Dead Sea treatment - principle for outpatient use in atopic dermatitis: safety and efficacy of synchronous balneophototherapy using narrowband UVB and bathing in Dead Sea salt solution

ARTICLE

Since the fifties patients with AD have been treated successfully at the Dead Sea [1-10]. Ultraviolet sunlight and salt water form the two main components of the treatment principle. To use the therapeutic advantages also for outpatients in Europe, a special treatment system simulating these conditions was developed. As light source, narrowband TL-01 lamps (311 nm) by Philips (Eindhoven, Netherlands) were selected in view of clinical studies, which have tended to report good improvement in AD [11-14]. The British Photodermatology Group concluded no greater risk for TL-01 lamps in comparison to conventional UVB light sources [15], which was also confirmed by other authors [16, 17]. In addition, the antiinflammatory effect of hypertonic solutions - elution of human leucocyte elastase and reduction of Langerhans cells in the skin - [18, 19], and the special effect of Dead Sea salt caused by its high Mg²⁺-ions have already been shown [18, 20, 21]. To analyse the efficacy of this treatment for outpatients while keeping to their daily routine, we set up an uncontrolled multicenter trial under GCP conditions at private practices. Our setting allowed us to investigate an expected loss of efficacy from switching a clinical trial to daily routine. To achieve data for a clinical trial and daily routine situations, two populations were compared: patients strictly treated atp with no protocol deviations (data usually published in clinical trials) and all patients participating the trial who received treatment at least once (itt as model for daily practice).

In the present study we report the data of one of the largest multicenter trials performed in 615 itt- and 143 atp-patients with AD using a standardized treatment system imitating the Dead-Sea-treatment-principle for outpatient use.

Material and methods

Treatment system

The treatment system consists of two components: an anatomically-shaped bathtub with a computer controlled purification system for salt solution balneotherapy and a continuously adjustable and individually dispensable light console with a reflector system located above the tub using TL-01 UVB lamps (311 nm, Philips, Eindhoven, Netherlands and Okkaido-Vario-System Tomesa^® Alteglofsheim, Germany) for synchronous phototherapy. Each unit is a closed therapy system performing continuous purification of the bath solution. A 10% Tomesa^®-sole (in German: Totes Meer Salz, Dead Sea salt) is used as analogue to the ion-formation of the Dead-Sea. A reduction to 10% of the sole concentration is necessary due to the purification system. The irradiation dosages are measured periodically at the water surface using a gauged dosimeter (Kompaktdosimeter RM-21, Gröbel, Ettlingen, Germany).

Treatment strategy

Study protocol for outpatients was set up according to previous experiences with inpatient therapy [9, 20, 22]. All patients should perform 3-5 sessions per week and 35 sessions in total with an increasing treatment duration (bathing and irradiation) up to 30-35 minutes per session. Starting dosage was determined according to the skin type of the patient. During treatment, the dosage per treatment unit was increased by prolonging the bathing time. Incremental steps were dependant on the skin type of patients and the patients' individual acceptance. Patients started a treatment session with a short bath without phototherapy followed by the synchronous balneophototherapy. Frequent turnovers (every 4 minutes) guaranteed a constant covering of the irradiated skin with the solution and were necessary to reach all parts of the body. One treatment session took approximately 30-45 minutes including a shower before treatment.

Inclusion criteria

* Atopic dermatitis (AD),

* SCORAD score [23, 24] > 35 at baseline.

Exclusion criteria

* erosions, ulcers, viral or bacterial superinfection,

* severe systemic diseases,

* use of medication with effects on photosensitivity,

* concomitant or previous malignant skin tumors,

* under 16 years of age,

* pregnancy, lactating.

Evaluation of efficacy and side effects

As a primary parameter, treatment effects on AD were assessed by the SCORAD score at baseline, after 20 treatment sessions and at end of treatment (after 35 treatment sessions in patients treated according to protocol). Furthermore irradiation dosages, bathing duration, side effects and reasons for withdrawal were documented. Quality of life and contentment with treatment were assessed using especially developed questionnaires.

