Robust and accurate digital measurement for HER2 amplification in HER2 equivocal breast cancer diagnosis

www.nature.com/scientificreports OPEN Received: 31 March 2017 Accepted: 21 June 2017 Published online: 28 July 2017 Robust and accurate digital measurement for HER2 amplification in HER2 equivocal breast cancer diagnosis Yuefeng Wang1, Julia Y. S. Tsang2, Yongmei Cui1, Ji Cui3, Ying Lin4, Songli Zhao1,5, Patrick T. W. Law6, Sai Yin Cheung7, Enders K. O. Ng6, Gary M. K. Tse2 & Zunfu Ke1 Currently, there are no recommended alternative assays for HER2 cases deemed equivocal by immunohistochemistry and fluorescent in situ hybridization. Digital PCR (ddPCR), a highly accurate method to determine DNA copy number, could be a robust alternative for clinical HER2 diagnostics. HER2 and CEP17 copy numbers were quantified using two ddPCR platforms (QX200 and RainDrop) in 102 samples of invasive breast cancers. Compared to routine assays, ddPCR gave a sensitivity and specificity of 82.8% and 97.3% respectively, with a kappa value of 0.833 (p < 0.001). Moreover, the method proved to be robust as the results from two platforms was highly correlated (R2 = 0.91; Concordance rate = 97%; κ = 0.923, P < 0.001). Its performance was further tested on 114 HER2 equivocal cases in an independent validation cohort. 75% (21/28) of cases with HER2 amplification and 95% (82/86) of HER2 non-amplified case were classified as positive and negative by ddPCR respectively (κ = 0.709, P < 0.001). Notably, in the HER2 amplified cases, a lower percentage of HER2 positive cells could be related to the discordant results. Altogether, ddPCR is a robust alternative for clinical HER2 diagnostics. However, intratumoral heterogeneity of HER2 status still pose a challenge for HER2 analysis by ddPCR. Human epidermal growth factor receptor 2 (HER2) has been used as a predictive and prognostic marker in breast cancers. Its over-expression is predictive of response to HER2-targeted therapy, such as trastuzumab1. Evaluation of the HER2 status is now a standard of care for all primary breast cancers. The current techniques for HER2 testing include immunohistochemistry (IHC) for detection of protein expression and fluorescence in situ hybridization (FISH) for detection of gene amplification. Basing on the testing algorithm recommended by the American Society of Clinical Oncology (ASCO), in many laboratories, IHC is generally performed first. FISH is added if the result of IHC is equivocal2. Despite being widely used as a standard procedure, both tests have several limitations. Both are technically demanding, thus results may vary between different laboratories3. One of the underlying reasons is that they are easily affected by variations in testing conditions (such as fixation, reagent and assay protocol used2) thus compromising their reproducibility and accuracy. In addition, they are labor intensive and time consuming, especially for FISH. The manual and complicated procedure in FISH tests limit the number of patients that can be handled at any one time. Prolonged HER2 turn-around time was highlighted as a concern in the last national review in UK4. The manual scoring method could also be problematic. It is subjected to inter-observer variability, particularly in IHC which is done qualitatively rather than quantitatively5, 6. Another challenge is, in some cases, the test results can be indeterminate, either due to technical issues or the inherent features of the tumor. For instance, a proportion of the equivocal cases related to CEP17 aneusomy are due to focal amplification 1 Department of Pathology, The First Affiliated Hospital, Sun Yat-sen University, Guangzhou, China. 2Department of Anatomical and Cellular Pathology, Prince of Wales Hospital, The Chinese University of Hong Kong, Hong Kong, China. 3Department of Gastrointestinal Surgery, The First Affiliated Hospital, Sun Yat-sen University, Guangzhou, China. 4Breast Disease Center, The First Affiliated Hospital, Sun Yat-Sen University, Guangzhou, China. 5Department of Pathology, Sanshui General Hospital, Foshan, China. 6Pangenia Lifesciences Limited, Hong Kong, China. 7 Department of Pathology, Tuen Mun Hospital, Hong Kong, China. Yuefeng Wang, Julia Y. S. Tsang, Yongmei Cui and Ji Cui contributed equally to this work. Correspondence and requests for materials should be addressed to G.M.K.T. (email: ) or Z.K. (email: ) Scientific REPOrTS | 7: 6752 | DOI:10.1038/s41598-017-07176-x 1 www.nature.com/scientificreports/ Figure 1. HER2:CEP17 ratio by ddPCR in the development cohort. (A) dot plot showing the HER2:CEP17 ratio with different IHC score and (B) HER2− and HER2+ status. (C) ROC curve to determine the optimal cutoff. of the reference region rather than true polysomy. Inclusion of multiple controls can improve the results in those cases7. Standard FISH test, however, enumerated only CEP17 signal; therefore, a more robust alternative test which can perform multiplex analysis and deliver accurate, reliable and timely results is highly desirable. Digital PCR (dPCR) has been developed recently for absolute nucleic acid quantification and rare allele detection8. This technique is amenable to high throughput development and can be applied to FFPE tissue9, 10. It has also the power for multiplex analysis11. Additional controls/targets can be included and examined at the same time, allowing multitude of information to be obtained in a single run. The feasibility of dPCR in HER2 testing has been reported using frozen samples, FFPE samples and plasma DNA9, 10, 12, 13. A good concordance with the results from standard HER2 tests has been demonstrated. However, its performance in comparison to the standard techniques has not been well established. In these studies, only limited number of samples was included, thus precluding a more detailed and systematic analysis, particularly on HER2 equivocal cases. In this study, a systematic evaluation of the usefulness of dPCR in assessing HER2 status in breast cancers was performed. We have compared 2 different dPCR platforms. The accuracy of dPCR compared to the standard techniques was examined. Additionally, we have examined the usefulness of dPCR to determine HER2 amplification in IHC 2+ cases. Results Of 102 cases of invasive breast cancers in the developing cohort, 25, 26, 26 and 25 cases were tested IHC 0, 1+ , 2+ , and 3+ respectively. There were 22 FISH negative and four FISH positive cases among the IHC2+ samples. Based on the IHC and FISH data, 29 and 73 cases were HER2 positive and negative respectively. The results of RainDance Droplet Digital PCR (ddPCR) analysis for the development cohort were shown in Fig. 1. The median (range) ddPCR HER2:CEP17 ratio in patients with IHC 0, 1+ , 2+ and 3+ were 0.90 (0.36–1.36), 0.93 (0.82–2.00), 1.00 (0.32–2.30) and 4.79 (1.06–24.84) respectively (Table 1). There are significant differences for the ratio between IHC 0 to 2+ and 3+ cases (p > 0.001) (Fig. 1A). However, there was no difference between IHC 2+ and IHC 0 and 1+ . When HER2 status were defined with ddPCR and FISH, (...truncated)