METHODS OFFICIAL JOURNAL www.hgvs.org Competitive Amplification of Differentially Melting Amplicons (CADMA) Enables Sensitive and Direct Detection of All Mutation Types by High-Resolution Melting Analysis Lasse S. Kristensen, 1 ∗ Gitte B. Andersen, 1 Henrik Hager, 2 and Lise Lotte Hansen 1 1 Department of Biomedicine, Aarhus University, Denmark; 2 Department of Pathology, Aarhus University Hospital, Denmark Communicated by Mats Nilsson Received 10 February 2011; accepted revised manuscript 16 August 2011. Published online 7 September 2011 in Wiley Online Library (www.wiley.com/humanmutation).DOI: 10.1002/humu.21598 ABSTRACT: Sensitive and specific mutation detection is of particular importance in cancer diagnostics, prognostics, and individualized patient treatment. However, the major- ity of molecular methodologies that have been developed with the aim of increasing the sensitivity of mutation test- ing have drawbacks in terms of specificity, convenience, or costs. Here, we have established a new method, Com- petitive Amplification of Differentially Melting Amplicons (CADMA), which allows very sensitive and specific de- tection of all mutation types. The principle of the method is to amplify wild-type and mutated sequences simultane- ously using a three-primer system. A mutation-specific primer is designed to introduce melting temperature decreasing mutations in the resulting mutated amplicon, while a second overlapping primer is designed to amplify both wild-type and mutated sequences. When combined with a third common primer very sensitive mutation detec- tion becomes possible, when using high-resolution melting (HRM) as detection platform. The introduction of melt- ing temperature decreasing mutations in the mutated am- plicon also allows for further mutation enrichment by fast coamplification at lower denaturation temperature PCR (COLD-PCR). For proof-of-concept, we have designed CADMA assays for clinically relevant BRAF, EGFR, KRAS, and PIK3CA mutations, which are sensitive to, be- tween 0.025% and 0.25%, mutated alleles in a wild-type background. In conclusion, CADMA enables highly sen- sitive and specific mutation detection by HRM analysis. Hum Mutat 33:264–271, 2012. C  2011 Wiley Periodicals, Inc. KEY WORDS: cancer; CADMA; high-resolution melting; HRM; COLD-PCR This article was published online on 28 September 2011. Subsequently, an error in the annealing temperature of KRAS c.35G>C in Table 1 was identified and corrected on 18 November 2011. Additional Supporting Information may be found in the online version of this article. ∗ Correspondence to: Lasse S. Kristensen, Department of Biomedicine, Aarhus Uni- versity, Wilhelm Meyers All ´ e 4, DK-8000 Aarhus C, Denmark. E-mail: lasse@humgen .au.dk Contract grant sponsors: The Danish Cancer Society; The A.P. Møller and Chastine McKinney Møller Foundation; Max and Inge Wørzners Memorial Foundation. Introduction As the fields of cancer diagnostics and treatment move toward the era of personalized medicine, the identification of molecular mark- ers becomes increasingly important. Several well-characterized mu- tations have already been implemented in the clinics as markers of response to different treatment strategies and more are underway. However, the use of mutations as molecular markers in the clin- ics presents a challenge to the methodologies employed for their detection. When applying a particular methodology for routine diagnostic testing it must achieve a sufficient sensitivity without compromising specificity, while being convenient and cost efficient. Colorectal cancer (CRC) and non-small cell lung cancer (NSCLC) are among the most frequent causes of cancer deaths worldwide [Jemal et al., 2010]. However, new targeted therapies are contin- uously being developed and may contribute to increased survival rates among these patients in the years to come. The epidermal growth factor receptor (EGFR; MIM# 131550) is often overexpressed in CRC and NSCLC, and contributes to cancer development and progression by stimulating proliferation, angio- genesis, invasion, and survival of cancer cells. A subset of NSCLC patients carrying activating somatic mutations in the tyrosine ki- nase domain of EGFR show excellent response to EGFR tyrosine kinase inhibitors such as gefitinib and erlotinib [Lynch et al., 2004; Paez et al., 2004; Pao et al., 2004], and metastatic CRC patients with activating mutations in the KRAS gene (MIM# 190070) are unlikely to respond to treatment with monoclonal antibodies against EGFR such as panitumumab and cetuximab [Amado et al., 2008; Lievre et al., 2006]. Other oncogenic mutations, including the BRAF (MIM# 164757) c.1799T>A mutation and PIK3CA (MIM# 171834) mutations, as well as loss of PTEN (MIM# 601728) expression, may also be predictive markers of resistance to anti-EGFR monoclonal antibodies but require further evaluation before being incorporated in clinical practice [Lievre et al., 2010]. Activating mutations in KRAS and BRAF are found in approx- imately 40–50% and 10–15% of CRC patients, respectively, and found to be mutually exclusive [Rajagopalan et al., 2002]. How- ever, the frequency of detected KRAS mutations in clinical samples is influenced by the sensitivity of the method employed for their detection [Kobunai et al., 2010; Kristensen et al., 2010]. This may in part be caused by intratumor heterogeneity and contamination with wild-type DNA from normal cells, which typically are observed in infiltrating cancer types such as pancreatic, colorectal, breast, and lung cancer. Methods based on standard PCR and subsequent assays for mu- tation detection such as traditional Sanger sequencing suffer from a C  2011 WILEY PERIODICALS, INC.