Abstract

Background: Cetuximab improves activity of chemotherapy in metastatic colorectal cancer (mCRC). Gene copy number (GCN) of epidermal growth factor receptor (EGFR) has been suggested to be a predictive factor of response to cetuximab in patients (pts) with mCRC; on the contrary, K-ras mutation has been associated with cetuximab resistance.

Patients and methods: We have conducted a phase II study with cetuximab administered weekly for 3 weeks as single agent and then with 5-fluorouracil and radiation therapy as neo-adjuvant treatment for locally advanced rectal cancer (LARC). EGFR immunohistochemistry expression, EGFR GCN and K-ras mutation were evaluated on diagnostic tumor biopsy. Dworak's tumor regression grade (TRG) was evaluated on surgical specimens.

Results: Forty pts have been treated; 39 pts are assessable. TRG 3 and 4 were achieved in nine (23.1%) and three pts (7.7%) respectively; TRG 3–4 rate was 55% and 5.3% in case of high and low GCN, respectively (P 0.0016). Pts with K-ras mutated tumors had lower rate of high TRG: 11% versus 36.7% (P 0.12). In pts with wild-type K-ras, TRG 3–4 rate was 58.8% versus 7.7% in case of high or low GCN, respectively (P 0.0012).

Conclusions: In pts with LARC, EGFR GCN is predictive of high TRG to cetuximab plus 5-FU radiotherapy. Moreover, our data suggest that a wild-type K-ras associated with a high EGFR GCN can predict sensitivity to cetuximab-based treatment.

introduction

EGFR expression evaluated by immunohistochemistry (IHC) is associated with a higher risk of disease recurrence and death in colon cancer treated with adjuvant chemotherapy [1]. In locally advanced rectal cancer treated with preoperative chemoradiotherapy, baseline EGFR (IHC) expression predicts a poor tumor downstaging [2] and is also an independent prognostic factor for local recurrence after preoperative chemoradiation and surgery [3–5]. We have previously reported that, in locally advanced rectal cancer, EGFR (IHC) expression on residual tumor after preoperative chemoradiation is an independent poor prognostic factor for disease recurrence [6].

On the basis of these data, EGFR can be a potentially interesting therapeutic target in colorectal cancer. Clinical trials have shown that cetuximab, an anti EGFR chimeric IgG1 mAb, improves activity and efficacy of chemotherapy in metastatic colorectal cancer (mCRC) in first, second or third line of treatment; unfortunately, only a fraction of patients (pts) achieves an objective response and the median progression-free survival does not exceed 9 months [7–10].

We and others have shown that preoperative cetuximab in combination with fluoropyrimidines and radiation therapy in locally advanced rectal cancer is feasible; however, the rate of pathological complete remission is rather low (5%–8%) [11–13].

The identification of predictive factors for response to cetuximab and other EGFR-tyrosine kinase inhibitors is therefore of primary interest for the pts, the medical community and the health providers.

Unfortunately, EGFR protein expression does not predict cetuximab activity in colorectal cancer [9, 14]. On the contrary, gene copy number (GCN) of EGFR seems to be a predictive factor for response to cetuximab and panitumumab in pts with mCRC [15–17]. Moreover, K-ras mutation has been shown to be an independent predictive factor of poor outcome in pts with mCRC treated with cetuximab [18–22].

We have previously reported the results of a phase II trial of neo-adjuvant cetuximab + radiation therapy + 5-fluorouracil (5-FU) in pts with locally advanced rectal cancer; primary aim of this trial was the pathological response [13]. As correlative study, the trial included the analysis of EGFR IHC expression, EGFR GCN and K-ras mutation on diagnostic biopsy. In the present paper, we report the relationship between EGFR expression, EGFR GCN and K-ras mutation evaluated on diagnostic biopsy and the pathological response according to Dworak.

patients and methods

patients

Pts with histologically confirmed diagnosis of locally advanced (cT3–T4; N0–1) rectal carcinoma (within 15 cm from anal verge) and no evidence of distant metastases were accrued in a phase II clinical trial of neo-adjuvant treatment. Clinical stage was determined by transrectal ultrasound probe. The trial was conducted at Department of Oncology and Hematology, University Hospital, Modena, Italy, and at Division of Medical Oncology, National Institute for Cancer Research, Genoa, Italy. The trial as well as the correlative studies were approved by the Ethical Committees of the two institutions.

