Analysis of surrogate gene expression markers in peripheral blood of melanoma patients to predict treatment outcome of adjuvant pegylated interferon alpha 2b (EORTC 18991 side study)

ORIGINAL ARTICLE

Analysis of surrogate gene expression markers in peripheral bloodof melanoma patients to predict treatment outcome of adjuvantpegylated interferon alpha 2b (EORTC 18991 side study)

Antonia Busse • Jerome Rapion • Alberto Fusi • Stefan Suciu • Anika Nonnenmacher •

Mario Santinami • Wim H. J. Kruit • Alessandro Testori • Cornelis J. A. Punt •

Angus G. Dalgleish • Alan Spatz • Alexander M. M. Eggermont • Ulrich Keilholz

Received: 11 October 2012 / Accepted: 10 April 2013

� Springer-Verlag Berlin Heidelberg 2013

Abstract We analysed mRNA levels of interferon

response genes (ISG15, STAT1, CXCL10) of inhibitors of

the JAK/STAT pathway (STAT3, SOCS1, SOCS3) and of

cytokines (TNFa, IL10, TGFß1) in peripheral blood of 91

stage III melanoma patients enrolled in EORTC 18991 trial

to find biomarkers indicative for disease stage and pre-

dictive for efficacy of pegylated interferon alpha-2b (PEG-

IFNa-2b) therapy. mRNA levels were analysed at baseline

and after 6 months. Univariate and multivariate analyses

were performed to estimate the prognostic and predictive

role of mRNA levels for distant metastasis-free survival

(DMFS) and relapse-free survival (RFS). Compared to

healthy controls, melanoma patients showed significantly

higher TGFb1 mRNA levels. In a multivariate model,

increasing SOCS1 and SOCS3 mRNA levels were associ-

ated with worse RFS (P = 0.02 and P = 0.04, respec-

tively) and DMFS (P = 0.05 and P = 0.05, respectively)

due to negative correlation between, respectively, SOCS1/

SOCS3 mRNA levels and ulceration or Breslow thickness.

No impact of PEG-IFNa-2b on mRNA levels was observed

except for ISG15 mRNA levels, which decreased in the

treatment arm (P = 0.001). It seems that patients with a

decrease [60 % of ISG15 mRNA levels during 6 months

PEG-IFNa-2b had inferior outcome.

Keywords RT-PCR � Melanoma � Adjuvant pegylated

interferon alpha-2b � EORTC � Immunosuppressive

cytokines � Interferon response genes

Abbreviations

PEG-IFNa-2b Pegylated interferon alpha-2b

SOCS Suppressors of cytokine signalling

DMFS Metastasis-free survival

RFS Relapse-free survival

pB Peripheral blood

Introduction

Adjuvant interferon (IFN)-a therapy in melanoma is still a

matter of debate due to its toxicity and costs in relation to

A. Busse (&) � A. Fusi � A. Nonnenmacher � U. Keilholz

Department of Medicine III, Charite-CBF, Hindenburgdamm 30,

12200 Berlin, Germany

e-mail: [email protected]

J. Rapion � S. Suciu

European Organisation for Research and Treatment of Cancer

Headquarters, Brussels, Belgium

M. Santinami

Istituto Nazionale dei Tumori, Milan, Italy

W. H. J. Kruit � A. M. M. Eggermont

Erasmus MC-Daniel den Hoed Cancer Center, Rotterdam,

The Netherlands

A. Testori

Istituto Europeo di Oncologia, Milan, Italy

C. J. A. Punt

Academic Medical Center, University of Amsterdam,

Amsterdam, The Netherlands

A. G. Dalgleish

St. Georges Hospital, London, UK

A. Spatz

Jewish General Hospital, Montreal, Canada

A. M. M. Eggermont

Cancer Institute Gustave Roussy, Villejuif-Paris, France

123

Cancer Immunol Immunother

DOI 10.1007/s00262-013-1428-4

benefit of treatment outcome. Several randomized con-

trolled clinical trials examined the role of IFNa-2b and

pegylated IFNa-2b (PEG-IFNa 2b) as an adjuvant treat-

ment for melanoma patients who have undergone resection

of high-risk stage III (nodal disease) or stage IIb (primary

tumour thickness [4 mm, node-negative) lesions. These

trials have shown an effect on relapse-free survival (RFS),

but without a clinically significant effect on overall sur-

vival [1–3]. To select for patients most susceptible to

respond to IFN therapy, it is important to understand the

mechanisms that influence the efficacy of IFN treatment

and to find biomarkers predicting response to therapy. Post

hoc meta-analysis of the EORTC18952 and 18991 trials

including more than 2,500 patients indicated that both

tumour stage and ulceration were not only strong prog-

nostic factors but are also predictive factors for the efficacy

of adjuvant IFN/PEG-IFN therapy. The greatest risk

reductions were observed in patients with ulceration and

stage IIb/III-N1. The efficacy of IFN/PEG-IFN was lower

in stage III-N2 patients with ulceration and uniformly

absent in patients without ulceration [3–6]. Post hoc meta-

analysis of adjuvant IFNa trials by Wheatley et al and

McMasters et al support the hypotheses that ulceration is a

predictive marker for the efficacy of adjuvant IFN therapy

in patients with microscopic nodal involvement [7, 8].

However, up to now, the biological basis of this observa-

tion is unclear.

