Journal of Dermatological Treatment

ISSN: 0954-6634 (Print) 1471-1753 (Online) Journal homepage: https://www.tandfonline.com/loi/ijdt20

Efficacy of Doxycycline in the Treatment of Early

Stages of Mycosis Fungoides, a Randomized
Controlled Trial

H. El Sayed, S. Shalaby, M.R.E. Abdel-Halim, D M. Aboelfadl & N. Samir

To cite this article: H. El Sayed, S. Shalaby, M.R.E. Abdel-Halim, D M. Aboelfadl & N. Samir
(2019): Efficacy of Doxycycline in the Treatment of Early Stages of Mycosis Fungoides, a
Randomized Controlled Trial, Journal of Dermatological Treatment

To link to this article: https://doi.org/10.1080/09546634.2019.1667474

Accepted author version posted online: 17

Sep 2019.

Submit your article to this journal

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=ijdt20
Title: Efficacy of Doxycycline in the Treatment of Early Stages of Mycosis Fungoides, a

Randomized Controlled Trial

Key words: Randomized controlled trial, Mycosis fungoides, Doxycycline, Apoptosis, Bcl-2

Word count:

t
ip
Abstract: 198

Manuscript: 3445

cr
Number of figures: 2

us
Number of tables: 4
an
Authors: H. El Sayed*, MBBCh, MSc; S. Shalaby*, MD; M.R.E. Abdel-Halim*, MD, Dip

Dermpath (ICDP-UEMS); D M. Aboelfadl**; and N. Samir*, MD

M
From:

*Dermatology Department, Faculty of Medicine, Cairo University, Kasr Al Aini University

e d

Hospital.
pt

**Pathology Department, National Research Institute, Cairo, Egypt.

ce

Corresponding author:
Ac

Suzan Shalaby, MD

Address: 6 Street 25 Takseem El-Nasr Lelsayarat, Cairo, Egypt

Telephone number: +2-01224019459

Fax number: +20 2 23641088

Email:suzyandnls@gmail.com
No funding was received for this work

The authors declare no conflicts of interest

Limitation: Small number of patients and short treatment duration.

Register number: NCT03454945 (www.clinicaltrials.gov)

Abstract

t
ip
Background: Mycosis fungoides is the most common type of primary cutaneous T cell

cr
lymphomas. Doxycycline promoted apoptosis in different human malignant cell lines and in-vivo

us
models.

an
Objectives: To test for the therapeutic efficacy of doxycycline in comparison to PUVA in early

stages of classic MF and its effect on T cell apoptosis.

M
Methods: Thirty-six patients were randomized into either: doxycycline 200 mg daily (n=18) or
d

PUVA (3 weekly sessions) (n=18) for 12 weeks. The primary outcome (therapeutic efficacy)
e

was defined in terms of objective response rate (ORR) which was measured according to
pt

changes in the modified severity weighted assessment tool (mSWAT).

ce

Results: Doxycycline achieved significantly less ORR (partial response) in comparison to

Ac

PUVA (11.1%, 50% respectively, p= 0.016). The percent reduction in mSWAT, CAILS,

histopathology score and CD3 expression was significantly less in the doxycycline group

(p=0.001, p=0.001, p˂0.001, and p=0.004 respectively). Within the doxycycline group, changes

in mSWAT and CAILS showed no correlation with changes in the CD3 or Bcl-2 expression.

Gastric upset was significantly more encountered in the doxycycline group (p=0.001).
Conclusion: Doxycycline is not suitable as a sole agent in the treatment of early stages of classic

MF, acting mainly by anti-inflammatory rather apoptotic function

Introduction

Mycosis fungoides (MF) is the most common form of primary cutaneous T cell

lymphomas (CTCLs). The widely accepted etiopathogenic theory of MF implies that chronic

t
antigenic stimulation renders T cells resistant to apoptosis. The NF-κB pathway which induces

ip
the antiapoptotic protein Bcl-2 was found to be activated in CTCLs and many alterations in

cr
apoptosis regulating genes (p53, Fas and Bcl-2) have been reported in patients with MF [1-4].

us
Despite the indolent behavior of early disease stages (stages IA-IIA), prompt treatment is
an
essential to prevent progression to more aggressive forms, where the median survival drops from

10-35 years in early stages to only 1.5 years in advanced stages [5,6]. Skin directed therapy is the
M
standard approach for treating early stages. It includes topical steroids, phototherapy, topical

nitrogen mustard, topical bexarotene, local radiation and total skin electron beam therapy
d

(TSEBT). Refractory early stage patients are candidates for combination therapies; phototherapy
e
pt

with low dose interferon or oral bexarotene [7].

Avoiding long term side effects especially in younger patients, and in those receiving
ce

repeated cycles of treatment is the main concern when choosing a therapeutic modality [8]. This
Ac

has triggered the continuous search for safer, more tolerable and effective therapies directed

towards early stages of MF [9].

Doxycycline is a member of the tetracycline group of antibiotics, which is available,

affordable and generally well tolerated with a low side effect profile [10]. In addition to its

antimicrobial action, doxycycline exerts also many non-antimicrobial effects, the most important
of which are the anti-inflammatory effects exerted by vast mechanisms mostly through inhibition

of matrix metalloproteinases (MMPs) and various inflammatory cytokines [11,12].

