ARTICLE
Auteur(s) : Claas Ulrich, Antje Johannsen,
Joachim Röwert-Huber, Martina Ulrich, Wolfram Sterry, Eggert
Stockfleth
Skin Cancer Centre Charité, Department of Dermatology,
Charité Universitätsmedizin, Charité-Platz 1, 10117 Berlin,
Germany
accepté le 18 Mars 2010
Solid organ transplantation has been performed in over one
million patients worldwide and increasing numbers are benefiting
from transplantation. As increasingly effective immunosuppressive
therapies are developed, the overall survival-times of patients
receiving grafts have almost doubled within the last 20 years
[1].
Paralleling the increasing survival time under chronic
immunosuppression, cancers are among the most frequent
complications registered in organ transplant recipients (OTRs).
Non-melanoma skin cancers (NMSC), especially invasive squamous cell
carcinomas (SCC), strongly outnumber all other malignancies in
OTRs. A study of renal allograft recipients found the risk of
NMSC was 20 times greater compared to the general population,
showing a cumulative incidence of NMSC of 7.43% after 3 years
post transplantation [2].
In OTRs, this increased incidence of NMSCs and especially SCC,
their multifocal development and increased aggressiveness with a
metastasising rate of 6-9% is mainly explained by impaired
cutaneous immunosurveillance that allows dysplastic keratinocytes
to proliferate and spread at an accelerated rate [3-5]. Because of
the great numbers of invasive SCC accumulating from earlier forms
of dysplasia in sun exposed areas, adequate tools for the
comprehensive clearance of early, subclinical and advanced (actinic
keratoses – AK) forms of dysplasia within these “dysplastic fields”
are needed to prevent further morbidity and mortality [6].
Diclofenac is a non-steroidal anti-inflammatory drug (NSAID),
reducing the production of prostaglandins by inhibiting the
inducible cyclo-oxygenase 2 (COX-2) enzyme. Sun-damage, AK and
invasive SCC have been linked with increasing levels of
prostaglandins and COX-2 activity, paralleling increased levels of
dysplasia [7]. Diclofenac has been shown to inhibit murine
angiogenesis, regulate apoptosis, induce cell cycle arrest and has
a significant anti-tumour effect in murine colon-26 growth [8].
Furthermore, diclofenac may act via an over expression of
metalloproteinases, which would have keratolytic and collagenolytic
effects [6]. Topical diclofenac 3% gel (SolarazeTM; in
2.5% hyaluronic acid (HA)) is licensed for the treatment of actinic
keratoses by the FDA and in many countries in Europe.
Randomised, double blind, placebo-controlled studies have shown,
depending on the treatment duration, 30-50% total clearance rates
or significant improvement around 80% using topical diclofenac 3%
gel (Solaraze™) for the treatment of AK in immuno-competent
patients, thus indicating a possible benefit for the use of this
drug in OTRs as well [9].
Recently, Nelson et al. presented 12 month follow-up
results of diclofenac 3% HA treated immuno-competent patients. 79%
of the treated target lesions stayed clear over this period
[10].
The aim of this small-size pilot study was to investigate the
effect and graft related safety of diclofenac 3% gel
on clearance rates of multiple AK lesions in organ transplant
patients. The second goal was to evaluate any prophylactic effect
against invasive SCC within a 24 month follow up period of the
verum group.
Patients and methods
Study design
The primary objective of this randomized, double-blind,
placebo-controlled trial was to evaluate the efficacy, graft
related safety and tolerability of diclofenac 3% gel
(SolarazeTM; in 2.5% HA, Almirall S.A., Spain) in
comparison with vehicle gel for the treatment of AK in organ
transplant patients. Patients with one of three organ transplant
types (kidney, liver, heart) were randomized 3:1 (active:vehicle)
in this vehicle-controlled, double-blind, parallel group design.
Patients applied either topical 3.0% diclofenac in 2.5% hyaluronan
gel or vehicle as study medication 2 times per day for
16 weeks to a defined treatment area of 50 cm [2]. The
complete clearance rate was defined as the proportion of patients
at the 4-week post treatment visit who had no evidence of AK on the
histology results of a target biopsy lesion site and no clinically
visible lesions in the remainder of the treatment area. The partial
clearance rate was defined as the proportion of patients at the
4-week post treatment visit who obtained at least 75% reduction in
the number of lesions counted at baseline in the treatment area.