IFNa is a multifunctional regulatory cytokine with

immunomodulatory, antitumour, and antiproliferative

effects [9]. Murine models have shown that the indirect

immunomodulatory activity of IFN may be more important

than direct cytotoxic, proapoptotic, or antiangiogenic

effects [10–12], which was also suggested by several

clinical studies [13–15]. Recent reports suggest that

downregulation in IFN signalling in T lymphocytes is an

important contributor to immune dysfunction in patients

with metastatic melanoma and distinguishes IFN respond-

ers and poor IFN responders [16, 17]. Interaction of IFNawith a cell surface receptor leads to a series of intracellular

reactions through JAK-STAT signalling (STAT1/STAT2)

that result in transcriptional induction of antiviral and

immunoregulatory molecules, so-called interferon-stimu-

lated genes (ISG) [18, 19]. IFN-a-activated cellular sig-

nalling is negatively regulated by a number of inhibitory

factors, including STAT3, protein tyrosine phosphatases,

protein inhibitors of activated STATS, and the suppressors

of cytokine signalling (SOCS) that are rapidly induced by

IFN and form a classical negative feedback loop to regulate

JAK-STAT signalling [18, 20]. Overexpression of SOCS1

and SOCS3 has been shown to lead to reduced induction of

ISG and antitumour activity of immune effector cells in a

melanoma mice model [21]. Immunosuppressive cytokines

like IL10 or TGFß secreted by tumour cells itself or by

immunosuppressive cells may also be involved in the

mechanism by which IFN signalling is inhibited in lym-

phocytes in cancer patients as they are able to induce

expression of the negative regulators of IFN signalling [22–

25]. Moreover, recent studies have shown a high degree of

variability in the activation of STAT1 [17] and induction of

ISG [26–28]. Taken together, differences in response to

IFN therapy might be related to clinical, immunological,

and interferon cell signalling factors.

Therefore, here we analysed the mRNA expression

levels of known interferon response genes (ISG15, STAT1,

CXCL10) of inhibitors of the JAK/STAT pathway

(STAT3, SOCS1, SOCS3) and of cytokines (TNFa, IL10,

TGFß1) in peripheral blood (pB) of stage III melanoma

patients that were included in the EORTC trial 18991 [3].

This trial investigated the effect of long-term administra-

tion of PEG-IFNa-2b in patients with stage III melanoma,

for a maximum of 5 years. We asked whether the tran-

scriptional profile would be altered by PEG-IFN-a2b

therapy and would be indicative for disease stage and

treatment outcome.

Materials and methods

Patients

Blood samples from 91 high-risk melanoma stage III

patients (Tx, N1 [microscopic non-palpable nodes]/N2

[palpable nodes], M0, with the exception of intransit

metastasis) were obtained. All patients were enrolled into

the multicenter prospective randomized phase III EORTC

18991 trial (NCT00006249) and were recruited from

multiple institutions [3]. This side study was approved by

the institutional review boards. Blood samples obtained at

the time of the enrolment and after 6 months were ana-

lysed. Informed consent from patients for translational

research for the identification of prognostic and predictive

factors in association with the EORTC 18991 trial was

obtained. Controls were blood samples of 12 healthy vol-

unteers who had no evidence of any clinically detectable

disease at the time of blood withdrawal.

Processing of blood samples

Blood (10 ml) was collected in EDTA-containing tubes

and processed within 2 h. Blood was centrifuged at

1,6009g for 10 min. Plasma was removed, and the cell

pellet was resuspended in 5 ml of guanidinium thiocya-

nate (GTC) buffer and stored at -80 �C. Samples were

shipped to the central laboratory (Charite—Campus

Benjamin Franklin, Berlin, Germany) in GTC buffer on

dry ice.

Cancer Immunol Immunother

123

RNA extraction and reverse transcription

The method is described by us in detail elsewhere [29]. In

brief, total cellular RNA was extracted from 10 ml of EDTA-

stabilized blood by acid guanidinium thiocyanate/phenol

chloroform isolation. For reverse transcription, Omniscript

Reverse Transcriptase kit (Qiagen, Hilden, Germany) was

used following the manufactures instructions.

Quantitative real-time PCR

Quantitative real-time RT-PCR was done by LightCycler

Technology (Roche Diagnostics) as described [29]. Primer

and probe sequences were designed using the LightCycler

Probe Design software, version 1.0. Sequences and PCR

conditions are listed in Table 1. About 2 ll of each cDNA

was diluted to a volume of 20 ll PCR mix (LightCycler

Fast Start DNA Master Hybridization Probes, Roche

Diagnostics) containing the appropriate MgCl2 concentra-

tion, 0.5 pmol of each primer, and 0.2 pmol of each probe.

For each target, an initial denaturation cycle at 95 �C for

10 min, followed by 50 cycles with a final extension of

2 min at 72 �C, was used. The specificity of the PCR

products was confirmed by melting curve analysis as well

as by gel electrophoresis. For quantification, PCR products

generated from the proteasome subunits cDNAs and from

HMBS cDNAs were cloned into the vector pCR2.1-TOPO

(Invitrogen, Groningen, the Netherlands). A standard curve

with 3 dilutions of the appropriate plasmid in duplicates

was included in each PCR run. Analysis of RT-PCR

expression data was done with the LightCycler software

(version 3). All samples were analysed in duplicate. The

average value of both duplicates was used as a quantitative

value. To correct for differences in cDNA amount on a per-

sample basis, results were provided as ratio to house-

keeping gene HMBS expression.

Statistical analysis

mRNA expression levels have been transformed to LOG2

(mRNA levels) in order to obtain variables which had a

normal distribution. To compare LOG2 of mRNA expres-

sion levels of melanoma patients (at baseline) with that of

healthy controls, a generalized linear model (GLM) has

Table 1 Primer/hybridization probe sequences, amplicon length, MgCl2 concentration, and annealing temperature (AT) for the real-time RT-

PCR; (P, dephosphorylated; X, Fluorescein; Y, LC Red 640)

