European Journal of Cancer (2014) 50, 1855– 1863

Available at www.sciencedirect.com

ScienceDirect journal homepage: www.ejcancer.com

Cytokine regulation by epidermal growth factor receptor inhibitors and epidermal growth factor receptor inhibitor associated skin toxicity in cancer patients q Tanusree Paul a, Christian Schumann b, Stefan Ru¨diger b, Stefan Boeck c, Volker Heinemann c, Volker Ka¨chele d, Michael Steffens e, Catharina Scholl e, Vivien Hichert e, Thomas Seufferlein d, Julia Carolin Stingl e,f,⇑ a

Institute of Pharmacology of Natural Products and Clinical Pharmacology, University of Ulm, Ulm, Germany Department of Internal Medicine II, University of Ulm, Ulm, Germany c Department of Internal Medicine III and Comprehensive Cancer Center, Ludwig-Maximilians-University of Munich, Munich, Germany d Department of Internal Medicine I, University of Ulm, Ulm, Germany e Research Division, Federal Institute of Drugs and Medical Devices (BfArM), Bonn, Germany f Faculty of Medicine, University of Bonn, Bonn, Germany b

Received 31 March 2014; accepted 26 April 2014 Available online 23 May 2014

KEYWORDS EGFR Erk 1/2 EGFRI Skin toxicity CXCL8/IL-8 Keratinocytes

q

Abstract Aim: Epidermal growth factor receptor inhibitor (EGFRI) induced skin toxicity has a prognostic value suggesting skin toxicity can be a useful surrogate marker for successful epidermal growth factor receptor (EGFR) inhibition, improved response and survival. But the pathophysiology of EGFRI induced skin toxicity remains elusive. However the involvement of immunological mechanisms has been speculated. This study investigates the possible underlying mechanism of EGFR inhibition associated cytokine production in keratinocytes as well as in patients after treatment with epidermal growth factor receptor inhibitors (EGFRIs). Methods: Normal human epidermal keratinocytes (NHEK) were incubated for 2 and 24 h with different concentrations of EGFRI (erlotinib) for Western blot analysis and cytokine expression analysis, respectively. Inhibition of EGFR, extracellular-signal-regulated kinase 1/2 (Erk 1/2) and c-Jun was examined by Western blot analysis. Cytokine concentrations were measured by enzyme-linked immunosorbent assay (ELISA) in the NHEK cell supernatant and also in the serum of 186 cancer patients who are followed up for EGFRI induced skin rash. Results: A significant inhibitory effect of EGFRI was seen on EGFR (Y845), Erk 1/2 and c-Jun in a dose dependent manner. Further downstream, increased CC-chemokine ligand 2 (CCL2),

This study was supported by a grant from the Wilhelm Sander Foundation (Grant No. 2008.017.1).

⇑ Corresponding author: Address: Federal Institute of Drugs and Medical Devices (BfArM), Kurt-Georg-Kiesinger-Allee 3, 53175 Bonn,

Germany. Tel.: +49 228 99307 3570; fax: +49 228 99307 3573. E-mail address: [email protected] (J.C. Stingl). http://dx.doi.org/10.1016/j.ejca.2014.04.026 0959-8049/Ó 2014 Elsevier Ltd. All rights reserved.

1856

T. Paul et al. / European Journal of Cancer 50 (2014) 1855–1863

CC-chemokine ligand 5 (CCL5) and decreased interleukin-8 (IL-8) or CXCL8 expression was observed in keratinocytes. In EGFRI treated patients, low levels of serum CXCL8 corresponding to stronger EGFR inhibition were associated with a higher grade of skin toxicity (p = 0.0016) and a prolonged overall survival (p = 0.018). Conclusions: The results obtained in this study indicate that EGFRI can regulate cytokines by modulating EGFR signalling pathway in keratinocytes. Moreover, serum levels of CXCL8 in EGFRI treated patients may be important for individual EGFRI induced skin toxicity and patient’s survival. Ó 2014 Elsevier Ltd. All rights reserved.

