JEADV ISSN 1468-3083

ORIGINAL ARTICLE
Blackwell Publishing Ltd

Narrowband UVB phototherapy for early-stage mycosis

fungoides: evaluation of clinical and histopathological changes
G Gökdemir,*† B Barutcuoglu,† D Sakız,‡ A KöÍlü†
†Dermatology Clinic and ‡Pathology Clinic, Sisli Etfal Research and Training Hospital, Istanbul, Turkey

Keywords Abstract
early-stage mycosis fungoides, mycosis
Background Early-stage (IA, IB, IIA) mycosis fungoides (MF) has long been
fungoides, narrowband UVB phototherapy,
therapy treated with various agents including topical potent steroids, nitrogen mustard,
carmustine, oral psoralen plus UVA (PUVA), broadband UVB, electron-beam
*Corresponding author, Fulya mah, Mevlüt radiotherapy, interferon-α and retinoids. However, each of these modalities is
Pehlivan sok, Ali Sami Yen Apt, A Blok No: 8/9, associated with various side-effects. Narrowband UVB (NB-UVB) therapy has
Sisli, Istanbul, Turkey, the same effect but is safer to use than the other methods.
tel. +90 212 231 22 09;
Objective Our purpose in this prospective study was to determine the effects
fax +90 212 234 11 21;
of NB-UVB in early-stage MF both clinically and histopathologically.
E-mail: goncagokdemir@hotmail.com
Materials and methods Twenty-three patients (20 men, three women, aged
Received: 13 March 2005, 27 – 78 years) with clinically and histologically confirmed MF were enrolled.
accepted 27 July 2005 Patients received NB-UVB therapy three times a week. Clinical and histological
responses, cumulative doses, total number of treatments, side-effects and
DOI: 10.1111/j.1468-3083.2006.01635.x duration of remission period were noted.
Results Six patients had stage IA MF, 15 patients stage IB and two patients
stage IIA. Eighteen patients had patch stage and five patients had plaque stage
histopathologically. All of the patients in the patch group had a complete
response (CR). In the plaque group, three patients (60%) had a CR and
two (40%) had partial (PR) or no clinical response (NR). The clinical response
between patch and plaque groups was statistically significant. Regarding the
histopathological findings, 17 (94.4%) had complete clearing and only one
(5.6%) patient had a partial improvement in the patch group. In the plaque
group, one (20%) patient had complete clearing and four (80%) patients had
partial or no improvement. The difference between the two groups was
statistically significant. In the patch group, the mean cumulative dose was
90.15 J/cm2 and the mean number of treatments was 35.33. In the plaque
group, the mean cumulative dose was 90.67 J/cm2 and the mean total number
of treatments was 39.40. The differences were not statistically significant, either
between the mean cumulative dose or the mean number of treatments. The
mean duration of follow-up was 10.87 months (range 1–25 months). Only one
of the patients had a relapse.
Conclusions NB-UVB therapy for patients with early-stage MF is an effective
and safe treatment with the effect lasting for months. We suggest that clinical
clearance correlates with histological improvement except for patients in the
plaque stage.

stages of mycosis fungoides: patch, plaque and tumour

Introduction stage. Histologically, the disease is typified by the
Mycosis fungoides (MF) is the most common form of proliferation of small cerebriform lymphocytes showing
cutaneous T-cell lymphoma (CTCL). There are three epidermotropism.1 Prognosis and survival in MF are

