Introduction

The Epidermal Growth Factor (EGF), one of the ligands of EGF receptor (EGFR), was first isolated from submandibular glands of male mice1. Known to stimulate the growth of several types of epithelial tissue, possesses strong mitogenic activity on tumor cells that converts this factor in an attractive target for designing antitumor strategies2. One of these EGF-targeted therapies is the Cuban vaccine CIMAvax-EGF® 3, a proven effective treatment for advanced non-small cell lung cancer (NSCLC). The vaccine induces anti-EGF antibodies that recognize the EGF in circulation, preventing its binding to EGFR, and disrupting this way the associated signal transduction cascade in cancer patients and ultimately cell proliferation. Studies of serum EGF concentrations ([EGF]) in Cuban patients treated with this vaccine revealed that high serum EGF levels are a factor of bad prognosis for NSCLC and at the same time a predictive biomarker of CIMAvax-EGF® efficacy4,5,6. However, high EGF concentrations in serum have been able to explain the bad prognosis and the good response to this vaccine not in all patients. Furthermore, the question about the capacity of EGF concentrations to discriminate between NSCLC patients and healthy individuals (its diagnostic value), suggestive of dependency on EGF in patients (tumors) and hence indicative of possible response to therapies directed to this growth factor or its receptor, has not been reliably answered by the scientific community. Thus far there are only a few reports available on this topic, some of which have published discrepant findings 7,8.

This manuscript is a mini-review of a recently published study, where the standardization of methodologies for blood and sera processing, along with the normalization of the estimates of EGF by platelets count, achieved the control of the key factors that influence the concentration of EGF in serum, helping to clarify its diagnostic value.

Discussion

Serum EGF Concentrations: The Causes of its Variability

The variability in published serum EGF levels and studies is presumably caused by the lack of harmonization and sometimes of standardization in the methodologies used for blood processing, sera collection, and EGF quantification. This methodological deficiency has a variety of expressions. Essentially, the majority of reports are unaware of the known dependency between serum EGF concentrations and the process of clot formation, during the incubation of blood for sera separation. Do not control the duration of this step9,10, neither the type of tubes employed for blood collection, which affects clotting times and then the release of EGF by platelets11,12,13, not even the temperature at which blood coagulates, which also influences this process. Another aspect that has probably increased the variability in reported values, thus contributing to the observed differences, is the single nucleotide polymorphism (SNP)14,15,16 of the EGF promotor gen. This polymorphism is functional, and modulates the expression levels of the molecule, thus provoking natural differences among individuals, which are not associated to their healthy condition or illness. This SNP, although extensively approached in specialized literature and linked to the risk of suffering the disease14,15,17,18, its severity19,20,21, prognosis20,21,22 and response to treatment23, in LC particularly, has not been considered in the comparison of serum EGF concentrations in patients, for the evaluation of its potential in prognosis and prediction; neither for its comparison with the respective
concentrations in healthy individuals, for the estimation of its possible diagnostic value. Differences in the selection of controls for the different cohorts of patients (the lack of control of confounding factors as age and gender) have also contributed to discrepant results8. Environmental factors could additionally contribute according to Pantsulaia et al.24.

Finally, the analysis of published data is more complicated due to the coexistence of different quantification platforms (ELISA, LUMINEX, microarrays, among others), some of which produce not comparable results.

Other studies have also revealed similar problems for several potential predictive biomarkers, which have slowed the functional transition of these markers to the oncology clinic25,26. The identified problems include, among others, inadequate attention to: the details of specimen collection; the definition of standard operating procedures; the requirements for analytical validation of assays and the statistical evaluation of the sources of assay variability27. Only the overcoming of these issues will facilitate the identification of predictive biomarkers and the development of predictive tests, capable of guiding novel systemic therapies of fundamental importance for advancing in the field of precision oncology.