1. Introduction Targeting epidermal growth factor receptor (EGFR) with small molecule tyrosine kinase inhibitors or monoclonal antibodies has offered promising treatment success in several solid malignancies [1,2]. At present, two small-molecule tyrosine kinase inhibitors, erlotinib and gefitinib and two monoclonal antibodies, cetuximab and panitumumab have been approved for clinical practice [3]. Treatment with epidermal growth factor receptor inhibitors (EGFRIs) is associated with less systemic side-effects than with standard chemotherapeutic drugs. Despite these benefits, skin toxicity has been identified as class specific toxicity affecting the majority of patients with the increasing clinical use of EGFRIs [4]. Skin toxicities may cause decreased quality of life that might impair treatment adherence and clinical outcomes of the EGFRIs [4–6]. On the other hand, a positive association between the severity of EGFRI associated skin toxicity and treatment response, progression-free and overall survival has been consistently observed with all EGFRIs approved for clinical use [4,7]. Therefore developing skin toxicity has been used as biomarker for treatment and dosing with EGFR inhibitors [8]. The EGFRI induced skin rash is a very common toxic reaction because skin depends on EGFR signalling for normal function. Keratinocytes represent 95% of epidermal cells. Even though the main function of keratinocytes is to provide the structural integrity and barrier for the epidermis, they also play an important role in the initiation and continuation of skin inflammatory reactions and the pathogenesis of inflammatory skin diseases [9]. Abrogation of EGFR signalling has detrimental effects on keratinocytes by inducing growth arrest and apoptosis, decreasing cell migration, increasing cell attachment, cell differentiation and stimulating inflammatory chemokine expression, all of these incidences may lead to unique cutaneous manifestation [10]. Sustained epidermal growth factor receptor (EGFR) activation is associated with enhanced CXCL8 expression. In contrast, an opposite pattern has been observed by inhibition of EGFR activation [11]. EGFR activation leads to the sustained induction of mitogen-activated protein kinase (extracellular-signal-regulated kinase 1/2

(Erk 1/2)) pathway [12], which further modulates the EGFR-driven epidermal proliferation [13]. The EGFR signalling is also involved in the control of chemokine expression in epidermal keratinocytes via Erk 1/2 (extracellular-signal-regulated kinase 1/2) mechanisms [11]. Interleukin-8 (IL-8) or CXCL8 is a cytokine of the CXC chemokine family, it is a potent angiogenic factor. CXCL8 is the best characterised chemoattractant which is active in neutrophil cell recruitment in inflammation as well as in epithelial and endothelial cell proliferation [14]. CC-chemokine ligand 2 (CCL2) is also referred to as monocyte chemotactic protein 1 (MCP-1), belongs to the CC chemokine family. CCL2 recruits monocytes/ macrophages, memory T cells and dendritic cells to the sites of inflammation [15,16]. Moreover, epidermal CCL2 involvement was also found in the early response to skin injury or irritants and T-cell mediated skin disorders [14,17,18]. CC-chemokine ligand 5 (CCL5) is also known as RANTES (Regulated on Activation, Normal T cell Expressed and Secreted), and belongs to the CC chemokine family. CCL5 is a potent chemoattractant and plays a crucial role in T-cell responses [19]. The precise functions of CCL2 and CCL5 in tumour biology are still elusive. However, it has been shown that inflammatory cell infiltration occurs in the skin of EGFRI treated patients [20]. Inhibition of the EGFR signalling pathway up-regulates pro-inflammatory cytokines like CCL2, CCL5 and C-X-C motif chemokine 10 (CXCL10)/interferon gamma-inducible protein 10 (IP-10) in keratinocytes, while down-regulates in response of EGFR stimulation [21]. Although the aetiology of skin rash is still not exactly elucidated but the inhibition of EGFR tyrosine kinase activity seems to be crucial for the pathogenesis of skin toxicity [22]. Recently Lichtenberger et al. confirmed that under physiological conditions the EGFR signalling in the epidermal keratinocytes plays a pivotal role in controlling inflammation, barrier function and innate host defence and impairment of it leads to characteristic cutaneous inflammatory phenotype in both humans and mice [23]. In this study, we investigated the possible interaction between EGFR inhibition and the cytokine (CXCL8, CCL2 and CCL5) production after EGFR-tyrosine kinase inhibitor (erlotinib) treatment in keratinocytes.

T. Paul et al. / European Journal of Cancer 50 (2014) 1855–1863

In addition, we also studied the correlation between serum pro-inflammatory cytokine concentrations (CXCL8, CCL5 and CCL2) and clinical development of skin toxicity in EGFRI treated cancer patients. 2. Materials and methods

1857

CCL2 and CXCL8 levels were assessed by enzymelinked immunosorbent assay (ELISA) using the Duo set ELISA development kit for Human CCL5/RANTES (R&D Systems, Minneapolis, USA, cat. no. DY278), CCL2/MCP-1(R&D Systems, Minneapolis, USA, cat. no. DY279) and CXCL8/IL-8 (Life Technologies, cat. no.CHC1303) following manufacturer’s instructions.