804 JEADV 2006, 20, 804– 809 © 2006 European Academy of Dermatology and Venereology
Gökdemir et al.
related to the stage of the disease as well as response to the
therapy. The patient’s life expectancy is not adversely
affected in stage IA. Patients with stage IB/IIA disease have
73 – 86% or 49 – 73% overall 5-year survival, respectively.
However, patients with stage IIB have a 40 – 65% 5-year
survival.2 Early-stage (IA, IB, IIA) MF has long been
treated with various agents including topical potent
corticosteroids, topical nitrogen mustard, topical carmu-
stine, oral psoralen plus UVA (PUVA), broadband UVB,
electron-beam radiotherapy, interferon-α, retinoids and
topical bexarotene.3–8 Although these modalities have
been reported to have similar efficacies for early stages of
MF, each one is associated with various side-effects.4
Narrowband UVB (NB-UVB) phototherapy has been used
successfully for the treatment of inflammatory and
pigmentary skin disorders including psoriasis, atopic
dermatitis, vitiligo and polymorphous light eruption.9 NB-
UVB phototherapy has only recently been described as a
first-line treatment for early-stage MF. There are many
advantages of NB-UVB phototherapy in comparison to
the above-mentioned treatments in MF.4 We report our
experience with 23 patients with MF treated with NB-
UVB. We also aimed to determine the effect of NB-UVB in
early-stage MF both clinically and histologically.
Patients and methods
Patients
A total of 23 patients with early-stage MF were recruited
from our outpatient clinic between 2002 and 2004. The
diagnosis of MF was made according to clinical and
histological findings. Stage of disease was classified based
on the TNM classification of the National Cancer Insti-
tute Workshop on MF.10 None of the patients had
organomegaly. There were no abnormalities in routine
biochemical investigations, red or white blood cell counts.
Computed tomography scans of thorax and abdomen
were normal in all patients. Only two patients had
inguinal lymphadenopathy. Histopathological find-
ings of these lymphadenopathies were indicated as
dermatopathic.11 These patients were evaluated as
stage IIA MF. Of the other 21 patients, five had less than
10% total skin surface involved (stage IA), and another
11 patients had greater than 10% involvement (stage IB).
Biopsy specimens were taken before and after therapy
and during the follow-up. Pretreatment skin biopsies
were considered diagnostic for MF if groups of atypical
lymphocytes formed small epidermal abscesses
(Pautrier’s microabscesses) or if a dense infiltrate of
numerous atypical lymphocytes with irregular nuclei and
mitosis was seen in the epidermis (epidermotropism) and
dermis.
JEADV 2006, 20, 804– 809 © 2006 European Academy of Dermatology and Venereology
Narrowband UVB phototherapy for early-stage mycosis fungoides
Light source and treatment regimen
Treatments were carried out in an irradiation cabinet,
Cosmedico GP-42K (Cosmedico Medizintechnik, Villingen-
Schweningen, Germany). Twenty-one Phillips TL-01/100 W
NB-UVB lamps were placed in the cabinet. Irradiation
was measured by a standard intrinsic UV meter. The
average irradiation was 2.80–3.20 mW/cm2 during the
course of the study.
Patients were given total body exposure NB-UVB
phototherapy three times a week. The initial UVB dose was
adjusted to the skin type of the patient, starting with an
irradiation dose of 0.300 J/cm2 for skin type II, 0.400 J/
cm2 for skin type III, and 0.500 J/cm2 for skin type IV. The
treatment with fixed dose increments was applied accord-
ing to skin phototype, 0.300 J/cm2 for skin type II,
0.400 J/cm2 for skin type III, and 0.400 J/cm2 for skin type
IV. If slight erythema was seen, the next exposure dose
was increased as mentioned above. If moderate erythema
was achieved, the same exposure dose was used again.
If marked erythema occurred, the treatment dose was
decreased by the amount of the initial dose. If severe
erythema and burning occurred, no more therapy was
performed until regression.12 Therapy was given three times
a week until more than 90% clearing of the patient’s skin
lesions had occurred. Once the clinical and histopatholo-
gical responses were achieved, treatment was continued
two times a week for 4 weeks and once a week for
4 weeks (another 8 weeks). If complete clinical response
was achieved but the patient had poor histological
improvement, maintenance treatment was started. Dur-
ing the therapy and the follow-up period, only emollients
were permitted for topical skin care.
Determination of clinical and histopathological
response and statistical analysis
Clinical response was defined as follows: complete
response (CR), more than 90% reduction in skin lesions;
partial response (PR), 50–90% reduction in the lesions;
and no response (NR), less than 50% reduction in the skin
lesions. Before and after therapy, a skin biopsy specimen
was taken from each patient for histological examination.
The specimens were processed routinely and stained
with haematoxylin–eosin (H&E) for light microscopic
examination. Histopathological response was determined
as follows: CR, the absence of epidermotropism and
Pautrier microabscesses and marked reduction in a dense
infiltrate of atypical lymphocytes with irregular nuclei
and mitosis in the epidermis and dermis; PR, marked
reduction in epidermotropism and sparsely scattered
atypical lymphocytes in the epidermis and dermis; NR, no
changes in histological findings.
805
Narrowband UVB phototherapy for early-stage mycosis fungoides Gökdemir et al.