treatment

Cetuximab was administered as single agent for three times at the doses of 400 mg/m2 (loading dose), then 250 mg/m2 weekly, followed by cetuximab 250 mg/m2 weekly plus 5-FU (225 mg/m2/day as i.v. continuous infusion 7 days a week for 5 weeks) concurrently to radiation therapy as neo-adjuvant treatment. Radiotherapy was delivered with 15–18 MV photon beams, at 1.8–2 Gy/fraction up to 50–50, 4 Gy in 25–28 daily fractions for 5 days a week, depending on treatment policy of the center.

tissue samples and pathology assessment

All pts underwent tumor biopsy for diagnostic purpose before starting the treatment. Several 4-μg thick sections were obtained from a representative formalin-fixed and paraffin-embedded block. At the time of surgery, pathologic evaluation in the resection specimens included tumor–node–metastasis categories, stage grouping, number of examined/involved lymph nodes and tumor differentiation. Tumor regression was semiquantitatively determined by the amount of viable tumor versus the amount of fibrosis, as described by Dworak et al. [24] and validated by Rodel et al. [25]. Tumor regression grade (TRG) 0 was defined as no regression; TRG1, minor regression (dominant tumor with fibrosis in ≤25% of the tumor mass); TRG2, moderate regression (dominant tumor with fibrosis in 26%–50% of the tumor mass); TRG3, good regression (>50% tumor regression) and TRG4, total regression (no viable tumor cells, only fibrotic mass).

EGFR expression IHC

IHC was carried out in the local laboratory using the avitin–biotin complex technique. Briefly, 5-μm thick sections were cut from formalin-fixed paraffin-embedded tissue, deparaffinized and rehydratated. The IHC procedure was carried out by an automated immunostainer (Benchmark, Ventana, Tucson, AZ). A tumor sample was categorized as positive when the percentage of EGFR-positive cells was ≥1%. All slides were reviewed by two observers (LL and BG) and interobserver disagreement was discussed in an interinstitutional meeting. All samples had enough diagnostic material for molecular determinations.

EGFR FISH

The technical procedures were carried out at the local laboratories of the two participating institutions and were predefined by an agreement among the pathologists. FISH studies were carried out on selected sections of paraffin-embedded tissue areas, containing representative malignant cells, using the LSI EGFR SpectrumOrange/CEP7SpectrumGreen probe (Vysis Inc., Downer's Grove, IL). Tissue sections of 3 μm were placed on electrostatically charged slides, air dried and baked overnight at 56°C. The slides were dewaxed in xylene for 2 × 15 min, immersed in 100% ethanol for 2 × 5 min and in 95% ethanol for 2 × 5 min. Air-dried tissue sections were treated with a Paraffin Pretreatment Kit (Vysis Inc). The slides were briefly incubated in 0.2 mol/l HCl for 20 min, washed with wash buffer, incubated for 30 min at 80°C with Pretreatment Solution (NaSCN), washed with Wash Buffer and finally treated in a protease I solution (0.5 mg/ml protease buffer; pH 2) for 10–12 min at 37°C.

After adding 10 μl of the hybridization probe and placing a coverslip, denaturation and hybridization of DNA was carried out using the metal block of a thermocycler (HyBrite, Vysis Inc). The denaturation was effect at 83°C for 3 min and the hybridization was carried out overnight at 37°C. After hybridization, the excess of the probes was washed in 2× SSC/0.3%NP40 at 73°C for 2 min. The nuclei were counterstained with 1000 ng/ml DAPI/antifade (4.6-diamidine-2-phenyl indole; Vysis Inc). For the scoring, a Zeiss Axioscope fluorescence microscope (Carl Zeiss Inc., Jena, Germany) was used, equipped with a specially designed filter combination; the EGFR sequence was visualized with a Orange filter, the chromosome 7 centromere sequence was visualized with a Green filter and the nuclei were identified with a DAPI filter. A triple band pass filter (Orange, Green and DAPI; Vysis Inc.) was also used. Hybridization signals were scored in at least 200 intact nonoverlapping nuclei and FISH analysis was carried out independently by two observer used constant adjustment of microscope focus because signals were located at different focal plans. Representative images of each specimen were acquired with a high-performance CCD camera in monochromatic layers that were subsequently merged by the Quips PathVysion Software (Vysis Inc).

