Review
Prognostic versus predictive value of biomarkers in oncology

https://doi.org/10.1016/j.ejca.2008.03.006Get rights and content

Abstract

Numerous options are currently available for tumour typing. This has raised intense interest in the elucidation of prognostic and predictive markers. A prognostic biomarker provides information about the patients overall cancer outcome, regardless of therapy, whilst a predictive biomarker gives information about the effect of a therapeutic intervention. A predictive biomarker can be a target for therapy. Amongst the genes that have proven to be of relevance are well-known markers such as ER, PR and HER2/neu in breast cancer, BCR–ABL fusion protein in chronic myeloid leukaemia, c-KIT mutations in GIST tumours and EGFR1 mutations in NSCLC. Several reasons for the difficult elucidation of new markers will be addressed including the involvement of cellular pathways in tumour biology instead of single genes and interference in disease outcome as a result of anticancer therapies. Future perspectives for the development of prognostic and predictive markers will be given.

Introduction

With the availability and application of various treatment modalities, survival amongst cancer patients has improved over the past decades. However, there are still many patients who receive anticancer therapy from which they do not benefit whilst they do experience toxicity. In recent years, a widespread search for new, tumour biology driven therapeutics has started. This has raised intense interest in the elucidation of corresponding prognostic and predictive biomarkers in order to improve outcome by better patient selection for an anticancer treatment. A biomarker is defined as a characteristic that is objectively measured and evaluated as an indicator of normal biologic processes, pathogenic processes or pharmacological responses to a specified therapeutic intervention.1 Biomarkers can be determined in numerous ways, for example, in easy obtainable body fluids serving as surrogate biological assay, like plasma, serum or urine. But also more invasive techniques requiring tumour tissue for immunohistochemistry as well as DNA and RNA analyses are widely used. A prognostic biomarker provides information about the patients overall cancer outcome, regardless of therapy. The presence or the absence of such a prognostic marker can be useful for the selection of patients for a certain treatment, but does not predict the response to this treatment. Prognostic biomarkers can be separated in two groups: biomarkers that give information on recurrence in patients who receive curative treatment and biomarkers that correlate with the duration of (progression free) survival in patients with metastatic disease. According to a NIH Consensus Conference, a clinical useful prognostic marker must be a proven independent, significant factor, that is easy to determine and interpret and has therapeutic consequences.2 A biomarker with predictive value gives information on the effect of a therapeutic intervention in a patient. A predictive biomarker can also be a target for therapy. One can distinguish upfront and early predictive markers. The first can be used for patient selection and the second provides information early during therapy.

The current interest in marker determination is boosted by the discovery of genes that have proven to be of clinical relevance such as the oestrogen receptor (ER), the progesterone receptor (PR) and HER2/neu in breast cancer, BCR–ABL fusion protein in chronic myeloid leukaemia, c-KIT mutations in gastrointestinal stromal tumours (GIST) and epidermal growth factor receptor 1 (EGFR1) mutations in non-small cell lung cancer (NSCLC). These genes all seem to be key regulators of development, growth and proliferation in the respective tumour types. Euphoria is now somewhat tempered because the discovery of other the so-called promising markers translates rather slowly into clinical applicability. One reason for this is the fact that the course of most of malignancies is the consequence of a number of essential alterations in tumour cells rather than a single mutation.3 In addition, the limited size of most studies and variable techniques used for marker determination plays a role. Often initially reported promising results are not reproducible. In an attempt to optimise biomarker studies, Hayes and colleagues proposed a tumour marker utility grading system (TMUGS). For each biomarker a grade of utility is assigned, accompanied by a level of evidence (LOE) that scores the quality of the research. The LOE categories range from I to V. Level V evidence is obtained from case reports and clinical experience and is considered weak, whilst level I evidence is derived either from at least one prospective randomised controlled trial specifically designed to test the marker or from a meta-analysis and/or overview of level II or III studies and is considered definitive.4 In addition, a consortium of the National Cancer Institute-European Organisation for Research and Treatment of Cancer (NCI-EORTC) reported in several journals a guideline for reporting tumour marker prognostic studies (REMARK).5

In this review, the progress in the development of biomarkers in solid tumours will be addressed such as, involvement of cellular pathways in tumour biology instead of single genes and interference in disease outcome as a result of anticancer therapies. Examples of both well-known biomarkers and potential new discoveries (summarised in Table 1) will form the basis for a discussion below on the present knowledge and new avenues for the development of prognostic and predictive biomarkers.