2.1. Keratinocyte cell culture 2.4. Patient samples Normal human epidermal keratinocytes (NHEK) were obtained from Promocell (Heidelberg, Germany) and maintained in keratinocyte growth medium 2 (Promocell, Heidelberg, Germany). Keratinocyte growth medium 2 was supplemented with cytokines, growth factors and supplements according to the manufacturer’s instructions. 100 U/ml penicillin and 100 lg/ml streptomycin were also added to the medium. Keratinocytes were cultured at 37 °C in humidified atmosphere containing 5% CO2. 2.2. Immunoblotting and antibodies Keratinocytes were cultured to near confluence. After overnight serum starvation cells were stimulated with 4 nM EGF (PeproTech, United States of America (USA)) for 10 min followed by incubation with different concentrations of erlotinib (0.01, 0.1, 0.5, 1, 5, and 10 lM) (LC Laboratories, USA) for 2 h. Cells were lysed in RIPA (radioimmunoprecipitation assay) lysis buffer (50 mM Tris–HCl, 0.5% sodium deoxycholate, 0.1% sodium dodecyl sulphate (SDS), 150 mM NaCl, 5 mM ethylenediaminetetraacetic acid (EDTA), 10% glycerol, 2.5 mM MgCl2, protease and phosphatase inhibitors) and cell lysates were centrifuged at 14,000 rpm for 10 min at 4 °C. Protein samples were separated on 10% SDS–polyacrylamide gel and transferred to PVDF (polyvinylidene fluoride) membranes. For the immunoblotting following antibodies were used: anti-phospho-EGFR (Tyr845) (#2231), antiEGFR (#2232), anti-phospho-p44/42 MAPK (Erk1/2) (Thr202/Tyr204) (#9101), anti-p44/42 MAPK (Erk1/2) (#9102), anti-phospho-c-Jun (Ser63) (54B3) rabbit mAb (#2361) and anti-c-Jun (60A8) rabbit mAb (#9165) (Cell Signaling Technology, Inc). Mouse anti-b-actin antibody was obtained from Sigma–Aldrich (St. Louis, USA). Specific signals were visualised by an enhanced chemiluminescence detection reagent (Super Signal West Dura, Thermo Scientific). 2.3. Cytokine enzyme-linked immunosorbent assay (ELISA) Overnight serum starved keratinocytes were incubated with different concentrations of erlotinib for 24 h. After 24 h, cells were incubated with or without 4 nM EGF for 10 min. Then cell supernatants were collected and ELISA was performed. Circulating CCL5,

From September 2008 to August 2013, 186 patients treated with EGFRI alone or in combination with other chemotherapeutic drugs were included in this multicenter prospective study as described in a previous publication [3]. We followed up patients for 12 months after initiation of EGFRI therapy. Inclusion criteria were histologically confirmed cancer disease and first-time treatment with an EGFRI (monoclonal antibody or small molecule). Written informed consent was obtained from all participating subjects before inclusion. The study was reviewed by the ethical review boards of the Ulm University and the Ludwig-Maximilians-University of Munich. Severity of skin rash was rated according to the common toxicity criteria of the National Cancer Institute (NCI CTC version 3.0 criteria) [24]: grade1, macular or papular eruption or erythema without associated symptoms; grade 2, macular or papular eruption or erythema with pruritus or other associated symptoms and localised desquamation or other lesions covering 50% of body surface area; grade 3, severe, generalised erythroderma or macular, papular or vesicular eruption and desquamation covering P50% body surface area; grade 4, generalised exfoliative, ulcerative or bullous dermatitis. Appearance and severity of skin toxicity and diarrhoea (as control for an unspecific but not EGFRI mediated side-effect) were rated weekly (or biweekly) within the first 4 weeks after treatment initiation, and highest reported toxicity grade was used for analysis. In addition, pharmacological interventions against skin toxicity and/or diarrhoea were recorded. Patient characteristics are summarised in Table 1. The sample consists of patients diagnosed with non-small-cell lung cancer (NSCLC) (n = 110), pancreatic cancer (n = 44), colorectal cancer (n = 20) and squamous epithelium cancer (n = 12). The majority of the patients received oral erlotinib daily (n = 121); NSCLC patients were treated with 150 mg, pancreatic cancer patients received 100 mg erlotinib. 49 patients were treated with 250 mg/m2 cetuximab weekly (or biweekly) and one patient received 500 mg/m2 biweekly. Only a few patients were treated with 250 mg gefitinib daily (n = 10) or 6 mg/ kg panitumumab biweekly (n = 6). 2.5. Preparation of serum samples Serum samples were collected after 4 weeks from the beginning of treatment. Serum samples were obtained

1858

T. Paul et al. / European Journal of Cancer 50 (2014) 1855–1863

Table 1 Patient characteristics. Characteristics

3. Results Total (N = 186)

Applied epidermal growth factor receptor inhibitor (EGFRI) Erlotinib, n (%) 121 (65.1%) Gefitinib, n (%) 10 (5.4%) Cetuximab, n (%) 49 (26.3%) Panitumumab, n (%) 6 (3.2%) Disease Non-small-cell lung cancer (NSCLC), n (%) Pancreatic cancer, n (%) Colorectal cancer, n (%) Squamous epithelium cancer, n (%)

110 (59.1%) 44 (23.7%) 20 (10.8%) 12 (6.5%)

Age Median age, years (range)

69 (43–86)

Sex Male, n (%) Female, n (%)

121 (65.1%) 65 (34.9%)

Skin rash (grade) 0 1 2 3

38 70 68 10

Diarrhoea (grade) 0 1 2 3 4 Short bowel syndrome

116 (62.4%) 40 (21.5%) 22 (11.8%) 6 (3.2%) 1 (0.5%) 1 (0.5%)

