Fc gamma receptor IIIa polymorphisms in advanced colorectal cancer patients correlated with response to anti-EGFR antibodies and clinical outcome

Background Anti-EGFR monoclonal antibodies have shown efficacy in the treatment of metastatic colorectal cancer (mCRC). One of the mechanism is the antibody-dependent cell-mediated cytotoxicity (ADCC) in which Fc region of the antibody binds to the Fc gamma receptors (FcγR) expressed by immune cells. The present study investigated the association between single nucleotide polymorphisms of FcγRIIa and FcγRIIIa and clinical outcome in mCRC treated with anti-EGFR antibodies. Methods Seventy-four consecutive patients with mCRC were analyzed. The genotypes for FcγRIIa-131 histidine (H)/arginine (R), FcγRIIIa-158 valine (V)/phenylanaline (F) polymorphisms were evaluated by directly sequencing. Multiplex allele-specific polymerase chain reaction was performed for FcγRIIIa-158 valine (V)/phenylanaline (F). Correlations between FcγR polymorphisms, baseline patient and tumor features were studied by contingency tables and the chi-square test. The Kaplan-Meier product limit method was applied to the progression-free survival (PFS) curves. Univariate analysis was performed with the log-rank test. Cox proportional-hazards regression was used to analyze the effect of multiple risk factors on PFS. Results FcγRIIIa polymorphisms were significantly associated with response to anti-EGFR-based therapy in 49 patients with kras wt tumors (p=0.035). There was not association with response for FcγRIIa polymorphisms. Furthermore, obtained results suggested that prognosis is particularly unfavorable for patients carrying the FcγRIIIa-158F/F genotype (median PFS V/V, V/F, F/F: 18.2 vs 17.3 vs 9.4 months). No prognostic ability was identified for FcγRIIa polymorphisms. Conclusions In mCRC patients the presence of FcγRIIIa-F can predict resistance to anti-EGFR therapy and unfavorable prognosis.


Background
Metastatic colorectal cancer (mCRC) is the second most common cause of cancer death in the Western world accounting for 40-50% of newly diagnosed patients [1]. Despite therapeutic advances, the prognosis for patients with mCRC remains poor. However, the addition of drugs such as irinotecan and oxaliplatin to 5-fluorouracil (5-FU) has almost doubled the median survival from 12 months to 21 months [2]. Monoclonal antibodies (mAbs) binding to the vascular endothelial growth factor (VEGF) (bevacizumab) or the epidermal growth factor receptor (EGFR) (cetuximab and panitumumab) have shown efficacy in the treatment of mCRC increasing the life expectancy of patients by more than 2 years [3]. While bevacizumab is administered in combination with chemotherapy as a first-line treatment, anti-EGFR mAbs find place in later-line treatments.
Cetuximab is an IgG1a chimeric mAb while panitumumab is a fully human IgG2 mAb; they bind to EGFR and block the binding of its natural ligands, preventing ligand dependent homodimerization and activation of intracellular cascades that control cellular proliferation, adhesion, angiogenesis, and apoptosis. Anti-EGFR mAbs have proven to be effective in combination with chemotherapy or as single agents for treatment of mCRC [3]. Recent evidences showed that mCRC responds differently to EGFR-targeted agents on genetic basis that involve also the EGFR downstream effectors (i.e. kras, braf, PIK3CA and PTEN) [4]. Although largely unexplored, monoclonal antibodies also induce antibody-dependent cell-mediated cytotoxicity (ADCC) [5][6][7][8]. ADCC is induced through the interaction of the Fc region of the mAb with the Fc gamma receptor (FcγR) expressed by effector cells (i.e. natural killer-NK-lymphocytes, monocytes/macrophages). Polymorphisms have been demonstrated on genes encoding for the activating receptors FcγRIIa (CD32, mainly expressed on macrophages) and FcγRIIIa (CD16, expressed on NK cells and macrophages) [9], affecting their affinity to human IgG: a histidine (H)/arginine (R) polymorphism at position 131 for FcγRIIa and a valine (V)/phenylalanine (F) polymorphism at position 158 for FcγRIIIa. Based on the different affinities, patients harboring FcγRIIa-131H/H and FcγRIIIa-4 158V/V genotypes would be expected to mediate a more efficient ADCC antitumor response. Clinical studies utilizing rituximab in the treatment of B-cell non-Hodgkin's lymphoma have shown that FcγRIIa-131H/H and FcγRIIIa-158V/V genotypes were associated with better clinical outcome [10,11]. Patients with 158V/V and/or 131 H/H had a significantly higher response rate than patients without either genotype (59% vs 18%). The progression-free survival (PFS) estimate of patients with 158V/V and/or 131H/H allele was also significantly longer, with median PFS of 445 and 140 days for the two groups, respectively [11]. Nevertheless it was shown that when CT is added to Rituximab the predictive value of FCGR polymorphisms was lost probably due to the high efficacy of CT [12]. In trastuzumab-treated metastatic breast cancer, ADCC analysis showed that the combination of 158 V/V and/or 131 H/H had a significantly higher trastuzumab-mediated cytotoxicity than other genotypes in addition to higher response rate and a longer PFS [13]. Contrasting results have been reported on the role of FcγR polymorphisms in mCRC [14,15]. Recently, it was described that FcγRIIa-131H/H and FcγRIIIa-158F/F polymorphisms associated with better PFS in a series of EGFR-expressing mCRC patients treated with single-agent cetuximab [14]. Conversely, Bibeau et al. demonstrated a favourable effect on PFS only for the FcγRIIIA-158V/V genotype unrelated to the kras status [15]. The goal of our study was to explore the association between FcγRIIa and FcγRIIIa polymorphisms and the outcome of mCRC patients treated with anti-EGFR-based therapies (cetuximab and panitumumab).