Since 2002, many researchers have been shedding light also on the anti-cancer effects of

doxycycline, which were mainly attributed to its pro-apoptotic, anti-angiogenic, anti-MMPs,

radio-sensitizing and quorum sensing inhibitory functions. Its selective toxicity on cancer cells

compared to chemotherapeutic agents represents a great advantage. However, most of these

t
ip
researches, whether in solid malignancies or in hematological malignancies (including

leukemias, diffuse large B cell lymphoma (DLBCL), MF and Sézary Syndrome), are still

cr
experimental studies conducted on different human malignant cell lines or in vivo models [13-

us
26].
an
A recent phase II study was reported where doxycycline was used to treat canine multi-

centric B cell lymphoma [27]. The only available reports in literature where doxycycline was
M
used to treat humans with lymphoma included a case of primary cutaneous CD4+ small- to

medium sized pleomorphic T cell lymphoma which achieved complete remission after receiving
e d

200mg doxycycline for 6 weeks and cases of ocular adnexal lymphoma where almost two thirds
pt

of the patients remained progression free 5 years after treatment with one or two cycles of
ce

doxycycline 100 mg bid for 3 weeks [28,29].

Accordingly, this study aimed to test for the therapeutic efficacy of doxycycline in
Ac

comparison to PUVA in early stages of classic MF and its effect on the apoptosis of T

lymphocytes by studying its effect on their Bcl-2 expression.

Methods

This randomized controlled trial (RCT) was conducted at the Dermatology Out Patient

Clinic (OPC), Dermatology Department, Faculty of Medicine, Cairo University (Kasr Al Aini

University Hospital) during the period from January 2017 to July 2018. It was approved by the

Department Scientific and Ethical Committee Boards and informed consents were signed by all

participants.

t
ip
Recruitment and enrollment

Forty adult patients with MF presenting consecutively to the Dermatology OPC were

cr
assessed based on certain inclusion and exclusion criteria for their eligibility to be enrolled in the

us
study. Assessment parameters included: medical history [history of relevant allergies,
an
photosensitivity, medical illnesses and drug history], type of MF skin lesions [patches, plaques or

tumors], body surface area (BSA) affected (rule of 9) [30], clinical variant of MF,
M
histopathological variant of MF [provided that diagnostic biopsies were taken within a maximum

of 30 days, otherwise, recent biopsies were taken to document the current histopathological
e d

state/variant of the disease], staging work up [complete blood count (CBC), cervical, axillary and
pt

inguinal lymph node ultrasound, lymph node biopsy (if indicated), chest X-ray and abdomino-
ce

pelvic ultrasound] and laboratory work-up [liver function tests (AST, ALT, bilirubin and

albumin), kidney function tests (urea and creatinine) and pregnancy test for females in the child
Ac

bearing period].

Included in the study were adults (age above 18) of both sexes with early stage MF (IA-

IIA). Excluded from the study were patients with any clinical or histopathological variant of MF

other than the classic variant, advanced stages of classic MF (IIB, III or IV), pregnant or
lactating females, patients with autoimmune diseases, patients with solid or hematological

malignancies and patients with any contraindications for doxycycline, phototherapy or psoralens.

Accordingly, and based on a predefined calculated sample size, 36 of the 40 recruited

patients were enrolled in the therapeutic phase. A wash out period of at least 4 weeks from any

relevant topical or systemic treatment that can affect the course of MF was adopted. Females in

the child bearing period were stressed upon keeping a reliable method of contraception.

t
ip
Randomization

cr
All 36 patients were randomly allocated with 1:1 ratio, using blinded cards, into either

us
one of the two parallel arms of the trial: active treatment group (18 patients) who received
an
Doxycycline 100 mg oral capsules (twice daily) or a control group (18 patients) who received

Psoralen-UVA (PUVA) [0.5mg/kg 8-methoxypsoralen, starting dose: 1-3 joules/cm2 according

M
to Fitzpatrick's skin type and increments of increase every other session of 0.5 joules/ cm 2

according to the degree of erythema and based on the guidelines of the Phototherapy Unit,
e d

Dermatology Department, Faculty of Medicine, Cairo University, Kasr Al Aini University

pt

Hospital]. No concomitant treatment was allowed except for Vaseline topical application (if
ce

needed).

End of study
Ac

End of study (EOS) was defined as achieving complete response or a maximum of 3

months (12 weeks) of treatment.

Clinical assessment at the onset of therapy (day 0) and at EOS

At both day 0 and at EOS, all 36 randomized patients were clinically assessed by a

blinded senior investigator using standardized clinical scoring systems for MF which included:
modified severity weighted assessment tool (mSWAT) and composite assessment of index lesion

severity (CAILS) [31]. Moreover, pruritus was assessed using a visual analog scale from 0-10,

where 0=no pruritus and 10=worst imaginable pruritus and patients’ satisfaction was assessed

using dermatology life quality index (DLQI) [32].

Histopathological and immunohistochemical assessments at the onset of therapy (day 0) and

t
ip
at EOS.

Slides and paraffin blocks of the pre-treatment diagnostic biopsies of the 36 randomized patients

cr
were retrieved. At EOS, another 4mm punch biopsy was taken close to the site of the original

us
biopsy. Biopsied lesions were ensured to be among the 5 lesions scored in the CAILS. Biopsies
an
were fixed in 10% neutral buffered formalin, routinely processed, embedded in paraffin and

stained with hematoxylin and eosin (H&E) [33]. Additional sections were obtained for
M
immunohistochemical staining using rabbit monoclonal antibody against CD3 (catalog number

BSB 790-4341, Ventana, USA) and mouse monoclonal antibody against Bcl-2 (catalog number
e d

790-4464, Ventana, USA). Automated immunohistochemical staining was carried out using
pt

BenchMark XT autostainer (Ventana Medical Systems, Inc. Tucson, Arizona, USA) and
ce

according to manufacturer's protocol.

