Epidermal growth factor receptor as a biomarker for cervical cancer
- T. Soonthornthum1,
- H. Arias-Pulido1,
- N. Joste2,
- L. Lomo2,
- C. Muller3,
- T. Rutledge3 and
- C. Verschraegen1,*
- 1Department of Internal Medicine
- 2Department of Pathology
- 3Department of Gynecology University of New Mexico Cancer Research and Treatment Center, Albuquerque, USA
- *Correspondence to: Dr C. Verschraegen, Department of Medicine, University of New Mexico Cancer Research and Treatment Center, 1201 Camino de Salud, Albuquerque, NM 87131, USA. Tel: +1-505-925-0429; Fax: +1-505-925-0408; E-mail: cverschraegen{at}salud.unm.edu
- Received July 30, 2010.
- Revision received November 5, 2010.
- Accepted November 5, 2010.
Abstract
This review focuses on the different modes of expression of the epidermal growth factor receptor (EGFR). All methods used to assess EGFR expression are critically analyzed and insights into the use of inhibitors of EGFR for treatment of cervical cancer are discussed. Currently, expression of EGFR as a biomarker for prognosis or for treatment of cervical cancer is not defined for clinical use.
Key words
introduction
Cervical cancer, a preventable disease, is the second most common malignancy in women worldwide and a major cause of morbidity and mortality particularly in developing countries where it accounts for 15% of all new female cancers, with an estimated 500 000 new cases and 275 000 deaths [1]. Eighty percent of incident cervical cancer cases and deaths occur in developing countries because of socioeconomic challenges, pattern of health care delivery, and societal factors. Cervical cancer is diagnosed at advanced stages in >80% of women in developing countries [1]. Affected women are usually young, working, and raising children, which creates substantial social problems.
Unlike the majority of other gynecologic malignancies, cervical cancer is clinically staged. Prognostic factors include clinical stage at time of diagnosis, tumor size, lymphovascular invasion, and parametrial and lymph node involvement. Early stages (stage IA to IB2) are often treated surgically and later stages require multidisciplinary treatments with either concurrent chemoradiation (the predominant treatment in the United States) or neoadjuvant chemotherapy followed by surgery [2]. Central recurrences can be cured with directed pelvic exenteration, but distant recurrences remain uniformly fatal. Active research is gaining insight into the molecular characterization of cervical cancer, which may help better define outcome and personalize treatment with novel therapeutic targets. Many biological factors that help regulate cell cycle control, apoptosis, angiogenesis, or invasive or metastatic potential have been proposed as prognostic determinants of cervical cancer. Examples include the epidermal growth factor receptor (EGFR) family, vascular endothelial growth factor, microvessel density, hypoxic mechanisms, and expression of COX-2. This review focuses on EGFR. A literature search was carried out using the PubMed and Google engines, with the terms EGFR, cervical cancer, prognosis, method of detection, and the Internet sites of laboratory companies manufacturing antibodies to research the epitopes.
EGFR is a 170-kDa transmembrane glycoprotein receptor encoded by the Her-1 proto-oncogene located on chromosome 7p12. EGFR functions through dimerization that activates a tyrosine kinase domain to regulate multiple functions such as cell growth, differentiation, gene expression, and development. EGFR is present in many normal tissues and expressed in a wide variety of solid tumors, including cervical cancer (Figure 1) [3]. In normal cervical mucosa, EGFR is normally expressed in the cytoplasm and the membrane of the cells within the basal layer, and as cells differentiate, there is a shift toward the cytoplasm. EGFR expression is associated with human papillomavirus (HPV) infection as EGFR cytoplasmic expression increases with increasing grade of intraepithelial neoplasia but is not correlated with the HPV type [4]. HPV is considered an etiologic factor in the development of cervical cancer [5]. The E5 protein of HPV type 16 can activate EGFR through binding to a subunit of the protein pump ATPase leading to a reduced degradation of EGFR receptors, an increase in EGFR recycling, and an overexpression of EGFR [6–8]. Expression of high-risk HPV E6 has also been linked to an increase in the EGFR levels [9, 10], and changes in functional levels of the HPV E6/E7 proteins may alter the growth rate of cervical cancer cell lines by reducing the stability of EGFR at the post-transcriptional level [11]. Because of the genomic instability induced by HPV [12], it was hypothesized that EGFR could be mutated. However, we and others [13,14] have shown that EGFR mutations are uncommon in high-grade cervical lesions and invasive cervical cancer. This suggests that high-risk HPV proteins affect EGFR at the protein but not the genomic level.