(20.4%) (37.6%) (36.6%) (5.4%)

following our standard procedures: 7.5 ml blood was collected in a blood sampling tube (S-MonovetteÒ 7.5 ml, 92  15 mm) for collection of serum. The whole blood samples were taken in serum separator tubes without anticoagulant. The tubes were allowed to clot for 60 min at room temperature. After 60 min the blood sampling tubes were centrifuged at 1992g for 10 min at 4 °C. Supernatants (blood serum) were collected and immediately stored in aliquots at 20 °C. Samples were transferred to 80 °C for long-term storage. Serum CCL5, CXCL8 and CCL2 levels were successfully measured in 186 patients. 2.6. Statistical analysis Statistical analyses were performed using SPSS version 20 and R version 3.01. Differences of chemokine concentrations in keratinocytes with and without erlotinib treatment were tested using Student’s t-test. All data are expressed as mean ± standard deviation (SD). The association between the serum pro-inflammatory cytokine concentrations and the effect of EGFRI treatments in terms of skin rash were assessed using the Kruskal– Wallis test. Two-tailed p-values < 0.05 were considered as significant. Box-and-whisker plots provide information about the variability of the data: median, first and third quartile, and 1.5-times the interquartile range from the box. Survival curve analysis was performed as described previously [3].

3.1. Phosphorylation inhibition of EGFR by erlotinib in human keratinocytes To confirm the EGFRI’s (erlotinib) effects, a normal human keratinocyte cell line was used as an experimental model. We compared the differences between the effect of erlotinib on the phosphorylation of EGFR, Erk 1/2, Akt, STAT3 and c-Jun in keratinocytes in the presence and absence of 4 nM EGF. With stimulation with 4 nM EGF the phosphorylation of EGFR(Y845), Erk 1/2 and c-Jun was enhanced (Fig. 1A–C). However, a significant inhibitory effect of erlotinib was also seen on EGFR (Y845), Erk 1/2 and c-Jun in a dose dependent manner (Fig. 1A–C). No significant phosphorylation inhibition effect on Akt and STAT3 was observed (data not shown). Our results indicate that the EGFR tyrosine kinase inhibitor erlotinib can inhibit phosphorylation of the downstream molecules of EGFR like mitogen-activated protein kinase (Erk 1/2) and c-Jun in normal human keratinocytes. 3.2. Effects of EGFR inhibition on inflammatory cytokines in keratinocytes We observed differential production of cytokines in EGFRI (erlotinib) treated keratinocytes. The production of CXCL8 decreased at higher concentration of erlotinib (Fig. 2A). On the other hand, increased production of CCL2 and CCL5 was seen in a dose dependent manner after incubation with erlotinib (Fig. 2B, 2C). Each experiment was performed in triplicate. 3.3. Severity of skin rash and diarrhoea in EGFRI treated patients 186 patients were included in this study. 38 (20.4%) patients did not show any skin rash, 70 (37.6%) patients experienced grade 1, and 78 (41.9%) patients developed grade 2 or higher grade of skin toxicity (Table 1). Diarrhoea was less frequent than skin rash. A total of 116 (62.4%) patients did not develop diarrhoea, 40 (21.5%) patients developed grade 1 and 29 (15.6%) patients developed grade 2 or higher grade of diarrhoea and 1 patient was with short bowel syndrome (SBS) (Table 1). No significant associations between the prevalence of skin rash or diarrhoea with age, gender, tumour type or clinical stage were observed. 3.4. Association between pro-inflammatory cytokines, skin toxicity and diarrhoea We analysed the association between pro-inflammatory cytokine concentrations 4 weeks after the beginning of treatment and the development of adverse drug effects

T. Paul et al. / European Journal of Cancer 50 (2014) 1855–1863

1859

Fig. 1. Inhibitory effect of erlotinib (epidermal growth factor receptor-tyrosine kinase inhibitor) on EGFR signal transduction pathway. Overnight serum starved keratinocytes were treated with different concentrations of erlotinib for 2 h. DMSO was used as vehicle control. After 2 h 4 nM EGF was added for 10 min. Cell lysates were immunoblotted with following antibodies. (A) phospho-EGFR, anti-EGFR, (B) phospho-extracellularsignal-regulated kinase 1/2 (Erk 1/2), anti-Erk 1/2 and (C) phospho-c-Jun and anti-c-Jun. b-Actin was used as experimental control.