Target Fwd/Rev primer LC/FL probe Amplicon/

bp

MgCl2 AT/�C

CXCL10 50-TCTAGAACTGTACGCTGTACC

50-GATTCAGACATCTCTTCTCACC

50-Y-TCCACGTGTTGAGATCATTGCTACAATG-P

50-AAACTTGAAATTATTCCTGCAAGCCAATTT-X

157 5 63

HMBS 50-TGCAGGCTACCATCCATGTCCCTGC

50-AGCTGCCGTGCAACATCCAGGATGT

50-Y-TGTGGGTCATCCTCAGGGCCATCTTC-P

50-CGTGGAATGTTACGAGCAGTGATGCCTACC-X

187 4 65

IL10 50-CCTTCCAGTGTCTCGG

50-TGGAGTACAGGGGCAT

50-Y-AGGTCAGGAGTTCCTAACCAG-P

50-AGGCGGGTGGATCACT-X

357 4 60

ISG15 50-GCCAGTACAGGAGCTTG

50-GCCGATCTTCTGGGTGA

50-Y-CAACGAATTCCAGGTGTCCCTGAGC-P

50-GACGGTGAAGATGCTGGCG-X

159 4 62

SOCS3 50-ATGTAGCAGCGATGGA

50-GTGCAAAGTTTGACTTGG

50-Y-CTACTCAGGCCCTTGAGGC-P

50-CTCCAGTCCTCTCTCCTCCC-X

325 4 61

SOCS1 50-CCGTGTTATTTTGTTATTACTTGC

50-AGATACTGGGTATATGTAAACATGA

50-Y-GGAGCCAGGACCTGAACT-P

50-TGGTTGTTGTAGCAGCTTAACTGTAT-X

267 5 57

STAT1 50-GAATGAGGGTCCTCTCATCGTTA

50-CGCCACCAGCATGTTGTA

50-Y-CCAACGTCAGCCAGCTCCC-P

50-TCTCTGCCCGTTGTGGTGAT-X

181 2 60

STAT3 50-TACTAAAGTCAGGTTGCTGGT

50-TGGATTCTTCCATGTTCATCAC

50-Y-TCCCGGAAATTTAACATTCTGGGCACAAA

CAC-P

50-GGGACGTTGCAGCTCTCAGAG-X

161 4 64

TGFb1 50-CCCACAACGAAATCTATGAC

50-GCTAAGGCGAAAGCCC

50-Y-AGAGTGGTTATCTTTTGATGTCACCG-P

50-GGCACCCAGCGACTCG-X

314 5 57

TNFa 50-CCTCCACCCATGTGC

50-CTCCCAGATAGATGGGCTCATA

50-Y-CCTGCCAGAGGGAGACCC-P

50-AACCTCCTCTCTGCCATCAAGAG-X

156 4 63

Cancer Immunol Immunother

123

been used, adjusted by age and sex. Association with

clinicopathologic features has been analysed by variance

(ANOVA). Clinicopathologic features included age (in 3

classes: \35 years, 35–49 years, and C50 years), sex,

presence of ulceration, tumour stage (N1 for microscopic

non-palpable nodes; N2 for palpable nodes), number of

lymph nodes involved (0–1/2–4/5?), and Breslow thick-

ness (\1.5 mm/1.5–4 mm/[4 mm). Adjusting of p-values

because of multiple testing was done with Benjamini and

Hochberg’s method. To analyse change of mRNA levels

between baseline and after 6 months therapy, for each

mRNA marker, the percentage of variation from baseline

of the LOG2 of mRNA level between 6 months and base-

line was determined as follows: %Diff of LOG2

(mRNA) = [LOG2 (mRNA 6 months)—LOG2 (mRNA base-

line)] 9 100/LOG2 (mRNA baseline). Two time-to-event

endpoints were considered: relapse-free survival (RFS) and

distant metastasis-free survival (DMFS) from randomiza-

tion in the study. Events for each these endpoints included:

any type of disease recurrence and death due to any cause

(for RFS), distant metastasis and death due to any cause

(for DMFS). For these two endpoints, patients were cen-

sored at the last follow-up date if the event of interest did

not occur. Relationship between mRNA markers and sur-

vival was examined in univariate analyses and subse-

quently by multivariate analysis using a Cox regression

model. Survival curves were plotted according to the

Kaplan–Meier method and were compared with the two-

tailed log-rank test.

Statistical analyses were performed with SAS 9.2 soft-

ware (SAS Institute, Cary, NC).

Results

Patients’ characteristics

A total of 91 patients out of 1,256 patients enrolled in the

EORTC 18991 trial were evaluable for the RT-PCR study.

Forty-four patients had been randomly assigned to the

treatment arm and 47 to the control arm. The median

actuarial follow-up was 7.7 years. Clinical characteristics

of patients in the RT-PCR study are listed in Table 2.

There were no marked differences between the entire

patient population and the patient subset included in the

RT-PCR study for the most important variables (treatment

arm, type of nodal involvement, and number of positive

lymph nodes; data not shown) [3]. However, there were

more patients with microscopic nodal involvement in the

RT-PCR-study population compared to the entire popula-

tion (54.5 vs. 43.2 %, respectively). Within 6 months from

randomization, one patient had a disease recurrence and 6

patients discontinued treatment due to toxicity.

mRNA expression levels at baseline

mRNA expression levels at baseline have been determined

for the interferon response genes ISG15, STAT1, CXCL10;

inhibitors of the JAK/STAT pathway STAT3, SOCS1,

SOCS3; and for the cytokines TNFa, IL10 and TGFb1. As

crude levels had a quite skewed distribution, each mRNA

level has been transformed to LOG2 (mRNA expression

level) in order to obtain variables, which had a distribution

close to a normal one. Analysis of variance (ANOVA) for

each LOG2 transformation of mRNA levels with all clini-

copathologic features revealed that ISG15 mRNA levels

were higher in N2 compared to N1 patients; in contrast,

patients with higher Breslow thickness had lower mRNA

expression levels of the IFN inducible genes STAT1 and

CXCL10. Moreover, patients with ulcerations or with high

Breslow thickness showed lower SOCS1 and SOCS3

Table 2 Baseline characteristics and outcome in the study population

Treatment arm Total

Observation PEG-IFNa-

2b

Number of LN at random, n (%)

1 30 (63.8) 25 (56.8) 55 (60.4)

2–4 12 (25.5) 16 (36.4) 28 (30.8)

5? 4 (8.5) 3 (6.8) 7 (7.7)

pNx 1 (2.1) 0 (0.0) 1 (1.1)

Stage of disease at random, n (%)

N1 (microscopic

involvement)

26 (55.3) 24 (54.5) 50 (54.9)

N2 (palpable nodes) 21 (44.7) 20 (45.5) 41 (45.1)

Breslow thickness, n (%)

\1.5 9 (19.1) 8 (18.2) 17 (18.7)

1.5–3.99 21 (44.7) 21 (47.7) 42 (46.2)