Statistical methods

We performed an open, uncontrolled, multicenter trial designed for documentation of safety and efficacy of synchronous balneophototherapy in AD under outpatient conditions close to daily practice. Cooperation with the primary health insurance companies in Bavaria, Germany, ensured that all their insured patients treated with this new treatment modality are documented accordingly and therefore selection bias could be minimized. Analysis was based on two groups:

1. Patients treated according to the study protocol without study withdrawal, non-compliance or other major protocol violations (atp-population, "clinical trial").

2. All patients participating in the study who received active treatment at least once (itt-population, "daily routine").

To ensure that all patients included in the study could be analysed considering clinical efficacy, missing values for SCORAD were replaced by the last value available from the patient for the itt-population. This "last observation carry forward principle" represents a worst case analysis commonly used in clinical trials. Since patients with early study withdrawal are a subgroup of itt-population efficacy data considering SCORAD-values were handled accordingly. Relative improvement in the primary parameter was calculated individually, mean values are presented in the paper. Tests between the atp- and itt-population could not be performed due to dependant data (atp is a subpopulation of itt).

Descriptive statistics

Descriptive statistics are presented for each group concerning demographics, treatment characteristics, and disease severity measurements at each assessment time.

Confirmatory statistics

Non-parametric statistics were used since the primary parameter (SCORAD) was non-normally distributed. Treatment effects over time were tested using the Wilcoxon-Rank-Sum, criteria influencing treatment efficacy (atp versus early study withdrawal) were detected by H-test by Krustal and Wallis.

Quality assurance and monitoring

The study protocol was established by the Department of Dermatology of the University of Regensburg as the scientific study center, which also trained the investigators and monitored closely onsite according to GCP-criteria.

Results

Study centers

The multicenter trial was commenced at ten different centres in Bavaria in January 1996. During study progress, further centers could be recruited. This analysis is based on all patients included in the study up to April 1999 with 55 participating centers.

Atp- and itt-population, patients with withdrawal

615 patients with AD were admitted to the study; 143 (23%) finalized atp. 77% (472 patients) had to be excluded from atp due to protocol violations which however will frequently occur in daily routine (Fig. 1): early study termination with less than the forseen 35 treatment sessions was observed in 47% of the patients. The rate of patients withdrawing the study due to "lack of time" was markedly high (16%). Analysis of the age distribution showed that 53% of these patients were 40 years of age or below. Obviously it is more difficult for younger people who are trying to build up their working career to regularly undergo a treatment which takes at least 45-60 minutes per session. Further reasons for early withdrawal were non-compliance (12%), sufficient relief (7%), lack of efficacy (6%), intercurrent illness (4%), and side effects (3%) (Table I). 27% completed less than 3 treatment sessions per week. Repeated admittance to the study in 8% of patients was the third reason for exclusion of patients from atp-analysis.

Patient characteristics

Patient characteristics are summarized for the itt-, atp-population, and patients with early study withdrawal in Table II. Here are the data of the itt-population: 46% were men, 54% women; mean age 37 years. All patients were Caucasian. Only 0.5% of the patients in employment were declared unfit for work due to the treatment.

64% of the patients had Fitzpatrick skin type II, 33% type III and 3% type IV. The clinical appearance before treatment was classified as chronic type with lichenification in 51%, acute type in 14% and combination type in 35% (chronic and acute lesions). 30% had complained about their lesions for more than one year, 12% between 6 and 12 months, 20% for 3 to 6 months, 26% for 1 to 3 months and 12% for less than 1 month.

Treatment characteristics

143 patients were treated atp (itt: 615) with 35 treatment sessions (mean value for itt: 26) with 3.8 sessions per week (itt: 3.1) (Table II). 72% of atp and 97% of itt-patients used an additional skin care during treatment, but there was no statistically significant difference in treatment outcome between patients with and without additional ointment.