The clearance rate of individual AK lesions was defined as the
percentage reduction of lesions from baseline to the 4-weeks post
treatment visit.
Patients
32 out of 40 organ transplant patients screened at our
specialized transplant dermatology outpatient clinic were found
eligible and were randomized to either active treatment (24) or
vehicle (8). Major reasons for the 8 patients’ ineligibility
were not meeting the inclusion/exclusion criteria due to abnormal
lab values (5), invasive SCC (2) or seborrhoic keratosis (1) in the
pre-therapy punch biopsy from the treatment area. This study was
conducted in compliance with all applicable national and EU
regulations. Local Ethical Committee (Charité Universitätsmedizin,
Berlin, Germany) review and approval was obtained and all patients
provided written informed consent.
Patient selection
Patients with kidney (± pancreas), liver, or heart transplantation
within 3 years and stable status of the transplanted graft in
the 12 months prior to entering the study, who had ≥ 3 AK
lesions in a contiguous 50 cm2 area on the face,
forehead, hands or balding scalp were eligible for inclusion in the
study. Criteria for determining the stability of grafts were
specific to each transplant type. Immunosuppressive therapy must
have been stable within the previous 6 months before enrolment
and during therapy with the study drug. Patients were excluded if
they had severe renal or hepatic impairment or had any evidence of
graft rejection. Ongoing treatments for AK or evidence of invasive
skin cancers also were exclusion criteria. Further exclusion
criteria were evidence of unstable and severe cardiovascular,
immunological, hematologic, hepatic, neurological, renal,
endocrine, collagen-vascular, gastrointestinal or non-study related
skin abnormalities or disease. In addition, patients could not have
malignant tumours of the skin within the treatment area
6 months before enrolling in the study. Any evidence of
systemic cancer or any systemic cancer chemotherapy or radiation
therapy within 6 months of study treatment initiation were
exclusion criteria. Patients could not have received other systemic
treatments, including retinoids, interferons or investigational
drugs, within 4 weeks of study initiation. Vitamin
A usage > 15,000 units per day was also excluded. Females
of childbearing potential could not be pregnant or nursing, and
must have been willing to use medically accepted methods of
contraception. Patients were also excluded if they had a
history of hypersensitivity or allergy to any of the ingredients of
active drug or vehicle or other non-steroidal anti-inflammatory
drugs.
Study procedures
The total duration of patient participation in this study was a
maximum of 20 weeks. Patient eligibility was determined at the
pre-study visit, at which time a 3 mm punch biopsy of a lesion
was performed to confirm the presence of AK in the treatment area.
Baseline safety and efficacy data were collected, and the patients
randomized to study drug. Patients returned to the clinic for
evaluations throughout the treatment period (weeks 4, 8, 12, 16).
Photographs of the treatment area were taken at each study visit.
Final evaluations, including another 3-4 mm punch biopsy, were
performed at the 4-week post treatment visit (week 20).
Because safety throughout the whole study time was a primary
focus, treatment continued for a full 16 weeks regardless of
clinical evidence of lesion clearance. After completing this first
part of the study, all patients with AK not responding to the study
drug, including those from the vehicle group, were treated with
various active, open label therapies such as topical diclofenac 3%
in HA, PDT, topical 5-FU or imiquimod 5% cream. Patients
successfully completing the verum arm of the study and showing
complete response rates were recruited for a 24 month follow
up phase of their study areas.
Patient disposition
Among the 24 receiving treatment, 2 patients discontinued
treatment, whereas 2 of 8 discontinued from the vehicle
arm (table 1).