EGFR gene status was scored as the average number of EGFR red signals per nucleus and as the ratio between EGFR red signals and CEP7 green signals. Polisomy of EGFR gene was defined as an increase of EGFR red signals (three or more signals per nucleus) paralleled by the same increase of chromosomes 7 (where the EGFR gene is located) as measured by the number of CEP 7 green signals per nucleus. High GCN was defined as EGFR/nuclei ratio ≥2.9 or a EGFR/CEP7 polysomy >3 in at least 50% of the cells.

DNA extraction and K-ras mutation analysis

K-ras mutation status was analyzed centrally at the Laboratory of Cell Biology of the Department of Pathology, University of Modena and Reggio Emilia, Modena, Italy.

Five-μm thick, hematoxylin–eosin-stained sections from a representative paraffin-embedded block were applied on noncover-slipped slides for microdissection and DNA extraction. Briefly, microdissection was carried out under direct observation with an inverted microscope using a sterile needle. Each microdissected sample was directly transferred to an eppendorf tube containing digestion buffer [2 mg/ml proteinase K in 50 mM Tris (pH 8.5), 1 mM EDTA, 0.5% Tween 20]. The tubes were then incubated overnight at 56°C, followed by 10 min of incubation at 95°C to eliminate any remaining proteinase K activity. PCR was carried out in 20 μl reactions containing 2.0 μl DNA, 2 μl of commercial PCR buffer (Applied Biosystems, Foster City, CA), 2.0 mM of MgCl2, 200 mM of each deoxynucleotide triphosphate, 20 pmol of each primer and three units of AmpliTaq Gold polymerase (Applied Biosystems). PCR reaction was carried out on Uno II Thermoblock (Biometra, Gottingen, Germany). Initial denaturation at 95°C for 10 min was followed by 41 cycles and a final extension step (10 min at 72°C). The cycles included denaturation at 95°C for 1 min, annealing at 52°C for 1 min and extension at 72°C for 2 min. Exon 2 of K-ras was PCR amplified using intron-based primers in order to investigate the mutational status of K-ras codons 12 and 13 because frequently founded mutated in colorectal cancer. The forward and reverse oligonucleotide primers used to amplify K-ras exon 2 are forward 5′-CAT GTT CTA ATA TAG TCA CA-3′, reverse 5′-AAC AAG ATT TAC CTC TAT TG-3′.

The amplified DNA was electrophoresed on 2% agarose gel for 1 h at 110 V. The amplification products were then purified using MinElute PCR purification Kit (Qiagen, Valencia, CA) as indicated by the manufacturer's instructions. PCR products were then sequenced in both directions with ABI Prism BigDye Terminator v1.1 Cycle Sequencing kit (Applied Biosystems), using the same primers as those employed for PCR. PCR products were finally purified by Centri-Sep Spin Columns (Applied Biosystems) and subsequently ran on the ABI Prism 310 automatic sequencer (Applied Biosystems). The data were analyzed with the Sequencing Analysis 5.2 Software (Applied Biosystems).

statistical analysis

Associations between tissue biomarkers and pathological response were evaluated by using logistic regression models. Odds ratio (OR) and their relative 95% confidence intervals (CIs) were estimated as well as sensitivity and specificity. Results were classified as statistically significant if their P values were <0.05.

results

pts characteristics

Forty pts have been treated in the phase II trial: median age was 61 years (28–77); ultrasound stage was T3N0, T3N1 and T4N1 in 30%, 62% and 8% of the pts, respectively. Median number of cetuximab administrations was 8 (range 2–8) with 70% of the pts completing the planned doses. Thirty-three pts (83%) received a dose of radiation therapy of 50 Gy and seven pts (17%) received a dose of 50.4 Gy. Thirty-eight (95%) pts underwent radical surgery: median interval from the end of neo-adjuvant treatment to surgery was 49 days (range 26–72 days). Two pts were not operated for early progression (one patient) or refusal (one patient). At surgery, a Dworak TRG 3 and 4 was achieved in nine (23.1%) and three pts (7.7%), respectively; 12 pts (30.8%) achieved a TRG 2 and 14 pts (35%) a TRG 1. The clinical data from this study have been previously reported in detail [13]. Baseline tumor samples for biomarker studies were available from all the pts entered into the study (Table 1). No significant correlation was observed between molecular markers profile and pretreatment clinical TN stage.