Standardization and Normalization: the Solution for a Comprehensive Harmonization

Recently, the evaluation of serum EGF levels and platelets counts in 25 NSCLC patients (at diagnosis and after first-line therapy)28, employing an standardized methodology for separation of the sera10 and its quantification29, along with the normalization of estimated concentrations by platelets count, has allowed to elucidate platelet´s contribution to serum EGF concentrations in healthy individuals and NSCLC patients, aiding to clarify the diagnostic value of EGF levels in NSCLC. For quantification it was employed a validated ELISA29, calibrated against the EGF international standard 91/530 from National Institute for Biological Standards and Control (NIBSC). It is noteworthy that the UMELISA EGF® kit exhibits similar characteristics to other commercially available assays, in terms of precision, accuracy and dynamic range29. Furthermore, its estimations correlated very well with those obtained with the Human EGF Immunoassay Quantikine® ELISA (R&D Systems, Minneapolis, MN, USA), which is probably the most widely used kit for EGF estimations at the moment.

In this study each phlebotomy provided two sera, separated at 1h and 4h after venipuncture, and therefore two EGF concentrations: [EGF]1h and [EGF]4h, respectively28. The [EGF]1h, was interpreted as a good estimate of the actual concentration of free EGF in blood circulation, while the [EGF]4h represented a good measure of the average total stock of EGF in the blood sample of an individual. From these primary variables were constructed and studied several EGF-related variables, which were also interpreted in a simple manner from the biological point of view. The variable ratio r=[EGF]1h/[EGF]4h was interpreted as the EGF fraction from the total stock which is available in circulation. The stratification of patients using the variable r removes the variability associated to stratification by absolute serum EGF concentrations, which is derived from the natural differences among individuals, and inherent even to measurements obtained under standardized procedures. Due to these inter-individual differences, the concepts high/low regarding to serum EGF levels are actually relative (r). These concepts contain information which is not included in absolute EGF concentrations10,28. An estimated EGF concentration could be considered high/low depending on the percentage it represents from the total EGF of the individual. The variable difference d=[EGF]4h-[EGF]1h, which offers different but complementary information to variable r, was interpreted as the EGF stored in platelets (not available to circulation). Among the variables normalized by platelets count, [EGF]1h/platelets/L was interpreted as the average EGF contributed to circulation per platelet, and d/platelets/L as the average EGF stored per platelet (not in circulation).

Diagnostic Capacity of Studied EGF-Related Variables

In the commented study28 several variables achieved a successful discrimination between healthy individuals and NSCLC patients, at diagnosis and after chemoradiotherapy, when were evaluated by ROC (Receiver Operating Characteristic) analysis30,31. Differences were found between NSCLC patients and healthy individuals regarding the accessibility of EGF to circulation, but not regarding the total stock of EGF. It was observed a higher fraction of free EGF in the circulation of patients (r) and consequently a lower amount of EGF stored in platelets (d). Similarly, the analysis of normalized variables showed that the EGF per platelet accessible to circulation ([EGF]1h/platelets/L) was significantly higher in patients, before/after chemoradiotherapy. Conversely, the average total stocks of EGF per platelet ([EGF]4h/platelets/L) were equal in healthy controls and patients, also before/after chemoradiotherapy.

The comparison of cohorts through the normalized variables made more evident the differences between them, suggesting an altered relationship between EGF and platelets in NSCLC patients, as contrasted with healthy controls. Overall the results of this study suggest that the increase in free/accessible EGF in blood circulation is relevant to the biology of NSCLC, most likely because it reflects a higher accessibility to this tumoral growth factor.

Inference of EGF-Dependency in NSCLC

According to Rodriguez´s et al. results6, those NSCLC patients with high EGF concentrations have a poor prognosis and respond better to therapy with the CIMAvax-EGF® vaccine, which reduces the free EGF in the blood of treated patients. This suggests the existence of NSCLC variants with different underlying biology of the EGF/EGFR system, and patients with different levels of dependency on the availability of EGF in serum. Inspired in these findings, in Gonzalez-Perez et al. work28 the studied EGF-related variables were used for stratification purposes, trying to infer the dependency on EGF in different NSCLC patients. It was reasoned that those variables with a higher capacity for discrimination between patients and healthy controls might better capture the aberrant EGF biology in cancer patients. Therefore, these variables might be better for the identification of those patients probably more sensitive to therapies attempting to normalize EGF/EGFR interactions.