Table 1 Patient Disposition
|
Variable
|
Liver
|
Kidney
|
Heart
|
|
(Diclofenac n = 5)
|
(Vehicle n = 1)
|
(Diclofenac n = 13)
|
(Vehicle n = 5)
|
(Diclofenac n = 6)
|
(Vehicle n = 2)
|
|
Sex
|
|
|
|
|
|
|
|
Female
|
1 (20%)
|
0 (0%)
|
1 (7.7%)
|
1 (20%)
|
0 (0%)
|
0 (0%)
|
|
Male
|
4 (80%)
|
1 (100%)
|
12 (92.3%)
|
4 (80%)
|
6 (100%)
|
2 (100%)
|
|
Age (years)
|
|
|
|
|
|
|
|
Mean ±SD
|
64.8 ± 6.06
|
72
|
62 ± 8.44
|
54.2 ± 13.55
|
65.7 ± 7.84
|
63.0 ± 9.90
|
|
Range
|
56 - 70
|
72
|
49 - 73
|
39 - 71
|
54 - 77
|
56 - 70
|
|
Race
|
|
|
|
|
|
|
|
White
|
5 (100%)
|
1 (100%)
|
13 (100%)
|
5 (100%)
|
6 (100%)
|
2 (100%)
|
|
Treatment Area Location
|
|
|
|
|
|
|
|
Face
|
5 (100%)
|
0 (0%)
|
11 (84.6%)
|
0 (0%)
|
6 (100%)
|
2 (100%)
|
|
Scalp
|
0 (0%)
|
1 (100%)
|
1 (7.7%)
|
4 (80%)
|
0 (0%)
|
0 (0%)
|
|
Hands
|
0 (0%)
|
0 (0%)
|
1 (7.7%)
|
1 (20%)
|
0 (0%)
|
0 (0%)
|
|
Completed treatment
|
5 (100%)
|
1 (100%)
|
13 (100%)
|
3 (60%)
|
4 (67%)
|
2 (100%)
|
|
Completed post-treatment
|
5 (100%)
|
1 (100%)
|
13 (100%)
|
3 (60%)
|
4 (67%)
|
2 (100%)
|
Efficacy assessments
At each visit (weeks 4, 8, 12, 16) and the post treatment visit
(week 20), a clinical count supported by a transparent grid was
made of the number of visible AK lesions in the treatment area.
Histological verification was obtained through a 3-4 mm punch
biopsy of a target lesion mapped at the initiation visit.
Safety monitoring
Adverse events: Safety variables included transplant rejection
status, laboratory results, adverse events, local skin reactions,
vital signs measurements, and the dosage of immunosuppressive
medications.
Clinical laboratory analysis: All transplant recipients were
monitored for their serum levels of immunosuppressive medication in
the therapeutic range. Renal transplant recipients were monitored
for abnormal levels in serum creatinine, C-reactive protein, and
proteinuria. Liver transplant recipients were monitored for levels
of gamma glutamyl-transpeptidase, glutamic-pyruvic transaminase,
glutamic-oxalacetic transaminase, and bilirubin. Heart transplant
recipients were monitored specifically for GOT and GPT, white cell
blood count, serum creatinine, hemoglobin, and signs of heart
failure.
At the post-treatment visit (week 20 in the study),
investigators assessed the skin quality of the treatment area with
regard to skin surface, hyperpigmentation, hypopigmentation, the
degree of scarring, and any atrophy.
Statistical analysis
Sample size for the study was based on the assumption that 0-1
spontaneous rejection events might be expected to occur in
40 patients observed over the length of the study. All
patients were included in the safety population. Efficacy analysis
was performed on the per-protocol population. Descriptive
statistics were used to summarize the safety variables for
transplant rejection status, local skin reactions, adverse events,
laboratory values and skin quality and the efficacy variables for
complete and partial clearance rates by transplant type and
treatment. No formal statistical comparisons between transplant
types or treatment groups were made.
Results
87% (n = 28/32) of the patients completed the 16 weeks
treatment phase and presented for final evaluation 4 weeks
after the end of treatment (table 1, figures 1,2, 3).
The immunosuppression was kept stable regarding type and dosage
throughout the whole time of the study.
Safety
No patient in either the diclofenac 3% gel group or the vehicle
group experienced a systemic reaction or impaired transplant
function relatable to the study drug. No meaningful trends were
observed in laboratory results in creatinine, transaminases and
C-reactive protein in the treatment group. Fluctuation in
creatinine, being of specific importance in all patients with
exposure to COX-inhibitors and potentially impaired renal function,
was within 9% of pre-study baseline. However, in one patient an
initially unexplainable increase of the creatinine level in week
4 was the reason to exclude him from the study. A workup
of the case showed that the creatinine-increase was related to a
previously unperceived calcineurine-toxicity related nephropathy.
