HER2 Overexpression as a Poor Prognostic Determinant in Resected Biliary Tract Cancer

Abstract

Background

HER2 overexpression has been investigated as a potential biomarker and therapeutic target in biliary tract cancer (BTC), but a prognostic role of such alteration has not been demonstrated yet.

Materials and Methods

We retrospectively evaluated HER2 protein expression by immunohistochemistry (IHC) in 100 patients with radically resected BTC. HER2 gene amplification was assessed by fluorescence in situ hybridization (FISH) in 2+ and 3+ cases at IHC. High HER2 protein expression was defined as either IHC 3+ or 2+ associated with FISH positivity. The primary objective of the study was to evaluate the prognostic role of HER2 overexpression in terms of disease‐free survival (DFS) and overall survival (OS). Secondary endpoints were the prevalence of HER2 overexpression and the possible correlation with other clinicopathological features.

Results

HER2 overexpression was identified in 11 patients and was not related to other clinicopathological factors. DFS was significantly shorter in HER2‐positive compared with HER2‐negative patients (10.6 vs. 20.9 months, log‐rank p = .017). HER2 confirmed its prognostic value for DFS at multivariate analysis (hazard ratio 2.512; 95% confidence interval, 1.232–5.125; p = .011) together with nodal stage (p < .001), resection margin (p = .027), and tumor site (p = .030). There was no difference in OS between HER2‐positive and ‐negative patients (p = .068).

Conclusion

HER2 overexpression represents an independent prognostic factor for disease recurrence in patients with BTC treated with potentially curative surgery.

Implications for Practice

HER2 overexpression may play an independent role in promoting an aggressive behavior in resectable biliary tract cancer. This evidence could be helpful in improving prognostic stratification after resection and, primarily, should endorse the rationale to investigate HER2 as a therapeutic target in biliary tract cancer.

Short abstract

Studies suggest that the HER2 pathway could play a role in the development and growth of biliary tract cancer. This study investigated the incidence and effect of HER2 expression on the risk of recurrence in a large series of resected biliary tract cancer cases at a single Institution.

Introduction

Biliary tract cancer (BTC) represents an infrequent and heterogeneous group of neoplasms characterized by poor prognosis 1. The only curative option consists of surgical resection, even though the disease often presents as locally advanced or metastatic, thus making chemotherapy the only treatment possibility in the majority of patients.

Despite the advances in the knowledge about carcinogenesis and molecular alterations characterizing different subtypes of BTC 2, 3, reliable molecular markers of prognosis are lacking. Both in clinical trials and in routine practice, BTC undergoing radical surgery are stratified on the basis of site of origin and several clinicopathological factors (such as nodal involvement, surgical margins, and tumor grade), which demonstrated an impact on prognosis in different series of resected patients 4, 5, 6. Among resected patients, the role of adjuvant treatment has been recently established in randomized studies. A few years ago, a meta‐analysis of 20 trials supported the use of adjuvant treatment, especially in the case of nodal involvement or positive margins after resection 7. Nonetheless, limitations in the design, size, and treatment regimens used for the included studies undermined the results of the metanalysis. Recently, two prospective trials evaluated the role of adjuvant chemotherapy compared with observation in patients with resected BTC: whereas the combination of gemcitabine and oxaliplatin failed to demonstrate a benefit in relapse‐free survival (RFS) 8, capecitabine monotherapy showed to improve both RFS and overall survival (OS) at the price of acceptable toxicity 4.

However, the usefulness of the identified molecular alteration in guiding treatment decisions after resection is unclear.

Among the most investigated biomarkers, the human epidermal growth factor receptor 2 (HER2) is a member of the ErbB family and is involved in tumor proliferation by downstream signaling activation 9. Its overexpression, measured by immunohistochemistry (IHC) and/or fluorescence in situ hybridization (FISH), represents a predictive biomarker for HER2‐targeted agents in different tumor types, such as breast and gastric cancer 10, 11.

Preclinical studies suggest that the HER2 pathway could have a role in the development and growth of BTC: in transgenic mice, overexpression of ErbB‐2 in the basal layer of biliary tract epithelium led to the development of adenocarcinoma 12, 13. In cholangiocarcinoma cell lines, HER2 expression was associated with increased invasiveness, motility, and proliferation through AKT/p70S6K pathway 14.