Stratification of Patients with Study Variables

Patients were stratified using the optimal cut-off values obtained according to Youden32, in ROC analysis for each study variable. Patients were predicted as highly EGF-dependent or vice versa, in each specific case, depending on its ranking with respect to the selected thresholds (cut-off values). To compare alternative stratifications, its percentages of overlapping in predictions were calculated by pairs of variables28. For the sake of comparison, the stratification method reported by Rodriguez et al.6 for the identification of patients more benefitted from CIMAvax-EGF® vaccine, was also included. In Rodriguez´s method, patients with [EGF] above the median of the studied population appear to carry tumors apparently more EGF-dependent. Making a parallel with Rodriguez´s method, in the revised study the cut-off values were also set according to the medians of either the [EGF] at 1h and 4h. Although in Rodriguez´s study the time of sera separation was not controlled, it was likely close to the 4h processing in González-Pérez et al. study, given the similarity between the corresponding reported medians of [EGF] after chemoradiotherapy (873 pg/mL and 829 pg/mL, respectively).

Interestingly, the classifications by the medians of [EGF] were quite different to those obtained with the normalized variables at diagnosis. However, the variable [EGF]1h/platelets/L appears to classify patients quite similarly to variables d/platelets/L and [EGF]1h. After first-line chemoradiotherapy, the normalized variables showed a remarkably high coincidence in patient´s classification, and moderate overlappings with the classification by the median of [EGF]1h. However, the classification by the median of [EGF]4h showed a very low overlapping with the selections of any other study variable, including the classification achieved by the median of [EGF]1h. Therefore, the normalized variables are quite complementary and therefore will provide similar classifications of patients, but different to those obtained when the median of [EGF] is used as cut-off, as proposed Rodriguez et al.6, especially when using [EGF]4h, a variable representing a good measure of the average total stock of EGF in the blood sample of individuals, which was not able to discriminate cases from controls in the revised study. Therefore, although in Rodriguez´s approach [EGF]4h could explain in some measure the prognosis of patients and the vaccine´s efficacy after chemoradiotherapy6, the normalized variables, which were able to discriminate in that scenery, might be more valuable than [EGF]4h for these purposes.

Conclusions

By accepting that the EGF levels in serum are influenced by several factors, which are mainly expressed in the process of sera separation, is comprehensible that the standardization of this procedure is crucial to guarantee results valid, reliable and comparable between laboratories. Therefore, besides a validated quantitative assay, calibrated against the current approved international standard for EGF, standardization and harmonization of key procedures are needed.

The methodology applied in the discussed manuscript for the separation of the sera, the estimation of serum EGF levels, its normalization and interpretation of results, allowed to elucidate the diagnostic value of EGF in NSCLC. Additionally, the study revealed that patients suffering NSCLC differ from healthy individuals not by the total stock of EGF, but by its higher accessibility to serum. Furthermore, the proposed approach added value to the efforts of finding an efficacy biomarker for CIMAvax-EGF® vaccine, so far approved as a second-line therapy. Overall, the study revealed that the normalized variables might be potential biomarkers in NSCLC, and good candidate biomarkers of efficacy for CIMAvax-EGF® immunotherapy. Further studies are needed to evaluate the usefulness of these markers on its predictive value to select good responders to treatment with therapies targeting the EGF/EGFR system, and also to estimate its effectiveness in prognosis, monitoring of therapy and evaluation of response, in NSCLC and other epithelial cancers. Finally, the proposed methodology, and particularly the normalization of EGF levels by platelets count, might help to better understand the role of EGF in other diseases where this growth factor is also involved^{33,34,35,36,37,38,39}.