C4 12 (25.5) 10 (22.7) 22 (24.2)

Unknown 5 (10.6) 5 (11.4) 10 (11.0)

Ulceration, n (%)

No 24 (51.1) 23 (52.3) 47 (51.6)

Yes 11 (23.4) 10 (22.7) 21 (23.1)

Unknown 12 (25.5) 11 (25.0) 23 (25.3)

Age (years), n (%)

\35 7 (14.9) 9 (20.5) 16 (17.6)

35–49 17 (36.2) 19 (43.2) 36 (39.6)

C50 23 (48.9) 16 (36.4) 39 (42.9)

Sex, n (%)

Male 29 (61.7) 34 (77.3) 63 (69.2)

Female 18 (38.3) 10 (22.7) 28 (30.8)

RFS

% at 5 years (95 % CI) 51 (36, 64) 59 (43, 72) 55 (44, 64)

DMFS

% at 5 years (95 % CI) 60 (44, 72) 63 (47, 76) 61 (51, 71)

Cancer Immunol Immunother

123

mRNA levels. IL10 mRNA levels decreased with age

(Fig. 1), but we did not see any influence of age or sex on

TGFb1 mRNA expression levels. However, after adjusting

p-values because of multiple testing, none of the associa-

tions remained significant.

Compared to 12 healthy controls (6 males, 6 females;

median age 36 years, age range 27–93 years), melanoma

patients showed significant higher TGFb1 mRNA levels

(P \ 0.0001) and lower IL10 mRNA levels (P = 0.023,

Fig. 2). After adjustment for multiple testing, this

remained only significant for TGFb1 mRNA levels

(P = 0.0005).

Prognostic impact of baseline mRNA levels

on outcome

None of the transformed mRNA expression levels at

baseline had influence on RFS and DMFS and RFS when

taken alone. However, as some of these variables were

correlated with initial clinical/disease features, a

Fig. 1 Boxplots showing the

association of the LOG2 of

mRNA levels with

clinicopathologic features. The

relative amount was calculated

as ratio marker [(pg/ml)/HMBS

(pg/ml)] and then LOG2-

transformed in order to obtain

variables which have a normal

distribution. Horizontal lineinside the boxes = median

values; upper and lowerboundaries of the boxes = first

and third quartiles of the

distribution; whiskers = mild

outliers, open dots outliers

Cancer Immunol Immunother

123

multivariate Cox model including not only basal log-

transformed mRNA levels but also stage, sex, Breslow

thickness, number of lymph nodes, and ulceration was set

up. In this multivariate model, increasing SOCS1 mRNA

levels and SOCS3 mRNA levels adjusted by the other

baseline factors were associated with worse RFS

(P = 0.02, Hazard Ratio [95 % CI] = 1.43 [1.07–1.91]

and P = 0.04, Hazard Ratio [95 % CI] = 1.27

[1.01–1.61], respectively) and DMFS (P = 0.05, Hazard

Ratio [95 % CI] = 1.35 [1–1.83] and P = 0.05, Hazard

Ratio [95 % CI] = 1.27 [1–1.62], respectively). This

influence is due to the negative correlation between,

respectively, SOCS1/SOCS3 mRNA levels and ulceration,

and Breslow thickness.

Impact of PEG-IFNa-2b therapy versus observation

on the change of mRNA expression levels

To analyse whether PEG-IFN therapy had any influence on

mRNA levels, mRNA levels have also been determined

after 6 months in PEG-IFNa-2b treated and untreated

patients. Data of 33 patients in the observation arm and 22

patients in the PEG-IFNa-2b arm were available. For each

mRNA marker, the percentage of variation from baseline

of the LOG2 (mRNA) expression level between 6 months

and baseline was determined.

Comparing the values of the PEG-IFNa-2b arm with

that of the observation arm by ANOVA showed a decrease

in IL10 mRNA levels (P = 0.029, not significant after

Fig. 2 Expression levels of IL10- and TGFb1-specific mRNA in

melanoma patients (n = 91) compared to healthy controls (n = 12).

The relative amount was calculated as ratio marker [(pg/ml)/HMBS

(pg/ml)] and then LOG2-transformed in order to obtain variables

which have a normal distribution. Dotted line indicates mean value:

LOG2 (TGFb1): -4.8 in melanoma patients versus -15.8 in healthy

controls; LOG2 (IL10): -16.2 in melanoma patients versus -14.9 in

healthy controls

Fig. 3 Change of ISG15- and IL10-specific mRNA in melanoma

patients after 6 months treatment with pegylated interferon a-2b

(PEG-IFNa-2b) compared to observation. The relative amount was

calculated as ratio marker [(pg/ml)/HMBS (pg/ml)] and then LOG2-

transformed in order to obtain variables, which have a normal

distribution. Differences are expressed as % variation from baseline

levels

Cancer Immunol Immunother

123

adjustment for multiple testing, Fig. 3); however, TGFb1

mRNA levels as well as TNFa mRNA levels remained

constant in both treatment groups. Moreover, we observed

no impact of PEG-IFNa-2b on mRNA levels of STAT or

SOCS proteins, nor on mRNA levels of the IFN inducible

gene CXCL10. However, mRNA levels of ISG15 surpris-

ingly decreased in the PEG-IFNa-2b arm (P = 0.0001,

after adjustment for multiple testing P = 0.001), Fig. 3).

We were wondering whether the decrease in ISG15

mRNA levels had a prognostic influence on the outcome.

Two distinct populations were extracted from the distri-

bution graph, which represented the variation from the

baseline of ISG15 mRNA levels. One group represented

the range from no decrease up to 60 % decrease in ISG15

mRNA expression levels after 6 months PEG-IFNa-2b

therapy. The other group represented the range of over

60 % decrease. Patients who relapsed before 6 months

post-randomization and patients who stopped in PEG-

IFNa-2b due to toxicity or treatment refusal were excluded

from the analysis. Interestingly, those 5 patients, for whom

a decrease of more than 60 % of ISG15 mRNA was

observed, had a worse RFS (Hazard Ratio [99 %

CI] = 3.72 [0.65, 21.15], P = 0.039) and DMFS (Hazard

Ratio (99 % CI) = 3.64 [0.63, 20.91]; P = 0.042) com-

pared to those with none or up to 60 % decrease in ISG15

mRNA expression levels (Fig. 4).