The most commonly reported adverse event was an application site
reaction which seemed not to be related to an increased level of
clinical clearance. Local side effects in the treatment area,
including erythema and erosion, were mild in most patients (figure 2). In
two patients of the treatment-group, itching and discomfort in the
treatment area were the reasons for stopping the therapy after
4 weeks of application. Another patient stopped the study
shortly after week 12 due to personal reasons. A kidney
transplant patient in the diclofenac-group developed a basal cell
carcinoma within the study area.
All patients in the treatment group had excellent and
cosmetically appealing results 4 weeks after the end of the
study.
Efficacy
Treatment outcome
In the diclofenac 3% gel treatment group, a complete clearance of
AK lesions was achieved in 4/13 (30.7%) of patients in the kidney
transplant group, 2/5 (40%) of patients in the liver
transplant group and 3/4 (75%) of patients in the heart transplant
group. The overall complete clearance result in the diclofenac-arm
of our study was 41% (9/22) compared to 0/6 (0%) in the vehicle
group (figure 3).
In the diclofenac treatment group, partial clearance – as
defined as a clearance rate of ≥ 75% of baseline lesions – was
achieved in 7/13 (53.8%) of patients in the kidney transplant
group, 2/5 (40%) of patients in the liver transplant group and 4/4
(100%) of patients in the heart transplant group. Combining the
transplant group results, diclofenac 3% gel partial clearance
rate was 59% (13/22) compared to 1/6 (16%) in the vehicle
group (figure 4).
When comparing lesion counts made at baseline to lesion counts
made at the 4 week post treatment visit, an average 53%
reduction in the individual lesion count was observed in all of the
patients treated with diclofenac 3% gel, whilst among patients
randomized to vehicle, the number of lesions increased during the
course of the study by 17% in average (figure 5).
Follow-up outcome
45% of the previously cleared patients stayed completely free of
new non-melanoma skin cancer, including AK, in the treated area for
the whole 24 month follow-up period. After an average of 9.3
months, 55% of the previously completely cleared patients developed
new AK (histological type I/II) in the study area. However, none of
these patients developed AK histological type III or invasive SCC
within 24 months of follow-up in the previously treated study
area. In a representative area outside the study area, 63% of
these patients developed AK (II + III), 5% Bowen's disease, 15%
invasive SCC and 21% BCC.
Discussion
To our knowledge this is the first double blinded study examining
the safety and efficacy of diclofenac 3% gel for the local
treatment of AK lesions in organ transplant recipients. Unlike
clinical observations in immuno-competent patients,
immuno-suppressed patients appear to have a highly accelerated rate
of AK development and progression. This is also shown in this
study, where the number of AK lesions increased on average by 17%
in the placebo arm. Aggressive treatment of AK is essential to
prevent its progression into invasive SCC [11]. AK and subsequently
resulting invasive SCC usually develop in large numbers in sun
exposed areas (“field carcinogenesis”). Unspecific, destructive,
lesion-directed therapies are applied only to clinically visible AK
and leave out fields of subclinical dysplasia. They therefore
usually fail to clear the complete area of substantial photodamage,
including subclinical AK, resulting in comparable high recurrence
rates in the treatment areas. This simple observation might help to
explain the significantly increased risk of AK reoccurrence after
cryotherapy, laser or curettage, leading towards steadily
increasing preferences for field-adaptable therapies. The ideal
therapy for NMSC, as suggested by Stasko and colleagues, would
selectively target both malignant and pre-malignant lesions, should
be easy and unobtrusive to use, cause no treatment-related
morbidity, have no deleterious effects on the user's health –
including the viability of the grafted organ – would not interfere
with other medications or therapies, and would have a perfect
clearance rate of cancerous and precancerous lesions with no
subsequent recurrences [12].
There is increasing evidence that cyclooxygenase-2 (COX-2) plays
an important role during the development and progression of
non-melanoma skin cancers [13]. COX-2 is normally undetectable in
most epithelial tissues. However, growth factors and
pro-inflammatory cytokines both may result in its overexpression as
it has been documented in actinic keratoses and invasive squamous
cell carcinoma [14, 15]. In a retrospective paired
immunohistochemical analysis of normal skin, actinic keratosis
(AK), Bowen's disease (BD) and invasive SCC among
35 individuals, Nijsten et al. found the COX-2 expression
and angiogenesis increased from 0% (normal epidermis), 31% (AK),
22% (BD) and up to 40% (SCC) in this multistage continuum leading
towards invasive SCC [16]. The authors were furthermore able to
show COX-2 immuno-positivity correlates with hypoxia and higher
proliferating endothelial cell fractions, indicating an involvement
of COX-2 in skin tumour angiogenesis. A correlation between
COX-2 levels with vascular endothelial growth factor expression and
tumour vascularization has previously been shown [17, 18]. Most
interestingly, studies on human cell lines revealed a causal, Bcl-2
dependent linkage, between COX-2 inhibition and anti-apoptosis
which, especially for pre-invasive SCC, could be of specific
importance [18]. Interestingly, no COX-2 expression was found in
basal cell carcinoma [19]. Various observational studies have shown
an association between regular consumption of non-steroidal
anti-inflammatory drugs and lower incidences of gastrointestinal
cancers but also SCC of the lung and esophagus [20, 21].