Pre-treatment and post-treatment H&E stained slides were scored according to the
Ac

scoring system described by Guitart and colleagues [34]. The immunohistochemically-stained

sections were analyzed using the Leica Qwin 3.5.1 2008 Image Analyzer (LEICA Imaging

Systems Ltd, Cambridge, England) Immunoreactive cells were counted in 5 non-overlapping

high-power (X400) fields.

 Follow up and safety assessment during the study

Patients were followed up at week 4 and week 8 where mSWAT, CAILS and pruritus

scoring were done. In addition, any possible side effects were assessed, reported and managed

accordingly.

Definition of treatment outcomes at EOS

t
ip
The primary treatment outcome at EOS was to compare the therapeutic efficacy of

doxycycline with PUVA. This was defined in terms of the objective response rate (ORR) and

cr
was measured according to changes in mSWAT from baseline (Table 1) [31].

us
Secondary treatment outcomes at EOS was to compare the effect of doxycycline with
an
PUVA as regards time to response (TTR) [the time from the first dose till the patient first met a

50% decrease in mSWAT] as well as regards changes in CAILS, pruritus score, DLQI score,
M
H&E score of MF, CD3 count and Bcl-2 count.
e d

Management of doxycycline patients after EOS

pt

Patients with progressive disease were shifted to PUVA, patients with partial response
ce

continued the same dose for another 12 weeks while patients with stable disease received 400mg

daily for another 12 weeks. At the end of the extended treatment period, patients with the same
Ac

or worsened response were shifted to PUVA.

Statistical methods

Power calculation

A statistical power analysis was performed for sample size estimation. According to

Cohen's criteria [35], the effect size was arbitrarily selected at d=1 due to absence of prior studies

or pilots. With a two tailed alpha error probability α set at 0.05, a sample size of 36 with 18 per

group was calculated. To be able to reject the null hypothesis that the difference is zero,

t
ip
probability power was estimated at 0.8 using Mann Whitney test for independent non-parametric

samples. Sample size was calculated using G power statistical software.

cr
Statistical analysis

us
Data were coded and entered using the statistical package SPSS (Statistical Package for
an
the Social Sciences) version 25. For quantitative data, data were described using mean and

standard deviation (SD) if homogenous (normally distributed) and median and range if non-
M
homogenous (non-normally distributed). Categorial data were presented using frequency (count)

and relative frequency (percentage). Comparisons of quantitative variables between study

e d

groups were done using the student t-test in homogenous data and the non-parametric Mann-
pt

Whitney test in non-homogenous data. For comparisons of serial measurements within each
ce

patient, paired t-test was used in homogenous data and the non-parametric Wilcoxon signed rank

test was used in non-homogenous data [36]. For comparing categorical data, Chi square (χ2) test
Ac

was performed. Exact test was used instead when the expected frequency is less than 5 [37].

Correlations between quantitative variables were done using Spearman correlation coefficient

[38]. P-values less than 0.05 were considered as statistically significant.

 Since no patient was lost to follow up, both primary treatment outcome [based on

intention to treat analysis (ITT)] and secondary treatment outcomes [based on per protocol

analysis] used n=18 in their calculation [39,40].

Results

Patient's flow diagram is presented in Fig. 1 [39]. At baseline, both groups were

statistically homogenous apart from gender distribution and CD3 count (Table 2).

t
ip
Results of the RCT

cr
Evaluation of primary treatment outcome, percent change in mSWAT and mSWAT at EOS

us
Objective response rate was significantly less in the doxycycline group than in the PUVA
an
group (p= 0.016) with partial response achieved in 11.1% and 50% of the cases respectively (Fig.

2). Complete response was not achieved in either group. Although the percent reduction in
M
mSWAT score in the doxycycline group was significantly less than in the PUVA group

(p=0.001), the mSWAT score at EOS showed no significant difference between both groups
e d

(p=0.389) (Table 3).

pt

Evaluation of secondary treatment outcomes at EOS

ce

The percent reduction in CAILS score in the doxycycline group was significantly less
Ac

than in the PUVA group (p=0.001) with significantly higher CAILS score at EOS (p=0.017)

(Table 3).

The percent reduction in H&E score of MF and CD3 count in the doxycycline group was

significantly less than in the PUVA group (p˂0.001 and p=0.004 respectively) with significantly

higher H&E score and CD3 count at EOS (p=0.003 and p ˂0.001). A significant difference
existed between both groups as regards the percent change in Bcl-2 expression (p=0.012), with

significantly higher Bcl-2 count at EOS in the doxycycline group (p=0.031) (Table 3).

No significant difference existed between both groups as regards TTR or changes in

pruritus and DLQI scores (Table 3).

Adverse events

Adverse events were mild and did not necessitate discontinuation of therapy in either

t
ip
group. Adverse effects, mainly gastric upset, were significantly more encountered in the

cr
doxycycline group (p=0.001) (Table 3).

us
Further analysis within the doxycycline group
an
Within the doxycycline group, mSWAT, CAILS, pruritus score, DLQI score and H&E

score of MF were reduced significantly from baseline (p=0.013, p=0.001, p=0.001, p=0.01, and
M
p=0.007 respectively) (Table 4).
d

The percent reductions in mSWAT and CAILS showed no significant correlation with the
e

percent changes in the H&E score of MF (p=0.948; r=-0.017, p=0.581; r=0.139 respectively),
pt

CD3 count (p=0.906; r=-0.03, p=0.839; r=0.052) or Bcl-2 count (p=0.072; r=-0.447, p=0.194;
ce

r=-0.331).
Ac

Results of patients continuing doxycycline after EOS

Two patients with partial response at EOS continued the same dose of doxycycline for

another 12 weeks. No further improvement was achieved and one developed new lesions. Both

were shifted to PUVA. Fourteen patients with stable disease at EOS received doxycycline (400

mg daily) for another 12 weeks. Three didn’t show any further improvement and 2 developed
new lesions. All 5 patients were shifted to PUVA. The 9 remaining patients were lost to follow-

up.