Section snippets

ER/PR in breast cancer

ER and/or PR expression is an independent prognostic factor in breast cancer. Patients with ER and/or PR positive tumours have a better survival than hormone receptor negative tumours, with a 5-year overall survival (all stages) of 83% in the ER+/PR+ group versus 69% in the double negatives (LOE III).6 High cellular expression of ER and PR predicts benefit from endocrine therapy in the adjuvant and metastatic setting (LOE I).7 Tumour hormone receptor status is, therefore, routinely assessed in

HER2/neu in breast cancer

Another relevant biomarker in breast cancer patients is HER2/neu. The HER2/neu gene amplification leads to overexpression of its receptor on the cell membrane. This results in increased proliferation and angiogenesis, and inhibition of apoptosis. HER2/neu positive tumours are more aggressive and have, therefore, a worse prognosis compared to negative tumours. In this respect, HER2/neu in node positive breast cancer is of prognostic value (LOE II). For the node negative, HER2/neu positive group

Prognostic biomarkers for the relapse of breast cancer

Decision making about adjuvant systemic treatment for breast cancer is based on nodal status, tumour grade, tumour size, tumour hormone receptor and HER2/neu status, age and co-morbidity. Prognostic biomarkers that could provide better information on risk of relapse could spare many patients chemotherapy toxicity without compromising survival. Amongst several initiatives Buyse and colleagues validated a 70-gene signature for node negative breast cancer patients that has independent prognostic

c-KIT in GIST

Several features are evaluated over the last few years to determine malignant behaviour of GISTs. Of known relevance are tumour size and mitotic index, which are used to classify the biologic behaviour of GIST.30 The majority of GISTs are characterised by mutations in either the proto-oncogene c-KIT or the platelet-derived growth factor receptor alpha (PDGFRα). Interestingly, patients with mutation in the c-KIT-gene in exon 11 have a better prognosis as compared to those who lack a mutation or

EGFR1 and K-ras in NSCLC and colorectal cancer (CRC)

In NSCLC and CRC, biomarkers of interest are EGFR1 and the K-ras oncogene. EGFR1 is overexpressed in multiple cancer types and is one of the targets in the treatment of NSCLC and metastatic CRC. The EGFR pathway plays a role in several cellular functions, including regulation of cell proliferation, migration and differentiation (Fig. 1). The prognostic value of EGFR1 protein expression is extensively studied in NSCLC and CRC patients but no definitive association between EGFR1 expression and

TRAIL receptors

Tumour necrosis factor (TNF)-related apoptosis-inducing ligand (TRAIL or Apo2L) induces apoptosis in a wide variety of tumour cell lines without causing toxicity to normal cells and is, therefore, a potential attractive agent. TRAIL binds the death receptors TRAIL-R1 (DR4) and TRAIl-R2 (DR5) and initiates the apoptotic pathway. DR4 and DR5 are expressed on most tumour cells. In contrast to, e.g., HER2 and EGFR1, this receptor has to be activated and not inhibited in order for cells to go into

VEGF and renal cell carcinoma (RCC)

Even very small tumours require angiogenesis to provide nutrients and oxygen for survival. There is a close interaction between tumour cells that produce pro-angiogenic growth factors, like vascular endothelial growth factor (VEGF) and PDGF, and endothelial cells expressing growth factor receptors. The stimulation of endothelial cells results in proliferation and migration and eventually in the formation of new vessels. Clear cell RCC provides a unique model for studying angiogenesis because of

Drug induced toxicity as a predictive biomarker

Interestingly a number of studies showed that the effect of a drug on normal tissues can be used as a biomarker. In both a phase II and a phase III study evaluating the antitumour activity of cetuximab in metastatic CRC, skin rash was strongly related to response and survival.68, 69 Similar results were found for erlotinib in NSCLC.70 Toxicity might thus be used to titrate drugs to effective doses as is done in the EVEREST study in CRC patients. Patients with no or mild skin toxicity after 22

Discussion

A confusing mix-up exists of the terms prognostic and predictive biomarkers. This is partially due to the fact that predictive and prognostic biomarkers are frequently exchanged. In addition, during therapy or as a result of therapy initial factors can vary in their presence and actual levels, e.g. a strong prognostic factor can be neutralised as a consequence of treatment (HER2/neu).

Despite a growing number of publications about biomarkers that give information on disease outcome, the best

Conflict of interest statement

None declared.

References (77)

  • Biomarkers Definitions Working Group

    Biomarkers and surrogate endpoints: preferred definitions and conceptual framework

    Clin Pharmacol Ther

    (2001)
  • NIH consensus conference. Treatment of early-stage breast cancer. JAMA...
  • D.F. Hayes et al.

