Predicting Tissue HER2 Status Using Serum HER2 Levels in Patients with Metastatic Breast Cancer

Transcription

1 Clinical Chemistry 52: (2006) Proteomics and Protein Markers Predicting Tissue HER2 Status Using Serum HER2 Levels in Patients with Metastatic Breast Cancer Sun-Young Kong, Byung-Ho Nam, Keun Seok Lee, Youngmee Kwon, Eun Sook Lee, Moon-Woo Seong, Do Hoon Lee, and Jungsil Ro* Background: Immunohistochemistry (IHC) and fluorescent in situ hybridization (FISH) are reliable ways to identify overexpression or amplification of the HER-2/ neu (HER2, symbol ERBB2) gene, but each technique requires a high-quality tissue sample, which may not be available. We investigated whether serum concentrations of the HER2 extracellular domain (ECD) can be used as an alternative to tissue HER2 status in metastatic breast cancer, and we defined an optimal decision-level concentration of serum HER2 for prediction of tissue HER2 status. Methods: In 195 patients with metastatic breast cancer, we determined HER2 expression by IHC and performed FISH analysis on tumors for which IHC staining was graded as 2. We measured serum HER2 by immunoassay and used ROC curve analysis to determine optimal serum HER2 ECD concentrations for differentiation between positive and negative HER2 status. Results: IHC results were 0/1 for 30 (15%) of the patients, 2 for 89 (46%), and 3 for 76 (39%). FISH revealed HER2 amplification in 19 (21%) of the IHC 2 tumors. Mean (SE) serum HER2 ECD was 22.2 (5.1) g/l in the tissue HER2-negative group, significantly lower than the concentration of 363 (96) g/l in the tissue HER2-positive group (P <0.0001). ROC curve analysis showed 95% specificity and 62% sensitivity for tissue HER2 positivity at 37 g/l of serum HER2. Conclusion: To use serum HER2 concentration as an alternative to direct determination of tissue HER2 status, we suggest 37 g/l as a cutoff for predicting positive tissue HER2 with 95% specificity. Sensitivity, however, is low American Association for Clinical Chemistry 1 Research Institute & Hospital, National Cancer Center, Goyang-si, Gyeonggi-do, Korea. * Address correspondence to this author at: Center for Breast Cancer, National Cancer Center, 809, Madu-dong, Ilsan-gu, Goyang-si, Gyeonggi-do, , Republic of Korea. Fax ; jungsro@ncc.re.kr. Received January 31, 2006; accepted May 24, Previously published online at DOI: /clinchem The HER-2/neu (HER2) 1 protooncogene (symbol ERBB2) encodes a transmembrane glycoprotein with tyrosine kinase activity that mediates proliferation and differentiation in normal epithelial cells (1). As a result of gene amplification, 20% 30% of breast cancers overexpress HER2 protein, and such overexpression is associated with poor prognosis (2 5). Metastatic breast cancers that test positive for HER2 amplification and/or overexpression are candidates for anti-her2 trastuzumab (Herceptin) therapy, so it is essential to evaluate HER2 status accurately (6, 7). Tissue HER2 status is generally determined by immunohistochemistry (IHC) or fluorescence in situ hybridization (FISH) (8). IHC is the method of choice, with 0/1 signifying HER2 negative status and 3 signifying HER2 positive status. FISH is performed when IHC results are indeterminate (2 ), and cancers that show HER2 amplification are classified as HER2 positive (9). Because IHC and FISH require a high-quality tissue sample, which is not always available, and because FISH is a costly procedure more than 10 times more expensive than IHC in Korea we sought an alternative method for evaluating HER2 status. We based our method on the finding that the HER2 extracellular domain (ECD) may be cleaved and shed from the surface of breast cancer cells and released into the circulation (10, 11). This leads to increased serum HER2 ECD concentrations in 30% 40% of patients with metastatic breast cancer and serves as a marker for prognosis and resistance to hormone and alkylating therapy (12 14). Serum HER2 ECD concentration correlates with tissue HER2 status (by IHC or FISH) (15), and the possibility of using it as an alternative to tissue analysis has recently been examined (16). In this study, we investigated that possibility further in view of the discordant positive rates between tissue HER2 and serum HER2 in 1 Nonstandard abbreviations: HER2, HER-2/neu protooncogene; IHC, immunohistochemistry; FISH, fluorescence in situ hybridization; ECD, HER2 extracellular domain; ROC, receiver operating characteristic; CI, confidence interval; CEA, carcinoembryonic antigen. 1510