Different retrospective analyses have been conducted in BTC, reporting a rate of HER2 overexpression or amplification depending on tumor site (more than 19% in extrahepatic and about 5% in intrahepatic cholangiocarcinoma) 15 and etiology (10.4% in Fluke‐positive and 2.7% in Fluke‐negative cholangiocarcinoma) 16. However, a prognostic role of such alteration has not been demonstrated yet. In this study, we investigated the incidence and explore the impact of HER2 expression on the risk of recurrence in a large series of BTC cases resected at a single Institution.

Materials and Methods

Patients

We retrospectively collected a cohort of 100 consecutive patients with radically resected BTC followed at the Pisa University hospital between 2006 and 2015. Inclusion criteria were age > 18 years; diagnosis of intrahepatic or extrahepatic cholangiocarcinoma, gallbladder carcinoma, or ampullary carcinoma; resectable disease at presentation, treated by surgery with curative intent; availability of paraffin‐embedded tumor sample; and access to clinical information, pathological features, and survival data. Data have been collected and analyzed assuring patient anonymization. The analyses included in this study were performed in accordance with the Declaration of Helsinki and were approved by the Ethics Committee of the Pisa University Hospital. Written informed consent from the patients for research use of data was obtained before the investigation.

HER2 protein expression was evaluated by IHC in all samples. HER2 gene amplification was assessed by FISH only in cases judged 2+ and 3+ at IHC. High expression of HER2 protein was defined as either IHC 3+ or IHC 2+ associated with FISH positivity.

HER2 IHC Analysis

Three‐micrometer‐thick tumor sections were stained with HER2/neu (4B5) Rabbit Monoclonal Primary Antibody Pathway (Ventana Medical Systems, Inc., Oro Valley, AZ). Staining was done on an automated IHC/ISH (chromogenic in situ hybridization) slide staining system (BenchMark XT, Ventana Medical Systems, Inc.) using the ultraView DAB Detection Kit (Ventana Medical Systems, Inc.).

HER2 scoring was conducted according to criteria used for gastric cancer 17. The key feature of the scoring criteria was the inclusion of complete or incomplete membrane staining as HER2‐positive if ≥10% of cells from surgical samples were stained. As previously described, HER2 IHC was reported as a score ranging from 0 to 3+. In particular, IHC was defined as follows: IHC 3+ staining is any membranous staining visible at low magnification (×2.5–5). Lateral or U‐shaped membranous staining is typically seen at cell‐cell junctions. Immunohistochemistry 2+ membranous staining is visible at ×10–20 magnification. Immunohistochemistry 1+ staining is visible only with ×40 magnification and should be considered immunohistochemistry negative. No staining is defined as IHC 0.

HER2 FISH Analysis

Immunohistochemistry 2+ and 3+ cases were retested using FISH. The FISH analysis was performed on 4–6‐μm‐thick paraffin sections of tumor tissues using Paraffin Pretreatment Reagent Kit II and PathVysion HER2 DNA Probe Kit II (Abbott Molecular, Abbott Park, IL). Probe, composed of Vysis CEP17 SpectrumGreen and Vysis LSI HER2/neu SpectrumOrange; Abbott Molecular), was used to detect gene amplification involving the HER2 gene as per the manufacturer's instructions. FISH results were expressed as the ratio between the number of copies of the HER2 gene and the number of copies of chromosome 17 within the nucleus counted in at least 20 cancer cells. The definition of FISH positivity was HER2:chromosome 17 ratio of ≥2.0 18.

Statistical Analysis

The primary objective of the study was to evaluate the prognostic role of HER2 overexpression in patients with resected BTC in terms of disease‐free survival (DFS) and OS. Secondary endpoints were the prevalence of HER2 overexpression in resected BTC and the possible correlation with other clinic‐pathological features.

DFS was measured from the date of surgery to the date of disease recurrence or death, whichever occurred first. Patients who were free of recurrence after resection were censored at the date of the last radiological assessment. OS was measured from the date of surgery to the date of death or the last follow‐up visit. DFS and OS were estimated using the Kaplan‐Meier product‐limit method. Median values with corresponding 95% confidence intervals (95% CIs) were reported.

The following factors were included in the analyses for DFS and OS: HER2 status (negative vs. positive), age (≤65 vs. >65 years), gender, Eastern Cooperative Oncology Group (ECOG) performance status (PS) (0–1 vs. >1), tumor location (gallbladder vs. intrahepatic vs. extrahepatic vs. ampullary), T stage (T1 vs. T2 vs. T3–4), N stage (N0 vs. N1/Nx), resection margin (R0 vs. R1/R2), tumor grading (G2 vs. G3), lymphadenectomy at the time of primary tumor resection (yes vs. no), adjuvant chemotherapy (yes vs. no), and adjuvant radiotherapy (yes vs. no).