References

1. Cohen S. Isolation of a mouse submaxillary gland protein accelerating incisor eruption and eyelid opening in the new-born animal. J Biol Chem., 1962; 237:1555-1562.
2. Arteaga C. Targeting HER1/EGFR: a molecular approach to cancer therapy. Semin Oncol., 2003; 30:3-14.
3. Crombet T, Camilo P, Neninger E, et al. A CIMAvax EGF (EGF-P64K) vaccine for the treatment of non-small-cell lung cancer. Expert Rev Vaccines., 2015; early online 1-9.
4. Hall R, Gray J, Chiappori A. Beyond the Standard of Care: A Review of Novel Immunotherapy Trials for the Treatment of Lung Cancer. Cancer Control., 2013; 20:22-31.
5. McCarthy F, Roshani R, Steele J, Hagemann T. Current clinical immunotherapy targets in advanced nonsmall cell lung cancer (NSCLC). J Leukocyte Biol., 2013; 94:1-6.
6. Rodriguez P, Popa X, Martínez O, Mendoza S, Santiesteban E, et al. A Phase III Clinical Trial of the Epidermal Growth Factor Vaccine CIMAvax-EGF as Switch Maintenance Therapy in Advanced Non–Small Cell Lung Cancer Patients. Clin Cancer Res., 2016; 22:3782-3790.
7. Lemos-González Y, Rodríguez-Berrocal F, Cordero O, Gómez C and Páez de la Cadena M. Alteration of the serum levels of the epidermal growth factor receptor and its ligands in patients with non-small cell lung cancer and head and neck carcinoma. Br J Cancer., 2007; 96:1569-1578.
8. Blanco-Prieto S, Vázquez-Iglesias L, Rodríguez-Girondo M, Barcia-Castro L, Fernández- Villar A, et al. Serum Calprotectin, CD26 and EGF to Establish a Panel for the Diagnosis of Lung Cancer. PLoS ONE., 2015; 10(5):e0127318.
9. Pesonen K, Viinikka L, Myllyla G, Kiuru J and Perheentupa J. Characterization of Material with Epidermal Growth Factor Immunoreactivity in Human Serum and Platelets. J Clin Endocrinol Metab., 1989; 68:486-491.
10. Gonzalez-Perez I, Cáceres H, Carr A, León K. Measurement of serum EGF levels, a methodological approach: learning what means “low-/high- concentration of EGF in serum.” Some clinical implications. J Mol Biomark Diagn., 2017; 8: 3.
11. Ramström S. Clotting time analysis of citrated blood samples is strongly affected by the tube used for blood sampling. Blood Coagul Fibrin., 2005; 16:447-452.
12. Banks R, Stanley A, Cairns D, Barrett J, Clarke P, et al. Influences of blood sample processing on low-molecular-weight proteome identified by surface-enhanced laser desorption/ionization mass spectrometry. Clin Chem., 2005; 51:1637-1649.
13. Yucel A, Karakus R, Cemalettin A. Effect of blood collection tube types on the measurement of human epidermal growth factor. J Immunoassay Immunochem., 2007; 28:47-60.
14. De Mello R. EGF+61 A>G Polymorphisms and Lung Cancer Risk: Future Directions. Pulmonar Respirat Med., 2012; S5: e001.
15. Li T, Ren K, Liu P. Meta-Analysis of Epidermal Growth Factor Polymorphisms and Cancer Risk: Involving 9,779 Cases and 15,932 Controls. DNA Cell Biol., 2012; 31:568-574.
16. Shahbazi M, Pravica V, Nasreen N, et al. Association between functional polymorphism in EGF gene and malignant melanoma. Lancet., 2002; 359:397-401.
17. Zhu J, Meng X, Yan F. A functional epidermal growth factor (EGF) polymorphism, EGF serum levels and renal cell carcinoma risk in a Chinese population. J Hum Genet., 2010; 55:236-240.
18. Costa B, Ferreira P, Costa S, et al. Association between Functional EGF+61Polymorphism and Glioma Risk. Clin Cancer Res., 2007; 13:2621-2626.
19. Samy M, Al-Haddad O, Nada A, et al. Association of serum levels of epidermal growth factor with disease severity in patients with unresectable hepatocellular carcinoma. Hepatoma Res., 2016; 2:18-25.
20. Bhowmick D, Zhuang Z, Wait S, et al. A Functional Polymorphism in the EGF Gene Is Found with Increased Frequency in Glioblastoma Multiforme Patients and Is Associated with More Aggressive Disease. Cancer Res., 2014; 64:1220-1223.