Fig. 2. Abrogation of epidermal growth factor receptor (EGFR) signalling by erlotinib leads to differential production of chemokine in keratinocyte supernatants. Chemokine release in the keratinocyte supernatants were assessed by enzyme-linked immunosorbent assay (ELISA). After overnight serum starvation cells were incubated with different concentrations of erlotinib for 24 h, then cells were incubated with or without 4 nM EGF for 10 min. After EGF treatment cell supernatants were collected for ELISA. Escalating doses of erlotinib (A) lowered the production of interleukin-8/CXCL8 but (B) enhanced the production of CC-chemokine ligand 2 (CCL2) and (C) CC-chemokine ligand 5 (CCL5).

observed after EGFRI treatment. Low serum CXCL8 levels were associated with higher severity of skin toxicity (p = 0.0016, Fig. 3, Table 2). On the other hand, high levels of serum CXCL8 were associated with higher severity of diarrhoea (p = 0.016). We also analysed the association between survival and pro-inflammatory cytokine concentrations. Kaplan–Meier survival

analysis showed that low serum CXCL8 was associated with longer survival in our patients (p = 0.018; Fig. 4). No association was observed between CCL2 and CCL5 concentrations and survival. In our study we observed a significant association between patients with skin toxicity compared with patients without skin toxicity using Kaplan–Meier survival curve analysis

1860

T. Paul et al. / European Journal of Cancer 50 (2014) 1855–1863

Fig. 3. Box-and-Whisker plot illustrating the spread of data between serum interleukin-8/CXCL8 and the grade of epidermal growth factor receptor inhibitor (EGFRI) induced skin toxicity. Serum CXCL8 levels after EGFRI treatment showed significant correlation with the severity of skin rash (p-value 0.0016). The width of each box plot is drawn proportional to the square root of the number of observations in the groups.

(log-rank p-value 2.4e 05; Fig. 5A). Additionally, we also found a significant correlation between the overall survival and the severity of skin toxicity (log-rank p-value 3.3e 05; Fig. 5B). On the other hand, we did not find any significant association between the presence and severity of diarrhoea with the survival of patients. Among 186 patients we presented survival data for 160 patients, the remaining patients are still under follow up.

4. Discussion It has been previously presented that cell cycling, motility and rescue from apoptosis are mainly maintained by EGFR mediated Erk 1/2 activation and can be induced by EGFR ligands [25]. In our study we also observed that EGFR inhibition by erlotinib leads to inhibition of Erk 1/2 and c-Jun (downstream effector of Erk 1/2) in keratinocytes. The impairment of EGFR and EGFR downstream molecules by erlotinib was associated with decreased CXCL8 and enhanced CCL2 and CCL5 production, similar effects have been

Fig. 4. Kaplan–Meier survival curves for overall survival with serum levels of interleukin-8/CXCL8 (log-rank p-value 0.018) from patients treated with epidermal growth factor receptor inhibitor (EGFRI). Patients with a CXCL8 concentration lower than 12 pg/ml had a median survival of 338 days, whereas patients with concentration larger than 12 pg/ml had only a chance of 50% to live beyond 215 days.

reported with the irreversible EGFR tyrosine kinase inhibitor PD168393 [11]. PD168393 is a potent, cell permeable, ATP (adenosine triphosphate)-competitive and selective EGFR inhibitor. It is used as an excellent anti-tumour agent in vitro. Mascia et al. indicated that the inhibition of EGFR signalling might exert anti-cancer activity by favoring the recruitment of inflammatory cells together with down regulation of CXCL8, which is an important growth factor for malignant epithelial cells [21]. CXCL8 or interleukin-8 (IL-8) is a multifunctional CXC chemokine that affects human neutrophil functions, chemotaxis, enzyme release and the expression of surface adhesion molecules. It has been reported to be expressed and secreted in many types of human cancer cells [26] and was identified to induce angiogenesis and metastasis in colorectal [27], Glioblastoma [28], prostate [29] and breast [30] carcinoma. Elevated serum levels of CXCL8 have been shown to be an independent negative prognostic factor for overall survival in soft tissue sarcoma patients [31] and could be a new predictive factor of disease progression and prognosis in malignant melanoma patients [32]. Luppi et al. reported that

Table 2 Severity of skin rash and serum cytokines (pg/ml). Skin rash (Grade)

N

Interleukin-8/CXCL8 [pg/ml] mean (SD)a

CC-chemokine ligand 2 (CCL2) [pg/ml] mean (SD)a

CC-chemokine ligand 5 (CCL5) [pg/ml] mean (SD)a

0 1 2 3 p-valueb

38 70 68 10

34.8 (45.0) 33.4 (55.3) 13.7 (15.4) 13.9 (18.3) 0.0016

172 (203) 213 (265) 190 (225) 239 (211) 0.59

4710 5241 5233 5075 0.78

a b

Mean and standard deviation (SD) for each chemokine (CXCL8, CCL2, CCL5) and the grade of skin rash. p-values are based on the null hypothesis of the Kruskal–Wallis test.

(1746) (872) (1149) (2072)

T. Paul et al. / European Journal of Cancer 50 (2014) 1855–1863

Fig. 5. Kaplan–Meier survival curves for overall survival, comparing (A) presence or absence of skin rash (log-rank p-value 2.4e 05) or (B) grade of skin rash (log-rank p-value 3.3e 05) after epidermal growth factor receptor inhibitor (EGFRI) treatment. The median survival of patients with skin rash was 336 days compared to 179 days for patients without any signs of skin rash. A further partition of patients with skin rash grade resulted in survival times of 236 days for patients with grade 1 and 361 days for patients with grade 2 skin rash.