Correlation between HER2 status and other clinicopathological features was performed by chi‐square test.

Kaplan‐Meier method was used to estimate survival curves that were compared by the use of the log‐rank test. We performed a univariate assessment of the effect of HER2 and the other variables on DFS and OS. Hazard ratios and 95% CIs were calculated with the Cox proportional hazard regression model. A two‐sided p value of ≤.05 was considered significant at univariate analysis. A multivariate analysis was carried out using stepwise Cox proportional hazards regression modeling, stratifying for the stage and setting statistical significance at p ≤ .05 for a two‐sided test. Statistical analyses were carried out using the statistical software package SPSS 19.0 (SPSS, Chicago, IL).

Results

Patient Characteristics

One hundred patients who met the eligibility criteria were included in the study. Baseline characteristics are reported in Table 1.

Table 1.

Patients' characteristics (n = 100)

Characteristic	All patients (n = 100), n	HER2− (n = 89), n	HER2+ (n = 11), n	p value
Sex				.536
Male	55	50	5
Female	45	39	6
Age, median (range)	68 (33–84)	69 (35–84)	65 (33–79)	.561
ECOG PS				.854
0	55	51	4
1	38	34	4
2/3	7	4	3
Tumor grade				.698
G2	59	52	7
G3	36	33	3
Gx	5	4	1
Tumor site				.901
Intrahepatic	36	32	4
Extrahepatic	28	24	4
Ampullary	23	21	2
Gallbladder	13	12	1
Resection margins				1.000*
R0	94	83	11
R1	5	5	0
R2	1	1	0
T stage				.051
T1	19	18	1
T2	28	24	4
T3	49	44	5
T4	3	3	0
Tx	1	0	1
N stage				.379
N0	39	35	4
N1	40	37	3
Nx	21	17	4
Adjuvant chemotherapy				.201
Yes	50	47	3
No	50	42	8
Type of adjuvant treatment				1.000**
Gemcitabine	44	41	3
Capecitabine	4	4	0
Other regimens	2	2	0
Adjuvant radiotherapy				.515
Yes	4	4	0
No	96	85	11
Type of first‐line chemotherapy				.145a
Gemcitabine	12	9	3
Gemcitabine + platinum	22	19	3
Fluoropyrimidines	2	2	0
Fluoropyrimidines + platinum	8	8	0
Other regimens	6	6	0
Best supportive care	12	8	4
Not available	6	5	1

^aCombination treatment versus monotherapy versus best supportive care.

^* Radical resection versus R1/R2 Resection.

^** Gemcitabine versus others.

Abbreviations: ECOG PS, Eastern Cooperative Oncology Group performance status; G, grade.

HER2 Analyses

High expression of HER2 protein at IHC was recorded in 33 patients (2+ in 24 and 3+ in 9 samples, respectively). In the other 67 patients, IHC score was negative (0 in 50 patients and 1+ in 17 patients; Fig. 1). FISH was performed on 28 of 33 high IHC‐positive samples (five cases were excluded because of inadequate tissue for analysis). Gene amplification was found in 2 out of 22 patients with IHC 2+ (9%) and in 5 out of 6 patients with IHC 3+ (83%). There was a statistically significant correlation between IHC expression 3+ and FISH amplification (p = .0012; supplemental online Table 1).

Figure 1.

Examples of HER2 immunohistochemistry (IHC) staining. IHC evaluation of HER2 in biliary tract cancer is scored as follows: score 3+ (A), score 2+ (B), score 1+ (C), score 0 (D).

According to combined IHC and FISH results, HER2 overexpression was identified in 11 patients (nine cases with IHC 3+ and two cases with IHC 2+ and FISH amplification).

As reported in Table 1, there was no statistically significant difference in baseline characteristics according to HER2 status.

Survival Analyses

At a median follow‐up of 82.8 months, 68 patients had experienced disease progression and 64 were expired. In the global population, median DFS and OS were 20.5 months (95% CI, 16.01–24.99) and 40.9 months (95% CI, 30.87–50.92), respectively. Data about treatment received after progression are listed in Table 1.