Discussion

With partial response rate of 11.1% compared to 50% in the PUVA group, it is obvious

that doxycycline can't substitute PUVA in treating early stages of MF. Although TTR showed no

t
ip
significant difference between both groups, the percent reduction in mSWAT and CAILS, was

significantly less in the doxycycline group (-11% and -20%) compared to PUVA (-45% and -

cr
47%).

us
Although doxycycline has caused significant reduction in both mSWAT and CAILS
an
scores from baseline, the percent reduction in CAILS score (-20%) was higher than the percent

reduction in mSWAT (-11%) indicating more influence on index lesion skin score than on the
M
actual disease extent. The CAILS score includes assessment of erythema, scaling and

pigmentation in addition to lesion elevation and size. It appears that the mild improvement
e d

achieved by doxycycline reflects mainly some effects on various parameters of this score.
pt

Angiogenesis has been found to be increased in lesional skin of early MF compared to normal
ce

skin reflecting the degree of erythema perceived in MF lesions [41]. Doxycycline has a well-

known anti-angiogenic effect through its inhibitory effect on vascular endothelial growth factor
Ac

Moreover, the inhibitory effects of doxycycline on inflammatory cytokines and MMPs may also

explain these mild improvements especially that MMP-9, an important mediator of

inflammation, was found to be significantly elevated in early lesions of MF compared to normal

skin where it mediates degradation of the extracellular matrix and basement membrane

facilitating epidermotropism [11,12,43-45].

 On the other hand, PUVA reduced both CAILS and mSWAT to almost a similar degree

(-47% and -45% respectively) reflecting its potent multi-target effect in MF thus it retains its

position as the cornerstone in the treatment of early disease stages [46]. It has achieved complete

remission rates of 90%, 76% and 78% in stages IA, IB and IIA respectively [47].

Although histopathological assessment is not standard in assessing skin response in

clinical trials of MF [31], we assessed, as secondary outcomes, changes in the H&E score of MF,

t
ip
CD3 and Bcl-2 counts in an attempt to study the effect of doxycycline on apoptosis of T cells in

comparison to PUVA. With 3% and 8% reductions in the H&E score and CD3 count in the

cr
doxycycline group compared to 88% and 65% in the PUVA group respectively, one can clearly

us
conclude that doxycycline has a very weak T cell apoptotic effect in early stage MF patients.
an
Within the doxycycline group, the H&E score was reduced significantly from baseline to EOS,

however the CD3 count showed no significant difference. The H&E score used in this study
M
assesses several parameters, one of them includes the density of the dermal infiltrate. Reactive

inflammatory cells are known to be present in the dermal infiltrate of MF especially in early
e d

stages and may include B cells and other non-T cells [48,49]. Accordingly, doxycycline via its
pt

anti-inflammatory effect might have minimally affected the density of the dermal infiltrate in
ce

general rather than significantly inducing T cell apoptosis. The superior effect of PUVA in

reducing the H&E score and CD3 count can be explained by the multiple mechanisms by which
Ac

photochemotherapy induces T cell apoptosis [50,51].

In the current RCT, a significant difference existed between the doxycycline group and

the PUVA group as regards the percent change in Bcl-2 count with Bcl-2 increasing in the

doxycycline group and decreasing in the PUVA group. The decreased expression by PUVA is

consistent with the results of other studies which explained that this Bcl-2 reduction is an
important pathway of induction of cell death through loss of the anti-apoptotic proteins [52]. The

increased expression of Bcl-2 in the doxycycline group, although not significant, may explain the

resistance to apoptosis of T cells which was manifested by the lack of significant reduction in the

CD3 count within the doxycycline group and the significantly less percent reduction compared to

PUVA.

This appears contradictory to our understanding of the role of doxycycline as a pro-

t
ip
apoptotic agent in malignancy. The only previous study assessing the apoptotic effect of

doxycycline in CTCLs was an in vitro study conducted on multiple different cell lines from

cr
patients with MF and Sézary syndrome [26]. They found that doxycycline achieved a significant

us
dose dependent apoptosis of T cells through inhibiting TNF induced NF-κB activation and
an
through reactive oxygen species. It appears that what is achieved in an in vitro setting cannot be

expected to be the same inside the human body where many players interact together creating a
M
completely different milieu. Also, the stage of the disease is a crucial factor, where the blood

resident T cell lymphocyte subtypes of late stage MF might be more vulnerable to the
e d

doxycycline apoptotic actions than those of the early stage skin disease.
pt

Moreover, some researches revealed that doxycycline prevents chemotherapy induced

ce

apoptotic injury [53]. Previous work on neu2ral tissue suggested anti-apoptotic actions of

tetracyclines attributed to increased Bcl-2 in certain conditions of tissue injury [54]. These
Ac

studies might advocate that doxycycline can activate both apoptotic and anti-apoptotic pathways

depending on surrounding cellular status.