Discussion

Although adjuvant therapy with IFNa for high-risk cuta-

neous melanoma is frequently used, only a minority of

patients seems to benefit. Identification of biomarkers

predicting outcome is a major challenge to select for

patients most susceptible to profit from IFN therapy. Sev-

eral prognostic and predictive factors have been studied in

adjuvant IFN trials, like autoantibodies [14, 30, 31],

SB100B [32] or ferritin, and CRP serum levels [33] which

Fig. 4 Kaplan–Meier curves

with respect to the change of

ISG15 mRNA levels after

6 months. RFS: relapse-free

survival (RFS), DMFS: distant

metastasis-free survival. PEG-

IFN interferon a-2b: pegylated

interferon a-2b, N total number

of patients, and O observed

events

Cancer Immunol Immunother

123

have at its best a prognostic value, but no predictive value

for IFN outcome. Here, we studied the prognostic and

predictive value of mRNA levels of interferon response

genes (ISG15, STAT1, CXCL10) of inhibitors of the JAK/

STAT pathway (STAT3, SOCS1, SOCS3) and of cytokines

(TNFa, IL10, TGFß1) in pB of stage III melanoma patients

receiving adjuvant PEG-IFNa-2b therapy within the EO-

RTC trial 18991 [3].

IFNa can act directly on tumour cells and vasculature,

but in melanoma, immunomodulatory properties may play

a more important role [21, 34]. Dysfunction of the immune

system arises already during the early stages of melanoma

and increases throughout progression to metastatic disease:

Low IL-2 and high IL-5 serum levels were observed in

stage III melanoma patients and Treg cell frequencies as

well as accumulation of non-mature DCs in lymph nodes

seem to increase with stage and tumour thickness,

respectively [35–37]. In accordance with those results, we

observed in our patient cohort of stage III melanoma

patients higher TGFb1 mRNA levels in pB compared to

healthy controls. However, differences in IL10 mRNA

expression levels between melanoma patients and healthy

controls were not significant, although we cannot exclude

subtle differences in mRNA expression levels as the cohort

of control samples was quite small.

We were wondering whether immune dysregulation is

related to tumour stage and outcome of IFNa therapy.

However, we found no association of mRNA levels of

immunosuppressive cytokines like TGFb1 or IL10 with

clinicopathological features like lymph node status, Bre-

slow thickness, or ulceration. Yurkovetsky et al. showed

that in melanoma patients treated with high-dose IFNa,

pre-treatment levels of the proinflammatory cytokines and

chemokines were significantly higher in high-risk patients

and that inflammatory cytokines positively correlate with

RFS [38]. STAT1, CXCL10, and ISG15 were reported to

be decreased in stage IV melanoma patients associated

with reduced pSTAT1 [16]. In our cohort of stage III

melanoma patients, we observed no difference in mRNA

levels of proinflammatory cytokines or ISGs compared to

healthy controls, but CXCL10 and STAT1 were negatively

associated with Breslow thickness, although not significant

after adjustment for multiple testing.

Recent studies have shown a high degree of (inter-

patient) variability in the activation of STAT1 [17] and

induction of ISGs [26–28]. This includes also the induction

of negative regulators of the immune response like SOCS1

and SOCS3, which have been shown to play a major role in

antitumour activity [21, 34]. In a multivariate model, when

other variables like ulceration and Breslow thickness were

taken into consideration, SOCS1 and SOCS3 mRNA levels

had a negative influence on RFS and DMFS. However,

they were not predictive factors. Expression of SOCS

proteins within immune effector cells may not only influ-

ence response to exogenous IFNa but also to endogenous

IFNa activity. Sustained SOCS1 and SOCS3 expressions

have been shown to play important roles in immune tol-

erance in a mice tumour model [39].

This influence of SOCS1/SOCS3 mRNA levels on RFS

and DMFS is due to their negative correlation with ulcer-

ation and Breslow thickness. As SOCS1/SOCS3 are

induced by IFN-a but also by various other inflammatory

cytokines in context of a negative feedback loop, their low

expression levels in blood cells of ulcerated and thick

melanomas might simply mirror a different cytokine

microenvironment [37]. Interestingly, it has been shown

that SOCS1 expression significantly downregulates

expression of VEGF, bFGF, and MMP-2 in melanoma cells

[40]. Downregulation of SOCS1/SOCS3 not only in

tumour cells but also in stromal and immune effector cells

may promote lymphovascular invasion and metastasis [40].

In vitro experiments have shown that high doses of

IFNa were able to overcome the signalling defects of T

cells of metastatic melanoma patients and were able to

induce effective polarization of dendritic cells [16].

Immunoregulatory effects associated with tumour regres-

sion have also been observed in vivo using the high-dose

regimen of IFNa-2b [15, 35, 38, 41] and were also linked

to restoration of impaired pSTAT1 signalling capacity as

shown by Simons et al in a small exploratory study

including 14 stage IIIB-C melanoma patients [42]: Patients,

who had a clinical response to high-dose IFN-a2b therapy,

showed a lower PBL interferon signalling capacity in T

cells at baseline than non-responders, but had a significant

increase in STAT1 activation in peripheral blood T cells

over the 4-week induction phase of neoadjuvant therapy.

Interestingly, only patients who displayed modest aug-

mentation had good clinical outcome; patients who had

minimal or negative changes in pSTAT1 response, but also

patients who had ‘‘hyper’’ IFN signalling responses had

poor outcome.

Although in our cohort, IL10 mRNA levels decreased

slightly after 6 months PEG-IFNa-2b therapy, elevated

TGFb1 mRNA levels did not. We also did not observe any

increase in mRNA levels of ISGs [27]. Surprisingly, we

rather detected a decrease in ISG15 mRNA levels after

6 months PEG-IFNa-2b therapy. mRNA levels of

CXCL10 or STAT1, which are early response genes, might

be only transiently induced and do not reach a steady-state

level over the treatment period as observed in PEG-IFNaand Ribaverin-treated Hepatitis C patients [27]. However,

the reason for the decrease of ISG15 is not clear, because

ISG15 induction is primarily triggered by type I IFNs.