2 Clinical Chemistry 52, No. 8, metastatic breast cancer, and we suggest a method that optimizes predictive power of serum HER2 ECD concentration. Materials and Methods patients and samples This retrospective study included 195 patients with metastatic breast cancer who were treated at the National Cancer Center, Goyang, Korea, from February 2002 to May During this period, metastatic breast cancer was diagnosed in 320 patients; however, we excluded 125 patients because we could not obtain their serum samples from the time when their diagnoses were made. Clinicopathologic features such as age, disease site, number of metastases, and estrogen and progesterone receptor status were evaluated. We typically do not perform tissue biopsy for metastatic breast cancer that has advanced from primary breast cancer, so prior tissue samples were obtained from 114 patients. We performed tissue IHC and FISH analysis on tissue samples obtained from 114 patients at the time of primary breast cancer (stage I, II, III) diagnosis and from 81 patients at the time of metastatic breast cancer diagnosis. Thirty-four of 81 patients had metastatic breast cancer at the initial diagnosis (stage IV disease), and in 47 of 81 patients, original primary breast tissue could not be retrieved. The tissue specimens were fixed in formaldehyde and embedded in paraffin by standard methodology. All serum samples were obtained at the time of metastatic breast cancer diagnosis and before chemotherapy or Herceptin therapy. immunohistochemistry Paraffin-embedded tissue was cut into 4- m sections and mounted on glass slides. The sections were deparaffinized in xylene and rehydrated through alcohols to distilled water. Sections were incubated for 15 min in 10 mmol/l citrate buffer, ph 6.0, at 98 C. Polyclonal rabbit antihuman c-erbb-2 oncoprotein (Code A0485; Dako Corp.) was applied at 1:100 dilutions. The slides were stained in an ES automatic stainer (Ventana Medical Systems Inc.), and the degree of staining was scored by a pathologist, according to the method recommended for the Dako HercepTest. fish Paraffin-embedded tissue was cut into 4- m sections, mounted onto positively charged slides, baked at 56 C overnight, and deparaffinized. FISH analysis was performed using a PathVysion HER2 DNA Probe Kit (Vysis) according to the manufacturer s recommendations. Briefly, the tissues were pretreated and subjected to protease digestion. A hybridization solution containing directly labeled probes of SpectrumGreen for the chromosome 17 centromere (CEP) and SpectrumOrange for the HER2 gene locus were applied, and the denaturation and hybridization procedures were perfomed in a HYBrite system (Vysis). The slides were washed in 2 SSC buffer (0.3 mol/l NaCl and 0.03 mol/l sodium citrate; ph 7.2) at 72 C for 2 min and counterstained with diaminopyrolylindole. Two observers assessed each slide with a Nikon Eclipse E600 fluorescence microscope equipped with narrow band-pass filters. The observers scored at least 60 cells per specimen for HER2 (ERBB2) and CEP signals. We set a HER2:CEP ratio 2.0 as evidence of HER2 (ERBB2) amplification. serum her2 ecd assay Blood was drawn into serum separator tubes and centrifuged at 2000g for 10 min at room temperature. Serum samples were stored at 70 C until used. To measure the serum HER2 ECD concentration, we used a 2-site chemiluminescence sandwich immunoassay (ADVIA Centaur System, Bayer Diagnostics) with a detection limit of 0.5 g/l; the interassay CV in our laboratory was 5.9%. Following the manufacturer s recommendation, we used 15 g/l as the serum HER2 ECD cutoff to discriminate normal concentrations from those in patients with breast cancer (17). Serum HER2 ECD measurement was carried out in a blinded manner without knowledge about the results of IHC and FISH. When clinically indicated, we measured carcinoembryonic antigen (CEA) using a cutoff value of 5 g/l (n 65) (ARCHITECT i2000, Abbott Diagnostics) and CA15.3 using a cutoff value of 38 units/ml (n 126) (ADVIA Centaur System). statistical analysis We used SAS version 8 software for statistical analysis. To examine the relationship between HER2 status and other clinicopathologic variables, we used the 2 test and logistic regression analysis. To measure concordance of IHC, FISH, and serum HER2 results, we changed the serum HER2 value to a dichotomous variable by the use of the normal cutoff concentration and estimated the Spearman correlation. We used the t-test to compare the serum HER2 concentration with tissue HER2 status in each group. We performed ROC curve analysis to evaluate the discriminating ability of serum HER2 and to arrive at a serum HER2 cutoff measure for predicting tissue HER2 positivity. Also, we separately analyzed ROC curves according to tissue origins (tissue at the site of primary diagnosis vs tissue at the site of metastasis). We analyzed CA15.3 and CEA ROC curves. Results Table 1 shows the clinicopathologic characteristics and tissue HER2 status of all 195 cases. The median age of the patients was 49 years (range, years). IHC results were 0/1 for 30 (15%) of the patients, 2 for 89 (46%), and 3 for 76 (39%). FISH revealed HER2 amplification in 19 (21%) of the IHC 2 tumors. Tissue HER2 positivity was significantly (P 0.001) associated with negative estrogen receptors, negative progesterone receptors, liver metastases, and brain metastases. A flow diagram for the study is shown in Fig. 1. Serum HER2 was 15 g/l in 119 (61%) of the

