Long-term results in the treatment of vitiligo with oral khellin plus UVA

ARTICLE

For the past 20 years, oral psoralen plus UVA (PUVA) has been the first-line therapy for patients with extensive vitiligo, even though it induces phototoxic reactions, nausea, and elevation of liver transaminases. However, PUVA treatment is also a well-known carcinogen in psoriasis patients and has recently been associated with skin cancer in PUVA-treated vitiligo patients [1, 2]. As long ago as 1982, Abdel-Fattah et al. [3] sought a safer therapy in the form of khellin, a furochromone whose chemical structure closely resembles that of psoralen 8-MOP, and sunlight. In a placebo-controlled study, Abdel-Fattah et al. found the khellin-plus-sunlight regimen to be an effective treatment for vitiligo. Later, Ortel et al. [4] explored the effects of oral khellin plus artificial UVA (KUVA) in patients with vitiligo and reported cosmetically satisfactory results and good repigmentation of vitiliginous areas in a large proportion of patients. More importantly, their results were comparable to those obtained with PUVA therapy but without the phototoxic side effects. We now wish to report on our long-term experience in 28 vitiligo patients treated with KUVA during the last 14 years.

Patients and methods

Patients

A review of the records of the Department of Dermatology, University of Graz, Austria, revealed 28 cases in which patients with generalized, universal, or localized vitiligo of more than 6 months duration were treated with KUVA between 1986 and 1998 (Table I).

Methods

The retrospective review revealed that all 28 patients underwent routine blood chemistry tests. In addition, laboratory thyroid function tests were done and antibodies against thyroid constituents (i.e. antithyroglobulin), antinuclear antibodies, and antiparietal cell antibodies were assessed in most patients. Liver enzymes were tested before starting therapy; at 2, 4, 6, 8, and 12 weeks; and thereafter in regular 8-week intervals during the entire treatment period.

All 28 patients were treated with KUVA two to three times weekly. The KUVA regimen consisted of taking 100 mg of khellin (packed along with lactose as excipient into gelatin capsules) 2.5 hrs before exposure to UVA light. To avoid nausea, the patients were advised to eat yogurt with the khellin capsules. The light source used was a Sellas 12,000 UVA tanning system equipped with metal halide lamps (Sellas Medizinische Geräte GmbH, Gevelsberg, Germany). In most patients the initial UVA treatment dose was 5 J/cm² for skin types I and II and 10 J/cm² for skin types III to V. In most cases, these UVA doses were increased after 1 to 2 weeks of therapy to 10 J/cm² in skin types I and II and to 15 J/cm² in skin types III to V. If less than 5% of vitiliginous areas became repigmented after 3 months of KUVA therapy, therapy was discontinued. Otherwise, patients continued treatment for 1 to 3 years, with interruptions only during the summer months from June to August.

Evaluation of treatment

The response to treatment (i.e., repigmentation of depigmented areas) was estimated retrospectively by comparing photographs of vitiliginous areas taken before therapy and at the end of therapy. Responses were rated by the investigators as good (i.e., repigmentation of 70-100%), medium (i.e., repigmentation of 30-70%), moderate (i.e., repigmentation of 5-30%), and none (repigmentation of less than 5%).

Statistics

Depending on the type of data involved, statistical analysis was performed by means of the Spearman rank test, unpaired two-tailed Student's t test, or khi² test using StatView program version 5.0.1 (SAS Institute Inc., Cary, NC) on a Power Macintosh (Apple Computer Inc., Cupertino, CA). A difference was considered to be statistically significant when the P-value was smaller than 0.05.

Results

Response to KUVA

Follicular repigmentation of the vitiliginous lesions had been found after 3 to 10 weeks of therapy in 19 of 28 patients (68%). Of 17 patients who had continued therapy for more than 3 months, 7 (41%) had a good response, 3 (18%) had a medium response, and 7 (41%) had a moderate response at the end of therapy (Tables II and III). The good response came after a mean of 194 treatments (range, 69-386) and a mean cumulative UVA dose of 2,036 J/cm² (range, 690-4,411 J/cm²) (Table III). The medium response came after a mean of 96 treatments (range, 69-114 treatments) and a mean cumulative UVA dose of 1,027 J/cm² (range, 890-1,140 J/cm²). The moderate response came after a mean of 89 treatments (range, 48-150 treatments) and a mean cumulative UVA dose of 872 J/cm² (range, 445-1,665 J/cm²). No patient achieved complete repigmentation of all vitiliginous lesions. However, 5 patients did achieve total repigmentation of some vitiliginous lesions on the face and trunk, whereas no patient achieved complete repigmentation of lesions of the fingers or distal dorsal surfaces of the hands or feet. There was no significant difference in response between patients with vitiligo associated versus not associated with autoimmune thyroiditis. Patients with short-term vitiligo responded no better than patients who suffered from vitiligo longer than 3 years. In three patients, therapy was discontinued after 3 months because of poor response.