ISG15 is an ubiquitin-like protein and modifier. ISG15

conjugation might promote degradation of targeted pro-

teins via a non-proteasomal pathway, but might also

Cancer Immunol Immunother

123

regulate enzymatic or DNA binding activity of target

proteins similar to other ubiquitin-like modifiers. Several

key regulators of signal transduction are targeted by ISG15.

Some of them are type I IFN-induced proteins, such as

PKR and RIG-I, others are key regulators that are involved

in IFN signalling, such as JAK1 and STAT1, implicating

the role of ISG15 and its conjugates in type I IFN-mediated

innate immune responses [43], that is, ISG15 enhances

IRF3 activity by ISGylation of IRF3 to prevent its ubiq-

uitylation and degradation [44] and mice with a massive

accumulation of ISG15 conjugates in various tissues,

exhibited a hypersensitivity to IFN stimulation [45].

Interestingly, patients with a decrease of more than 60 % of

ISG15 mRNA level compared to the respective ISG15

mRNA level at baseline had an inferior outcome. This is in

line with studies in Hepatitis C patients, reporting higher

levels of ISG15 in Kupffer cells of IFN responders [46].

However, our patient population was quite small; therefore,

it needs to be confirmed in further studies, if the course of

ISG15 mRNA levels in fact is a predictive factor.

In conclusion, melanoma stage III patients had an

immunosuppressive phenotype characterized by signifi-

cantly higher TGFb1 mRNA levels in peripheral blood.

Adjuvant PEG-IFN-a therapy did not resolve this immu-

nosuppressive phenotype. High SOCS1 and SOCS3 mRNA

expression levels were negatively associated with ulcera-

tion and Breslow index and had a significantly negative

prognostic influence on RFS and DMFS in a multivariate

COX model. Moreover, our results indicate that the course

of ISG15 mRNA levels during PEG-IFNa-2b therapy

might be of potential value to predict outcome; however,

this needs to be confirmed in larger studies.

Acknowledgments This study was supported by Sonnenfeld Stif-

tung from Germany, Berliner Krebsgesellschaft from Germany, and

Fonds Cancer (FOCA) from Belgium.

Conflict of interest The authors declare that they have no conflict

of interest.

References

1. Wheatley K, Ives N, Hancock B, Gore M, Eggermont A, Suciu S

(2003) Does adjuvant interferon-alpha for high-risk melanoma

provide a worthwhile benefit? A meta-analysis of the randomised

trials. Cancer Treat Rev 29(4):241–252

2. Kirkwood JM, Manola J, Ibrahim J, Sondak V, Ernstoff MS, Rao

U (2004) A pooled analysis of eastern cooperative oncology

group and intergroup trials of adjuvant high-dose interferon for

melanoma. Clin Cancer Res 10(5):1670–1677

3. Eggermont AM, Suciu S, Santinami M, Testori A, Kruit WH,

Marsden J, Punt CJ, Sales F, Gore M, Mackie R, Kusic Z,

Dummer R, Hauschild A, Musat E, Spatz A, Keilholz U (2008)

Adjuvant therapy with pegylated interferon alfa-2b versus

observation alone in resected stage III melanoma: final results of

EORTC 18991, a randomised phase III trial. Lancet 372(9633):

117–126

4. Anaya DA, Xing Y, Feng L, Huang X, Camacho LH, Ross MI,

Gershenwald JE, Lee JE, Mansfield PF, Cormier JN (2008)

Adjuvant high-dose interferon for cutaneous melanoma is most

beneficial for patients with early stage III disease. Cancer

112(9):2030–2037

5. Eggermont AM, Suciu S, MacKie R, Ruka W, Testori A, Kruit

W, Punt CJ, Delauney M, Sales F, Groenewegen G, Ruiter DJ,

Jagiello I, Stoitchkov K, Keilholz U, Lienard D (2005) Post-

surgery adjuvant therapy with intermediate doses of interferon

alfa 2b versus observation in patients with stage IIb/III melanoma

(EORTC 18952): randomised controlled trial. Lancet 366(9492):

1189–1196

6. Eggermont AM, Suciu S, Testori A, Kruit WH, Marsden J, Punt

CJ, Santinami M, Sales F, Schadendorf D, Patel P, Dummer R,

Robert C, Keilholz U, Yver A, Spatz A (2012) Ulceration and

stage are predictive of interferon efficacy in melanoma: results of

the phase III adjuvant trials EORTC 18952 and EORTC 18991.

Eur J Cancer 48(2):218–225

7. McMasters KM, Edwards MJ, Ross MI, Reintgen DS, Martin RC,

2nd, Urist MM, Noyes RD, Sussman JJ, Stromberg AJ, Scoggins

CR (2010) Ulceration as a predictive marker for response to

adjuvant interferon therapy in melanoma. Ann Surg 252(3):

460–465; discussion 465–466

8. Wheatley KIN, Eggermont A (2007) Interferon-a as adjuvant

therapy for melanoma: an individual patient data meta-analysis of

randomised trials. ASCO Annual Meeting Proceedings Part 1.