UVB dosage

Patients were irradiated individually according to their skin type and acceptance. The total dose of irradiation per patient was 16.3 J/cm² in mean (itt: 11.9 J/cm²/patients with early study withdrawal: 9.1 J/cm²).

Primary parameter

Upon admission and prior to the commencement of treatment, the average SCORAD was 59.4 for all patients (itt) and 60.1 for patients treated atp. Baseline values for patients with early study withdrawal were also comparable (57.8). After 20 treatment sessions, mean SCORAD decreased to 44.5 in all patients and 37.5 in patients treated atp. At the end of study after 35 sessions, SCORAD was significantly reduced in both itt and atp: mean SCORAD 35.2 for all patients and 27.1 for atp. This corresponds to a relative improvement in SCORAD of 41% in all patients and 55% for patients treated atp (Fig. 2). For the subgroup of protocol-violating patients with early study withdrawal (mean number of sessions: 15), mean SCORAD decreased to 42.6, a relative improvement of 26% (Table II). Decrease of SCORAD-values after 20 sessions and at end of treatment in comparison to SCORAD-valus at baseline was statistically significant in atp-, itt-population and in patients with early study withdrawal (p < 0.05).

Concerning clinical efficacy in itt-population, age and sex of patients had no statistically significant influence on improvement of SCORAD-values. Besides the number of treatments in total and per week, the following criteria evaluated at baseline influenced treatment efficacy: type of AD, duration of symptoms, and distress due to problems in personal relations and/or employment. Distributions of these criteria were statistically significantly different between atp-population and patients with early withdrawal (p < 0.05): in withdrawals, 26% of patients had acute type of AD (in atp-population only 16%), 41% had lesions of up to three months duration before starting the study (atp-population only 31%), and 29% said they had "severe" or "very severe distress" at baseline (atp-population only 10%). Therefore it can be suggested that this treatment is more appropriate for patients with chronic type of AD and less effective especially in patients already suffering from severe distress pre-treatment since out-patient synchronous balneophototherapy probably increases distress with the important time consumption.

Secondary parameters

For evaluation of impairment in quality of life patients were asked before and after treatment "how much they feel impaired" using an ordinal scale. Data of 591 itt-patients could be analyzed: before treatment, 15% of patients stated to be "very much" impaired by the disease, 38% complained about "severe", 31% about "moderate" and 11% about "slight impairment" whereas 5% felt "not impaired at all" by the disease. At finalization of therapy, patients were asked about any improvements in their quality of life: 54% acknowledged an improvement, whereas 37% did not recognize any change after treatment, 9% felt worse than at the beginning of treatment schedule.

Safety

76 patients (12.3%) developed side effects during treatment. 17 patients (3%) had to quit the treatment due to side effects. Most frequently observed side effects were erythema induced by phototherapy in 45 patients (7.3%) and burning of the skin due to salt solution in 22 patients (3.6%). Due to these side effects, a short treatment interruption and concomitant treatment with topical steroids became necessary in 2% each. Further side effects were chlorine incompatibility, circulation disorders and claustrophobia in less than 1% of patients each (Table III).

Discussion

Safety and efficacy of the Dead-Sea-treatment-principle could be proven under outpatient conditions for a representative number of patients suffering from AD. Single center studies and studies with a small number of patients can be biased due to pre-selected patients "suitable for the study protocol". Analysis according to protocol may further exaggerate treatment effects. Therefore it is of special interest that a statistically significant improvement of SCORAD could be seen in both the atp- and itt-population. We could provide valuable data which help to answer the question whether the efficacy as observed in a clinical trial is transferable to daily routine. Our results clearly show a difference in efficacy between using a treatment modality atp (data mainly published for clinical trials) and itt (our model for daily routine).