Correlation analysis revealed that the response to therapy was statistically significantly linked to the number of KUVA treatments (Spearman rank test, r = 0.833, P ¾ 0.001) and to the total cumulative UVA dose (r = 0.840, P ¾ 0.001) (Table III). There was no significant correlation between clinical response and age or duration of disease before treatment and no association between patient's sex or presence of thyroid disease and clinical response.

Short-term side effects of KUVA

In total eight patients (29%) reported episodes of mild nausea after the ingestion of khellin, particularly in the first week(s) of therapy, two patients (7%) had elevated liver enzymes, two patients (7%) had gastritis, and one patient (4%) reported a temporary reduction in visual acuity that resolved after KUVA therapy. In total KUVA had to be stopped in two of the patients with nausea after khellin administration and in one of the patients with elevation of liver enzymes, and in another patient with presumably therapy-unrelated massive sinusitis. In three patients KUVA treatment was stopped prematurely due to non-response at 3 months of treatment. In addition, four patients discontinued KUVA within 3 months after start of treatment for personal reasons.

Long-term follow-up

Follow-up assessments were done in 1996 and in 1998 (for those patients who were treated after 1996) (Table II). Eighteen of 28 patients (64%) were reexamined in our department; 5 (18%) were interviewed on the telephone; and 5 (18%) were lost to follow-up. At a mean of 40 months (range, 4-110 months) after the end of KUVA treatment, 3 patients (11%) reported further improvement; 8 (29%) reported no further progression, 4 (14%) reported stable disease; and 8 (29%) reported disease progression. One patient with universal vitiligo reported progression of the disease to total depigmentation (except of the hairs) and spontaneous repigmentation a few years later. No skin cancers or actinic damage of vitiliginous skin were found in any patient.

Discussion

We report here the effects and side effects of KUVA in 28 patients with vitiligo treated within the last 14 years in our department in Graz (Table I). Of 61% (17/28) of these patients who had continued KUVA therapy for more than three months, seven (41%) had a good response (i.e., repigmentation of more than 70% of vitiliginous skin) (Table II). The response to KUVA was clearly related to the number of treatments and total UVA dose applied (Table III). Our data compare well with those of Ortel et al. [4], who reported having comparable success (i.e., repigmentation of more than 70%) with KUVA and PUVA, provided that KUVA therapy was continued long enough. For instance, their "good" response rate of 41% (5/12) in vitiligo patients who received more than 100 treatments compares with our 55% (5/9). Abdel-Fattah et al. [3], who investigated the effect of khellin plus sunlight in 30 patients with vitiligo, reported a response of more than 50% repigmentation in 40% of patients (12/30) and a response of less than 25% or no repigmentation in 60% (18/30). However, none of the 30 control subjects in that study, all of whom received placebo and sunlight, showed any repigmentation, indicating that khellin and exposure to natural sunlight was effective but not sunlight itself. Interestingly, experimental work by Nordlund et al. [5] has indicated that, at least in mice, best melanoyte proliferation was obtained by PUVA with low doses of UVA. However, consistent with another clinical study [4] on KUVA in vitiligo, we used medium doses of UVA (of 5 to 15 J/cm²) after khellin administration for treatment.

In a recent meta-analysis, Njoo et al. [6] compared the response to common treatment modalities for localized and generalized vitiligo. They found that repigmentation of more than 75% was achieved in roughly half of the 349 patients with generalized vitiligo in 6 series, who were treated with PUVA versus less than 20% of 65 patients in 5 series (including the 30 patients of Abdel-Fattah) who were treated with KUVA. This relatively low success rate of KUVA compared with PUVA may have been due to short therapy duration (e.g., less than 4 months in the study by Abdel-Fattah [3]). However, any direct comparison is difficult because up to now there have been no controlled studies comparing KUVA with PUVA therapy in patients with vitiligo.

In our retrospective study, we also investigated the short-term and long-term side effects of KUVA. The most common short-term side effect was nausea, occurring in 8 of 28 patients (29%), and mainly in the first week(s) of treatment. However, only 2 patients discontinued KUVA because of nausea. In their meta-analysis of the literature, Njoo et al. [6] reported that nausea and vomiting were found in 29% of the patients treated with PUVA (80/277) but only in 9% of the patients treated with KUVA (6/65). Two of our 28 patients (7%) experienced elevations in liver transaminases; consequently, one of them had to discontinue therapy. This was not unexpected since temporary elevation in liver transaminases has been reported in response to both khellin and methoxsalen [4, 7]. Ortel et al. [4] observed a mild elevation in liver transaminases to up to three times the normal limit in 7 of 28 patients (25%), however, as in our patients, these values returned to normal in all cases within weeks of discontinuing KUVA therapy. This is in agreement with the fact that in the 1940s and 1950s khellin was prescribed as a vasodilator for the treatment of angina pectoris and was found to be a safe drug [8]. It was also given for long continuous periods in doses of 300 mg daily and even higher dosages without causing severe side effects.

During KUVA therapy, none of our patients developed the phototoxic reactions or freckling of the skin that are often seen in patients undergoing long-term PUVA therapy. This agrees with the literature, in which there have been no reports of phototoxic reactions to KUVA but reports of such reactions in 25% of psoralen (8-MOP)-treated patients (69/277) [6].