J Clin Oncol 25(suppl 18S):abstract 8526

9. Ferrantini M, Capone I, Belardelli F (2007) Interferon-alpha and

cancer: mechanisms of action and new perspectives of clinical

use. Biochimie 89(6–7):884–893

10. Lesinski GB, Anghelina M, Zimmerer J, Bakalakos T, Badgwell

B, Parihar R, Hu Y, Becknell B, Abood G, Chaudhury AR,

Magro C, Durbin J, Carson WE 3rd (2003) The antitumor effects

of IFN-alpha are abrogated in a STAT1-deficient mouse. J Clin

Investig 112(2):170–180

11. Badgwell B, Lesinski GB, Magro C, Abood G, Skaf A, Carson W

3rd (2004) The antitumor effects of interferon-alpha are main-

tained in mice challenged with a STAT1-deficient murine mela-

noma cell line. J Surg Res 116(1):129–136

12. Astsaturov I, Petrella T, Bagriacik EU, de Benedette M, Uger R,

Lumber G, Berinstein N, Elias I, Iscoe N, Hammond C, Hamilton

P, Spaner DE (2003) Amplification of virus-induced antimel-

anoma T-cell reactivity by high-dose interferon-alpha2b: impli-

cations for cancer vaccines. Clin Cancer Res 9(12):4347–4355

13. Ascierto PA, Kirkwood JM (2008) Adjuvant therapy of melanoma

with interferon: lessons of the past decade. J Transl Med 6:62

14. Gogas H, Ioannovich J, Dafni U, Stavropoulou-Giokas C, Frangia

K, Tsoutsos D, Panagiotou P, Polyzos A, Papadopoulos O,

Stratigos A, Markopoulos C, Bafaloukos D, Pectasides D,

Fountzilas G, Kirkwood JM (2006) Prognostic significance of

autoimmunity during treatment of melanoma with interferon.

N Engl J Med 354(7):709–718

15. Moschos SJ, Edington HD, Land SR, Rao UN, Jukic D, Shipe-

Spotloe J, Kirkwood JM (2006) Neoadjuvant treatment of

regional stage IIIB melanoma with high-dose interferon alfa-2b

induces objective tumor regression in association with modula-

tion of tumor infiltrating host cellular immune responses. J Clin

Oncol 24(19):3164–3171

16. Critchley-Thorne RJ, Yan N, Nacu S, Weber J, Holmes SP, Lee

PP (2007) Down-regulation of the interferon signaling pathway in

T lymphocytes from patients with metastatic melanoma. PLoS

Med 4(5):e176

17. Lesinski GB, Kondadasula SV, Crespin T, Shen L, Kendra K,

Walker M, Carson WE 3rd (2004) Multiparametric flow

Cancer Immunol Immunother

123

cytometric analysis of inter-patient variation in STAT1 phos-

phorylation following interferon Alfa immunotherapy. J Natl

Cancer Inst 96(17):1331–1342

18. Schindler C, Plumlee C (2008) Inteferons pen the JAK-STAT

pathway. Semin Cell Dev Biol 19(4):311–318

19. Theofilopoulos AN, Baccala R, Beutler B, Kono DH (2005) Type

I interferons (alpha/beta) in immunity and autoimmunity. Annu

Rev Immunol 23:307–336

20. Dalpke A, Heeg K, Bartz H, Baetz A (2008) Regulation of innate

immunity by suppressor of cytokine signaling (SOCS) proteins.

Immunobiology 213(3–4):225–235

21. Zimmerer JM, Lesinski GB, Kondadasula SV, Karpa VI, Lehman

A, Raychaudhury A, Becknell B, Carson WE 3rd (2007) IFN-

alpha-induced signal transduction, gene expression, and antitu-

mor activity of immune effector cells are negatively regulated by

suppressor of cytokine signaling proteins. J Immunol 178(8):

4832–4845

22. Ito S, Ansari P, Sakatsume M, Dickensheets H, Vazquez N,

Donnelly RP, Larner AC, Finbloom DS (1999) Interleukin-10

inhibits expression of both interferon alpha- and interferon

gamma- induced genes by suppressing tyrosine phosphorylation

of STAT1. Blood 93(5):1456–1463

23. Cassatella MA, Gasperini S, Bovolenta C, Calzetti F, Vollebregt

M, Scapini P, Marchi M, Suzuki R, Suzuki A, Yoshimura A

(1999) Interleukin-10 (IL-10) selectively enhances CIS3/SOCS3

mRNA expression in human neutrophils: evidence for an IL-10-

induced pathway that is independent of STAT protein activation.

Blood 94(8):2880–2889

24. Park IK, Shultz LD, Letterio JJ, Gorham JD (2005) TGF-beta1

inhibits T-bet induction by IFN-gamma in murine CD4 ? T cells

through the protein tyrosine phosphatase Src homology region 2

domain-containing phosphatase-1. J Immunol 175(9):5666–5674

25. Fox SW, Haque SJ, Lovibond AC, Chambers TJ (2003) The

possible role of TGF-beta-induced suppressors of cytokine sig-

naling expression in osteoclast/macrophage lineage commitment

in vitro. J Immunol 170(7):3679–3687

26. Brassard DL, Delorenzo MM, Cox S, Leaman DW, Sun Y, Ding

W, Gavor S, Spond J, Goodsaid F, Bordens R, Grace MJ (2004)

Regulation of gene expression by pegylated IFN-alpha2b and

IFN-alpha2b in human peripheral blood mononuclear cells.

J Interferon Cytokine Res 24(8):455–469

27. Taylor MW, Tsukahara T, Brodsky L, Schaley J, Sanda C, Ste-

phens MJ, McClintick JN, Edenberg HJ, Li L, Tavis JE, Howell

C, Belle SH (2007) Changes in gene expression during pegylated

interferon and ribavirin therapy of chronic hepatitis C virus dis-

tinguish responders from nonresponders to antiviral therapy.

J Virol 81(7):3391–3401

28. Zimmerer JM, Lehman AM, Ruppert AS, Noble CW, Olencki T,

Walker MJ, Kendra K, Carson WE 3rd (2008) IFN-alpha-2b-

induced signal transduction and gene regulation in patient

peripheral blood mononuclear cells is not enhanced by a dose

increase from 5 to 10 mega units/m2. Clin Cancer Res 14(5):

