followed by light doses of 70-120J/cm2. Their tumor responses were characterized by full field necrosis of normal skin. It has been reported previously that light doses exceeding 25-50J/cm2 with Photofrin doses of 2.0mg/kg produce no selectivity and may lead to whole field necrosis of the skin[10,11,24]. We have previously reported on our strategy to maximize the therapeutic index for treatment by lowering the Photofrin dose with escalating light dose  as was done in this study. In this way, the unacceptable cosmetic results reported by Hebeda et al. could be largely avoided while maintaining good tumor response. However, we would caution against using PDT even at the Photofrin and light doses reported here for large areas of contiguous disease. This can lead to a poor outcome, often with ulceration and infection as was found in a patient treated by us before this trial.
In conclusion, PDT with Photofrin offers safe palliative relief for cutaneous KS without apparent further compromise of the patient‚s immune status.
The authors thank S. Shanler, D. Buscaglia, C. Conti, D. Babich, W. R. Potter, M. Cooper, C. Sommer, D. Blaird-Wagner, L. Wood, K. Weishaupt, A. Sumlin, J. Felski and D. Donaldson for their assistance during this study.
1. Wang CE, Schroeter AL, Su WPD. Acquired immunodeficiency syndrome-related Kaposi‚s sarcoma. Mayo Clin Proc
2. Chah LY, Gill PS, Levine AM, et al. Radiation therapy for acquired immunodeficiency syndrome-related Kaposi‚s sarcoma. J Clin Oncol
3. Evans LM, Itri LM, Campion M, et al. Interferon alpha 2a in the treatment of acquired immunodeficiency syndrome-related Kaposi‚s sarcoma. J Immunother
4. Gill PS, Lunardi-Iskandar Y, Louis S, et al. The effects of preparations of human chronic gonadotropin on AIDS-related Kaposi‚s sarcoma. New Engl J Med
5. Lightdale CJ, Heier SK, Marcon NE, et al. Photodynamic therapy with porfimer sodium versus thermal ablation therapy with Nd:YAG laser for palliation of esophageal cancer: A multicenter randomized trial. Gastrointest Endos
6. Fisher AM, Murphree Al, Gomer CJ. Clinical and preclinical photodynamic therapy. Lasers Surg Med
7. Dougherty TJ, Marcus SL. Photodynamic therapy. Eur J Cancer
8. Webster GF. Local therapy for mucocutaneous Kaposi‚s sarcoma in patients with acquired immunodeficiency syndrome. Dermatol Surg
9. Dougherty TJ, Cooper MT, Mang TS. Cutaneous phototoxicity occurrences in patients receiving Photofrin®. Lasers Surg Med
10. Wilson BD, Mang TS, Stoll H, Jones C, Cooper MT, Dougherty TJ. Photodynamic therapy for treatment of basal cell carcinoma. Arch Dermatol
11. Khan S, Dougherty TJ, Mang TS. Evaluation of photodynamic therapy in management of cutaneous metastatic breast cancer. Eur J Cancer
12. Tappero JW, Berger TG, Kaplan LD, Valberding PA, Kohn PA. Cryosurgery for cutaneous Kaposi‚s sarcoma associated with acquired immunodeficiency syndrome. J Am Acad Dermatol
13. Gill PS, Bernstein-Singer M, Espina BM, et al. Adriamycin, bleomycin and vincristine chemotherapy with recombinant granulocyte-macrophage colony-stimulating factor in the treatment of AIDS-related Kaposi‚s sarcoma. AIDS
14. Gill PS, Rarick M, McCutchan LA, et al. Systemic treatment of AIDS-related Kaposi‚s sarcoma: results of a randomized trail. Am J Med
15. Northfelt DW, Dezube BJ, Thommes JA, et al. Efficacy of peglylated-liposomal doxorubicin in the treatment of AIDS-related Kaposi‚s sarcoma after failure of standard chemotherapy. J Clin Oncol
16. Gill PS, Wernz J, Scadden DT, et al. Randomized phase III trial of liposomal daunorubicin versus doxorubicin, bleomycin, and vincristine in AIDS-related Kaposi‚s sarcoma. J Clin Oncol
17. Real FX, Oettgen HF, Krown SE. Kaposi‚s sarcoma and the acquired immunodeficiency syndrome: treatment with high and low doses of recombinant leukocyte A interferon. J Clin Oncol
18. Krown SE, Gold JWM, Real FX, et al. Interferon alpha-2a ± vinblastine (VLB) in AIDS-associated Kaposi‚s sarcoma (KS/AIDS); therapeutic activity, toxicity and effects on HTLV-III/LAV viremia [abstract]. J Interferon Res
19. Bower M, Fife K, Landau D, Gracie F, Phillips RH, Gazzard BG. Phase II trial of 13-cis-retinoic acid for poor risk HIV-associated Kaposi‚s sarcoma. Int J STD AIDS
20. Kick G, Messer G, Goetz A, Plewig G, Kind P. Photodynamic therapy induces expression of interleukin 6 by activation of AP-1 but not NF- B DNA binding. Cancer Res
21. Krosl G, Korbelik M, Dougherty GJ. Induction of immune cell infiltration into murine SCCVII tumor by photofrin-based photodynamic therapy. B J Cancer
22. Nseyo UO, Whalen RK, Duncan MR, Berman B, Lundahl BS. Urinary cytokines following photodynamic therapy for bladder cancer. Urology
23. Hebeda KM, Huizing, MT, Brower PA, et al. Photodynamic therapy in AIDS-related cutaneous Kaposi‚s sarcoma. J Acquir Immune Def Syndr Hum Retrovirol
24. Mang TS, Dougherty TJ, Potter WR, Boyle DG, Somer S, Moan J. Photobleaching of porphyrins used in photodynamic therapy and implications for therapy. Photochem Photobiol