CXCL8 induces tumour proliferation in NSCLC and this effect can be blocked by an EGFRI inhibitor [33]. Arenberg et al. showed, inhibition of CXCL8 can reduce tumourigenesis of human NSCLC in SCID mice [34]. Our in vitro study with keratinocytes as well as the clinical data in patients also indicated that pro-inflammatory cytokines were affected by EGFRI. Lower levels of serum CXCL8 were observed in patients after EGFRI treatment who had higher severity of skin rash. Thus, EGFR inhibition may have led to stronger inhibition of CXCL8 finally leading to more severe skin rash with lower serum concentrations of CXCL8. Decreased levels of serum CXCL8 also significantly correlated with longer survival of patients corresponding to the finding that less severe skin toxicity as a marker of less effective EGFR inhibition being associated with poorer survival. We also observed a clear and positive association

1861

between the presence and severity of skin rash and the overall survival of patients in our study like in a previous study [3]. Mostly patients who developed any grade of skin toxicity showed longer survival compared with patients who did not show any skin toxicity at all. Interestingly, HuMab-10F8, a neutralising human antibody against CXCL8/IL-8, has been shown to ameliorate the EGFRI associated dermatological adverse events in healthy volunteers [35]. These findings indicate that local injection of antibodies to CXCL8 may be a potential target for manipulating EGFRI induced skin rash without affecting the systemic efficacy of the treatment. We also observed differential production of two other CC chemokines RANTES/CCL5 and CCL2/MCP-1 in patients’ serum. Our in vitro studies with keratinocytes showed increased production of CCL2 and CCL5 upon abrogation of EGFR-tyrosine kinase with erlotinib. Mascia et al. showed that CCL5 and CCL2 were increased in mouse skin treated with EGFRI [21]. Pastore et al. also reported similar phenomena in keratinocytes [11]. The exact roles of CCL5 and CCL2 in cancer biology are still being understood. However, CCL5 is important to trigger and amplify the anti-tumour host response. CCL5 is a potent chemoattractant for T-cells, human eosinophils and basophils and plays an active role in recruiting leukocytes into inflammatory sites [36]. Therefore, production of CCL5 is important for inducing proper immune responses against tumours. On the other hand, it has been also reported that CCL5 (RANTES) is associated with cancer progression and metastasis by promoting tumour cell survival, proliferation and invasion [37]. Serum monocyte chemotactic protein (MCP-1) or CCL2 has been investigated in patients with ovarian cancer and increased levels of CCL2 were found to be correlated with a higher tumour grade [38]. CCL2 is also detected in tumour tissue and plays a critical role in tumourigenesis and metastasis which has been shown in prostate cancer [39]. Increased levels of tumour and serum CCL2 in breast and prostate cancer were associated with poor prognosis [40]. In our study we did not observe any significant association of CCL5 and CCL2 serum levels with clinical efficacy or EGFRI mediated toxicity. The discrepancy in the association of CCL5 and CCL2 levels and clinical efficacy or EGFRI mediated toxicity observed in our study with aforementioned results from literature can be explained by the design of our study. Our study group comprised of patients with heterogeneous tumours and tumour stages. Treatment with an EGFRI in a combination drug therapy resulted a heterogeneous group of patients with different drugs, dosage and treatment duration which limited sample size. As EGFRI induced skin toxicity is class specific, our main intention was to estimate the EGFRI

1862

T. Paul et al. / European Journal of Cancer 50 (2014) 1855–1863

mediated cytokine regulation regardless of above mentioned heterogeneity. In addition we analysed cytokines only from serum samples that were collected after EGFRI treatment. Evaluation of the cytokine concentrations before and after EGFRI treatment was not done. For the in vitro experiments only normal human keratinocytes were used. Keratinocytes from EGFRI treated patients were not available. These facts could limit our results and conclusions. Despite the heterogeneity of our patient population, a clear association between skin rash and survival as well as between CXCL8, skin rash and patients’ survival was observed independent of tumour type, therapy and stage, indicating that the individual extent of EGFR inhibition is a prognostic factor and rather a host phenomenon than a tumour specific phenomenon. We investigated the expression of cytokines upon EGFRI treatment in vitro and in vivo in order to elucidate the role of cytokines on clinical outcome and skin toxicity. To understand the exact underlying mechanism of EGFRI mediated cytokine regulation and skin toxicity, future work will be necessary. In conclusion, our data suggest that abrogation of EGFR signalling by erlotinib leads to inhibition of EGFR, Erk1/2 and its downstream effector c-Jun in keratinocytes. Impairment of the mitogen-activated protein kinase (MAPK) pathway may exert a profound effect on inflammatory process by increasing or decreasing certain chemokines. In our study, erlotinib lowered the production of CXCL8 in keratinocytes. Our clinical study with EGFRI treated patients also supported our findings in keratinocytes. Decreased levels of serum CXCL8 were associated with more severe skin rash indicating stronger EGFR inhibitory effect observed after 4 weeks of treatment. In addition, lower CXCL8 concentrations and severe skin rash were associated with a prolonged overall survival in our patient population. Furthermore, our findings may be crucial for further development of both treatment and research strategy of EGFRIs and EGFRI mediated skin toxicity. Since skin toxicity as a host effect of EGFR inhibition has a prognostic value, and we could show a positive correlation between serum CXCL8 concentrations and the severity of skin rash with longer survival, CXCL8 may be suited for the development of a functional biomarker for effective EGFR inhibition in the future.