We found no significant correlation within the doxycycline group between the percent

change in the clinical scores (mSWAT and CAILS) and the percent change in the H&E score,

CD3 or Bcl-2 counts, further supporting the idea that the minimal doxycycline achieved effects
in early stage MF in the current study are mainly mediated through its anti-inflammatory effect

rather than actually achieving significant apoptosis of T lymphocytes.

From patients' symptoms and psychological perspective, the current study showed that

the overall patient satisfaction (pruritus and DLQI scores) with doxycycline was comparable to

that of PUVA. Doxycycline was a more feasible mode of treatment to most patients compared to

PUVA therapy which necessitated 3 weekly visits to the phototherapy unit. On the other hand,

t
ip
the doxycycline associated side effects in the form of GIT upset and nausea were significantly

higher when compared to PUVA.

cr
The choice of the 200 mg daily dose in the current study was based on the standard

us
guidelines for acne [55]. According to data available from a registered clinical trial
an
NCT02341209,53677 where a 400mg daily dose of doxycycline is given for 5months-1year for

relapsed cases of CTCL (results not available yet), patients with partial response in the current
M
study were given doxycycline 200mg daily for another 3 months and patients with stable disease

were given doxycycline 400mg daily for another 3 months. However, neither dosage
e d

modification has resulted in more satisfactory results.

pt

The lack of satisfactory therapeutic efficacy in the current study compared to the
ce

successful treatment of a case of primary cutaneous CD4+ small/medium sized pleomorphic T

cell lymphoma, after 6 weeks of 200 mg doxycycline daily therapy can be explained by the
Ac

current understanding of this type of T cell proliferation which was downgraded in the current

WHO classification of hematological malignancies to just a lymphoproliferative disorder not an

actual lymphoma based on limited clinical risk, localized disease, and similarity to clonal drug

reactions [28,56]. And so, doxycycline seems to have acted mainly in this case as an anti-

inflammatory agent.
 Moreover, the satisfactory effect of doxycycline in treating cases of ocular adnexal

lymphoma is explained by the fact that most of these cases are marginal zone lymphomas

associated with Chlamydophila psittaci infection for which doxycycline has been suggested as a

treatment option mainly for its anti-bacterial action [29].

In conclusion, doxycycline can't be used as a sole agent in the treatment of early stages of

MF since it has achieved significantly less therapeutic efficacy compared to PUVA in a 3 months

t
ip
treatment duration. Changes achieved are mainly attributed to its anti-inflammatory effects. The

potential of promoting T cell apoptosis achieved in vitro is questionable in MF patients. The

cr
small number of patients and short treatment duration were limitations in this study. Moreover,

us
the potential risk of disease progression while on an experimental treatment prevented us from
an
trying doxycycline for periods longer than 6 months. The value of using doxycycline as an

adjuvant to PUVA in early stages of MF awaits further studies.

M

References
e d

1. Sors A, Jean-Louis F, Pellet C et al. Down-regulating constitutive activation of the NF-

pt

kappa B canonical pathway overcomes the resistance of cutaneous T-cell lymphoma to

ce

apoptosis. Blood 2006; 107:2354–2363.

2. Oeckinghaus A, Ghosh S. The NF-kappa B family of transcription factors and its

Ac

regulation. Cold Spring Harb Perspect in Boil 2009; 1:1-14.

3. Karin M. NF-κB as a Critical link between inflammation and cancer. Cold Spring Harb

Perspect in Boil 2009; 1(5): a000141.

4. Knadolf Sekulovic L, Cikota B, Jovic M et al. The role of apoptosis and cell-proliferation

regulating genes in mycosis fungoides. J Dermatol Sci 2009; 55: 53–56.

5. Willemze, R., Jaffe, E., Burg G et al. WHO-EORTC classification for cutaneous

lymphomas. Blood 2005; 105:3768–3785.

6. Kim Y, Martinez G, Varghese A et al. Topical nitrogen mustard in the management of

mycosis fungoides: update of the Stanford experience. Arch Dermatol 2003; 139: 165-

173.

7. Jawed S, Myskowski P, Horwitz S et al. Primary cutaneous T-cell lymphoma (mycosis

t
ip
fungoides and Sézary syndrome): Part II. Prognosis, management, and future directions. J

Am Acad Dermatol 2014; 70: 223.e1-223.e17.

cr
8. Hoot J, Wang, L, Kho T et al. The effect of phototherapy on progression to tumors in

us
patients with patch and plaque stage of mycosis fungoides. J Dermatolog Treat 2018;

29:272-276.
an
9. Prichard M, Harris T, Williams M et al. Treatment strategies for relapsed and refractory
M
aggressive non-Hodgkin's lymphoma. Expert Opin Pharmacother 2009; 10:983-995.

10. Smith K, Leyden JJ. Safety of doxycycline and minocycline: A systematic review. Clin
e d

Ther 2005; 27:1329-1342.

pt

11. Pirard B. Insight into the structural determinants for selective inhibition of matrix
ce

metalloproteinases. Drug Discov Today 2007;12: 640–646.

12. Henehan M, Montuno M, De Benedetto A. Doxycycline as an anti-inflammatory agent:

Ac

updates in dermatology. J Eur Acad Dermatol Venereol 2017; 31:1800-1808.

13. Lokeshwar BL, Selzer MG, Zhu BQ et al. Inhibition of cell proliferation, invasion, tumor

growth and metastasis by an oral non-antimicrobial tetracycline analog (COL-3) in a

metastatic prostate cancer model. Int J Cancer 2002; 98:297-309.

14. Pulvino M, Chen L, Oleksyn D et al. Inhibition of COP9-signalosome (CSN)

deneddylating activity and tumor growth of diffuse large B-cell lymphomas by

doxycycline. Oncotarget 2015; 6:14796–14813.