3 1512 Kong et al.: Serum HER2 as an Alternative to Tissue HER2 Study Table 1. Clinicopathologic characteristics and tissue HER2 status of 195 patients with metastatic breast cancer. Characteristic Patients, n (%) Tissue HER2 negative, n (%) Tissue HER2 positive, n (%) P Estrogen receptor Negative 90 (46.2) 28 (31.1) 62 (68.9) Positive 105 (53.8) 72 (68.6) 33 (31.4) Progesterone receptor Negative 117 (60.0) 46 (39.3) 71 (60.7) Positive 75 (38.5) 53 (70.7) 22 (29.3) Unknown 3 (1.5) 1 (33.3) 2 (66.7) Serum HER2 ECD Not increased 74 (37.9) 63 (85.1) 11 (14.9) Increased 121 (62.1) 37 (30.6) 84 (69.4) Sites of metastasis Bone 119 (61.0) 56 (47.1) 63 (52.9) Liver 93 (47.7) 36 (38.7) 57 (61.3) Lung 90 (46.2) 48 (53.3) 42 (46.7) Pleura 46 (23.6) 24 (52.2) 22 (47.8) Skin 27 (13.8) 12 (44.4) 15 (55.6) Brain 22 (11.3) 3 (13.6) 19 (86.4) No. of disease sites (27.7) (29.2) (43.1) Tissue HER2 positivity was significantly associated with negative estrogen receptors, negative progesterone receptors, liver metastases, and brain metastases. patients, including 35 from the tissue HER2-negative group (n 100) and 84 from the tissue HER2 positive group (n 95). The serum HER2 results were correlated with the IHC (P 0.01) and FISH (P 0.01) results and showed significant (P 0.01) correlation with progesterone receptor negativity and liver metastases. Mean (SE) serum HER2 ECD in the tissue HER2- negative group was 22.2 (5.1) g/l, significantly lower than the concentration of 363 (96) g/l in the tissue HER2-positive group (P ) (Fig. 2.). Mean (SE) CEA [13.9 (6.3) vs 40.3 (14.1)] and CA15.3 [120 (22) vs 231 (48)] concentrations did not differ significantly (P 0.05) between the 2 groups. ROC curve analysis showed 62% specificity and 88% sensitivity for tissue HER2 positivity at 15 g/l serum HER2. A 90% specificity and 65% sensitivity were seen at 30 g HER2/L [AUC, 0.87; 95% confidence interval (CI), ] (Fig. 3). At 37 g/l serum HER2, specificity was 95% and sensitivity 62%. When we used 30 g/l as the decision level, 71 (36.4%) of the patients showed increased serum HER2; at 37 g/l, 64 (32.8%) of the patients showed increased serum HER2. When we analyzed only data from patients in whom the primary site of cancer was studied (n 114), a cutoff of 33 g/l was associated with 95% specificity and 66% sensitivity (AUC, 0.87; 95% CI, ). The corresponding cutoff (that is, for 95% specificity) derived from metastatic tissue data (n 81) was 47 g/l, at which cutoff the sensitivity was 47% and the AUC was 0.85 (95% CI, ). The cutoff (for 95% specificity) derived from the data on metastatic tissue was slightly higher than that for primary tissue data, but the distributions of serum HER2 in the negative (primary tissue vs metastatic tissue) and positive tissue HER2 group showed no significant difference by t-test (P 0.56 in tissue HER2 negative group, and P 0.63 in tissue HER2 positive group). Because the serum HER2 results showed significant (P 0.01) correlation with liver metastases, we analyzed ROC curve for the patients with liver metastasis vs without liver metastasis. The cutoff for 95% specificity in patients with liver metastasis was 133 g/l, at which the sensitivity was 51% and the AUC 0.84 (95% CI ); the corresponding cutoff for patients without liver metastasis was 32 g/l which had a 55% sensitivity and AUC of 0.89 (95% CI ). For CEA and CA15.3, the optimal cutoffs were 38.0 g/l (AUC, 0.70; 95% CI, ; P 0.006), and 342 ku/l (AUC, 0.59; 95% CI, ; P 0.09), respectively. Discussion This study, like previous ones (15, 16), demonstrated a good correlation between serum HER2 ECD concentration and IHC and FISH results in metastatic breast cancer. Here, in addition, we determined that 37 g/l serum HER2 ECD showed 95% specificity and 62% sensitivity in predicting tissue HER2, and we suggest that that value can be used as the cutoff because we focused on a cutoff concentration at which the false-positive rate for tissue positivity was low.