Conflict of interest statement Stefan Boeck received research funding and honoraria for scientific presentations from Roche. Tanusree Paul is a fellow of International Graduate School in Molecular Medicine Ulm, Ulm University, Germany. The authors have no other relevant affiliations or financial involvement with any organisation or entity with a financial interest in or financial conflict with the subject

matter or materials discussed in the manuscript apart from those disclosed. References [1] Mendelsohn J, Baselga J. Status of epidermal growth factor receptor antagonists in the biology and treatment of cancer. J Clin Oncol 2003;21:2787–99. [2] Hynes NE, Lane HA. ERBB receptors and cancer: the complexity of targeted inhibitors. Nat Rev Cancer 2005;5:341–54. [3] Parmar S, Schumann C, Ru¨diger S, Boeck S, Heinemann V, Ka¨chele V, et al. Pharmacogenetic predictors for EGFR-inhibitor-associated skin toxicity. Pharmacogenomics J 2013;13:1–8. [4] Pere´z-Soler R, Saltz L. Cutaneous adverse effects with HER1/ EGFR-targeted agents: is there a silver lining? J Clin Oncol 2005;23:5235–46. [5] Li T, Perez-Soler R. Skin toxicities associated with epidermal growth factor receptor inhibitors. Target Oncol 2009;4:107–19. [6] Molinari E, De Quatrebarbes J, Andre´ T, Aractingi S. Cetuximab-induced acne. Dermatology 2005;211:330–3. [7] Wacker B, Nagrani T, Weinberg J, Witt K, Clark G, Cagnoni PJ. Correlation between development of rash and efficacy in patients treated with the epidermal growth factor receptor tyrosine kinase inhibitor erlotinib in two large phase III studies. Clin Cancer Res 2007;13:3913–21. [8] Brahmer JR, Lee JW, Traynor AM, Hidalgo MM, Kolesar JM, Siegfried JM, et al. Dosing to rash: a phase II trial of the first-line erlotinib for patients with advanced non-small-cell lung cancer an Eastern Cooperative Oncology Group Study (E3503). Eur J Cancer 2014;50:302–8. [9] Miodovnik M, Koren R, Ziv E, Ravid A. The inflammatory response of keratinocytes and its modulation by vitamin D: the role of MAPK signaling pathways. J Cell Physiol 2012;227: 2175–83. [10] Kari C, Chan TO, De Quadros MR. Targeting the epidermal growth factor receptor in cancer: apoptosis takes center stage. Cancer Res 2003;63:1–5. [11] Pastore S, Mascia F, Mariotti F, Dattilo C, Mariani V, Girolomoni G. ERK1/2 regulates epidermal chemokine expression and skin inflammation. J Immunol 2005;174:5047–56. [12] Pearson G, Robinson F, Beers Gibson T, Xu BE, Karandikar M, Berman K, et al. Mitogen-activated protein (MAP) kinase pathways: regulation and physiological functions. Endocr Rev 2001;22:153–83. [13] Iordanov MS, Choi RJ, Ryabinina OP, Dinh T, Bright RK, Magun BE. The UV (Ribotoxic) stress response of human keratinocytes involves the unexpected uncoupling of the Rasextracellular signal-regulated kinase signaling cascade from the activated epidermal growth factor receptor. Mol Cell Biol 2002;22:5380–94. [14] Gillitzer R, Goebeler M. Chemokines in cutaneous wound healing. J Leukoc Biol 2001;69:513–21. [15] Carr MW, Roth SJ, Luther E, Rose SS, Springer TA. Monocyte chemoattractant protein 1 acts as a T-lymphocyte chemoattractant. Proc Natl Acad Sci 1994;91:3652–6. [16] Xu LL, Warren MK, Rose WL, Gong W, Wang JM. Human recombinant monocyte chemotactic protein and other C–C chemokines bind and induce directional migration of dendritic cells in vitro. J Leukoc Biol 1996;60:365–71. [17] Giustizieri ML, Mascia F, Frezzolini A, De Pita` O, Chinni LM, Giannetti A, et al. Keratinocytes from patients with atopic dermatitis and psoriasis show a distinct chemokine production profile in response to T cell-derived cytokines. J Allergy Clin Immunol 2001;107:871–7. [18] Kim BH, Lee YS, Kang KS. The mechanism of retinol-induced irritation and its application to anti-irritant development. Toxicol Lett 2003;146:65–73.