EGFR immunohistochemistry (T = tumor, N = normal cervix mucosa). Clone 31G7, a mouse monoclonal derived from A431 cells, which recognizes the extracellular domain, was used (Zymed; Invitrogen, Carlsbad, CA). These examples are from our series of 69 cervical cancer cases (unpublished data). The normal cervical epithelium and the tumor cells expressed EGFR both in the cytoplasm and on the membrane. Of the 69 cases, 64% expressed EGFR below the level seen in normal basal layer of the cervix epithelium and 36% stained more strongly (by Remmele score). EGFR, epidermal growth factor receptor.
In cervical cancer, the expression of EGFR varies from 6% to 90%, depending on the study methodology. Overexpression has been associated with poor prognosis in some studies but not in others (reviewed in [2, 15]). Dysregulation of EGFR causes oncogenicity, but the exact biological mechanisms promoting cell growth are not completely understood. The roles of coexpression of receptor ligands [EGF, transforming growth factor-α (TGF-α), amphiregulin], gene amplification, decreased levels of phosphatase, heterodimerization and cross-talk between other members of the erbB receptor family, and interaction with downstream and other cell signaling pathways and viral proteins have been implicated in the oncogenic process and remain an active subject of investigation (reviewed in [16]).
current clinical use of EGFR inhibition for cancer
EGFR inhibitors have proven efficacy in some clinical trials in patients with lung, colon, pancreas, or head and neck cancer [17–20]. There are two different ways to pharmacologically target EGFR: anti-EGFR monoclonal antibodies (cetuximab, panitumumab) and specific inhibitors of the EGFR tyrosine kinase domain (erlotinib, lapatinib). Cetuximab is approved as a single agent or in combination with irinotecan for treatment of metastatic colorectal cancer [21] and as single agent or in combination with radiation for head and neck cancer [18, 22]. The addition of cetuximab to radiation demonstrated a statistically significant prolongation of overall survival (OS) (54 versus 28 months; P = 0.02) and improved locoregional control at 2 years (56% versus 48%; P = 0.02) in 424 patients with locally advanced squamous cell cancer of the head and neck [22]. This favorable outcome in head and neck cancers suggests that a similar approach, with or without chemotherapy, may be warranted in other epithelial EGFR-dependent and radiation-sensitive tumors such as cervical cancer.
It is clear from numerous clinical trials that only a subset of patients responds to EGFR inhibition. Yet, a common predictor of response has not been clearly defined. In colon cancer, a KRAS wild-type status is predictive of response to cetuximab or panitumumab [23–25]. Furthermore, in this subset of patients treated with cetuximab, high epiregulin messenger RNA (mRNA) expression is solely predictive of OS [26]. In contrast, KRAS mutation status does not predict response to erlotinib in pancreatic cancer. In lung cancer, EGFR mutations are associated with response to erlotinib [27]. These examples illustrate the complexity of EGFR biology, which is not yet understood.