Table 1.

Patient characteristics

No. of patients 40 
Median age, years 61 (range 28–77) 
Characteristic n 
Gender   
    Male 34 85 
    Female 15 
PS (ECOG scale) 100 
Clinical stage (by TRUS)   
    uT3N0 12 30 
    uT3N1 25 62 
    uT4N1 
EGFR IHC   
    Positive 14 35 
    Negative 26 65 
EGFR/nuclei ratio   
    ≥2.9 21 52.5 
    <2.9 19 47.5 
CEP7 polisomy   
    >3 copies 22 55.0 
    ≤3 copies 18 45.0 
K-ras   
    Wild type 31 77.5 
    Mutated 22.5 
Pathological response (Dworak's grade)   
    TRG 4 7.7 
    TRG 3 23.1 
    TRG 2 12 30.8 
    TRG 1 14 35 
    TRG 0 
    Not evaluable 2a 
No. of patients 40 
Median age, years 61 (range 28–77) 
Characteristic n 
Gender   
    Male 34 85 
    Female 15 
PS (ECOG scale) 100 
Clinical stage (by TRUS)   
    uT3N0 12 30 
    uT3N1 25 62 
    uT4N1 
EGFR IHC   
    Positive 14 35 
    Negative 26 65 
EGFR/nuclei ratio   
    ≥2.9 21 52.5 
    <2.9 19 47.5 
CEP7 polisomy   
    >3 copies 22 55.0 
    ≤3 copies 18 45.0 
K-ras   
    Wild type 31 77.5 
    Mutated 22.5 
Pathological response (Dworak's grade)   
    TRG 4 7.7 
    TRG 3 23.1 
    TRG 2 12 30.8 
    TRG 1 14 35 
    TRG 0 
    Not evaluable 2a 
a

One patient had a clinical progression and was considered as treatment failure.

PS, performance status; ECOG, Eastern Cooperative Oncology Group; IHC, immunohistochemistry; TRG, tumor regression grade.

EGFR status and pathological response

EGFR was evaluated for protein expression by IHC and GCN by FISH. EGFR expression was positive in 14 pts (40%). After preoperative treatment, 3 of 14 (21.4%) EGFR IHC-positive pts achieved a TRG 3–4 versus 9 of 25 (36%) EGFR-negative pts (P 0.34).

Median EGFR/CEP7 ratio was 1.06 (range 0.71–4.8). EGFR gene amplification was observed in only one patient (2.5%). Twenty-one pts (52.5%) had a EGFR/nuclei ratio ≥2.9; moreover 22 pts (55.0%) had EGFR/CEP7 polysomy >3 in at least 50% of the cells. The concordance between the two tests measured by Cohen's kappa test was 95%.

Eleven of the 20 assessable pts with EGFR/nuclei ratio ≥2.9 (55%) achieved a TRG 3–4 compared with 1 of 19 pts (5.3%) with EGFR/nuclei ratio <2.9 (OR 22; 95% CI 2.44–198.14; P 0.0016). Moreover, 12 of 21 (57.1%) pts with EGFR/CEP7 polysomy >3 in at least 50% of the cells obtained a TRG 3–4 compared with no response in the 18 pts with EGFR/CEP7 polysomy ≤3 (P 0.0003). The sensibility (95% CI) and specificity (95% CI) of tests were 91.7% (0.61–0.99) and 66.7% (0.46–0.84), respectively, for EGFR/nuclei ratio and 100% (0.74–1) and 66.7% (0.46–0.84), respectively, for EGFR/CEP7 polysomy (Table 2).

Table 2.

Pathological regression according to biological characteristics

Marker Assessable patients, n = 39 TRG 3–4
 
Sensitivity (95% CI) Specificity (95% CI) 
No. (%)
 
12 (30.8) 
EGFR IHC+ 14 3 (21.4) 25% (0.05–0.57) 59.3% (0.39–0.78) 
EGFR IHC− 25 9 (36.0)   
P value  0.34   
EGFR/nuclei ratio ≥2.9 20 11 (55.0) 91.7% (0.61–0.99) 66.7% (0.46–0.84) 
EGFR/nuclei ratio <2.9 19 1 (5.3)   
P value  0.0016   
CEP7 polisomy >3 copies 21 12 (57.1) 100% (0.74–1) 66.7% (0.46–0.84) 
CEP7 polisomy ≤3 copies 18 0 (0)   
P value  0.0003   
K-ras wild type 30 11 (36.7) 91.67% (0.62–0.99) 70.4% (0.49–0.86) 
K-ras mutated 1 (11.0)   
P value  0.119   
Marker Assessable patients, n = 39 TRG 3–4
 