4 Clinical Chemistry 52, No. 8, Fig. 1. Flow diagram of patients in study. Tissue HER2 (ERBB2) status was determined by IHC and FISH. IHC 0/1 was classified tissue HER2-negative status, and 3 was classified HER2-positive status. FISH was performed when IHC results were 2 (indeterminate), and cancers that showed HER2 amplification were classified as tissue HER2-positive. A previous study determined 16 g/l as the cutoff point (16), but the specificity at 16 g/l in our study was only 72%. The discrepancy might result from differences in the assays, but if we consider the different positive rates 30% 40% when the serum HER2 ECD cutoff concentration was 15 g/l and 20% 30% in tissue HER2 studies in metastatic breast cancer the decision cutoff level for tissue HER2 positivity would be higher than the general cutoff point used to diagnose breast cancer. Thus, we chose a serum HER2 ECD level with higher specificity to achieve a more stringent clinical application of trastuzumab therapy, especially in cases for which no tissue HER2 data was available. Some cases with negative tissue HER2 in our study showed high serum HER2 ECD concentractions (up to g/l), which might be attributable to transformation of tissue HER2 status at disease recurrence, as has been reported (19, 20). To obtain optimal cutoff according to clinical variables, we considered tissue origin and liver metastasis and analyzed ROC curves separately. The cutoff determined from metastatic tissue data was slightly higher than that of primary tissue data, but we observed no significant difference of serum HER2 between each of the negative (primary tissue vs. metastatic tissue) and positive tissue HER2 groups. The cutoff values of serum HER2 according to the presence or absence of liver metastasis showed differences as wide as 32 g/l vs 133 g/l; thus, we

5 1514 Kong et al.: Serum HER2 as an Alternative to Tissue HER2 Study In conclusion, determination of serum HER2 ECD concentration in metastatic breast cancer may provide a useful index of tissue HER2 status, especially in light of its relatively easy and inexpensive methodology and the simplicity of sample collection. This work was supported by the National Cancer Center, Korea (Grant ). Fig. 2. ROC curve of serum HER2 ECD concentration for differentiation of tissue HER2-negative and -positive patients. AUC for serum HER2 ECD was 0.87 (95% confidence interval; ). Specificities for tissue HER2 positivity were 90% at 30 g/l and 95% at 37 g/l serum HER2 ECD. might consider a higher cutoff level of serum HER2 when assessing HER2 status in patients with liver metastasis. When we compared serum HER2 with CA15.3 and CEA, 2 other common tumor markers in metastatic breast cancer, serum HER2 showed a higher discrimination power, although CA15.3 and CEA were evaluated in a smaller number of patients. Serum HER2 ECD concentration was not only useful for assessing prognosis, predicting the response to trastuzumab, and reflecting tumor responses reported in previous studies (7, 14, 21, 22), but also, at the higher cutoff concentration, for reflected tissue HER2 status with enhanced specificity. Fig. 3. Serum HER2 ECD concentration according to tissue HER2 status in 195 patients with metastatic breast cancer. The mean (SE) HER2 ECD concentrations (horizontal lines) were 22.2 (5.1) g/l in the tissue-negative group and 363 (96) g/l in the tissue-positive group (P ). Positive tissue cases with serum HER2 ECD levels 1000 g/l (n 10) are not represented in this Fig. References 1. Bargmann CI, Hung MC, Weinberg RA. The neu oncogene encodes an epidermal growth factor receptor-related protein. Nature 1986; 319: Lal P, Salazar PA, Hudis CA, Ladanyi M, Chen B. HER-2 testing in breast cancer using immunohistochemical analysis and fluorescence in situ hybridization: a single-institution experience of 2,279 cases and comparison of dual-color and single-color scoring. Am J Clin Pathol 2004;121: Slamon DJ, Clark