From research to clinical application multi-parameter testing: Marker panels for the early detection of complex diseases

JMB 2009; 28 (4) DOI: 10.2478/v10011-009-0024-9 UDK 577.1 : 61 ISSN 1452-8258 JMB 28: 279–284, 2009 Review article Pregledni ~lanak FROM RESEARCH TO CLINICAL APPLICATION MULTI-PARAMETER TESTING: MARKER PANELS FOR THE EARLY DETECTION OF COMPLEX DISEASES OD ISTRA@IVANJA DO KLINI^KE PRIMENE: PANELI MARKERA ZA RANU DETEKCIJU SLO@ENIH BOLESTI Norbert Wild, Johann Karl, Bernhard Risse Department of New Technologies Reagents/Systems Roche Professional Diagnostics, Roche Diagnostics GmbH, Penzberg, Germany Summary: Multi Parameter Analysis can open novel diagnostic opportunities for the early diagnosis and screening of multimodal diseases like cancer. Single proteins have so far failed to describe such complex diseases. Being able to screen with a set of analytes is one promising way to overcome the present limitations. Various marker identification tools including proteomics approaches have been successfully applied to identify new screening markers for early detection of colorectal cancer (CRC). CRC is one of the most incident cancers worldwide and early detection is clearly a key factor in reducing mortality from CRC. Several screening methods are recommended, including colonoscopy, fecal occult blood testing (FOBT) and fecal DNA analysis. Of these annual stool testing with the guaiac based FOBT is most often applied, in spite of limitations such as low sensitivity and dietary influences. Though procedures with improved performance eg. immunological FOBT are available, a screening assay for CRC in serum that could easily be integrated in any health check-up would be highly welcome. A positive result of such a test would trigger a follow-up colonoscopy for an exact diagnosis. In this review we will cover aspects of marker identification strategies and describe a well structured marker validation process that is based on clinically characterized sample materials. Finally the value of analytical multi-parameter platforms enabling the combination of multi markers in routine diagnostics settings is outlined. An appropriate multi-parameter immunochemistry platform concept, currently developed under the working name »IMPACT« will be introduced. Kratak sadr`aj: Analiza vi{e parametara (multi-parameter analysis) mo`e doneti nove dijagnosti~ke mogu}nosti za ranu dijagnozu i testiranje multimodalnih bolesti poput raka. Pojedina~ni proteini dosad nisu uspeli da opi{u tako slo`ene bolesti. Mogu}nost testiranja pomo}u seta analita najavljuje prevazila`enje postoje}ih ograni~enja. Razli~ite alatke za identifikaciju markera uklju~uju}i proteomske pristupe uspe{no se primenjuju za identifikaciju novih markera za rano otkrivanje kolorektalnog kancera. Kolorektalni kancer je jedan od naj~e{}ih kancera u celom svetu a rana detekcija je klju~ni faktor za smanjenje mortaliteta od te vrste raka. Preporu~eno je nekoliko metoda za testiranje, uklju~uju}i kolonoskopiju, analizu okultnog krvarenja u stolici i analizu DNK u stolici. Naj~e{}e se primenjuje godi{nje testiranje stolice metodom gvajak zasnovano na analizi okultnog krvarenja u stolici, uprkos ograni~enjima kao {to su niska senzitivnost i uticaj ishrane. Mada postoje postupci s pobolj{anim performansama, npr. imunolo{ko testiranje okultnog krvarenja u stolici, skrining test na kolorektalni rak koji bi se lako integrisao u svaki zdravstveni pregled bio bi veoma korisan. Pozitivan rezultat takvog testa ukazao bi na potrebu za kolonoskopijom radi precizne dijagnoze. U ovom radu bi}e razmotreni aspekti strategija za identifikaciju markera i opisan valjano struktuiran postupak validacije markera zasnovan na klini~ki okarakterisanim uzorcima materijala. Najzad, bi}e nazna~ena vrednost analiti~kih platformi s vi{e parametara koje omogu}uju kombinovanje multimarkera u rutinskoj dijagnostici. Bi}e predstavljen i koncept odgovaraju}e imunohemijske platforme s vi{e parametara koja se trenutno razvija pod radnim nazivom »IMPACT«. Keywords: colorectal cancer, multi-parameter analysis, early diagnosis, marker combination, multivariate analysis, protein array Klju~ne re~i: kolorektalni kancer, analiza vi{e parametara, rana dijagnoza, kombinovanje markera, multivarijatna analiza, proteinski skup Address for correspondence: Bernhard Risse Department of New Technologies Reagents/Systems Roche Professional Diagnostics, Roche Diagnostics GmbH, Penzberg, Germany 280 Wild et al.: From research to clinical application Starting a substantial marker identification program, proteomics studies have been the preferred approach to identify promising marker candidates. In addition, there are alternatives that should also be considered. DNA and RNA methods could complement proteomics studies, while extended pathway analysis and information from public domain might identify additional markers. In the overall perception, proteomics is strongly associated with the discovery phase, that will only be the first step if the whole process of marker discovery is considered. Proteomics aims to identify proteins that are »different« in various states of a biological system classified as diseased and »healthy«. In consequence the discovery phase is overrated if a program in proteomics is limited to the identification of novel validated biomarkers with a clearly defined position in in vitro diagnostics. The absolute number of discovery hits cannot be considered the primary goal of proteomic approaches in the in vitro diagnostic setting, but the true challenge will be the carefully planned assessment of the selected candidates. Only the subsequent validation phase, not the discovery activities, will reveal whether a marker has the potential to fill the existing sensitivity and/or specificity gaps. The validation strategies will vary depending on the intended positioning of the marker. A screening marker will have to meet completely different performance criteria than, for example, a monitoring marker or an efficacy marker modulated by the drug being administered. New markers will be used to complement already existent products to significantly advance the specific intended use. This is also an economic necessity because novel replacements are bound to face high »entry barriers« into the market. Currently regulatory agencies give little advice as to the specific benchmark(s) against which the single novel marker must be compared and a real challenge for the future will be gaining approval for marker combinations created with mathematical models (»disease algorithms«) based on multi-parametric testing. Given the intricacy of some diseases, it is quite likely that a single protein on its own will not be sufficient. A »golden bullet approach« seems to be an unrealistic scenario. Thus, it is necessary to combine more than one biomarker in a disease algorithm. It is our goal that the biomarker should ultimately be detected in serum/plasma using immunological detection (ELISA) formats. If we assume that the marker candidate was identified by comparing sets of tumor and healthy (...truncated)