1438–1445

29. Keilholz U, Goldin-Lang P, Bechrakis NE, Max N, Letsch A,

Schmittel A, Scheibenbogen C, Heufelder K, Eggermont A, Thiel

E (2004) Quantitative detection of circulating tumor cells in

cutaneous and ocular melanoma and quality assessment by real-

time reverse transcriptase-polymerase chain reaction. Clin Cancer

Res 10(5):1605–1612

30. Bouwhuis MG, Suciu S, Collette S, Aamdal S, Kruit WH, Bas-

tholt L, Stierner U, Sales F, Patel P, Punt CJ, Hernberg M, Spatz

A, ten Hagen TL, Hansson J, Eggermont AM (2009) Autoim-

mune antibodies and recurrence-free interval in melanoma

patients treated with adjuvant interferon. J Natl Cancer Inst

101(12):869–877

31. Bouwhuis MG, Suciu S, Testori A, Kruit WH, Sales F, Patel P,

Punt CJ, Santinami M, Spatz A, Ten Hagen TL, Eggermont AM

(2010) Phase III trial comparing adjuvant treatment with pegy-

lated interferon Alfa-2b versus observation: prognostic signifi-

cance of autoantibodies–EORTC 18991. J Clin Oncol 28(14):

2460–2466

32. Bouwhuis MG, Suciu S, Kruit W, Sales F, Stoitchkov K, Patel P,

Cocquyt V, Thomas J, Lienard D, Eggermont AM, Ghanem G

(2011) Prognostic value of serial blood S100B determinations in

stage IIB-III melanoma patients: a corollary study to EORTC trial

18952. Eur J Cancer 47(3):361–368

33. Bouwhuis MG, Collette S, Suciu S, de Groot ER, Kruit WH, Ten

Hagen TL, Aarden LA, Eggermont AM, Swaak AJ (2011)

Changes of ferritin and CRP levels in melanoma patients treated

with adjuvant interferon-alpha (EORTC 18952) and prognostic

value on treatment outcome. Melanoma Res 21(4):344–351

34. Guenterberg KD, Lesinski GB, Mundy-Bosse BL, Karpa VI,

Jaime-Ramirez AC, Wei L, Carson WE 3rd (2011) Enhanced

anti-tumor activity of interferon-alpha in SOCS1-deficient mice is

mediated by CD4(?) and CD8(?) T cells. Cancer Immunol

Immunother 60(9):1281–1288

35. Ascierto PA, Napolitano M, Celentano E, Simeone E, Gentilcore

G, Daponte A, Capone M, Caraco C, Calemma R, Beneduce G,

Cerrone M, De Rosa V, Palmieri G, Castello G, Kirkwood JM,

Marincola FM, Mozzillo N (2010) Regulatory T cell frequency in

patients with melanoma with different disease stage and course,

and modulating effects of high-dose interferon-alpha 2b treat-

ment. J Transl Med 8:76

36. Cesana GC, DeRaffele G, Cohen S, Moroziewicz D, Mitcham J,

Stoutenburg J, Cheung K, Hesdorffer C, Kim-Schulze S, Kauf-

man HL (2006) Characterization of CD4 ? CD25 ? regulatory

T cells in patients treated with high-dose interleukin-2 for met-

astatic melanoma or renal cell carcinoma. J Clin Oncol 24(7):

1169–1177

37. Elliott B, Scolyer RA, Suciu S, Lebecque S, Rimoldi D, Gugerli

O, Musat E, Sharma RN, Lienard D, Keilholz U, Testori A,

Eggermont A, MacKie R, Robert C, Cook M, Thompson JF,

Angevin E, Spatz A (2007) Long-term protective effect of mature

DC-LAMP ? dendritic cell accumulation in sentinel lymph

nodes containing micrometastatic melanoma. Clin Cancer Res

13(13):3825–3830

38. Yurkovetsky ZR, Kirkwood JM, Edington HD, Marrangoni AM,

Velikokhatnaya L, Winans MT, Gorelik E, Lokshin AE (2007)

Multiplex analysis of serum cytokines in melanoma patients

treated with interferon-alpha2b. Clin Cancer Res 13(8):

2422–2428

39. Worschech A, Kmieciak M, Knutson KL, Bear HD, Szalay AA,

Wang E, Marincola FM, Manjili MH (2008) Signatures associ-

ated with rejection or recurrence in HER-2/neu-positive mam-

mary tumors. Cancer Res 68(7):2436–2446

40. Huang FJ, Steeg PS, Price JE, Chiu WT, Chou PC, Xie K,

Sawaya R, Huang S (2008) Molecular basis for the critical role of

suppressor of cytokine signaling-1 in melanoma brain metastasis.

Cancer Res 68(23):9634–9642

41. Gogas H, Paterakis G, Frangia K, Bafaloukos D, Pectasides D,

Kalofonos HP, Loukopoulos D, Stavropoulou-Giokas C, Ioan-

novich J, Mihm MC Jr (2002) Lymphocyte subpopulations and

interleukin levels in high-risk melanoma patients treated with

high-dose interferon A-2B. Am J Clin Oncol 25(6):591–596

42. Simons DL, Lee G, Kirkwood JM, Lee PP (2011) Interferon

signaling patterns in peripheral blood lymphocytes may predict

clinical outcome after high-dose interferon therapy in melanoma

patients. J Transl Med 5(9):52–60

43. Malakhov MP, Kim KI, Malakhova OA, Jacobs BS, Borden EC,

Zhang DE (2003) High-throughput immunoblotting. Ubiquitin-

Cancer Immunol Immunother

123

like protein ISG15 modifies key regulators of signal transduction.

J Biol Chem 278(19):16608–16613

44. Shi HX, Yang K, Liu X, Liu XY, Wei B, Shan YF, Zhu LH,

Wang C (2010) Positive regulation of interferon regulatory factor

3 activation by Herc5 via ISG15 modification. Mol Cell Biol

30(10):2424–2436

45. Malakhova OA, Yan M, Malakhov MP, Yuan Y, Ritchie KJ, Kim

KI, Peterson LF, Shuai K, Zhang DE (2003) Protein ISGylation

modulates the JAK-STAT signaling pathway. Genes Dev 17(4):

455–460

46. Chen L, Borozan I, Sun J, Guindi M, Fischer S, Feld J, Anand N,

Heathcote J, Edwards AM, McGilvray ID (2010) Cell-type spe-

cific gene expression signature in liver underlies response to

interferon therapy in chronic hepatitis C infection. Gastroenter-

ology 138(3):1123–1133, e1121–e1123

Cancer Immunol Immunother

123