    Tumor marker utility grading system: a framework to evaluate clinical utility of tumor markers

    J Natl Cancer Inst

    (1996)
  • L.M. McShane et al.

    Reporting recommendations for tumor marker prognostic studies (REMARK)

    J Natl Cancer Inst

    (2005)
  • V.R. Grann et al.

    Hormone receptor status and survival in a population-based cohort of patients with breast carcinoma

    Cancer

    (2005)
  • Early Breast Cancer Trialists’ Collaborative Group (EBCTCG)

    Effects of chemotherapy and hormonal therapy for early breast cancer on recurrence and 15-year survival: an overview of the randomised trials

    Lancet

    (2005)
  • A. Franco et al.

    Discordance in estrogen (ER) and progestin receptor (PR) status between primary metastatic breast cancer: a meta-analysis

    J Clin Oncol

    (2004)
  • T. Kuukasjarvi et al.

    Loss of estrogen receptor in recurrent breast cancer is associated with poor response to endocrine therapy

    J Clin Oncol

    (1996)
  • L. Harris et al.

    American Society of Clinical Oncology 2007 update of recommendations for the use of tumor markers in breast cancer

    J Clin Oncol

    (2007)
  • H.M. Linden et al.

    Quantitative fluoroestradiol positron emission tomography imaging predicts response to endocrine treatment in breast cancer

    J Clin Oncol

    (2006)
  • B.A. Gusterson et al.

    Prognostic importance of c-erbB-2 expression in breast cancer. International (Ludwig) Breast Cancer Study Group

    J Clin Oncol

    (1992)
  • A.K. Tandon et al.

    HER-2/neu oncogene protein and prognosis in breast cancer

    J Clin Oncol

    (1989)
  • N. Kroger et al.

    Prognostic and predictive effects of immunohistochemical factors in high-risk primary breast cancer patients

    Clin Cancer Res

    (2006)
  • G.M. Clark et al.

    Follow-up study of HER-2/neu amplification in primary breast cancer

    Cancer Res

    (1991)
  • M.C. Paterson et al.

    Correlation between c-erbB-2 amplification and risk of recurrent disease in node-negative breast cancer

    Cancer Res

    (1991)
  • J. Winstanley et al.

    The long term prognostic significance of c-erbB-2 in primary breast cancer

    Br J Cancer

    (1991)
  • E.H. Romond et al.

    Trastuzumab plus adjuvant chemotherapy for operable HER2-positive breast cancer

    New Engl J Med

    (2005)
  • C.L. Vogel et al.

    Efficacy and safety of trastuzumab as a single agent in first-line treatment of HER2-overexpressing metastatic breast cancer

    J Clin Oncol

    (2002)
  • H. Joensuu et al.

    Adjuvant docetaxel or vinorelbine with or without trastuzumab for breast cancer

    New Engl J Med

    (2006)
  • Y. Gong et al.

    Comparison of HER-2 status determined by fluorescence in situ hybridization in primary and metastatic breast carcinoma

    Cancer

    (2005)
  • S. Meng et al.

    HER-2 gene amplification can be acquired as breast cancer progresses

    Proc Natl Acad Sci USA

    (2004)
  • J. Zidan et al.

    Comparison of HER-2 overexpression in primary breast cancer and metastatic sites and its effect on biological targeting therapy of metastatic disease

    Br J Cancer

    (2005)
  • L.N. Harris et al.

    Comparison of methods of measuring HER-2 in metastatic breast cancer patients treated with high-dose chemotherapy

    J Clin Oncol

    (2001)
  • P.J. Perik et al.

    Indium-111-labeled trastuzumab scintigraphy in patients with human epidermal growth factor receptor 2-positive metastatic breast cancer

    J Clin Oncol

    (2006)
  • M. Scaltriti et al.

    Expression of p95HER2, a truncated form of the HER2 receptor, and response to anti-HER2 therapies in breast cancer

    J Natl Cancer Inst

    (2007)
  • M. Buyse et al.

    Validation and clinical utility of a 70-gene prognostic signature for women with node-negative breast cancer

    J Natl Cancer Inst

    (2006)
  • C.D. Fletcher et al.

    Diagnosis of gastrointestinal stromal tumors: a consensus approach

    Int J Surg Pathol

    (2002)
  • S. Singer et al.

    Prognostic value of KIT mutation type, mitotic activity, and histologic subtype in gastrointestinal stromal tumors

    J Clin Oncol

    (2002)
  • Cited by (266)

    View all citing articles on Scopus
    View full text