DFS significantly differed according to HER2 status; median DFS was 20.9 months in HER2‐negative and 10.6 months in HER2‐positive patients (log‐rank p = .017; Fig. 2). At univariate analysis, three other parameters resulted significantly associated with DFS: nodal status (p = .0005), ampullary tumor site (p = .009), and resection margin (p = .037). The results of univariate analyses are reported in Table 2.

Figure 2.

Disease‐free survival curves according to HER2 status.

Table 2.

Univariate analyses for disease‐free survival and overall survival

Characteristic	Disease‐free survival			Overall survival
	Median, months	HR (95% CI)	p value	Median, months	HR (95% CI)	p value
Age
>65 years	20.9	1.013 (0.626–1.640)	.957	39.7	1.288(0.779–2.129)
≤65 years	20.5	1		44.0		.323
Gender
Female	16.4	1.161 (0.720–1.874)	.540	43.4	1.097 (0.857–1.405)
Male	23.6	1		29.1		.463
ECOG PS
0/1	20.7	0.561 (0.224–1.402)	.216	42.7	0.349 (0.149–0.817)
>1	8.6	1		18.0		.015
Tumor site
Intrahepatic	11.2	1	.063	42.7	1	.008
Ampullary	64.2 18.7	0.379 (0.182–0.786)	.009	Not reached	0.353 (0.194–0.643)	.001
Gallbladder	13.3	0.837 (0.380–1.846)	.660	18.0	1.693 (1.004–2.856)	.048
Extrahepatic	18.7	0.931 (0.532–1.626)	.801	29.0	1.357 (0.892–2.064)	.154
T stage
T1	31.4	0.527 (0.264–1.054)	.079	Not reached	0.463 (0.261–0.996)	.101
T2	23.6	0.600 (0.337–1.069)	.070	43.4	0.690 (0.389–1.223)	.051
T3–4	16.2	1	.083	30.4	1	.204
Nodal stage
N0	57.9	0.387 (0.227–0.661)		75.3	0.417 (0.241–0.747)
N1/Nx	15.7	1	.0005	30.4	1	.002
Resection margins
R0	20.6	1		42.0	1
R1/R2	3.7	2.674 (1.059–6.753)	.037	14.1	3.135 (1.339–7.339)	.008
Grading
G2	23.6	0.847 (0.662–1.084)		43.4	0.879 (0.680–1.136)
G3	13.3	1	.188	30.4	1	.324
Lymphadenectomy
No	24.6	1.074 (0.810–1.426)		31.0	1.358 (1.031–1.790)
Yes	20.5	1	.619	43.4	1	.029
Adjuvant CT
No	15.7	1.113 (0.876–1.415)		37.4	1.178 (0.920–1.510)
Yes	20.7	1	.382	51.2	1	.194
Adjuvant RT
No	17.8	1.495 (0.739–3.023)		39.5	1.415 (0.699–2.864)
Yes	28.3	1	.263	51.2	1	.334
HER2
Negative	20.9	1		42.7	1
Positive	10.6	2.261 (1.134–4.507)	.020	27.9	1.861 (0.945–3.665)	.073

Abbreviations: HR, hazard ratio; CI, confidence interval; ECOG PS, Eastern Cooperative Oncology Group performance status; G, grade; CT, chemotherapy; RT, radiotherapy.

A multivariate regression analysis was conducted including these four parameters, and all of them confirmed their independent prognostic value (HER2 overexpression, p = .011; nodal stage, p < .001; resection margin, p = .027; and tumor site, p = .030; Table 3).

Table 3.

Multivariate analyses for disease‐free survival and overall survival

	Disease‐free survival		Overall survival
	HR (95% CI)	p value	HR (95% CI)	p value
HER2
Negative	1		—	—
Positive	2.512 (1.232–5.125)	.011
Nodal stage
N0	1		1
N1/Nx	2.969 (1.705–5.170)	.00012	3.516 (1.864–6.633)	.0001
Resection margin
R0	1		1
R1/R2	2.503 (1.109–5.651)	.027	1.826 (0.857–3.892)	.119
Tumor site
Intrahepatic	1	.030	1	.004
Ampullary	0.347 (0.165–0.728)	.005	0.235 (0.095–0.583)	.002
Gallbladder	0.882 (0.380–2.045)	.77	1.571 (0.716–3.451)	.260
Extrahepatic	1.017 (0.576–1.794)	.954	0.984 (0.504–1.922)	.963
ECOG PS
0/1	—	—	1
>1			2.933 (1.140–7.86)	.026
Lymphadenectomy
No	—	—	1
Yes			0.876 (0.621–1.237)	.453

ECOG PS and lymphadenectomy were not tested in the multivariate model for disease‐free survival because significance was not reported at univariate analysis (as was for HER2 status in the multivariate model for overall survival).