Our rate of early study withdrawals was high (47%). Nevertheless, only 3% of all patients withdrew due to side effects: the most important reason was burning of skin caused by the high salt solution. This problem occurs in the first minutes of bathing during initial sessions and could be probably avoided by reducing the salt concentration at beginning of treatment. While skin lesions already cleared sufficiently for patients before session 35 in 7%, increasing skin lesions during treatment were also observed in 6% of the patients. We therefore believe that a special group of patients does not benefit from this therapy. The treatment is time consuming with approximately 30-45 minutes per session including initial shower. For 16% this was the reason for discontinuing treatment, especially as most patients were in employment. The fact that 27% of the patients had to be excluded from the atp-population due to "less than 3 treatments per week" further pin points the time problem. Nevertheless reasons for withdrawals in our study appear to be comparable with data from the literature: "side effects" were observed by Kobyletzki and George et al. [11, 25] in 1-6% of their cases and in 3% in our study. "No change" or "deterioration" were obtained in 4-8% [11, 26, 27] which is quite similar to our study (6%). Dittmar et al. [26] found similar data in comparison to our study (11% versus 16%) considering time problems. The number of non-compliant patients in our study (12%) also seemed to be similar in comparison to other studies (4-10%) [12, 26], and was markedly lower in comparison to the use of topical corticosteroidal cremes (non-compliance 25%) [28]. Comparison of efficacy with inpatient data is difficult due to lack of SCORAD-evaluations in most studies [1-10]. Furthermore light sources and concentration of the salt solution are not really comparable and climatic conditions at the Dead Sea, which may positively influence treatment, could not be fully imitated. Therefore it is only possible to compare efficacy with studies using phototherapy alone or in combination with other treatments under outpatient conditions. However, important study criteria differ (e.g. severity of disease at baseline, number of treatment sessions). Grundmann-Kollmann [29] treated five patients with 15 treatment sessions with a cumulative dose of 9.2 J/cm², acheiving a SCORAD improvement of 50%. In our study, performed in 143 atp-patients, we could show an improvement of 55% with 35 sessions using 16.3 J/cm². Although the difference of SCORAD improvement to us seems rather low, it has to be considered that herein patients were much more affected (SCORAD at baseline: 33.6 versus 59.7 in our study). Der-Petrossian et al. [27] achieved a SCORAD improvement of 64% in a halfside trial (UVB 311 versus bath-PUVA) in 10 patients (17 treatments, 14 J/cm² UVB), which is higher than in our study. Although also the baseline scores were similar (67.9 versus 59.7 in our study) a comparison with our data is not really possible: Der-Petrossian used a modified SCORAD, which excludes skin lesions on the face and the criterion "sleeplessness". Furthermore, an application of 14 J/cm² in 17 treatments (16.3 J/cm² in 35 treatments in our study) suggests the use of a near erythemogenic regimen which might influence long-term side effects. Two authors used a combination of UVA and UVB broadband [25, 26, 30] as an arm in their phase III-studies using the SCORAD as primary parameter. In comparison with their results (relative SCORAD improvement 12.5% and 41%) the combination of UVB 311 nm with bathing in Dead sea salt solution seems to be superior (55%). Dittmar et al. [26] analyzed the efficacy of the combination of UVA and UVB broadband after bathing in 3-5% salt solution and achieved a SCORAD improvement of 47%. They treated 16 patients with a mean baseline score of 69.5 with 16 sessions and registered no burning or itching while bathing with this lower salt concentration. Frequent side effects reported by authors who used UVB 311 nm, combination of UVA and UVB or UVA1 [11, 12, 29, 31, 34], were "mild to moderate erythema", xerosis and induction of herpes labialis in some patients. There is no crucial difference of kind and extension of side effects in comparison to our study except burning of skin due to salt solution. Considering sanitary costs for this new treatment system a final price cannot be given at present since this study was performed under special study conditions (e.g. higher costs for documentation, monitoring, study centre, etc.). The primary health insurance companies in Bavaria have not yet decided on a reimbursement policy which surely will influence the price of this treatment system.