Sensitivity (95% CI) Specificity (95% CI) 
No. (%)
 
12 (30.8) 
EGFR IHC+ 14 3 (21.4) 25% (0.05–0.57) 59.3% (0.39–0.78) 
EGFR IHC− 25 9 (36.0)   
P value  0.34   
EGFR/nuclei ratio ≥2.9 20 11 (55.0) 91.7% (0.61–0.99) 66.7% (0.46–0.84) 
EGFR/nuclei ratio <2.9 19 1 (5.3)   
P value  0.0016   
CEP7 polisomy >3 copies 21 12 (57.1) 100% (0.74–1) 66.7% (0.46–0.84) 
CEP7 polisomy ≤3 copies 18 0 (0)   
P value  0.0003   
K-ras wild type 30 11 (36.7) 91.67% (0.62–0.99) 70.4% (0.49–0.86) 
K-ras mutated 1 (11.0)   
P value  0.119   

CI, confidence interval; IHC, immunohistochemistry; TRG, tumor regression grade.

K-ras mutation status and pathological response

Nine pts (9 of 39 = 23.1%;) showed a K-ras mutation. One patient with wild-type K-ras refused surgery and was not assessable. Eleven of 30 assessable pts with wild-type KRAS had a TRG 3–4 versus only 1 of 9 patient with K-ras mutation (36.7 versus 11%, respectively); OR 4.63; 95% CI 0.51–42.11 (P 0.15). Sensibility (95% CI) and specificity (95% CI) of the test were 91.67% (0.62–0.99) and 70.4 (0.49–0.86), respectively.

molecular biomarker profile and pathological response

On the basis of GCN and K-ras status, four groups of pts were identified: 18 pts with high GCN and a wild-type K-ras; three pts with high GCN and K-ras mutation, 13 pts with low GCN and wild-type K-ras and six pts with low GCN and K-ras mutation. The TRG 3–4 rates in the four groups were 58.8%, 33.3%, 7.7% and 0%, respectively. The odds of pathological response were higher in pts with high EGFR GCN and wild-type K-ras compared with those pts with low EGFR GCN regardless of K-ras mutation status (OR = 25.6; 95% CI 2.76–239.9, P = 0.0012) (Table 3).

Table 3.

Pathological regression according to high gene copy number, KRAS status and both

EGFR/nuclei ratio KRAS No. of assessable patients No. of TRG (3–4) OR (95% CI) P value 
≥2.9 – 20 11 55 22 (2.44–198.149) 0.0016 
<2.9 – 19 5.3   
– Wild type 30 11 36.7 4.63 (0.51–42.11) 0.119 
– Mutated 11.0   
≥2.9 Wild type 17 10 58.8 25.71 (2.76–239.94) 0.0012 
≥2.9 Mutated 33.3 9 (0.39–206.53)  
<2.9 Wild type 13 7.7 Referencea  
<2.9 Mutated   
EGFR/nuclei ratio KRAS No. of assessable patients No. of TRG (3–4) OR (95% CI) P value 
≥2.9 – 20 11 55 22 (2.44–198.149) 0.0016 
<2.9 – 19 5.3   
– Wild type 30 11 36.7 4.63 (0.51–42.11) 0.119 
– Mutated 11.0   
≥2.9 Wild type 17 10 58.8 25.71 (2.76–239.94) 0.0012 
≥2.9 Mutated 33.3 9 (0.39–206.53)  
<2.9 Wild type 13 7.7 Referencea  
<2.9 Mutated   
a

Reference was obtained lumping wild-type and mutated K-ras when EGFR gene copy number was low.

TRG, tumor regression grade OR, odds ratio; CI, confidence interval.

discussion

Cetuximab alone or in combination with chemotherapy is active in metastatic colorectal cancer; However, the results are relatively modest with response rates between 16.4% and 46.9% and median progression-free survivals ranging from 4.1 to 8.9 months [7–10]. In locally advanced rectal cancer, when cetuximab was used as preoperative treatment in combination with capecitabine or 5-FU and radiation therapy, the rate of pathological complete response was disappointingly low [11–13]. Several factors might contribute to this subadditive effect, including upregulation of cycline-dependent kinase p27 and G1 cell cycle arrest induced by the cetuximab [14], the sequence of cetuximab and chemoradiotherapy [26] and the redundancy of EGFR pathways [27]. Irrespectively of the resistance mechanisms, it is, however, clear that the availability of biomarkers of cetuximab sensitivity would greatly improve the cost–benefit ratio of this treatment.

Studies in mCRC have reported a nonsignificant relationship between EGFR protein expression determined by IHC and clinical response to cetuximab with ∼25% of the pts who do not express EGFR achieving a clinical response [9, 15].

The EGFR GCN determined by FISH has been proposed as a more reliable method to predict the response to cetuximab in mCRC. Several retrospective analysis have reported an high EGFR GCN in a 48%–69% of the pts and this biological characteristic has been significantly associated with clinical response to cetuximab [16, 17]. Moreover a high EGFR genomic gain was associated with a significantly longer time to progression (6.6 versus 3.5 months; P 0.02) and a positive trend in overall survival. (11.3 versus 8.5 months; P 0.8) in pts with mCRC treated with cetuximab [17, 18]. The cut-off value of EGFR GCN to discriminate high and low GCN was firstly proposed by Moroni et al. [16] as a EGFR/nuclei ratio of 2.47. Cappuzzo et al. [17] using a receiver operating characteristics analysis showed that the best cut-off was 2.92 EGFR copies per nucleus. Moroni et al. and Frattini et al. have also shown that EGFR amplification and chromosome 7-marked polisomy was associated with a higher probability of response to cetuximab [16, 28].

In our study, we have observed a significant relationship between EGFR high GCN and a high pathological TRG in locally advanced rectal cancer. A high EGFR GCN was defined on the basis of EGFR nuclei ratio and CEP7 polysomy and the cut-off values previously reported by other authors have been used [16, 17].

More recently, K-ras mutation has been reported as a negative predictor factor of response to cetuximab as single agent or cetuximab-containing regimens [21]. The negative prognostic value of K-ras mutation was maintained even after multivariate analysis [22]. Recently, Machiels and coworkers have reported data on a correlative analysis of molecular marker expressions on pathological response to preoperative chemoradiotherapy plus cetuximab in rectal cancer. This analysis included tissue markers, genomic and proteomic profile. Microarray gene expression analysis and proteomics showed downregulation of invasion and proliferation pathways and an upregulation of inflammatory pathways and EGFR ligands after the first dose of cetuximab. The immunohistochemically determined expressions of Ki-67 and TGF-α were significantly correlated with T-level downstaging. A trend (P 0.06) for better tumor regression was found in pts with wild-type K-ras [29]. In our study, the K-ras mutation was observed in 22.5% of the pts; a TRG 3–4 was achieved in only 1 of 9 of these pts (11%) versus 11 of 30 (36.7%) pts with wild-type K-ras. However, this difference was not statistically significant probably because of the low number of pts. The best probability to predict TRG 3–4 following cetuximab-based treatment was observed when GCN and K-ras status were combined: the coexpression of high level of EGFR genomic gain (EGFR/nuclei ratio ≥2.9) and wild-type K-ras was associated with a 58.8% of high pathologic tumor regression.

In conclusion, the results of the present study confirm that a high GCN, either a EGFR/nuclei ratio >2.9 or high chromosome 7 polisomy, can allow to select pts at higher probability of response to cetuximab-containing regimen. The concomitant presence of wild-type K-ras and a high EGFR genomic gain seems to identify a subgroup of pts with the best chance of cetuximab benefit. The evaluation of these parameters should be prospectively incorporated into clinical trials to confirm these observations.

funding

National Ministry of Health, Italy (RF-EMR-2006-361866); Regione Emilia-Romagna, Italy (DGR-ER-2242-2007).

Presented in part at the 44th ASCO Meeting, Chicago, IL.

References

1.
Watanabe
T
Wu
TT
Catalano
PJ
, et al.  . 
Molecular predictors of survival after adjuvant chemotherapy for colon cancer
N Engl J Med
 , 
2001
, vol. 
344
 (pg. 
1196
-
1206
)
2.
Kim
J-S
Kim
J-M
Li
S
, et al.  . 
Epidermal growth factor receptor as a predictor of tumor dowstaging in locally advanced rectal cancer patients treated with preoperative chemoradiotherapy
Int J Radiat Oncol Biol Phys
 , 
2006
, vol. 
66
 (pg. 
195
-
200
)
3.
Azria
D
Bibeau
F
Barbier
N
, et al.  . 
Prognostic impact of epidermal growth factor receptor (EGFR) expression on loco-regional recurrence after preoperative radiotherapy in rectal cancer
BMC Cancer
 , 
2005
, vol. 
5
 (pg. 
62
-
71
)
4.
Giralt
J
de las Heras
M
Cerezo
L
, et al.  . 
The expression of epidermal growth factor receptor results in a worse prognosis for patients with rectal cancer treated with preoperative radiotherapy: a multicenter, retrospective analysis
Radiother Oncol
 , 
2005
, vol. 
74
 (pg. 
101
-
108
)
5.
Li
S
Kim
JS
Kim
JM
, et al.  . 
Epidermal growth factor receptor as a prognostic factor in locally advanced rectal-cancer patients treated with preoperative chemoradiation
Int J Radiat Oncol Biol Phys
 , 
2006
, vol. 
65
 (pg. 
705
-
712
)
6.
Bertolini
F
Bengala
C
Losi
L
, et al.  . 
Prognostic and predictive value of baseline and dynamic molecular marker expression in locally advanced rectal cancer treated with neo-adjuvant chemo-radiotherapy
Int J Radiat Oncol Biol Phys
 , 
2007
, vol. 
68
 (pg. 
1455
-
1461
)
7.
Van Cutsem
E
Nowacki
M
Lang
S
, et al.  . 
Randomized phase III study of irinotecan and 5FU/FA with or without cetuximab in the first line treatment of patients with colorectal cancer: the Crystal study
Proc Am Soc Clin Oncol
 , 
2007
, vol. 
25
 pg. 
164s
  
(Abstr 4000)
8.
Sobrero
A
Maurel
J
Fehrenbacher
L
, et al.  . 
EPIC: phase III trial of cetuximab plus irinotecan after fluoropyrimidine and oxaliplatin failure in patients with metastatic colorectal cancer
J Clin Oncol
 , 
2008
, vol. 
26
 (pg. 
2311
-
2319
)
9.
Cunningham
D
Humblet
Y
Siena
S
, et al.  . 
Cetuximab monotherapy and cetuximab plus irinotecan in irinotecan-refractory metastatic colorectal cancer
N Engl J Med
 , 
2004
, vol. 
351
 (pg. 
337
-
345
)
10.
Saltz
LB
Meropol
NJ
Loehrer
PJ
, et al.  . 
Phase II trial of cetuximab in patients with refractory colorectal cancer that express the epidermal growth factor receptor
J Clin Oncol
 , 
2004
, vol. 
22
 (pg. 
1201
-
1208
)
11.
Machiels
JP
Sempoux
C
Scalliet
P
, et al.  . 
Phase I/II study of preoperative cetuximab, capecitabine, and external beam radiotherapy in patients with rectal cancer
Ann Oncol
 , 
2007
, vol. 
18
 (pg. 
738
-
744
)
12.
Rodel
C
Arnold
D
Hipp
N
, et al.  . 
Phase I–II trial of cetuximab, capecitabine, oxaliplatin and radiotherapy as preoperative treatment in rectal cancer
Int J Radiat Oncol Biol Phys
 , 
2008
, vol. 
70
 (pg. 
1081
-
1086
)
13.
Bertolini
F
Chiara
S
Bengala
C
, et al.  . 
Neo-adjuvant treatment with single agent cetuximab followed by 5-fu, cetuximab and pelvic radiotherapy: a phase II study in locally advanced rectal cancer
Int J Radiat Oncol Biol Phys
 , 
2008
 
(in press)
14.
Peng
D
Fan
Z
Lu
Y
, et al.  . 
Anti-epidermal growth factor receptor monoclonal antibody C225 upregulates p27kip1 and induces G1 arrest in prostatic cancer cell line DU145
Cancer Res
 , 
1996
, vol. 
56
 (pg. 
3666
-
3669
)
15.
Chung
KY
Shia
J
Kemeny
NE
, et al.  . 
Cetuximab shows activity in colorectal cancer patients with tumors that do not express the epidermal growth factor receptor by immunohistochemistry
J Clin Oncol
 , 
2005
, vol. 
23
 (pg. 
1803
-
1810
)
16.
Moroni
M
Veronese
S
Benvenuti
S
, et al.  . 
Gene copy number for epidermal growth factor receptor (EGFR) and clinical response to antiEGFR treatment in colorectal cancer: a cohort study
Lancet Oncol
 , 
2005
, vol. 
6
 (pg. 
279
-
286
)
17.
Cappuzzo
F
Finocchiaro
G
Rossi
E
, et al.  . 
EGFR FISH assay predicts for response to cetuximab in chemotherapy refractory colorectal cancer patients
Ann Oncol
 , 
2007
, vol. 
19
 (pg. 
717
-
723
)
18.
Sartore-Bianchi
A
Moroni
M
Veronese
S
, et al.  . 
Epidermal growth factor receptor gene copy number and clinical outcome of metastatic colorectal cancer treated with panitumumab
J Clin Oncol
 , 
2007
, vol. 
25
 (pg. 
3238
-
3245
)
19.
De Rook
W
Piessevaux
H
De Schutter
J
, et al.  . 
KRAS wild-type state predicts survival and is associated to early radiological response in metastatic colorectal cancer treated with cetuximab
Ann Oncol
 , 
2008
, vol. 
19
 (pg. 
508
-
515
)
20.
Lievre
A
Bachet
J-B
Le Corre
D
, et al.  . 
KRAS mutation is predictive of response to cetuximab therapy in colorectal cancer
Cancer Res
 , 
2006
, vol. 
66
 (pg. 
3992
-
3995
)
21.
Lievre
A
Bachet
J-B
Boige
A
, et al.  . 
KRAS mutations as an independent prognostic factor in patients with advanced colorectal cancer treated with cetuximab
J Clin Oncol
 , 
2008
, vol. 
26
 (pg. 
374
-
379
)
22.
Di Fiore
F
Blanchard
F
Charbonnier
F
, et al.  . 
Clinical relevance of KRAS mutation detection in metastatic colorectal cancer treated by cetuximab plus chemotherapy
Br J Cancer
 , 
2007
, vol. 
96
 (pg. 
1166
-
1169
)
23.
Khambata-Ford
S
Garrett
CR
Meropol
NJ
, et al.  . 
Expression of epiregulin and amphiregulin and K-ras mutation status predict disease control in metastatic colorectal cancer patients treated with cetuximab
J Clin Oncol
 , 
2007
, vol. 
25
 (pg. 
3230
-
3237
)
24.
Dworak
O
Keilholz
L
Hoffmann
A
, et al.  . 
Pathological features of rectal cancer after preoperative radiochemotherapy
Int J Colorectal Dis
 , 
1997
, vol. 
12
 (pg. 
19
-
23
)
25.
Rodel
C
Martus
P
Papadoupolos
T
, et al.  . 
Prognostic significance of tumor regression after preoperative chemoradiotherapy for rectal cancer
J Clin Oncol
 , 
2005
, vol. 
23
 (pg. 
8688
-
8696
)
26.
Morelli
MP
Cascone
T
Troiani
T
, et al.  . 
Sequence-dependent anti-proliferative effects of cytotoxic drugs and epidermal growth factor receptor inhibitors
Ann Oncol
 , 
2005
, vol. 
16
 
Suppl 4
(pg. 
iv61
-
iv68
)
27.
Scaltriti
M
Baselga
J
The epidermal growth factor receptor pathway: a model for targeted therapy
Clin Cancer Res
 , 
2006
, vol. 
12
 (pg. 
5268
-
5272
)
28.
Frattini
M
Saletti
P
Romagnani
E
, et al.  . 
PTEN loss of expression predicts cetuximab efficacy in metastatic colorectal cancer patients
Br J Cancer
 , 
2007
, vol. 
97
 (pg. 
1139
-
1145
)
29.
Machiels
JH
Debucquoy
A
Gevaert
O
, et al.  . 
Prediction of pathological response to preoperative chemoradiotherapy with cetuximab in rectal cancer
J Clin Oncol
 , 
2008
pg. 
26
  
(Abstr 4095)