preclinical treatment studies of EGFR inhibition in cervical cancer
There is very scant experimentation on EGFR inhibition in cervical cancer cell lines. Radiolabeled cetuximab is internalized in cells expressing the EGFR [28]. Cetuximab inhibits all EGFR-positive cell lines, with a percentage of inhibition varying from 37% to 58% (P < 0.05). All primary tumors and seven of eight (87.5%) established cervical cancer cell lines expressed EGFR-1 by flow cytometry [29]. Cell lines from recurrent/metastatic sites of disease expressed higher levels of EGFR when compared with those obtained from primary sites (P > 0.05). Minimal complement-dependent cytotoxicity but a high degree of antibody-dependent cellular cytotoxicity was detected in the majority of cervical cancer cell lines exposed to cetuximab in the presence of peripheral blood lymphocytes from either healthy donors or cervical cancer patients. Additionally, in the human epidermoid carcinoma A431 cells, which bear a large number of EGFRs, cetuximab exhibited effective cytolytic activity for A431 cells with human polymorphonuclear leukocytes but not with mononuclear cells [30]. EGFR signaling is dependent upon the differential information transmitted through receptor dimers. Differences in receptor combinations cause variance in cell culture growth. Homodimers are less mitogenic than heterodimeric combinations and have a lower transforming activity. Tumor necrosis factor-α can sensitize cervical cancer cells to matuzumab, a humanized monoclonal antibody against EGFR, which reduced tumor size [31]. The combination of cetuximab and chemoradiation, with trastuzumab or a MEK1/2 inhibitor, PD98059, inhibits proliferation of cervical cancer cells by decreasing mitogen-activate protein kinase phosphorylation independently of EGFR status [32]. Lastly, novel downstream substrates of the EGFR have been identified by phosphoproteomics. These proteins, endofin, DCBLD2, and KIAA0582, are part of the EGF phosphotyrosine signaling network and could be used as indirect targets [33].
EGFR pathway and mutations in cervical cancer
Two studies demonstrated the absence of EGFR mutation in the tyrosine kinase domain (exons 18–21) [13, 14]. The lack of mutation may explain the lack of activity of small-molecule tyrosine inhibitors, such as gefitinib and erlotinib. Other mutations such as EGFRvIII in the extracellular domain of EGFR may be clinically relevant. Although there are no data for cervical cancer, an EGFRvIII mutation is associated with gene amplification in other cancers, resulting in the overexpression of EGFR lacking amino acids 30–297, corresponding to protein domains I and II [34]. In lung cancer, EGFRvIII is found in 5% of squamous cells. In addition, lung tumors that have EGFRvIII are sensitive to an irreversible EGFR tyrosine kinase inhibitor, HKI-272, despite the fact that these tumors are relatively resistant to the tyrosine kinase inhibitors gefitinib and erlotinib [35]. A novel missense mutation in the extracellular domain of EGFR gene has been identified in glioblastoma and occurs independently of EGFRvIII. This mutation provides an alternative mechanism of EGFR activation in glioblastoma and confers sensitivity to small-molecule EGFR tyrosine kinase inhibitors [36, 37]. The absence of exon 18–21 mutations in cervical cancer does not rule out a possible activity of small molecules through other mechanisms related to undiscovered mutations that could affect other domains of EGFR. Additionally, in a study of 258 tissue samples of primary cervical cancer, 13.9% of adenocarcinomas had K-RAS mutations, whereas only 0.7% of squamous cell cancers did [38]. The incidence of K-RAS mutations in cervical cancer needs to be investigated further, as this could impact treatment with monoclonal antibodies as shown for colon cancer.
detection methods of EGFR expression in cervical cancer
EGFR protein levels have been reported by a variety of methods, including radioligand binding, RT-PCR, enzyme-linked immunosorbent assay (ELISA), flow cytometry, FISH, and immunohistochemistry (IHC). This variety of techniques and the lack of standardized assays and reagents may explain the wide range of EGFR expression found in cervical cancer, ranging from 6% to 100% (Table 1–3), and the discrepancies found in relation to EGFR as a predictor of patients outcome; while several studies reported a positive association with OS, no association was found in other studies. Thirty-seven studies have evaluated EGFR expression in cervical cancer between 1991 and 2009.
IHC of EGFR in cervical cancer
technical overview
IHC is the most commonly used technique (Table 1–3) [39–41, 43, 44, 46–51, 53–58, 61–65]. IHC is a convenient inexpensive method that can be standardized against known controls and can evaluate the level of activated EGFR, but there are many variations between published protocols.