CONCLUSION

In conclusion our study provides two important results: (1) safety and efficacy of out-patient synchronous balneophototherapy could be proven in both atp- and itt-population. A marked difference in efficacy between atp and itt shows the importance of evaluating itt-data which provide a more realistic assessment of a treatment modality in practice. (2) Clear recommendations for optimised use in daily practice can be given: type of AD, duration of complaints and distress pre treatment had a statistically significant influence on treatment efficacy. Therefore application seems to be most effective in patients with chronic type of AD showing mainly lichenification and pruritus. Patients should be pre-selected according to compliance in the past and time resources available for treatment. This should help to avoid early withdrawals.

Acknowledgements

The study was sponsored by the primary health insurance companies in Bavaria, Germany. The authors gratefully thank all responsibles for leaving the data for analysis and the rights for publication. We furthermore gratefully acknowledge assistance of Dr. A. Gläßl and Mr. T. Walther for monitoring the investigator centers and our secretaries (Mrs. S. Komm, Mrs. W. Roiger and Mrs. B. Fendl) at the study center in Regensburg.

Article accepted on 12/7/02

REFERENCES

1. Dostrovsky A, Sagher F, Even-Paz Z. Preliminary report: the therapeutic effect of the hot springs of Zohar (Dead Sea) on some skin diseases. Harefuah 1959; 57: 13-145.

2. Avrach W. Climatotherapy at the Dead Sea. In: Farber E, et al. eds. Psoriasis. Proceedings of the second International Symposium. Yorke Medical Books, New York, 1974: 258-61.

3. Avrach W, Niordson A. Psoriasis treatment at the Dead Sea. Ugeskrift for Laeger 1974; 136: 2687-90.

4. Abels DJ, Kattan-Byron J. Psoriasis treatment at the Dead Sea: a natural selective ultraviolet phototherapy. J Am Acad Dermatol 1985; 12: 639-43.

5. Duvic M. Possible mechanism of effectivness of Dead Sea balneotherapy. J Am Acad Dermatol 1986; 17: 1061.

6. Even-Paz Z, Shani J. The Dead Sea and psoriasis, historical and geographic background. Int J Dermatol 1989; 28: 1-9.

7. Abels DJ, Rose T, Bearman JE. Treatment of psoriasis at the Dead Sea dermatology clinic. Int J Dermatol 1996; 34: 134-7.

8. Abels DJ, Even-Paz Z, Efron D. Bioclimatology at the Dead Sea in Israel. Clin Dermatol 1996; 14: 653-8.

9. Even-Paz Z, Efron D, Kipnis V, et al. How much Dead Sea sun for psoriasis. J Dermatol Treat 1996; 7: 17-9.

10. Kushelevsky AP, Harari M, Kudish AI, et al. Safety of solar phototherapy at the Dead Sea. J Am Acad Dermatol 1998; 38: 447-52.

11. George SA, Bilsland DJ, Johnson BE, et al. Narrowband (TL-01) UVB air-conditiones phototherapy for chronic severe adult atopic dermatitis. Br J Dermatol 1993; 128: 49-56.

12. Collins P, Ferguson J. Narrowband (TL-01) UVB air-conditioned phototherapy for atopic eczema in children. Br J Dermatol 1995; 133: 653-67.

13. Hudson-Peacock MJ, Diffey BL, Farr PM. Narrow-band UVB phototherapy for severe atopic dermatitis. Br J Dermatol 1996; 135: 332.

14. Goldberg LH, Kushelevsky AP. Ultraviolet light measurements at the Dead Sea. Psoriasis: proceedings of the Second International Symposium. Yorke Medical Books, New York, 1976: 461-3.

15. British photodermatology group. An appraisal of narrowband (TL-01) UVB phototherapy. British photodermatology group workshop report (April 1996). Br J Dermatol 1997; 137: 327-30.

16. Taylor CR, Liebmann EJ, Ortel B. UVB Phototherapy. In: Roenigk HH. Maibach HI, eds.: Psoriasis. Marcel Dekker Inc, New York, Basel, Hong Kong, 1998: 529-41.