T. Paul et al. / European Journal of Cancer 50 (2014) 1855–1863 [19] Smeets A, Brouwers B, Hatse S, Laenen A, Paridaens R, Floris G, et al. Circulating CCL5 levels in patients with breast cancer: is there a correlation with lymph node metastasis? ISRN Immunol 2013;2013:1–5. [20] Lacouture ME. Mechanisms of cutaneous toxicities to EGFR inhibitors. Nat Rev Cancer 2006;6:803–12. [21] Mascia F, Mariani V, Girolomoni G, Pastore S. Blockade of the EGF receptor induces a deranged chemokine expression in keratinocytes leading to enhanced skin inflammation. Am J Pathol 2003;163:303–12. [22] Segaert S, Van Cutsem E. Clinical signs, pathophysiology and management of skin toxicity during therapy with epidermal growth factor receptor inhibitors. Ann Oncol 2005;16:1425–33. [23] Lichtenberger BM, Gerber PA, Holcmann M, Buhren BA, Amberg N, Smolle V, et al. Epidermal EGFR controls cutaneous host defense and prevents inflammation. Sci Transl Med 2013;5:199ra111. [24] Cancer Therapy Evaluation Program (CTEP). Common Terminology Criteria for Adverse Events v3.0 (CTCAE). U.S. Department of Health and Human Services, National Cancer Institute (NCI); 2006. p. 0–71. [25] Jost M, Huggett TM, Kari C, Boise LH, Rodeck U. Epidermal growth factor receptor-dependent control of keratinocyte survival and Bcl-xL expression through a MEK-dependent pathway. J Biol Chem 2001;276:6320–6. [26] Xie K. Interleukin-8 and human cancer biology. Cytokine Growth Factor Rev 2001;12:375–91. [27] Lee YS, Choi I, Ning Y, Kim NY, Khatchadourian V, Yang D, et al. Interleukin-8 and its receptor CXCR2 in the tumour microenvironment promote colon cancer growth, progression and metastasis. Br J Cancer 2012;106:1833–41. [28] Brat DJ, Bellail AC, Van Meir EG. The role of interleukin-8 and its receptors in gliomagenesis and tumoral angiogenesis. Neurooncol 2005;7:122–33. [29] Wilson C, Wilson T, Johnston PG, Longley DB, Waugh DJJ. Interleukin-8 signaling attenuates TRAIL- and chemotherapyinduced apoptosis through transcriptional regulation of c-FLIP in prostate cancer cells. Mol Cancer Ther 2008;7:2649–61.

1863

[30] Green AR, Green VL, White MC, Speirs V. Expression of cytokine messenger RNA in normal and neoplastic human breast tissue: identification of interleukin-8 as a potential regulatory factor in breast tumours. Int J Cancer 1997;72:937–41. [31] Rutkowski P, Kaminska J, Kowalska M, Ruka W, Steffen J. Cytokine serum levels in soft tissue sarcoma patients: correlations with clinico-pathological features and prognosis. Int J Cancer 2002;100:463–71. [32] Ugurel S, Rappl G, Tilgen W, Reinhold U. Increased serum concentration of angiogenic factors in malignant melanoma patients correlates with tumor progression and survival. J Clin Oncol 2001;19:577–83. [33] Luppi F, Longo AM, de Boer WI, Rabe KF, Hiemstra PS. Interleukin-8 stimulates cell proliferation in non-small cell lung cancer through epidermal growth factor receptor transactivation. Lung Cancer 2007;56:25–33. [34] Arenberg DA, Kunkel SL, Polverini PJ, Glass M, Burdick MD, Strieter RM. Inhibition of interleukin-8 reduces tumorigenesis of human non-small cell lung cancer in SCID mice. J Clin Invest 1996;97:2792–802. [35] Bangsgaard N, Houtkamp M, Schuurhuis DH, Parren PWHI, Baadsgaard O, Niessen HWM, et al. Neutralization of IL-8 prevents the induction of dermatologic adverse events associated with the inhibition of epidermal growth factor receptor. PLoS One 2012;7:e39706. [36] Fakieh RA, Taha S, Bakhiet M. The role of IL-18, CCL5 and their receptors as diagnostic markers in breast cancer. GERF Bull Biosci 2012;3:1–8. [37] Lapteva N, Huang XF. CCL5 as an adjuvant for cancer immunotherapy. Expert Opin Biol Ther 2010;10:725–33. [38] Hefler L, Tempfer C, Heinze G, Mayerhofer K, Breitenecker G, Leodolter S, et al. Monocyte chemoattractant protein-1 serum levels in ovarian cancer patients. Br J Cancer 1999;81:855–9. [39] Craig MJ, Loberg RD. CCL2 (Monocyte Chemoattractant Protein-1) in cancer bone metastases. Cancer Metastasis Rev 2006;25:611–9. [40] Soria G, Ben-Baruch A. The inflammatory chemokines CCL2 and CCL5 in breast cancer. Cancer Lett 2008;267:271–85.