15. Kao GD, Jiang Z, Fernandes AM et al. Inhibition of phosphatidylinositol-3-OH

kinase/Akt signaling impairs DNA repair in glioblastoma cells following ionizing

radiation. J Biol Chem 2007; 282: 21206-21212.

t
ip
16. Son K, Fujioka S, Iida T et al. Doxycycline induces apoptosis in PANC-1 pancreatic

cancer cells. Anticancer Res 2009; 29:3995-4003.

cr
17. Holmes N, Charles P. Safety and efficacy review of Doxycycline. Clin Med Insights Ther

us
2009; 1: 471-482.
an
18. Neckers L, Workman P. Hsp90 molecular chaperone inhibitors: Are we there yet? Clin

Cancer Res 2012;18: 64–76.

M
19. Wu Z, Gholami AM, Kuster B. Systemic identification of the HSP90 candidate

Regulated Proteome. Mol Cell Proteomics 2012; 11: M111.016675.

e d

20. Husain FM, Ahmad I. Doxycycline interferes with quorum sensing-mediated virulence
pt

factors and biofilm formation in Gram negative bacteria. World J Microbiol Biotechnol
ce

2013; 29: 949-957.

21. Song H, Fares M, Maguire K et al. Cytotoxic effects of tetracycline analogues

Ac

(doxycycline, minocycline and COL-3) in acute myeloid leukemia HL-60 cells. PLoS

One 2014; 9: e114457.

22. Yang B, Lu Y, Zhang A et al. Doxycycline Induces Apoptosis and Inhibits Proliferation

and Invasion of Human Cervical Carcinoma Stem Cells. PLoS One 2014; 10: e0129138.
23. Galván-Salazar H, Soriano-Hernández A, Montes-Galindo D et al. Preclinical trial on the

use of doxycycline for the treatment of adenocarcinoma of the duodenum. Mol Clin

Oncol 2016; 5:657-659.

24. Kashii-Magaribuchi K, Takeuchi R, Haisa Y et al. Induced Expression of Cancer Stem

Cell Markers ALDH1A3 and Sox-2 in Hierarchical Reconstitution of Apoptosis-resistant

Human Breast Cancer Cells. Acta Histochem Cytochem 2016; 49:149-158.

t
ip
25. Wang S, Zhao B, Liu Y et al. New application of an old drug: Antitumor activity and

mechanisms of doxycycline in small cell lung cancer. Int J Oncol 2016; 48:1353-1360.

cr
26. Alexander-Savino CV, Hayden M, Richardson C et al. Doxycycline is an NF-κB

us
inhibitor that induces apoptotic cell death in malignant T-cells. Oncotarget 2016;7:

75954-75967.
an
27. Hume KR, Sylvester SR, Borlle L et al. Metabolic abnormalities Detected in Phase II
M
evaluation of Doxycycline in Dogs with Multicentric B-cell lymphoma. Front Vet Sci

2018; 5:25.
e d

28. Toberer F, Hartschuh W, Hadaschik E. Primary cutaneous CD4+ small- to medium-sized

pt

pleomorphic T-cell lymphoma: temporary remission by oral doxycycline. JAMA

ce

Dermatol 2013; 149: 956-959.

29. Han JJ, Kim TM, Jeon YK et al. Long-term outcomes of first-line treatment with
Ac

doxycycline in patients with previously untreated ocular adnexal marginal zone B cell

lymphoma. Ann Hematol 2015; 94:575-581.

30. Hettiaratchy S, Papini R. Initial management of a major burn: assessment and

resuscitation. BMJ 2004; 329: 101–103.

31. Olsen E, Whittaker S, Kim Y et al. International Society for Cutaneous Lymphomas;

United States Cutaneous Lymphoma Consortium; Cutaneous Lymphoma Task Force of

the European Organisation for Research and Treatment of Cancer.Clinical End Points and

Response Criteria in Mycosis Fungoides and Se´zary Syndrome: A Consensus Statement

of the International Society for Cutaneous Lymphomas, the United States Cutaneous

Lymphoma Consortium, and the Cutaneous Lymphoma Task Force of the European

t
ip
Organisation for Research and Treatment of Cancer. J Clin Oncol 2011; 29: 2598-2607.

32. Finlay AY, Khan GK. Dermatology Life Quality Index (DLQI)--a simple practical

cr
measure for routine clinical use. Clin Exp Dermatol 1994; 19:210-216.

us
33. Fischer A, Jacobson K, Rose J et al. Hematoxylin and eosin staining of tissue and cell
an
sections. CSH Protoc 2008. Doi:10.1101/pdb.prot4986.

34. Guitart J, Kennedy J, Ronan S et al. Histologic criteria for the diagnosis of mycosis
M
fungoides: proposal for a grading system to standardize pathology reporting. J Cutan

Pathol 2001; 28:174-183.

e d

35. Cohen J. Statistical Power Analysis for the Behavioral Sciences. 2nd edn. New York:
pt

Lawrence Erlbaum Associates,1988.

ce

36. Chan Y. Biostatistics102: Quantitative Data – Parametric & Non-parametric Tests.

Singapore Med J 2003; 44: 391-396.

Ac

37. Chan Y. Biostatistics 103: Qualitative Data –Tests of Independence. Singapore Med J

2003; 44: 498-503.