The cumulative UVB dose and the number of treat-
ments for clearance were calculated. Disease status after
therapy was assessed. During the follow-up period, skin
biopsies were taken from each patient. Side-effects were
noted during the therapy. The statistical analysis was per-
formed with the Mann–Whitney U-test and Fisher’s exact
test. Significance was defined as P < 0.05.
Results
Twenty-three patients with MF (20 men and three
women; mean age 52.35 years, SD 13.70, range 27–78)
were enrolled in the study. Of these 23 patients, six had
stage IA MF, 15 had stage IB and two had stage IIA
according to the clinical findings. Regarding the his-
topathological findings, 18 (78.3%) patients had patch
stage and five (21.7%) patients had plaque stage MF.
The mean duration of the disease was 42.17 months
(SD 62.69, range 2 – 274 months). According to the
classification of Fitzpatrick, five patients had skin type II,
14 patients had skin type III and four patients had skin
type IV.10 At the end of the therapy, the mean cumulative
dose was 90.26 J/cm2 (SD 47.95, range 47 – 231 J/cm2)
and the mean number of treatments was 36.22 (SD
9.17, range 23–53). The mean duration of follow-up
was 10.87 months (SD 6.93, range 1– 25 months)
(Table 1).
Of the 23 patients, 21 (91.30%) had complete CR and
two (8.7%) (stage IB and plaque type) had PR and NR,
respectively (fig. 1). All of the patients in the patch group
had CR. In the plaque group, three patients (60%) had CR
and two (40%) patients had PR or NR. The clinical
responses between the patch and plaque stage groups
were statistically different (Table 2).
Skin biopsies before and after therapy from all of the
patients were taken. Eighteen (78.26%) of the patients
had complete histopathological improvement and five
(21.74%) had partial clearing or no histopathological re-
sponse. Seventeen (94.4%) of the 18 patients in the
patch group had CR in the biopsy findings and only one
(5.6%) patient had partial clearing. One (20%) of the five
patients in the plaque group had complete histological
clearing, two (40%) had partial histological clearing and
the other two (40%) had no any histological improve-
ment. The difference in histopathological responses
between the patch and plaque groups was statistically sig-
nificant (P = 0.001). There was no statistically significant
difference between the two groups regarding mean

Table 1 Clinical characteristics of patients

Duration of Patch/ Follow-up

Patient Skin disease plaque Total UVB No. of Clinical Histopathological Side- period
no./age/sex type (years) Stage stage dose (J/cm2) treatments response response effects (months)

1/58/M III 0.25 IB Patch 116.00 51 CR CR – 6

2/43/M III 3 IA Patch 58.00 28 CR CR – 6
3/47/M II 0.16 IB Patch 72.00 32 CR CR – 7
4/62/M III 0.33 IA Patch 53.00 24 CR CR – 4
5/76/M III 0.5 IB Patch 107.00 48 CR CR – 3
6/77/M III 3 IB Patch 50.00 28 CR CR – 4
7/35/F III 1 IB Patch 108.00 40 CR CR – 1
8/59/F IV 0.33 IA Patch 71.50 38 CR CR Pruritus 8
9/59/M III 25 IB Patch 67.20 37 CR CR PH 8
10/50/M III 1.5 IB Patch 60.40 33 CR CR – 9
11/78/M III 0.25 IB Patch 47.20 30 CR CR – 12
12/42/M II 2 IB Patch 73.20 28 CR CR Burning 11
13/27/M II 0.25 IIA Patch 56.40 28 CR CR Erythema 23
14/39/M III 3 IA Patch 136.40 45 CR CR – 24
15/54/F III 1 IIA Patch 231.38 53 CR PR PH 25
16/43/M IV 8 IB Patch 200.34 43 CR CR – 14
17/59/M III 10 IA Patch 49.20 23 CR CR – 16
18/49/M III 7 IB Patch 65.40 27 CR CR – 12
19/61/M IV 1 IB Plaque 66.00 32 NR NR – 3
20/53/M II 3 IB Plaque 63.30 31 PR NR Pruritus 16
21/36/M III 10 IA Plaque 78.80 40 CR CR – 19
22/59/M II 0.41 IB Plaque 62.40 43 CR PR – 16
23/38/M IV 2 IB Plaque 142.50 51 CR PR – 12

CR, complete response; PR, partial response; NR, no response; PH, postlesional hyperpigmentation.

806 JEADV 2006, 20, 804– 809 © 2006 European Academy of Dermatology and Venereology
Gökdemir et al. Narrowband UVB phototherapy for early-stage mycosis fungoides

fig. 1 (a) Patch-stage mycosis fungoides with

erythematous patches on the trunk and abdo-
men. (b) Complete clinical remission following
narrowband UVB phototherapy.

Table 2 Summary of response to narrowband UVB therapy

Discussion
Patch group Plaque group P
The extent and type of skin involvement and the pres-
2 ence of extracutaneous disease are the most important
Mean total UVB dose (J/cm ) (SD) 90.15 (52.68) 90.67 (29.16) 0.174*
Mean total UVB sessions (SD) 35.33 (9.43) 39.40 (8.26) 0.325* indicators predictive of prognosis in patients with MF. For
Clinical response patients with early-stage MF, life expectancy may not be
CR (%) 18 (100) 3 (60) adversely affected and very early-stage MF may have a
PR + NR (%) – 2 (40) 0.040† favourable long-term outcome. However, MF at a later
Histopathological response stage may have a worse prognosis.2,13 NB-UVB was
CR (%) 17 (94.4) 1 (20)
introduced recently for the treatment of early-stage MF
PR + NR (%) 1 (5.6) 4 (80) 0.001†
and found to be effective and safe.14 The mechanisms of
CR, complete response; PR, partial response; NR, no response. action of NB-UVB therapy in MF are still unknown. UVB
*Mann–Whitney U-test. has been shown to decrease the antigen-presenting
†Fisher’s exact test. capacity of Langerhans cells, and increase interleukin-2
and interkeukin-6 production by human keratinocytes,
and also increase tumour necrosis factor-α. As a result,
many authors have suggested that UVB light could
cumulative dose and total number of treatments suppress the function of the neoplastic population of
(P = 0.174 and P = 0.325) (Table 2). clonal T cells in the skin and serve as an up-regulator of
Acute adverse effects from NB-UVB included burning the immune system.15,16
(one patient), severe erythema (one patient), pruritus In the present study, we have demonstrated the clinical
(two patients) and postinflammatory pigmentation (two and histopathological efficacy of NB-UVB in 23 patients
patients). with MF. We found that 91.30% of patients achieved
In the follow-up period, all of the patients except complete response clinically. Eighteen patients (78.26%)
one were in remission. Only one patient had relapsed had complete histopathological improvement, three had
7 months after therapy. Three patients (patients 15, 22 partial improvement and two had no histological clearing.
and 23) who had partial improvement had a marked The differences between the patch and plaque groups
reduction in the inflammatory infiltrate and epidermotro- were statistically significant both clinically and histolo-
pism in their biopsy specimens 6 months after therapy gically. Hofer et al.17 previously studied six patients with MF
(fig. 2). Two patients who had no response both clinically and found complete clinical response after an average of
and histopathologically were still receiving therapy at the 20 treatments in five (83.3%) patients. However, relapses
end of the study. occurred in all patients within a mean time of 6 months.17

JEADV 2006, 20, 804– 809 © 2006 European Academy of Dermatology and Venereology 807
Narrowband UVB phototherapy for early-stage mycosis fungoides
fig. 2 (a) Plaque-stage MF pretreatment showing acanthosis, papillary
dermal mononuclear cell infiltrate containing abnormal lymphocytes that
demonstrate epidermotropism [haematoxylin–eosin (H&E), ×200]. (b)
Marked reduction in atypical lymphocytic infiltrate 6 months after therapy
(H&E, ×100).
Clark et al.18 and Gathers et al.4 revealed the efficacy of
NB-UVB therapy in patients with MF and found complete
clinical response in 75% and 54.2%, respectively. Com-
pared with those studies, our study showed slightly higher
relapse rates. There are two possible explanations for this
discrepancy. One reason was the application of mainten-
ance therapy for all patients. The second is that the mean
cumulative dose and the mean number of treatments
were higher than in the previous studies.
Skin biopsies before and after therapy were taken
from all of the patients in our study. There are a few data
in the literature about histopathological changes in NB-
UVB therapy for patients with MF. Abe et al.19 studied the
efficacy of NB-UVB in five patients with MF, and revealed
the loss of HLA-DR antigen expression by epidermal
Langerhans cells after NB-UVB on immunohistochemical
study. In our study, clinical improvement was an impor-
tant indication for changing the therapy protocol. The
patients in the patch stage had significantly more
808
Gökdemir et al.
complete clinical and histological clearing than those in
the plaque group, although this finding might have been
due to the limited ability of UVB to penetrate thicker
lesions.18
Only one of the patients in the study had a relapse
during the follow-up period. Patient 16 had relapsed
7 months after therapy. He had had previous treatment
with PUVA and a previous recurrence 6 months later.
Three patients (patients 15, 22 and 23) who had partial
improvement had marked reduction in infiltration and
epidermotropism in their specimens taken during the
follow-up period, indicating that the effect of NB-UVB had
continued for months. The relapse rate in our study was
low because of the short follow-up period.
Regarding the other therapies in early-stage MF, NB-
UVB seemed to have equal or greater effectiveness and
safety. Ramsay et al.20 and Resnik and Vonderheid21
reported results from patients with MF who were treated
with broadband UVB. They achieved complete response
in 83% and 85% of patients, respectively. However, it was
indicated that broadband UVB produced more irritation
and severe erythema compared with NB-UVB.
The benefit of PUVA in MF has been reported in many
studies and response rates of 79–88% in stage IA and 52–
59% in stage IB have been recorded.2 However, there
are many disadvantages in the use of PUVA: acute adverse
effects can be seen in systemic psoralen use; protective
glasses are needed after therapy; and PUVA cannot be
used during pregnancy.18 There is only one reported study
comparing the effects of NB-UVB and PUVA in MF.
Diederen et al.14 suggested that NB-UVB is as effective as
PUVA in early-stage MF and could be advisable as first-line
therapy over PUVA.
The other therapy methods in MF are topical steroids,
topical nitrogen mustard and total skin electron-beam
radiation. The adverse effects of these treatments are
allergic contact dermatitis, narrow suppression, severe
erythema and desquamation.2
Acute adverse effects of NB-UVB might be erythema
and pruritus. The patients in our study had minimal side-
effects, similar to those in other studies. Additionally,
it has been reported that the photocarcinogenic risk of
NB-UVB was less than that associated with PUVA and
broadband UVB.4,18
In our prospective study, we found that NB-UVB
therapy for patients with early-stage MF was an effective
and safe treatment in patients with darker skin types. We
suggest that clinical clearance correlates with histopatho-
logical improvement except for patients in the plaque
group. The effect of NB-UVB may continue for months.
We suggest that NB-UVB may be a first-line therapy
option in early-stage MF regarding its effects and
advantages.
JEADV 2006, 20, 804– 809 © 2006 European Academy of Dermatology and Venereology
Gökdemir et al.
References
1 Kazakov DV, Burg G, Kempf W. Clinicopathological
spectrum of mycosis fungoides. J Eur Acad Dermatol Venereol
2004; 18: 397–415.
2 Whittaker SJ, Marsden JR, Spittle M, Russell Jones R. Joint
British Association of Dermatologists and UK Cutaneous
Lymphoma Group guidelines for the management of
primary cutaneous T-cell lymphomas. Br J Dermatol 2003;
149: 1095–1107.
3 Roupe G, Sandström MH, Kjellström C. PUVA in early
mycosis fungoides may give long-term remission and delay
extracutaneous spread. Acta Derm Venereol (Stockh) 1996; 76:
475–478.
4 Gathers RC, Scherschun L, Malick F et al. Narrowband UVB
phototherapy for early-stage mycosis fungoides. J Am Acad
Dermatol 2002; 47: 191–197.
5 Duvic M, Apisarnthanarax N, Cohen DS et al. Analysis of
long-term outcomes of combined modality for cutaneous
T-cell lymphoma. J Am Acad Dermatol 2003; 49: 35–49.
6 Jones GW, Kacinski BM, Wilson LD et al. Total skin electron
radiation in the management of mycosis fungoides:
consensus of the European Organization for Research and
Treatment of Cancer (EORTC) Cutaneous Lymphoma
Project Group. J Am Acad Dermatol 2002; 47: 364–370.
7 Zackheim HS. Treatment of mycosis fungoides/Sezary
syndrome: the University California, San Francisco (UCSF)
Approach. Int J Dermatol 2003; 42: 53–56.
8 Singh F, Lebwohl G. Cutaneous T-cell lymphoma treatment
using bexarotene and PUVA: a case series. J Am Acad
Dermatol 2004; 51: 570–573.
9 Yashar SS, Gielczyk R, Scherschun L, Lim HW. Narrow-band
ultraviolet B treatment for vitiligo, pruritus, and
inflammatory dermatoses. Photodermatol Photoimmunol
Photomed 2003; 19: 164–168.
10 Latkowski JA, Heald P. Cutaneous T cell lymphoma. In:
JEADV 2006, 20, 804– 809 © 2006 European Academy of Dermatology and Venereology
Narrowband UVB phototherapy for early-stage mycosis fungoides
Freedberg IM, Eisen AZ, Wolff K et al., eds. Fitzpatrick’s
Dermatology in General Medicine, Vol. II, 6th edn.
McGraw-Hill, New York, 2003: 1537–1558.
11 McKee P. Pathology of the Skin, 2nd edn. Mosby-Wolfe,
London, 1996: 12.1–12.64.
12 Zanolli MD, Feldman SR, Clark AR, Fleischer AB.
Phototherapy Treatment Protocols. Bookcraft, New York, 2000:
105–117.
13 Koh HK. Is cutaneous T-cell lymphoma curable? Arch
Dermatol 1995; 131: 1081–1082.
14 Diederen PVM, van Weelden H, Sanders CJG et al.
Narrowband UVB and psoralen-UVA in the treatment of
early-stage mycosis fungoides: a retrospective study. J Am
Acad Dermatol 2003; 48: 215–219.
15 Duthie MS, Kimber I, Norval M. The effects of ultraviolet
radiation on the immune system. Br J Dermatol 1999; 140:
995–1009.
16 Mori M, Campolmi P, Mavilia L et al. Monochromatic
excimer light (308nm) in patch-stage IA mycosis fungoides.
J Am Acad Dermatol 2004; 50: 94945.
17 Hofer A, Cerroni L, Kerl H, Wolf P. Narrowband (311-nm)
UV-B therapy for small plaque parapsoriasis and early-stage
mycosis fungoides. Arch Dermatol 1999; 135: 1377–1380.
18 Clark C, Dawe RS, Evans AT et al. Narrowband TL-01
phototherapy for patch-stage mycosis fungoides. Arch
Dermatol 2000; 136: 748–752.
19 Abe M, Ohniski Kan C, Ishikawa O. Ultraviolet-B
phototherapy is successful in Japanese patients with
early-stage mycosis fungoides. J Dermatol 2003; 30:
789–796.
20 Ramsay DL, Lish KM, Yalowitz CB, Soter NA. Ultraviolet-B
phototherapy for early stage cutaneous T-cell lymphoma.
Arch Dermatol 1992; 128: 931–933.
21 Resnik KS, Vonderheid EC. Home UV phototherapy of early
mycosis fungoides: long-term follow-up observations in
thirty-one patients. J Am Acad Dermatol 1993; 29: 73–77.
809