C. Steiner (1,2), J.-C. Tille (2), J. Lamerz (1), S. Kux van Geijtenbeek (1), T.A. McKee (2), M. Venturi (1), L. Rubbia-Brandt (2), D. Hochstrasser (2), P. Cutler (1), P. Lescuyer (2) A. Ducret (1)
(1) F. Hoffmann-La Roche Ltd (2) Geneva University Hospitals
We describe in this study a selected reaction monitoring (SRM) assay for the quantification of HER2 in FFPE tissues. The assay’s six best candidate peptides showed a linear response over a calibration range of 0.012 to 100 fmol on column (R2: 0.99–1.00) and a lower limit of quantification of 0.155 fmol on column. HER2 peptides were quantified in a cohort of 40 breast tumors expressing different HER2 levels (FISH ratio: 0-2, 2-4, 4-10 and >10 respectively). The SRM assay showed good analytical performance and a high agreement with IHC and FISH data. Furthermore, after normalization for tissue sample size, SRM peptide measurements were able to correctly predict 90% of HER2 amplification status as defined by American Society of Clinical Oncology and College of American Pathologists.
Introduction: Formalin fixation is the gold standard for tissue storage in clinical pathology and formalin-fixed paraffin-embedded (FFPE) samples are therefore widely available. In addition, FFPE samples are often clinically annotated, which makes such material attractive for retrospective proteomic biomarker discovery studies. While several untargeted mass spectrometry (MS) studies have demonstrated the qualitative equivalence of the biological information obtained from FFPE compared to fresh-frozen tissues, the quantitative performance of MS applied to FFPE tissues still needs to be assessed. In this study, we developed a targeted MS assay for the quantification of the human receptor tyrosine-protein kinase erbB-2 (HER2) and demonstrated its applicability by quantifying HER2 in FFPE breast cancer samples with varying HER2 expression levels.
Methods: HER2 was chosen as a candidate protein because its overexpression is routinely assessed in breast tumors in order to determine susceptibility to anti-HER2 treatment. Peptide candidates were identified by untargeted MS analysis in relevant HER2-overexpressing cell lines, fixed with formalin, and in FFPE breast tumors. Peptides were extracted from formalin-fixed material using a heat-induced antigen retrieval procedure. A selected reaction monitoring (SRM) assay based on isotope dilution with an internal reference standard was developed for the six best performing peptides, which were further analytically evaluated in terms of linearity, reproducibility and lower limit of quantification (LLOQ). We then quantified HER2 in FFPE samples from a cohort of 40 patients with invasive breast carcinomas showing different levels of HER2 gene amplification (fluorescence in situ hybridization (FISH) ratio: 0-2, 2-4, 4-10 and >10 respectively). The SRM data were compared with immunohistochemistry (IHC) and FISH data using logistic regression analysis.
Results: Six of the 29 HER2 peptides identified in the untargeted MS approach were selected as candidates for the development of an SRM assay based on the following criteria: unique to HER2, absence of methionines and satisfactory chromatographic performance. Method evaluation showed a linear response over a calibration range from 0.012 to 100 fmol on column (R2: 0.99–1.00) and a LLOQ of 0.155 fmol on column for all peptides evaluated. Total individual peptide amounts extracted from 20 µm thick tumor slices spanned a range of 7.2 fmol to 4.7 pmol in the 40 breast tumor samples investigated in our study. The amounts of the six HER2 peptides quantified by SRM were highly and significantly correlated with each other, indicating that peptide levels can be used as surrogates of HER2 protein amounts in FFPE tissues. After normalization for tissue sample size, the agreement of SRM peptide measurements with IHC scores or FISH ratios was tested using ordinal logistic regression. Overall, SRM data predicted better IHC scores (77.5 % of IHC scores, κ = 0.874) than FISH ratios (47.5 % of FISH ratios, κ = 0.688). However, due to the large overlap between each group, SRM data predicted best tumors with IHC 0+ and IHC 3+ scores (correct classification rates of 88.9 % and 95.5 %, respectively) and FISH ratio <2 (non-amplified, 100% correct classification). Interestingly, the HER2 SRM assay correctly predicted 90% of HER2 amplification status as defined by the American Society of Clinical Oncology and College of American Pathologists (IHC score 0+, 1+ or 2+ non amplified by FISH is considered as HER2 non amplified; IHC score 3+ or 2+ amplified by FISH is considered as HER2 amplified). The assay’s two best performing peptides were independently able to differentiate the two classes of tumors from each other with AUC under the ROC curve above 0.96.
Conclusion: The developed assay showed good analytical performance and a high agreement with IHC and FISH data, confirming that SRM can be used to quantify protein expression in FFPE tissues. As exemplified in this HER2 study, distinction between low- and mildly over-expressers might represent more of a challenge due to the finite sensitivity of mass spectrometers and the use of different types of measurements (here, IHC scores and FISH ratios versus absolute peptide amounts). Increased measurement precision at lower LLOQ might be achieved, for example, by using the following generation of MS instruments. In parallel, increasing the sample number in the study might increase statistical significance and thus enable a better separation between the different IHC and FISH categories.