Furthermore, IHC is not strictly quantitative and there is no uniformly adopted scoring system; therefore, interpretation of the staining intensity remains observer dependent. Two common scores are used, the Reiner score [66] or the Remmele score [67], which uses the sum or the product of the stain intensity and the percentage of stained cells, respectively. In breast cancer studies, radio-IHC of EGFR is more quantitative than IHC [68]. This technique has not been used in cervical cancer. Automatic IHC imaging systems significantly correlate with a manual scoring system (P < 0.0001) and are practical, rapid, and inherently reproducible, providing improved interdepartmental and interinstitutional concordance of results [69]. Standardization of antibody preparations and detection method, as well as agreement on the scoring system, is necessary to produce an assay for EGFR evaluation that provides consistent and comparable results [70, 71]. Quantum dot, the combination of novel optical contrast agent and structured illumination, is a good alternative to the use for biological imaging in immunohistochemical study for detecting receptors but to date is not clinically practical [72, 73].
EGFR expression by IHC and prognosis
Table 1–3 shows EGFR expression, treatment received, and analyzed outcomes. Ten studies show that EGFR overexpression is significantly associated with poor disease outcomes in cervical carcinoma [39–41, 43, 44, 46–50]. Four studies associate EGFR overexpression with poor disease-specific OS [39, 40, 44, 48]. EGFR overexpression is also associated with poor prognostic factors such as increased tumor size, lymph node metastases, and recurrence of disease [39, 41, 44, 46, 50]. In contrast, a similar number of studies do not show any correlation between EGFR overexpression and OS or prognostic parameters [14, 51, 53–58, 61–63, 65]. Only one study demonstrates an association between EGFR overexpression and improved OS [hazard ratio (HR) 0.3; 95% confidence interval (CI) 0.12–0.86; P = 0.025]. In this study, the analyses of 78 tumor specimens for EGFR, HER-3, and HER-4 by IHC and HER-2 gene amplification by FISH show that (i) overexpression of HER-2 and HER-3 is associated with poor prognosis (P = 0.006 and 0.05, respectively) and (ii) a correlation is observed between EGFR and HER-4 receptor expressions (P = 0.019). The authors conclude that a cross-talk between EGFR and HER-4 receptor could be associated with a better prognosis [64].
In all studies that compared EGFR expression in squamous cell carcinoma versus adenocarcinoma and adenosquamous carcinoma, the squamous cell type always stains more strongly [41,44, 51, 57, 62]. The average percentage of squamous cell carcinoma overexpressing EGFR is 51% compared with 23% for adenocarcinoma and 38% for adenosquamous carcinoma.
The Zymed antibody, clone 31G7, was never associated with clinical outcome (four studies [14, 51, 53, 61]). The Dako clone H11 was associated with poor outcome (two studies [40, 49]). The polyclonal rabbit EGFR antibody and the monoclonal antibodies from clone 144.8 (Oncogene Science/EMD), clone 111.6 (Thermo Fisher Scientific), and clone E30 (Dako or Biogenex) gave mixed results (Table 1–3 for references). These discrepancies could be related to the scoring system. In studies looking for a high cut-off expression of EGFR (Table 2), there was usually no association of expression with outcome. Most studies associated with poor outcome used a lower cut-off for a positive value than the one used in the negative studies, i.e. including more patients with widespread expression (Table 1). Would this mean that very low expression or the absence of EGFR staining could confer a better prognosis? However, the only study that associates high expression with good outcome had a very high cut-off (Table 3) [64]. Perhaps a true meta-analysis of the patients in these studies using nondichotomized results may help understanding whether EGFR expression is related to clinical outcome.
Relationship between EGFR overexpression and poor outcome in cervical cancer by immunohistochemistry detection
Absence of relationship between EGFR overexpression and outcome in cervical cancer by immunohistochemistry detection
Relationship between EGFR overexpression and good outcome in cervical cancer by immunohistochemistry detection
EGFR overexpression and response of treatment
In 375 specimens of cervical cancer from patients treated with chemoradiation, a multivariate analysis showed that EGFR overexpression and phosphorylated EGFR expression were independent predictors of poor response to treatment (HR 1.84; 95% CI 1.20–2.82; P = 0.005 and HR 1.17; 95% CI 1.11–2.66; P = 0.016, respectively). Overexpression of EGFR was also associated to squamous cell histology and to shorten disease-free survival (DFS) and OS [48]. EGFR expression during radiation therapy may increase or decrease. EGFR up-regulation is not a universal response to radiation therapy and EGFR targeting may not necessarily become more efficient during radiotherapy treatment [53]. In 170 patients with locally advanced cervical cancer treated by concurrent chemoradiation, a multivariate analysis showed that isolated overexpression of EGFR or HER-2 conferred a poor DFS (HR 2.43; 95% CI 0.98–6, P = 0.05) and that coexpression of EGFR and HER-2 was an independent predictive factor of poorer DFS (HR 3.99; 95% CI 1.44–11.05; P = 0.007) [49]. Similarly, in another 68 patients treated with chemoradiation, a multivariate analysis found that coexpression of EGFR and COX-2 was also an independent predictive factor for poor overall DFS (relative risk 4; 95% CI 2.7–5.3; P = 0.03) [46]. The activity of the COX-2 promoter and the induction of COX-2 enzyme are known to be stimulated by receptor-mediated signaling triggered by EGF, TGF-α, and other ligands of EGFR [74].
EGFR amplification by FISH
In one study of EGFR by FISH, the ratio of the EGFR gene signals to chromosome 7 centromeric signals did not show gene amplification [75]. The authors could not provide evidence that alterations of chromosome 7 were related to EGFR overexpression. However, in this study, polysomy of chromosomes 3 and X defined the transition from high-grade squamous intraepithelial lesions to cervical carcinoma and could be of prognostic significance.
EGFR protein levels and ELISA
The ELISA has been used to detect the level of EGFR protein in tissue or in serum. The advantage of this method is that it requires smaller volumes of tissue and the technique is less sensitive to degradation of receptor protein than IHC. However, this method measures total receptor protein, thus providing no information on cellular localization. A careful microdissection of cancer cells from the stroma must also be carried out. Four studies use ELISA for detecting EGFR protein level in tissue or serum [76–79]. The cut-off point to determine EGFR overexpression varies for tumor specimens and serum (250 fmol/mg of protein and 100 fmol/ml of serum). There is no agreement on the cut-off level of EGFR as shown in Table 4. A strong correlation between overexpression of EGFR by ELISA and poor prognostic outcome is observed. Two studies totaling 143 patients show in multivariate analyses that EGFR overexpression significantly correlates with decreased DFS and OS, bigger tumor size, and advanced stage of disease [76, 78]. Finally, an analysis of EGFR overexpression in the serum of 32 patients treated with neoadjuvant chemotherapy does not find a correlation with response to treatment [77].
Relationship between EGFR overexpression and outcome in cervical cancer by ELISA detection
EGFR expression by radioligand binding assay
The radioligand binding assay (I 125–EGF) is used to determine the localization of accessible, unoccupied cell surface receptor. There are four studies by three scientific teams using this technique for cervical cancer [80–83]. Pfeiffer et al. [81] also developed a method of autoradiography to localize and quantify EGFR in cryostat section and compared their results with their radioligand assay and a flow cytometry technique with excellent correlations (r = 0.8 and 0.84). Binding capacity of EGFR usually decreases with cancer cell dedifferentiation. One pitfall is the amount of stromal tissue adherent to the tumor cells during the extraction of cellular membranes, which will dilute the binding capacity as stroma does not contain EGFR. These teams used different experimental conditions and different cut-off points to determine EGFR binding capacity on a Scatchard plot as shown in Table 5. In one study, increased EGFR binding capacity correlated with recurrence and poor OS [82], but in another study [83] no relationship between EGFR expression and OS was observed.
Relationship between EGFR binding capacity and prognostic outcome in cervical cancer by radioligand binding method
EGFR expression by flow cytometry
Flow cytometry allows precise quantitative determination of EGFR expression in individual tumor cells. A standard reference is required for quantification of EGFR expression. Selective determination of EGFR expression in tumor cells and in normal epithelial cells is accomplished by using microbeads to define the number of antigenic binding sites [84]. Using a novel multiparameter flow cytometric method, EGFR expression was found in 63%–72% of patients with cervical cancer [84, 85]. Cryopreserved cell suspensions are best used for this technique [86]. There was a good correlation between this method and autoradiography with I 125–EGF. Expression of EGFR is reduced in most cervical carcinomas compared with normal cervical epithelium, indicating that dedifferentiation of tumor cells may be associated with a reduction of EGFR expression instead of an increase [81]. Only 10% of cervical cancers show up-regulation by this method [84].
EGFR expression by RT-PCR
PCR is used to detect DNA amplification of EGFR, gene mutations, or gene polymorphisms as shown in Table 6. Specimens used for detection varied from tissue biopsies to circulating tumor cells in peripheral blood (mononuclear fraction) to cervical cytology. EGFR DNA amplification occurs in 12.5%–35.4% of cases with cervical cancer but does not correlate with EGFR overexpression [44, 56, 89]. Amplification is found on chromosomes 3q and 8q, but no amplification of the 7q chromosome has been demonstrated [90]. No correlation is found between EGFR gene alteration and poor clinicopathological parameters [14].
Relationship between EGFR DNA amplification and prognostic outcome in cervical cancer by PCR detection
EGFR mRNA expression levels measured by RT-PCR technique in blood were associated with advanced stages of cervical cancer (P = 0.049) [88]. This method is often associated with problems of RNA degradation and post-transcriptional modulations; therefore, EGFR levels based on mRNA measurement should be evaluated cautiously.
EGFR and treatment
To date, none of the information described previously has been clinically translated to rationally test EGFR inhibitors for the treatment of cervical cancer, with only a few small-scale clinical trials of some inhibitors and no correlative work to understand the clinical observations.
studies with single-agent cetuximab
There is only one prospective study of cetuximab in women with recurrent cervical cancer, which has accrued the first step of a Simon design and is not yet published (ClinicalTrials.gov identifier: NCT00499031). There was no response observed retrospectively in five refractory patients treated with cetuximab [91].
studies with single-agent tyrosine kinase inhibitors
Studies of gefitinib [92], imatinib [93], and lapatinib (a dual EGFR and HER-2 inhibitor) [94] were negative. The rationale for imatinib was based on the expression of the platelet-derived growth factor receptor [95]. These results are not unexpected given the lack of mutations on exons 18–21 [13, 14] and the lack of c-kit on cervical cancer cells [96].
studies of combinations of EGFR inhibitor with chemotherapy
The combination of cisplatin and cetuximab was studied in women with incurable advanced, persistent, or recurrent carcinoma of the cervix. Cisplatin 30 mg/m2 was administered on days 1 and 8 and cetuximab weekly. Of 69 assessable patients, the response rate was only 12% (8% for patients with prior chemotherapy and 17% for chemotherapy-naive patients) [97]. In a phase II trial, the combination of cisplatin and topotecan with cetuximab was terminated prematurely because of excessive toxicity. The overall response rate was 32% [98]. This response rate is similar to that of cisplatin-based doublets. Previous treatments and poor performance status of the patients may have contributed to the observed toxic effects, but a pharmacokinetic or pharmacodynamic interaction could not be excluded. A randomized phase II trial of carboplatin and paclitaxel with or without cetuximab, in advanced and/or recurrent cervical cancer (ClinicalTrials.gov identifier: NCT00997009), is ongoing.
studies of chemoradiation
Erlotinib combined with cisplatin and radiation for treatment of patients with locally advanced squamous cell cervical cancer induced a very high complete response rate of 91.3% [99]. The combination is feasible and well tolerated. The maximum tolerated dose was defined as 150 mg/day combined with weekly cisplatin at standard doses along with radiotherapy [100]. Studies are ongoing to assess the role of cetuximab in combination with chemoradiation (ClinicalTrials.gov identifier: NCT00104910, NCT00292955, and NCT00957411).
conclusions
In cervical cancer, EGFR expression as a biomarker of clinical behavior or a predictor of response to therapy is controversial. Half of the studies have shown a correlation with disease progression, poor prognosis, and poor response to therapy. However, the other half studies did not find such correlations.