21. Teixeira A, Ribeiro R, Cardoso D, et al. Genetic Polymorphism in EGF Is Associated with Prostate Cancer Aggressiveness and Progression-Free Interval in Androgen Blockade-Treated Patients. Clin Cancer Res., 2008; 14:3367-3371.
22. Yoshiya S, Fujimoto Y, Bekki Y, et al. Impact of epidermal growth factor single-nucleotide polymorphism on recurrence of hepatocellular carcinoma after hepatectomy in patients with chronic hepatitis C virus infection. Cancer Sci., 2014; 105:646-650.
23. Hu-Lieskovan S, Vallbohmer D, Zhang W, et al. EGF61 Polymorphism Predicts Complete Pathologic Response to Cetuximab-Based Chemoradiation Independent of KRAS Status in Locally Advanced Rectal Cancer Patients. Clin Cancer Res., 2011; 17:5161-5169.
24. Pantsulaia I, Trofimov S, Kobyliansky E, Livshits G. Heritability of circulating growth factors involved in the angiogenesis in healthy human population. Cytokine., 2004; 27:152-158.
25. McShane L, Cavenagh M, Lively T, Eberhard D, Bigbee W, et al. Criteria for the use of omics-based predictors in clinical trials: explanation and elaboration. BMC Medicine., 2013; 11:220-241.
26. Doroshow J, Kummar S. Translational research in oncology-10 years of progress and future prospects. Nature Reviews, Clin Oncol., 2014; 11:649-662.
27. Parkinson D, McCormack R, Keating S, Gutman S, Hamilton S, et al. Evidence of clinical utility: an unmet need in molecular diagnostics for patients with cancer. Clin Cancer Res., 2014; 20:1428-1444.
28. González-Pérez I, Cáceres H, Rodríguez P, Carr A, León K. Normalized serum EGF levels as a potential biomarker in non-small cell lung cancer: the role of platelets. J Mol Biomark Diagn., 2018; 9:5, DOI: 10.4172/2155-9929.1000402.
29. Castells E, Del Valle R, González E, Melchor A, Pérez P, González-Pérez I, et al. An enzyme immunoassay for determining epidermal growth factor (EGF) in human serum samples using an ultramicroanalytical system. J Immunoassay Immunochem., 2016; DOI:10.1080/15321819.2016.1236729.
30. DeLong E, DeLong D, Clarke-Pearson D. Comparing the areas under two or more correlated receiver operating characteristic curves: A nonparametric approach. Biometrics., 1988; 44:837-845.
31. Goksuluk D, Korkmaz S, Zararsiz G, Karaağaoğlu A. EasyROC: An Interactive Web-tool for ROC Curve Analysis Using R Language Environment. The R Journal., 2016; 8:213-230.
32. Youden W. “Index for rating diagnostic tests”. Cancer., 1950; 3:32-35.
33. Oxford G, Tayaria L, Barfoot M, Peck A, Tanaka Y. Salivary EGF levels reduced in diabetic patients. J Diabetes Its Complicat., 2000; 14:140-145.
34. Humpel C, Hochstrasser T. Cerebrospinal fluid and blood biomarkers in AD. World J Psychiatr., 2011; 1:8-18.
35. Marksteiner J, Kemmler G, Weiss E, Knaus G, Ullrich C. Five out of 16 plasma signaling proteins are enhanced in plasma of patients with mild cognitive impairment and Alzheimer’s disease. Neurobiol Aging., 2011; 32:539-540.
36. Hochstrasser T, Ehrlich D, Marksteiner J, Sperner-Unterweger B, Humpel C. Matrix metalloproteinase-2 and epidermal growth factor are decreased in platelets of AD. Curr Alzheimer Res., 2012; 9:982-989.
37. Plagg B, Marksteiner J. Platelet dysfunction in hypercholesterolemia mice, two Alzheimer’s disease mouse models and in human patients with Alzheimer’s disease. Biogerontology., 2015; 16: 543-558.
38. Tian W, Zhang J, Zhang K, Yang H, Sun Y. A study of the functional significance of epidermal growth factor in major depressive disorder. Psyc Genet., 2012; 22:161-167.
39. Suzuki K, Hashimoto K, Iwata Y, Nakamura K, Tsujii M, et al. Decreased serum levels of epidermal growth factor in adult subjects with high-functioning autism. Biol Psychiatry., 2007; 62:267-269.