Abbreviations: —, not performed; CI, confidence interval; ECOG PS, Eastern Cooperative Oncology Group performance status; HR, hazard ratio.

Median OS was 55.2 and 23.4 months in HER2‐negative and ‐positive patients, respectively (log‐rank p = .068; Fig. 3). The following covariates resulted associated with poorer OS: ECOG PS (p = .015), N stage (p = .002), and tumor site (p = .008 for intrahepatic, p = .001 for ampullary, and p = .048 for gallbladder tumors, respectively; Table 2). In particular, the median OS was not reached in ampullary tumors, 55.3 months in intrahepatic tumors, 34.7 months in extrahepatic tumors, and 18.1 months in gallbladder tumors.

Figure 3.

Overall survival curves according to HER2 status.

In the OS multivariate model, tumor site (p = .005) and nodal stage (p = .004) confirmed their prognostic role (Table 3).

Also excluding ampullary tumors the difference in DFS was statistically significant (median DFS of 10.1 months for HER2 positives vs. 17.8 months for HER2 negatives, p = .040), whereas the difference in OS remained not statistically significant (median OS 23.5 months vs. 36.1 for HER2 positives vs. negatives, respectively, p = .241).

Tumor Characteristics and Association with HER2 Status

HER2 status was not correlated with any of the clinicopathological features analyzed. In particular, HER2 overexpression was not associated with tumor site (14.3% of extrahepatic, 11.1% of intrahepatic, 8.7% of ampullary, and 7.7% of gallbladder tumors were HER2 positive: p = .9), tumor grade (p = .737), T stage (p = .713), or N stage (p = .712). In particular, HER2 overexpression was 8% in stage II cases, 0% in stage III cases, and 11% in stage IV cases (p = .717).

Discussion

BTC represents a group of rare cancers with a dismal prognosis both in localized and distant metastatic cases 19. Chemotherapy has shown to minimally improve outcome in this disease 20. For this reason, molecular characterization is becoming crucial for the identification of potential therapeutic targets and more accurate patient stratification beyond conventional clinicopathological features. Previous studies described the incidence of HER2 alterations in different cohorts of BTC, but the clinical implication of such alteration has not been clarified 21, 22.

Among 100 cases of resected BTC, we reported HER overexpression (defined as score 3+ by IHC or score 2+ confirmed by FISH amplification) in 11% of cases: this rate was in line with that reported by other authors 15. Moreover, in line with previous reports, in our cohort, the incidence of HER2 overexpression did not differ between tumor sites (p = .9) or with other clinical and pathologic factors 15. As described, we found a strong correlation between IHC expression and amplification (p = .0012) 23. These data confirm that a subset of BTC cases is defined by HER2 overexpression and that both IHC and FISH are reliable techniques for HER2 assessment in this setting, as already observed in other tumor types 17, 24.

Although previous studies had suggested a putative prognostic role for other members of the ErbB family in BTC such as EGFR and HER3 25, 26, data regarding the prognostic role of HER2 overexpression were not conclusive. To our knowledge, this is the first report describing that HER2 expression is significantly associated with prognosis in resected BTC. In particular, the prognostic role of HER2 overexpression for DFS emerged in univariate analysis and was confirmed in multivariate analysis. In our series, HER2‐positive patients experienced disease progression after a median of 10.6 months from surgery, whereas DFS was almost twice as high among HER2‐negative cases (median, 20.9 months). Other known prognostic determinants such as nodal status, resection margin, and tumor site were confirmed as factors influencing disease recurrence in resected patients, thus making our results more robust in terms of external validity 27. Although the number of patients in the HER2‐positive group is small, the DFS prognostic significance of HER2 is confirmed by descriptive statistics that show no difference in terms of other baseline characteristics according to the HER2 status.

Although a substantial difference in median OS between HER2‐positive and ‐negative patients was observed (23.4 and 55.2 months, respectively), this parameter did not reach statistical significance (p = .068), probably because of heterogeneity in metastatic spreading and treatment administered after relapse. Notably, our study confirmed at the multivariate analysis that PS, nodal stage, and tumor site are associated with OS 28. In particular, ampullary cancer confirmed its better prognosis compared with gallbladder and intrahepatic or extrahepatic cholangiocarcinoma 29.