Patient management and follow-up
Seventy-four stage IV CRC patients were studied at the Division of Abdominal Medical Oncology of the National Cancer Institute (Naples, Italy) from May 2007 to May 2009. Patients were eligible after specific discussion on the study. Informed consent from each patient was sought. The protocol was conducted according to a protocol approved by the institutional review board/independent ethics committee. Patients were routinely characterized for kras mutational status [16]. All patients underwent to sequential standard treatments based on chemotherapy and/or biologic therapies (bevacizumab, cetuximab, panitumumab). First and second-line chemotherapy (CT) included the association of fluoropyrimidines (capecitabine or 5-fluoruracile) with oxaliplatin or irinotecan. The chemotherapy regimen was based on patient's performance status, extent of disease, comorbidities, previous treatments and individual preferences. Some selected patients underwent pulmonary and/or liver metastasectomies as established in a multidisciplinary team discussion. Bone metastases were treated with palliative radiotherapy. Patients features are shown in Table 1. Cetuximab or panitumumab were administered only in patients with kras wilde-type (wt) tumors. All patients underwent first-line chemotherapy, 54 patients (72.9%) received a second-line chemotherapy and 22 (29.7%) a third-line chemotherapy. Eight patients received palliative radiotherapy. Eleven patients with advanced disease underwent to palliative resection of primary colonic tumor. Metastasectomies before or after chemotherapy were performed in 26 patients. Total body computed tomography scan and CEA monitoring were done every three   No  months. The response to therapy was evaluated by RECIST criteria. Patients with target metastatic lesions restaged at the Radiology Unit were considered for response evaluation. Complete response (CR) was defined as complete disappearance of all detectable evidence of disease on total body computed tomography. Partial response (PR) was defined as at least a 30% decrease in the sum of diameters of target lesions. Stable disease (SD) was defined as everything between 30% decrease and 20% growth of tumor size. Progressive disease (PD) was defined as at least a 20% increase in the sum of diameters of target lesions. Two patients were lost at follow-up.