38. Chan Y. Biostatistics 104: Correlational Analysis. Singapore Med J 2003; 44: 614-619.

39. Schulz KF, Altman DG, Moher D. CONSORT 2010 statement: updated guidelines for

reporting parallel group randomized trials. Ann Intern Med 2010; 152: 726-732.
40. Higgins JPT, Green S, eds. Cochrane handbook for systematic reviews of interventions

version 5.1.0 [updated March 2011], The Cochrane Collaboration.2011. Available

via:www.cochrane-handbook.org.

41. Bosseila M, Sayed Sayed K, El-Din Sayed SS et al. Evaluation of Angiogenesis in Early

Mycosis Fungoides Patients: Dermoscopic and Immunohistochemical Study.

Dermatology 2015; 231:82-86.

t
ip
42. Smith J, Lanier V, Braziel R et al. Expression of vascular endothelial growth factor and

its receptors in rosacea. Br J Ophthalmol 2007; 91:226-229.

cr
43. Nissinen L, Kähäri VM. Matrix metalloproteinases in inflammation. Biochim Biophys

us
Acta 2014; 1840:2571-2580.
an
44. Rasheed H, Tolba Fawzi MM, Abdel-Halim MR, Eissa AM et al. Immunohistochemical

study of the expression of matrix metalloproteinase-9 in skin lesions of mycosis

M
fungoides. Am J Dermatopathol 2010; 32:162-169.

45. Vacca A, Ribatti D, Ria R et al. Proteolytic Activity of Human Lymphoid Tumor Cells.
e d

Correlation with Tumor Progression. Dev Immunol 2000; 7: 77-88.

pt

46. Gettler S. Fung M. Efficacy of treatments for mycosis fungoides and Sézary syndrome:
ce

Nationwide survey responses. Dermatol Online J 2005; 11: 6.

47. Herrman JJ, Roenigk HH Jr, Hurria A, et al. Treatment of mycosis fungoides with
Ac

photochemotherapy (PUVA): Long-term follow-up. J Am Acad Dermatol 1995; 3: 234-

242

48. Papadavid E, Economidou J, Psarra A et al. The relevance of peripheral blood T-helper 1

and 2 cytokine patterns in the evaluation of patients with mycosis fungoides and Sezary

syndrome. Br J Dermatol 2003; 148:709-718.

49. Ferrara G, Crisman G, Zalaudek I et al. Free-Floating Collagen Fibers in Interstitial

Mycosis Fungoides. Am J Dermatopathol 2010; 32: 352-356.

50. Johnson R, Staiano-Coico L, Austin L et al. PUVA treatment selectively induces a cell

cycle block and subsequent apoptosis in human T lymphocytes. Photochem Photobiol

1996; 63:566-571.

51. Viola G, Fortunato E, Cecconet L et al. Central role of mitochondria and P53 in PUVA-

t
ip
induced apoptosis in human keratinocytes cell line nctc-2544. Toxicol Appl Pharmacol

2008; 277: 84-96.

cr
52. Werber-Bandeira L, Herdy AM, Pagani EA et al. Primary cutaneous T cell lymphomas:

us
photochemotherapy immunomodulation with analysis of the inflammatory-expansive
an
cellular dynamic. Dermatol Ther 2014; 27: 74–78.

53. Yeh YC, Lai HC, Ting CT et al. Protection by doxycycline against doxorubicin-induced
M
oxidative stress and apoptosis in mouse testes. Biochem Pharmacol 2007; 74:969-980.

54. Scarabelli TM, Stephanou, AE, Pasini E et al. Minocycline inhibits caspase activation
e d

and reactivation, increases the ratio of XIAP to smac/DIABLO, and reduces the
pt

mitochondrial leakage of cytochrome C and smac/DIABLO. J Am Coll Cardiol 2004; 43:

ce

865-874.

55. Bienenfeld A, Nagler AR, Orlow SJ. Oral antibacterial therapy for acne vulgaris: An
Ac

evidence-based review. Am J Clin Dermatol 2017; 18: 469–490.

56. Swerdlow SH, Campo E, Pileri SA et al. The 2016 revision of the World Health

Organization classification of lymphoid neoplasms. Blood 2016; 127:2375-2390.

Figure legends

t
ip
cr
us
an
M
e d

Fig. 1: Patient's flow diagram at EOS according to CONSORT guidelines for reporting RCTs.39
pt
ce
Ac
 t
ip
cr
us
an
M
Fig. 2: A patient in the doxycycline group achieving partial response at EOS
d

(a-d: at day 0, e-h: at EOS)

e
pt
ce
Ac
 Table 1: Objective response definitions based on changes in mSWAT (Olsen et al., 2011)
Response Definition
Complete response 100% clearance of skin lesions
50%-99% clearance of skin disease from
baseline
Partial response
without new tumors (T3) in patients with T1,
T2 or T4 only skin disease
˂ 25% increase to ≥ 50% clearance in skin
Stable disease disease from baseline without new tumors (T3)
in patients with T1, T2, or T4 only skin disease
≥25% increase in skin disease from baseline or

t
New tumors (T3) in patients with T1, T2 or T4

ip
only skin disease or Loss of response: in those
Progressive disease
with complete or partial response, increase of

cr
skin score of greater than the sum of nadir plus
50% baseline score

us
Any disease recurrence in those with complete
Relapse
response
an
M
e d
pt
ce
Ac
 Table 2: Baseline characteristics in doxycycline and PUVA groups before therapy§