17. Degitz K, Messer G, Plewig G, et al. Schmalspektrum-UVB 311 nm versus Breitspektrum-UVB. Neue Entwicklungen in der Phototherapie. Hautarzt 1998; 49: 795-806.

18. Gruner S, Zwirner A, Boonen H, et al. Der Einfluß einer Behandlung mit Salz des Toten Meeres (Tomesa-Therapie) auf epidermale Langerhanszellen - Eine klinische Studie. Z Hautk 1990; 65: 1146-51.

19. Wiedow O, Streit V, Christophers E, et al. Freisetzung humaner Leukozytenelastase durch hypertone Salzbäder bei Psoriasis. Hautarzt 1989; 40: 518-22.

20. Ständer M. Mineralsalze zur Unterstützung der selektiven Ultraviolett-Phototherapie. Hautarzt 1991; 42: 402.

21. Ludwig P, Petricj K, Schewe T, et al. Inhibition of eicosanoid formation in human polymorphonuclear leucocytes by high concentrations of magnesium ions. Biol Chem Hoppe-Seyler 1995; 376: 739-44.

22. Gambichler T, Schröpl F. Balneophototherapie bei Patienten mit Psoriasis vulgaris. Z Dermatol 1995; 181: 176-81.

23. Saurat JH. Severity scoring of atopic dermatitis: the SCORAD index. Consensus report of the European task force on atopic dermatitis. Dermatology 1993; 186: 23-31.

24. Kunz B, Oranje AP, Labreze L, et al. Clinical validation and guidelines for the SCORAD index: consensus report of the European Task Force on atopic dermatitis. Dermatology 1997; 195: 10-9.

25. Kobyletzki G, Freitag M, Herde M, et al. Phototherapy in severe atopic dermatitis. Comparison between current UVA1 therapy, UVA1 cold light and combined UVA-UVB therapy. Hautarzt 1999; 50: 27-33.

26. Dittmar HC, Pflieger D, Schempp CM, et al. Vergleichsstudie Solebäder plus UVA/B versus UVA/UVB-Monotherapie bei Patienten mit subakuter atopischer Dermatitis. Hautarzt 1999; 50: 649-53.

27. Der-Petrossian M, Seeber A, Hönigsmann H, et al. Half-side comparison study on the efficacy of 8-methoxypsoralen bath-PUVA versus narrow-band ultraviolet B phototherapy in patients with severe chronic atopic dermatitis. Br J Dermatol 2000; 142: 39-43.

28. Charman CR, Morris AD, Williams HC. Topical corticosteroids phobia in patients with atopic eczema. Br J Dermatol 2000; 142: 931-6.

29. Grundmann-Kollmann M, Behrens S, Podda M, et al. Phototherapy for atopic eczema with narrow-band UVB. J Am Acad Dermatol 1999; 40: 995-7.

30. Kobyletzki G, Pieck C, Hoffmann K, et al. Medium-dose UVA1 cold-light phototherapy in the treatment of severe atopic dermatitis. J Am Acad Dermatol 1999; 41: 931-7.

31. Krutmann J, Diepgen TL, Luger TA, et al. High-dose UVA1 therapy for atopic dermatitis: results of a multicenter trial. J Am Acad Dermatol 1998; 38: 589-93.

32. Kowalzick L, Kleinheinz A, Weichenthal M, et al. Low dose versus medium dose UV-A1 treatment in severe atopic eczema. Acta Derm Venereol (Stockh) 1995; 75: 43-5.

33. Abeck D, Schmidt T, Fesq H, et al. Long-term efficacy of medium-dose UVA1 phototherapy in atopic dermatitis. J Am Acad Dermatol 2000; 42: 254-7.

34. Zimmermann J, Utermann S. Photosoletherapie bei Patienten mit Psoriasis und Neurodermitis atopica. Hautarzt 1994; 45: 849-53.