Analysis of FcγRIIa-H131R, FcγRIIIa-V158F polymorphisms
Genomic DNA was extracted from white blood cells (WBCs) using a DNA extraction kit.(Qiagen,Valencia, CA) and stored at −20°. FcγRIIa genotyping was performed on genomic DNA by polymerase chain reaction (PCR) method adapted from a previously established protocol [17]. Briefly, PCR amplification was performed in 50 μl reaction mixture containing 100 ng genomic DNA, 0.4 mM of each primer, 0.4 mM dNTPs, 20 mM Tris-HCl, pH 9.0, 100 mM KCl, 0.1 mM EDTA, 1.0 mM DTT, 0.5% tween 20, and 1 U Taq DNA polymerase. The program was performed in the thermal Cycler 2770 by Applied Biosystems and consisted of an initial denaturation step at 95°C for 5 0 , followed by 36 cycles of 95°C for 30 seconds, 56°C for 40 seconds, 72°C for 40 seconds and a final elongation step at 72°C for 10 minutes. The primers used for PCR amplification were forward primer 5 0 -GGAGAAACCATCATGCTGAG-3 0 and reverse primer 5'-CAATTTTGCTGCTATGGGC-3 0 . The resulting PCR product (289bp) was purified with the Montage SEQ 96 Sequencing Reaction Cleanup Kits (Millipore) and prepared to sequence through a second PCR reaction using Big Dye Terminator v3. of the region of interest (96°C for 1 minutes, 25 cycle of 96°C for10 seconds, 56°C for 5 seconds and 60°C for 2 minutes). PCR product was purified with Montage SEQ 96 Sequencing Reaction Cleanup Kits (Millipore) and direct sequencing was run with Applied Biosystems3130 Genetic Analyzers (Figure 1). For FcγRIIIa-V158F polymorphism, allele-specific PCR method was followed. Briefly, 100 ng of genomic DNA was amplified using allele-specific common forward primer 5 0 -TCCAAAAGCCACACTCAAAGAC-3 0 and reverse primer 5 0 -CTGAAGACACATTTTTACTCCCAAAC-3 0 . PCR amplification was performed in 25 μl reaction mixture containing 100 ng genomic DNA, 0.3 mM of each primer, 0.2 mM dNTPs, 20 mM Tris-HCl, pH 9.0, 100 mM KCl, 0.1 mM EDTA, 1.0 mM DTT, 0.5% tween 20, and 1 U Taq DNA polymerase. The program was performed in the thermal Cycler 2770 by Applied Biosystems and consisted of an initial denaturation step at 95°C for 5 0 , followed by 35 cycles of 94°C for 30 seconds, 64°C for 30 seconds, 72°C for 30 seconds and a final elongation step at 72°C for 10 minutes. Three DNA samples previously sequenced FcgRIIIa-V/V158, FcgRIIIa-V/F158, FcgRIIIa-F/F158 were run in all reactions (Figure 1). The reaction products were run on 3% ethidium bromide-stained agarose gel. Seventy-three base pair PCR fragment either positive for valine (V) or F allele was visualized under UV light as reported previously [18]. To confirm FcγRIIIa genotype automatic sequencing was performed using forward pri-

Statistical analyses and data presentation
Associations between FcγR polymorphisms and clinical pathologic variables were evaluated by χ 2 test. p < 0.05 was considered statistically significant. Genotype data for FcγR polymorphisms and clinicpathological variables were retrospectively collected and associated with response to anti-EGFR-based therapy by χ 2 test with level of significance set at p < 0.05. Progression-free survival (PFS) was defined as the time elapsed from the treatment initiation and tumor progression or death from any cause. The Kaplan-Meier product limit method was applied to graph PFS. Univariate analysis was done with the log-rank test. Cox proportional hazards regression was used to analyze the effect of several risk factors on PFS. Risk factors (covariates) were: age, sex, grading, response to I°line chemotherapy, FcγR polymorphisms. Ninety-five percent confidence intervals of hazard ratios were also reported.
No attempts were done to analyze overall survival because of low events. Seventy-two patients were analyzed since two were lost at follow-up. Statistical analysis was performed using the MedCalc W 9.3.7.0 and Excel software.