Particular attention must be paid to the possible carcinogenic risk of long-term KUVA therapy in vitiligo because many patients are younger than 30 years old. In contrast to PUVA, little is known about the long-term side effects of KUVA. Khellin mainly forms monofunctional adducts and interstrand cross-links with DNA with a maximum action spectrum at 360 nm, but such activity requires 1.2 x 10² times more UVA irradiation than used in PUVA therapy [9-11]. Therefore, khellin may be less carcinogenic than psoralen. In fact, since the introduction of khellin into photochemotherapy, there have been no reports of skin cancer resulting from KUVA therapy in patients with vitiligo, whereas there have been reports of keratoses and squamous cell carcinomas resulting from PUVA therapy [1, 2]. However, the occurrence of squamous cell carcinoma in vitiligo after PUVA may be very rare. Indeed, the vitiligo patients from the two anecdotal case reports [1, 2] on squamous cell carcinoma after PUVA were both older than 60 years and their cumulative UVA doses did not exceed the maximum recommended safety dose of 1,000 J/cm². In addition, one of the patients had developed squamous cell carcinoma within 3 years after the start of PUVA, a time period that is relatively short for tumor induction in general. Taken together, this makes a causative relationship between PUVA and squamous cell carcinoma formation in those cases somewhat unlikely. In our 14-year experience with KUVA therapy for vitiligo none of our vitiligo patients has developed keratoses or skin cancer. Nor have we seen any of the skin changes reported for long-term PUVA therapy given at high cumulative UVA doses (i.e., freckling, mottling of the skin, and solar degeneration). However, only 18 of 28 patients (64%) were reexamined at our department in the follow-up after KUVA treatment, and, thus, we can not exclude that one or more of the remaining patients may have developed actinic skin changes and/or skin cancer after KUVA. Taken together the results nevertheless suggest that KUVA may be safer than PUVA in the long-term.

CONCLUSION

In summary, we conclude that oral KUVA may effectively induce repigmentation of vitiliginous areas of the skin at rates comparable to those of PUVA therapy, provided treatment is continued long enough.

However, controlled studies have to be done to determine how the responses to KUVA and/or PUVA compare directly with the response to narrow-band 311 nm UVB therapy, which has been recently used with great success in limited numbers of vitiligo patients, including children [12, 13].

Article accepted on 20/2/01

REFERENCES

1. Buckley DA, Rogers S. Multiple keratoses and squamous carcinoma after PUVA treatment of vitiligo. Clin Exp Dermatol 1996; 21: 43-5.

2. Takeda H, Mitsuhashi Y, Kondo S. Multiple squamous cell carcinomas in situ in vitiligo after long-term PUVA therapy. J Am Acad Dermatol 1998; 38: 268-70.

3. Abdel-Fattah A, Aboul-Enein MN, Wassel GM, El-Menshawi BS. An approach to the treatment of vitiligo by khellin. Dermatologica 1982; 165: 136-40.

4. Ortel B, Tanew A, Hönigsmann H. Treatment of vitiligo with khellin and ultraviolet A. J Am Acad Dermatol 1988; 18: 693-701.

5. Nordlund JJ, Ackles AE, Traynor FF. The prolifertive and toxic effects of ultraviolet light and inflammation on epidermal pigment cells. J Invest Dermatol 1981; 77:361-368.

6. Njoo MD, Spuls PI, Bos JD, Westerhof W, Bossuyt PMM. Nonsurgical repigmentation therapies in vitiligo: Meta-analysis of the literature. Arch Dermatol 1998; 134: 1532-40.

7. Duschet P, Schwarz T, Pusch M, Gschnait F. Marked increase of liver transaminases after khellin and UVA therapy. J Am Acad Dermatol 1989; 21: 592-4.

8. Galal E, Kandil A, Abdel-Latif M. Acute toxicity of Ammi visnaga principles. J Drug Res Egypt 1975; 7: 1-7.

9. Cassuto E, Gross N, Bardwell E, Howard-Flanders P. Genetic effects of photoadducts and photocross-links in the DNA of phage lambda exposed to 360 nm light and tri-methylpsoralen or khellin. Biochim Biophys Acta 1977; 475: 589-600.

10. Trabalzini L, Martelli P, Bovalini L, Dall`Acqua F, Sage E. Photosensitization of DNA of defined sequence by furochromones, khellin and visnagin. J Photochem Photobiol: B Biol 1990; 7: 317-36.

11. Abeysekera BF, Abramowski Z, Towers GHN. Genotoxicity of natural furochromones, khellin and visnagin and the identification of a khellin-thymine photoadduct. Photochem Photobiol 1983; 38: 311-5.

12. Westerhof W, Nieuweboer-Krobotova L. Treatment of vitiligo with UVB(311 nm) versus topical PUVA. Arch Dermatol 1997; 133: 1525-8.

13. Njoo MD, Bos JD, Westerhof W. Treatment of generalized vitiligo in children with narrow-band (TL-01) UVB radiation therapy. J Am Acad Dermatol 2000; 42: 245-53.