A series of placebo controlled studies on topical application
of diclofenac 3% gel for the treatment of AK has shown total
clearance rates up to 50%. In a recently published study, topical
diclofenac 3% gel showed similar efficacy comparable with the
topical chemotherapeutic agent – 5-fluorouracil 5%, which, despite
significant side effects, is still one of the most popular topical
forms of field management worldwide [22, 23]. The observation that
the number of AK lesions increased under placebo treatment in our
study underpins the mode of action of diclofenac, also inducing a
certain preventive effect. This should be evaluated in further
studies. Reported adverse events of diclofenac 3% gel were mild to
moderate eczemateous reactions with pruritus in our study, as well
as the studies published so far. To our knowledge, systemic adverse
events of the topical use of diclofenac 3% gel have not been
observed. In our study no systemic side effects, especially renal
or cardiovascular complications, were observed. Laboratory
parameters were carefully checked throughout the whole study period
and were found to be generally stable and unaffected by the study
drug. However, further studies are needed before recommending
NSAIDs for use on skin areas exceeding the study areas tested in
this trial.
The accelerated skin carcinogenesis seen in immuno-compromised
patients makes them the ideal population to study short term
efficacy rates in the clearance of AK and long term prevention of
invasive squamous cell carcinoma (SCC). Our study adds to the
findings of other trials that show the benefits of so called “field
or topical therapies” compared with non-topical therapies (e.g.
cryotherapy) that are unspecific, more destructive and provide
limited, short term efficacy in immunosuppressed patients with
actinic “field dysplasia”. In addition, this study suggests that
the transplant population would benefit from a self-applied
effective, convenient and safe method to treat clinical and
subclinical AKs. Our 24-month follow-up of these patients suggests
that field therapies of actinic dysplasia may not only provide
initial relief from clinical signs of AK but also may delay or even
prevent severe forms of AK and resulting invasive SCC in this
high-risk patients.
The findings of this study on OTRs are also likely to
have significance for other diseases with either
therapeutically induced, or directly disease-related lack of
immunosurveillance, like HIV; therapeutically-induced
immunosuppression in rheumatic and other auto-immune disorders;
chemotherapy for systemic-malignancies where high levels of NMSC
may represent a significant complication [24, 16]. The management
of skin diseases in OTRs remains an exciting and interesting area
providing rewarding opportunities for dermatologists within the
field of transplant medicine. Furthermore, organ transplant
recipients represent an excellent target population to study the
efficacy and sustainability of established as well as novel forms
of therapy for field management of actinically damaged skin
areas.
Acknowledgements
This paper and the study were supported by a scientific grant from
Shire Pharmaceuticals. Conflict of interest: none.
References
1 Hariharan S, Johnson CP, Bresnahan BA, et al.
Improved graft survival after renal transplantation in the United
States, 1988-1996. N Engl J Med 2000; 342: 605-12.
2 Kasiske BL, Snyder JJ, Gibertson DT,
et al. Cancers after kidney transplantation in the United
States. Am J Transplant 2004; 4: 905-13.
3 Veness MJ, Quinn DI, Ong CS, Keogh AM,
Macdonald PS, Cooper SG, et al. Aggressive cutaneous
malignancies following cardiothoracic transplantation: the
Australian experience. Cancer 1999; 85: 1758-64.
4 Euvrard S, Kanitakis J, Pouteil-Noble C,
Disant F, Dureau G, Finaz de Villaine J, et al.
Aggressive squamous cell carcinomas in organ transplant recipients.
Transplant Proc 1995; 27: 1767.