Characteristics of patients and tumors
Seventy-four patients seen from May 2007 to May 2009 were studied for the FcγRIIa and FcγRIIIa polymorphisms through direct sequencing and allele specific PCR as reported in Figure 1. Patients features are detailed in Table 1. Median age was 65 years. Genders were equally represented. Twenty-six tumors originated in the rectum 51.4% of patients had highgrade (G3) disease. The majority of lesions presented with a pT3 extent of invasion at diagnosis and 22 presented with pN+ disease. Fifty patients presented with stage IV disease, 16 with stage III and 8 with stage I/II. The majority of tumors (90.5%) did not  (Table 1).
FcγRIIIa but not FcγRIIa polymorphisms were significantly associated with response to anti-EGFR-based therapy in kras wt tumors Fifty patients were treated with anti-EGFR-based therapy and forty nine were evaluable for clinical response (according to RECIST criteria) and PFS. Forty-five patients were treated with cetuximab, five with panitumumab. The genotypic frequencies of FcgRIIIA and FcgRIIa detected within the analyzed population were 36% VV, 54% VF ,10% FF and 36% HH, 56% HR, 8% RR, respectively. The χ 2 test showed that there were no significant differences in the genotype frequencies (p=0.109 for V158F; p=0.183 for FcgRIIa) between patients and healthy controls. The genotypic distributions were in Hardy-Weinberg Equilibrium.
Objective responses according to FcγR polymorphisms were shown in Table 2. FcγRIIIa but not FcγRIIa polymorphisms were significantly associated with response to anti-EGFR-based therapy in kras wt tumors (p=0.035). The mean number of anti-EGFR therapy cycles were 15 (range: 5-27) considering panitumumab as single administration every two weeks (one adminis-tration=1 cycle) and cetuximab weekly (two administra-tions=1 cycle). To evaluate skin toxicity and its predictive role and correlation with FcgR polymorphisms (21), the skin related toxicity was evaluated versus the clinical response. A significant correlation was identified ( p= 0.005) between skin toxicity and clinical response ( Table 3) while no significant correlation was identified between skin toxicity and the genotype distribution (Table 4).

FcγR polymorphisms predict PFS in mCRC patients treated with anti-EGFR mAbs
The anti-EGFR treated patients were analyzed for PFS.  (Figure 2).

Discussion
In this manuscript the value of the FcγRIIa-FcγRIIIa polymorphisms was retrospectively correlated to the efficacy of anti-EGFR therapy in mCRC. FcγRIIIa polymorphisms were significantly associated with response to anti-EGFRbased therapy in 49 valuable patients with kras wt tumors. The results suggested that prognosis is particularly unfavorable for patients expressing the FcγRIIIa-158F/F genotype versus patients carrying a V allele (the FcγRIIIa-158F/V or the FcγRIIIa-158V/V genotypes). On this issue, conflicting results were previously described: Bibeau et al. showed a statistically significant difference in PFS in 69 mCRC patients treated with cetuximab plus irinotecan carrying the the FcγRIIIa-158V/V genotype compared to other combinations expressing an F allele while FcγRIIa polymorphisms did not affect prognosis [15]. Conversely, in a series of 39 EGFR-expressing mCRC patients treated with single-agent cetuximab, Zhang et al. [14] found that FcγRIIa-H131R and FcγRIIIa-V158F polymorphisms were independently associated with better PFS. However, against their hypothesis, FcγRIIIa-158V/V genotype was associated with more unfavorable clinical outcome. The authors suggest that variants of human IgG1-binding sites p=0.183 for FcgRIIa). Moreover we can also speculate that, although the studied population represent 74 consecutive mCRC patients coming to our observation, they all showed a good Performance Status (0-1) further validated by the neoadjuvant treatment for 23 patients (Fluopirimidine/ Irinotecan/ Bevacizumab) before hepatic metastasectomy.
The role of ADCC induced by EGFR-specific mAbs may prevent tumor outgrowth or metastasis in vivo, even in cancers insensitive to EGFR signaling inhibition [19]. In fact, evidences accumulated on a complex patients evaluation including the kras status but also EGF/EGFR polymorphisms and downstream pathway mutations. To date, unless a large number of patients have been treated with mAbs there are still two crucial issues: i) a small percent of kras mutated patients respond to cetuximab therapy [3,4] and, ii) although kras wt tumors are potentially sensitive to EGFR-targeted mAbs, not all respond to anti-EGFR therapy for multiple target expression, amplification or mutations downstream [4]. Blockade of signal transduction may not be the only mechanism of action  mediating clinical benefit of mAb-treated patients with colorectal cancer [20]. Cell-dependent lysis of target cells mediated by mAbs in vitro and in animal models is a crucial mechanism of action regulated by multiple factors (i.e. FcγR on Dendritic Cells, T-helper system, Tregs, B-cells, NK-cells inhibitory proteins, cytokines, etc.). Currently, we are studying the correlation between specific FCγR polymorphisms to in vitro ADCC efficacy (Trotta et al., manuscript in preparation).

Conclusions
Although the exact role of FcγRIIIa-V158F polymorphism and anti-EGFR therapy require more basic studies, the presence of one allele F of the FcγRIIIa in mCRC patients predicted poor response to anti-EGFR-based therapy and worsen the prognosis.