What is known:
-
Normal cervical epithelium does express EGFR, which is mainly confined to the basal layer (Figure 1), but the expression may vary from none to very strong.
-
EGFR is associated with HPV infection but not correlated with HPV type. HPV infection may change the biology of EGFR expression by preventing EGFR degradation.
-
Most carcinoma cells express less EGFR than normal epithelial cells (Figure 1).
-
EGFR is expressed in squamous cell carcinoma in ∼50%–70% of cases. The expression is less for adenocarcinomas and adenosquamous carcinomas.
-
Given the lack of mutations on the tyrosine kinase domain of EGFR, inhibitors of this ATPase (erlotinib, lapatinib, and gefitinib) will probably be inefficient, and this was demonstrated in small phase II studies [92–95]. Anti-EGFR monoclonal antibodies may be more useful, as cervical cancer cells do express EGFR.
It is clear that EGFR is present in normal and cervical tumors with varying degrees of EGFR expression, which makes this cancer amenable to targeted therapy. The various pathways downstream of EGFR (the RAS–BRAF pathway) and the cross-talk happening through dimerization with other receptors (HER-2, HER-3, and HER-4) need to be better understood before anti-EGFR treatment can be applied intelligently. Despite the study of over a thousand cervical cancer samples described in this review, no practical molecular marker emerged as a predictor of response to anti-EGFR-targeted therapy. The absence of EGFR mutations makes response to small-molecule inhibitors such as erlotinib unlikely. The biology of EGFR in cervical cancer must be studied further to understand how to predict response to targeted treatments and how to match individual patient characteristics to the appropriate EGFR inhibitor. A true meta-analysis of the published data on IHC could be appropriate to help understand the relationship of EGFR membranous expression to prognosis. If monoclonal antibodies are made available for clinical studies of cervical cancer, it is essential that biomarker analyses be carried out at the same time. Such biomarkers include other members of the EGFR family, such as HER-3 and HER-4, and downstream molecular targets such as RAS, AKT, STAT3, or survivin, which interact with HPV proteins.
- © The Author 2011. Published by Oxford University Press on behalf of the European Society for Medical Oncology. All rights reserved. For permissions, please email: journals.permissions@oup.com
References
- 1.↵
- 2.↵
- 3.↵
- 4.↵
- 5.↵
- 6.↵
- 7.
- 8.↵
- 9.↵
- 10.↵
- 11.↵
- 12.↵
- 13.↵
- 14.↵
- 15.↵
- 16.↵
- 17.↵
- 18.↵
- 19.
- 20.↵
- 21.↵
- 22.↵
- 23.↵
- 24.
- 25.↵
- 26.↵
- 27.↵
- 28.↵
- 29.↵
- 30.↵
- 31.↵
- 32.↵
- 33.↵
- 34.↵
- 35.↵
- 36.↵
- 37.↵
- 38.↵
- 39.↵
- 40.↵
- 41.↵
- 42.↵
- 43.↵
- 44.↵
- 45.↵
- 46.↵
- 47.↵
- 48.↵
- 49.↵
- 50.↵
- 51.↵
- 52.↵
- 53.↵
- 54.↵
- 55.↵
- 56.↵
- 57.↵
- 58.↵
- 59.↵
- 60.↵
- 61.↵
- 62.↵
- 63.↵
- 64.↵
- 65.↵
- 66.↵
- 67.↵
- 68.↵
- 69.↵
- 70.↵
- 71.↵
- 72.↵
- 73.↵
- 74.↵
- 75.↵
- 76.↵
- 77.↵
- 78.↵
- 79.↵
- 80.↵
- 81.↵
- 82.↵
- 83.↵
- 84.↵
- 85.↵
- 86.↵
- 87.↵
- 88.↵
- 89.↵
- 90.↵
- 91.↵
- 92.↵
- 93.↵
- 94.↵
- 95.↵
- 96.↵
- 97.↵
- 98.↵
- 99.↵
- 100.↵