Although anti‐HER2 agents failed to show activity in treating unselected patients with BTC 30, preliminary reports have shown interesting results in preclinical setting 31 and subsequently in patients with BTC harboring HER2 overexpression. A pivotal case report published by Law et al. in 2012 described the case of a patient with metastatic HER2‐positive BTC treated with paclitaxel and trastuzumab achieving a dramatic disease response 32. A subsequent retrospective analysis described eight cases of BTC treated with anti‐HER2 agents (trastuzumab, lapatinib, and pertuzumab) obtaining promising results in terms of response rate (50%) and duration of response (median 40 weeks) 33. Recently, preliminary data of two basket trials with HER2‐targeted agents (trastuzumab plus pertuzumab and trastuzumab emtansine) confirmed HER2 as a promising therapeutic target in advanced BTC 34, 35.

Although the retrospective nature of our study and the number of analyzed samples represent potential limitations, results reported in terms of DFS and OS according to HER2 status suggest that this biomarker may play a role in promoting a more aggressive behavior in a subset of BTC cases. Because of the exploratory nature of our study, our results should be confirmed in prospective validating cohorts. Intriguingly, as no correlation between HER2 status and the other clinical and pathological parameters has emerged, in future studies all patients with BTC should be screened for HER2 overexpression and could thus potentially benefit from anti‐HER2 therapy. Moving from our data as well as considering the preliminary results achieved with targeted agents in selected BTC cases, prospective studies are warranted in order to improve the results achieved with cytotoxic chemotherapy in first‐line or even in the adjuvant setting.

Conclusion

Our study shows that HER2 expression represents an independent prognostic factor in BTC cases treated with potentially curative surgery. This evidence could help in refining prognostic stratification after resection but, more importantly, should be the rationale to investigate HER2 as a therapeutic target in this challenging clinical situation.

Author Contributions

Conception/design: Caterina Vivaldi, Lorenzo Fornaro, Enrico Vasile

Provision of study material or patients: Caterina Vivaldi, Lorenzo Fornaro, Clara Ugolini, Cristina Niccoli, Gianna Musettini, Irene Pecora, Andrea Cacciato Insilla, Francesca Salani, Giulia Pasquini, Silvia Catanese, Monica Lencioni, Gianluca Masi, Daniela Campani, Gabriella Fontantini, Alfredo Falcone, Enrico Vasile

Collection and/or assembly of data: Caterina Vivaldi, Lorenzo Fornaro, Clara Ugolini, Cristina Niccoli, Gianna Musettini, Irene Pecora, Andrea Cacciato Insilla, Francesca Salani, Giulia Pasquini, Silvia Catanese, Monica Lencioni, Gianluca Masi, Daniela Campani, Gabriella Fontantini, Alfredo Falcone, Enrico Vasile

Data analysis and interpretation: Caterina Vivaldi, Lorenzo Fornaro, Clara Ugolini, Enrico Vasile

Manuscript writing: Caterina Vivaldi, Lorenzo Fornaro, Clara Ugolini, Enrico Vasile

Final approval of manuscript: Caterina Vivaldi, Lorenzo Fornaro, Clara Ugolini, Cristina Niccoli, Gianna Musettini, Irene Pecora, Andrea Cacciato Insilla, Francesca Salani, Giulia Pasquini, Silvia Catanese, Monica Lencioni, Gianluca Masi, Daniela Campani, Gabriella Fontantini, Alfredo Falcone, Enrico Vasile

Disclosures

Caterina Vivaldi: Eli Lilly & Co. (H), Bayer (other—travel expenses); Lorenzo Fornaro: Eli Lilly & Co. (H), Merck Sharp & Dohme (SAB), Gilead Sciences, Merck Sharp & Dohme (RF), Celgene (other—travel expenses); Alfredo Falcone: Bayer, Bristol, Eli Lilly & Co., Merck, Pierre‐Fabre, Roche, Servier (C/A, H), AstraZeneca, Bayer, Bristol, Eli Lilly & Co., Merck, Merck Sharp & Dohme, Novertis, Roche, Sanofi, Servier (RF—institutional).

(C/A) Consulting/advisory relationship; (RF) Research funding; (E) Employment; (ET) Expert testimony; (H) Honoraria received; (OI) Ownership interests; (IP) Intellectual property rights/inventor/patent holder; (SAB) Scientific advisory board

Supporting information

See http://www.TheOncologist.com for supplemental material available online.

Supplementary Table 1 Correlation between IHC and FISH results