5 Dantal J, Hourmant M, Cantarovich D,
et al. Effect of long-term immunosuppression in kidney-graft
recipients on cancer incidence: randomised comparison of two
cyclosporin regimens. Lancet 1998; 351: 623-8.
6 Kovach BT, Stasko T. Use of topical immunomodulators
in organ transplant recipients. Dermatol Ther 2005; 18: 19-27.
7 Nijsten T, Colpaert CG, Vermeulen PB,
et al. Cyclooxygenase-2 expression and angiogenesis in
squamous cell carcinoma of the skin and its precursors: a paired
immunohistochemical study of 35 cases. Br J Dermatol 2004;
151: 837-45.
8 Seed MP, Brown JR, Freemantle CN, et al.
The inhibition of colon-26 adenocarcinoma development and
angiogenesis by topical diclofenac in 2.5% hyaluronan. Cancer Res
1997; 57: 1625-9.
9 Jarvis B, Figgitt DP. Topical 3% diclofenac in 2.5%
hyaluronic acid gel: a review of its use in patients with actinic
keratoses. Am J Clin Dermatol 2003; 4: 203-13.
10 Nelson C, Rigel D. Long term follow up Diclofenac
Sodium 3% in 2.5% hyaluronic Acid gel for actinic keratosis. J Clin
Anaestetic Dermatol 2009; 2: 20-5.
11 Berg D, Otley CC. Skin cancer in organ transplant
recipients: Epidemiology, pathogenesis, and management. J Am Acad
Dermatol 2002; 47: 1-17.
12 Kovach BT, Stasko T. Use of topical
immunomodulators in organ transplant recipients. Dermatologic
Therapy 2005; 18: 19-27.
13 Fisher SM. Is cyclooxygenase-2 important in skin
carcinogenesis? J Environ Pathol Toxicol Oncol 2002; 21:
183-91.
14 Buckmann SY, Gresham A, Hale P, et al.
COX-2 expression is induced by UVB exposure in human skin:
implications for the development of skin cancer. Carcinogenesis
1998; 19: 723-9.
15 Kagoura M, Toyoda M, Masuri C,
Morohashi M. Immunohisochemical expression of cyclooxygenase-2
in skin cancers. J Cutan Pathol 2001; 28: 298-302.
16 Nijsten T, Colpaert CG, Vermeulen PB,
Harris AL, Van Marck E, Lambert J. Cyclooxygenase-2
expression and angiogenesis in squamous cell carcinoma of the skin
and its precursors: a paired immunohistochemical study of
35 cases. Br J Dermatol 2004; 151: 837-45.
17 William CS, Tsujii M, Reese J, Dey SK,
DuBois RN. Host cyclooxygenase-2 modulates carcinoma growth. J
Clin Invest 2000; 105: 1589-94.
18 Gallo O, Franchi A, Magnelli L, et al.
Cyclooxygenase-2 pathway correlates with VEGF expression in head
and neck cancer: implications for tumour angiogenesis and
metastasis. Neoplasia 2001; 3: 53-61.
19 Vogt T, McClelland M, Jung B, et al.
Progression and NSAID-induced apoptosis in malignant melanomas are
independent of cyclooxygenase II. Melanoma Res 2001; 11:
587-99.
20 Moyisch KB, Menezes RJ, Ronani A, et al.
Regular aspirin use and lung cancer risk. BMC Cancer 2002; 2:
31-8.
21 Corely D, Krlikowske K, Verma R,
Buffler P. Protective association of aspirin/NSAIDs and
esophageal cancer: a systematic review and meta-analysis.
Gastroenterology 2003; 124: 47-56.
22 Smith SR, Morhenn VB, Piaquadio DJ. Bilateral
comparison of the efficacy and tolerability of 3% Diclofenac Sodium
gel and 5% 5-Fluorouracil cream in the treatment of Actinic
Keratoses of the face and scalp. JDrugs in Dermatol 2006; 5:
156-9.
23 Loven K, Stein L, Furst K, Levy S.
Evaluation of the efficacy and tolerability of 0.5% fluorouracil
cream and 5% fluorouracil cream applied to each side of the face in
patients with actinic keratosis. Clin Ther 2002; 24: 100-990.
24 Honda KS. HIV and skin cancer. Dermatol Clin 2006; 24:
521-30.
|
Printable version
Version PDF