Doxycycline group (n=18) PUVA group (n=18) P value

Gender (M/F) Number (%) 12 (66.7%)/6(33.3%) 5 (27.8%)/13(72.2%) 0.019*

Age (years) Mean±SD 38.94 ± 13.55 43.28 ± 10.18 0.521

Range 1 - 26 2 - 25
Duration (years) 0.051
Median 5 9.50
Disease Ia 3 (16.7%) 2 (11.1%)
stage

t
Ib 13 (72.2%) 14 (77.8%) 1
Number

ip
(%) IIa 2 (11.1%) 2 (11.1%)
Range 5 - 92 7 - 78
mSWAT 0.563

cr
Median 36.5 46
CAILS Mean±SD 54.83 ± 16.35 56.5 ± 17.2 0.913
Range 0-8 0-5

us
Pruritus score 0.265
Median 2.5 2
Range 0 - 22 2 - 19
DLQI score 0.791
Median 6an 7
H&E score of MF Mean±SD 5.97 ± 0.88 6.61 ± 1.4 0.203
Range 228 - 924 46 - 803
CD3 count/5HPF ˂0.001*
Median 713 356.5
M
Range 0-273 2-217
Bcl-2 count /5HPF 0.433
Median 67 54
§
Data is presented as mean±SD if normally distributed and as range and median if non-normally
distributed; *p<0.05 is significant M: Males, F: Females; mSWAT: Modified Severity Weighted
d

Assessment Tool; CAILS: Composite Assessment of Index Lesion Severity; DLQI: Dermatology
e

Life Quality Index; H&E: Hematoxylin and Eosin; HPF: High Power Field.
pt
ce
Ac
 Table 3: Summary of the results of the RCT at EOS§
Doxycycline PUVA group
P value
group (n=18) (n=18)
Complete response 0 (0%) 0 (0%)
Partial response 2 (11.1%) 9 (50.0%)
ORR Stable disease 14 (77.8%) 9 (50.0%) 0.016*
Number (%) Progressive disease 2 (11.1%) 0 (0%)
Relapse 0 (0%) 0 (0%)
Percent change Range -51.1 - 61.3 -82.2 - (-10.2)
0.001*
in mSWAT score at EOS Median -11.05 -44.95
mSWAT
Range 6.5 - 83 5 - 70
mSWAT at EOS 0.389

t
Median 29 23

ip
TRR (weeks) Mean±SD 12 ± 0 10.22 ± 2.11 0.436
Percent change Range -67.1 - 16.1 -81 - (-11.1)
0.001*

cr
CAILS in CAILS score at EOS Median -20.3 -46.85
CAILS at EOS Mean±SD 42.28 ± 10.94 32.22 ±15.78 0.017*

us
Percent change Range -100 - 0 -100 - 50
0.719
Pruritus in pruritus score at EOS Median -48.55 -80
score Range 0-4 0-3
Pruritus score at EOS 0.481
Medianan 0 0
Percent change Range -83.3 - 66.7 -100 - 0
0.265
in DLQI score at EOS Median -26.1 -38.1
DLQI score
Range 0 - 12 0 - 14
M
DLQI score at EOS 0.355
Median 5 4
Percent change Range -85.7 - 0 -100 - 0
˂0.001*
H&E score in H&E score at EOS Median -3.85 -88.25
d

of MF Range 1-7 0-6

H&E score at EOS 0.003*
Median 5 1
e

Percent change Range -85.3 - 67.3 -96 - 60.2

0.004*
CD3 in CD3 count at EOS Median -8.3 -65.6
pt

count/5HPF Range 123 - 1171 20 - 572

CD3 count at EOS ˂0.001*
Median 548.5 90.5
ce

Percent change Range -83.65 - 475 -100 - 277.78

0.012*
Bcl-2 in Bcl-2 count at EOS Median 44.64 -58.97
count/5HPF Range 0 - 411 0 - 161
Bcl-2 count at EOS 0.031*
Median 81.5 39.5
Ac

Gastric upset 8 (44.4%) 0 (0%)

Phototoxicity 0 (0%) 5 (27.8%)
Side effects Gastric upset and 0.001*
0 (0%) 1 (5.6%)
Number (%) phototoxicity
No side effects 10 (55.6%) 12 (66.7%)

§
Data is presented as mean±SD if normally distributed and as range and median if non-normally
distributed, *p<0.05 is significant ; ORR: Objective Response Rate; mSWAT: Modified Severity
Weighted Assessment Tool; CAILS: Composite Assessment of Index Lesion Severity; DLQI:
Dermatology Life Quality Index; H&E: Hematoxylin and Eosin; HPF: High Power Field
 §
Table 4: Comparison between baseline and EOS values within the doxycycline group
Doxycycline (baseline) Doxycycline (at EOS) P value
(n=18) (n=18)
Range 5 - 92 6.5 – 83
mSWAT 0.013*
Median 36.5 29
CAILS Mean±SD 54.83 ± 16.35 42.28 ± 10.94 0.001*
Range 0-8 0-4
Pruritus score 0.001*
Median 2.5 0
Range 0 - 22 0 - 12
DLQI score 0.01*
Median 6 5
H&E score of
Mean±SD 5.97 ± 0.88 4.67 ± 1.75 0.007*
MF

t
CD3 Range 228 - 924 123 - 1171

ip
0.372
count/5HPF Median 713 548.5
Bcl-2 Range 0-273 0 - 411
0.478

cr
count/5HPF Median 67 81.5
§
Data is presented as mean±SD if normally distributed and as range and median if non-
normally distributed; *p<0.05 is significant ; mSWAT: Modified Severity Weighted Assessment

us
Tool; CAILS: Composite Assessment of Index Lesion Severity; DLQI: Dermatology Life
Quality Index; H&E: Hematoxylin and Eosin; HPF: High Power Field
an
M
e d
pt
ce
Ac