GM, Wong SG, Levin WJ, Ullrich A, McGuire WL. Human breast cancer: correlation of relapse and survival with amplification of the HER-2/neu oncogene. Science 1987;235: Andrulis IL, Bull SB, Blackstein ME, Sutherland D, Mak C, Sidlofsky S, et al. neu/erbb-2 amplification identifies a poor-prognosis group of women with node-negative breast cancer. Toronto Breast Cancer Study Group. J Clin Oncol 1998;16: Agrup M, Stal O, Olsen K, Wingren S. C-erbB-2 overexpression and survival in early onset breast cancer. Breast Cancer Res Treat 2000;63: Vogel CL, Cobleigh MA, Tripathy D, Gutheil JC, Harris LN, Fehrenbacher L, 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;20: Burstein HJ, Harris LN, Marcom PK, Lambert-Falls R, Havlin K, Overmoyer B, et al. Trastuzumab and vinorelbine as first-line therapy for HER2-overexpressing metastatic breast cancer: multicenter phase II trial with clinical outcomes, analysis of serum tumor markers as predictive factors, and cardiac surveillance algorithm. J Clin Oncol 2003;21: Yaziji H, Goldstein LC, Barry TS, Werling R, Hwang H, Ellis GK, et al. HER-2 testing in breast cancer using parallel tissue-based methods. Journal of the American Medical Association 2004;291: Diaz NM. Laboratory testing for HER2/neu in breast carcinoma: an evolving strategy to predict response to targeted therapy. Cancer Control 2001;8: Lin YZ, Clinton GM. A soluble protein related to the HER-2 proto-oncogene product is released from human breast carcinoma cells. Oncogene 1991;6: Zabrecky JR, Lam T, McKenzie SJ, Carney W. The extracellular domain of p185/neu is released from the surface of human breast carcinoma cells, SK-BR-3. J Biol Chem 1991;266: Colomer R, Montero S, Lluch A, Ojeda B, Barnadas A, Casado A, et al. Circulating HER2 extracellular domain and resistance to chemotherapy in advanced breast cancer. Clin Cancer Res 2000; 6: Lipton A, Ali SM, Leitzel K, Demers L, Chinchilli V, Engle L, et al. Elevated serum Her-2/neu level predicts decreased response to hormone therapy in metastatic breast cancer. J Clin Oncol 2002; 20: Carney WP, Neumann R, Lipton A, Leitzel K, Ali S, Price CP.

6 Clinical Chemistry 52, No. 8, Potential clinical utility of serum HER-2/neu oncoprotein concentrations in patients with breast cancer. Clin Chem 2003;49: Harris LN, Liotcheva V, Broadwater G, Ramirez MJ, Maimonis P, Anderson S, et al. Comparison of methods of measuring HER-2 in metastatic breast cancer patients treated with high-dose chemotherapy. J Clin Oncol 2001;19: Tse C, Brault D, Gligorov J, Antoine M, Neumann R, Lotz JP, et al. Evaluation of the quantitative analytical methods real-time PCR for HER-2 gene quantification and ELISA of serum HER-2 protein and comparison with fluorescence in situ hybridization and immunohistochemistry for determining HER-2 status in breast cancer patients. Clin Chem 2005;51: Cook GB, Neaman IE, Goldblatt JL, Cambetas DR, Hussain M, Luftner D, et al. Clinical utility of serum HER-2/neu testing on the Bayer Immuno 1 automated system in breast cancer. Anticancer Res 2001;21: Bossuyt PM, Reitsma JB, Bruns DE, Gatsonis CA, Glasziou PP, Irwig LM, et al. The STARD statement for reporting studies of diagnostic accuracy: explanation and elaboration. Clin Chem 2003 Jan;9: Zidan J, Dashkovsky I, Stayerman C, Basher W, Cozacov C, Hadary A. 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;93: Meng S, Tripathy D, Shete S, Ashfaq R, Haley B, Perkins S, et al. HER-2 gene amplification can be acquired as breast cancer progresses. Proc Natl Acad Sci USA 2004;101: Kostler WJ, Schwab B, Singer CF, Neumann R, Rucklinger E, Brodowicz T, et al. Monitoring of serum Her-2/neu predicts response and progression-free survival to trastuzumab-based treatment in patients with metastatic breast cancer. Clin Cancer Res 2004;10: Fehm T, Gebauer G, Jager W. Clinical utility of serial serum c-erbb-2 determinations in the follow-up of breast cancer patients. Breast Cancer Res Treat 2002;75: