Abstract
The therapeutic options for patients with metastatic medullary thyroid carcinoma (MTC) have recently increased due to the development of tyrosine kinase inhibitors (TKIs), some of which have achieved remarkable clinical responses in MTC patients. However, the molecular basis for the large variability in TKI responses is unknown. In this exploratory study, we investigated the expression of eight key TKI target proteins (EGFR, KIT, MET, PDGFRB, VEGF (VEGFA), VEGFR1 (FLT1), VEGFR2 (KDR), and VEGFR3 (FLT4)) by immunohistochemistry in 103 molecularly characterized MTC samples and identified the associated clinical and molecular features. A number of MTC samples exhibited a high expression of VEGFR2 and VEGFR3, which were overexpressed in 57 and 43% of the MTC samples respectively. VEGFR1, PDGFRB, VEGF, KIT, and MET were present in 34–20% of the cases, while EGFR was highly expressed in only 10% of the MTC samples. Some proteins exhibited large differences in expression between sporadic and familial cases, suggesting that different RET mutations may be associated with the immunohistochemical profiles. MTC samples with the C634 RET mutation exhibited a higher expression of VEGFR3 and KIT than the M918T RET-mutated and non-mutated RET tumor samples (P=0.005 and P=0.007 respectively) and a lower expression of VEGFR1 (P=0.04). Non-mutated RET MTC cases exhibited a lower expression of PDGFRB (P=0.04). Overall, this is the first study, to our knowledge, to show that multiple TKI targets are highly expressed in a subset of MTCs, suggesting that molecular stratification of patients may have the potential to improve TKI therapies for MTC.
Introduction
Medullary thyroid carcinoma (MTC) arises from the parafollicular cells of the thyroid gland. Sporadic MTC accounts for 75% of the cases, and the remaining 25% is inherited in nature as part of multiple endocrine neoplasia type 2 syndrome. The RET proto-oncogene plays a major role in MTC development, with 30–50% of the sporadic MTCs carrying somatic RET mutations (Uchino et al. 1999). RET mutations involved in MTC are gain-of-function alterations that increase RET kinase activity, resulting in a constant activation of downstream signaling pathways that ultimately lead to tumor growth (Schuffenecker et al. 1998, Randolph & Maniar 2000, Nagy et al. 2004). The presence of specific RET mutations determines the age of presentation and aggressiveness of the tumor, allowing for genetic screening and recommendations for preventive surgical management in familial cases (Brandi et al. 2001, Cote & Gagel 2003).
Patients with MTC undergo total thyroidectomy and lymph node dissection. However, because MTC is derived from neuroendocrine cells, it is unresponsive to radioiodine and TSH suppression, and it is unclear as to which is the most appropriate treatment for patients with residual or recurrent disease after primary surgery and for those with distant metastasis. Until recently, management of metastatic disease has primarily been oriented toward the relief of symptoms (Wells et al. 1982, Giraudet et al. 2007), but in the last few years, much effort has been devoted to developing clinical trials using targeted therapies.
In addition to RET mutations, vascular endothelial growth factor (VEGF)-mediated angiogenesis, leading to increased tumor growth and invasiveness, has been recognized as an important feature in MTC. Thus, targeted molecular therapies that inhibit oncogenic kinases such as RET and tyrosine kinase receptors involved in angiogenesis could be important for the treatment of metastatic or locally advanced MTC. In this regard, a number of small-molecule inhibitors that selectively inhibit tyrosine kinase receptors, such as vandetanib, sorafenib, sunitinib, axitinib, motesanib, and cabozantinib (XL184), have shown remarkable clinical responses in MTC patients, inducing partial responses and stabilization of the disease in a substantial number of patients (Cohen et al. 2008, Schlumberger et al. 2009, Carr et al. 2010, Lam et al. 2010, Robinson et al. 2010, Wells et al. 2010, 2012, Ahmed et al. 2011, Hong et al. 2011, Kurzrock et al. 2011). Although comprehensive clinical trials for MTC should ideally incorporate patient and tumor characteristics, most of these studies include only a small number of MTC patients who are not characterized or are heterogeneous with respect to the specific RET mutation. Some kinase receptors, such as EGFR and VEGFR2, have been shown to be overexpressed in subsets of primary MTC tumors and metastases (Rodriguez-Antona et al. 2010). However, currently such information is lacking for most of the tyrosine kinase inhibitor (TKI) targets, and it is unknown whether they are expressed in MTC and whether this expression is associated with specific clinical and molecular features. Therefore, we set out to investigate the expression of eight key TKI target proteins in an outstanding series of 103 molecularly characterized MTC cases. This is the first study, to our knowledge, to demonstrate that multiple TKI targets are highly expressed in a subset of MTCs in a RET mutation-dependent manner. This information might be critical for the inclusion of MTC patients in future clinical trials and, ultimately, for improving treatment response.
Subjects and methods
Human MTC samples
A total of 103 paraffin-embedded MTC samples from 101 patients were obtained from the Spanish National Cancer Centre in collaboration with the CNIO tumor Bank. Institutional Review Board approval was obtained for the study, and informed consent was obtained from all the patients. The patients were aged from 11 to 80 years (median age 50 years), and 56% were females. Of the 103 MTC samples, 92 corresponded to primary tumors and 8 to metastases (see Table 1). The mutational status of the RET proto-oncogene in exons 10, 11, 13, 14, 15, and 16 was assessed from genomic DNA using standard PCR conditions, primers, and automated sequencing as described previously (Ceccherini et al. 1993). The tumor samples corresponded in most cases to patients diagnosed as sporadic or familial, based on the analysis of the RET proto-oncogene in peripheral blood samples. Familial cases carried germline mutations in exon 10, 11, 14, or 15, while sporadic cases carried somatic RET mutations in exon 10, 11, 15, or 16, or were classified as ‘no mutation in RET’ when no somatic mutations were found in exon 10, 11, 13, 14, 15, or 16 (Table 1). Among the apparently sporadic cases (without RET mutations in blood samples), four tumors were classified as ‘undetermined RET mutation’ due to the failure of the PCR analysis caused by low tumor DNA quality. Among the familial cases, one tumor derived from a patient with a family history of multiple neoplasia type 2 (second-degree relative diagnosed with a pheochromocytoma and mother with a MTC) was classified as ‘undetermined RET mutation’ because no blood sample was available and the tumor DNA was of low quality. Five cases could not be classified as sporadic or familial (referred to as ‘unknown’) because blood samples were not available.
Characteristics of the 101 MTC patients included in the study and mutational status of the corresponding samples
| Characteristics | n | Percentage (%) |
|---|---|---|
| Age (years) | ||
| Median | 50 | |
| Range (min–max) | 38–59 (11–80) | |
| Gender | ||
| Female | 57 | 56 |
| Male | 40 | 40 |
| Unknown | 4 | 4 |
| Type of tumor | ||
| Primary | 92 | 89 |
| Metastatic | 8 | 8 |
| Unknown | 3 | 3 |
| Genetic featuresa | ||
| Sporadic | 69 | 68 |
| RET exon 10 (C618, F619, and C620) | 3 | 4 |
| RET exon 11 (C634, c.1894_1900 del GAGCTGT ins A) | 5 | 7 |
| RET exon 15 (A883) | 1 | 1 |
| RET exon 16 (M918T) | 17 | 25 |
| No mutation in RETb | 39 | 57 |
| Undetermined RET mutationc | 4 | 6 |
| Familial | 27 | 27 |
| RET exon 10 (C618) | 4 | 15 |
| RET exon 11 (C634) | 20 | 74 |
| RET exon 14 (V804) | 1 | 4 |
| RET exon 15 (S891) | 1 | 4 |
| Undetermined RET mutationc | 1 | 4 |
| Unknown | 5 | 5 |
| Undetermined RET mutationc | 5 | 100 |
Percentages in bold type refer to all the patients, whereas percentages in regular type refer to the specific subgroup (sporadic, familial, or unknown).
No mutations found in RET exons 10, 11, 13, 14, 15, and 16.
Tumors in which RET exon 10, 11, 13, 14, 15, or 16 could not be assessed.
Immunohistochemical study
Hematoxylin and eosin-stained sections of each tumor sample were examined by two pathologists to confirm the diagnosis and to select MTC areas representative of each tumor to construct tissue microarrays (TMAs). Three TMAs containing all the tumor samples (103) were constructed as described previously with two cores of each tumor placed at different positions in the TMAs (Cascon et al. 2005).
The three paraffin-embedded TMAs were used for the detection of EGFR, KIT, MET, PDGFRB, VEGF (VEGFA), VEGFR1 (FLT1), VEGFR2 (KDR), and VEGFR3 (FLT4) proteins by immunohistochemistry (IHC) using specific antibodies. The suppliers, dilutions, visualization systems, and immunostainers used for the antibodies are given in Supplementary Table 1, see section on supplementary data at the end of this article. Two independent experienced pathologists (I Muñoz-Repeto and M Cañamero) evaluated the intensity and extension of staining for all the antibodies by visual examination under a microscope. Not only the tumoral cells but also the stroma (fibroblasts, inflammatory cells, and blood vessels) were evaluated, taking into account the fact that these tumors are highly cellular with low stromal component, mostly consisting of blood vessels. Since each TMA included two different tumor cylinders from each case, immunohistochemical scoring was done after examining both samples.
The IHC scoring used was as follows: for EGFR, VEGFR2, and PDGFRB, tumor samples with moderate/strong staining were considered positive (Maderna et al. 2007, Rodriguez-Antona et al. 2010); for KIT, both the intensity and extent of staining were evaluated (Miliaras et al. 2004), and the intensity of staining was graded as 0 (absent), 1 (weak), 2 (moderate), or 3 (strong), and the extent of staining was evaluated semiquantitatively and categorized as 0 (0% of cells), 1 (<10%), 2 (between 10 and 50%), 3 (between 50 and 80%), and 4 (more than 80%), and aggregate scores were obtained for each case (range 0–7) and cases with scores >3 were regarded as KIT-positive; MET protein was considered positive when its expression was positive for 30% of the tumor cells with moderate/strong staining (Lee et al. 2010); for VEGF, the intensity of staining was estimated on a four-tiered scale encoded as 0 (absent), 1 (weak), 2 (moderate), and 3 (strong), and immunopositivity was defined by strong staining (Duncan et al. 2008); for VEGFR1, a tumor was considered positive if cytoplasmic expression was detected; for VEGFR3, the percentage of cells with positive staining was evaluated for each case and the median was calculated, and tumor samples with a percentage of positive cells higher than the median (50% of the positive tumor cells) were considered to have a high expression of VEGFR3. The staining intensity for each protein marker analyzed and the number of tumor samples included in each category are given in Supplementary Table 2, see section on supplementary data at the end of this article.
Clustering
Hierarchical unsupervised cluster analysis of the 92 primary tumor cases was carried out by the average linkage clustering method using GeneCluster 3.0 (mean centered; de Hoon et al. 2004) and viewed in a visualizer (TreeView) that displays cluster profiles and relevant cluster member information. Immunohistochemical results were represented by a range of colors from green to red, with the brightest green representing the lowest staining intensity for each marker and the brightest red the highest one (Fig. 1). Three samples in which more than three antibodies failed to yield results were excluded from the analysis.
Hierarchical clustering of 89 primary MTC samples. The staining intensity for each immunohistochemical marker is represented as a range of colors between the brightest green (lowest expression) and the brightest red (highest expression). Gray squares indicate that no data was available.
Citation: Endocrine-Related Cancer 20, 4; 10.1530/ERC-12-0316
Statistical analyses
All statistical analyses were carried out using SPSS version 17.0 statistical software. The χ2 test or Fisher's exact test was used to compare variables representing patient characteristics (male/female) and tumor types (primary tumors/metastases and sporadic/familial) with IHC protein expression (Tables 2 and 3). In addition, logistic regression analyses were carried out to obtain odds ratios (ORs) and 95% CI, which are given in Supplementary Table 4, see section on supplementary data given at the end of this article. For RET mutation analysis, we considered independently each of the major RET-mutated groups: C634, M918T, and no RET mutation. The C634 RET-mutated group included both familial and sporadic forms with germline and somatic mutations respectively; the M918T RET-mutated group corresponded exclusively to sporadic forms with M918T somatic mutation; and the ‘no mutation in RET’ group corresponded exclusively to sporadic forms in which germline mutations were discarded and no somatic RET mutation was detected (see Table 1). Other less frequent RET mutations were not considered for further analysis as they represented a heterogeneous group. To calculate the correlation between the expression of the different proteins evaluated, we used the Spearman's test. Since this is an exploratory study, no correction of P values due to multiple testing was performed, and bilateral P values <0.05 were considered significant.
Protein expression of EGFR, KIT, MET, PDGFRB, VEGF, VEGFR1, VEGFR2, and VEGFR3 according to the type of MTC
| Protein | Primary n (%) | Metastatic n (%) | P | Sporadica n (%) | Familiala n (%) | P |
|---|---|---|---|---|---|---|
| EGFR−b | 83 (91) | 5 (71) | NS | 58 (92) | 24 (89) | NS |
| EGFR+c | 8 (9) | 2 (29) | 5 (8) | 3 (11) | ||
| KIT− | 65 (75) | 8 (100) | NS | 50 (83) | 14 (54) | 0.0040 |
| KIT+ | 22 (25) | 0 (0) | 10 (17) | 12 (46) | ||
| MET− | 63 (83) | 4 (50) | 0.049 | 48 (84) | 14 (78) | NS |
| MET+ | 13 (17) | 4 (50) | 9 (16) | 4 (22) | ||
| PDGFRB− | 60 (70) | 3 (38) | NS | 46 (78) | 13 (50) | 0.010 |
| PDGFRB+ | 26 (30) | 5 (63) | 13 (22) | 13 (50) | ||
| VEGF− | 62 (69) | 4 (50) | NS | 41 (65) | 20 (77) | NS |
| VEGF+ | 28 (31) | 4 (50) | 22 (35) | 6 (23) | ||
| VEGFR1− | 45 (62) | 8 (100) | 0.046 | 31 (58) | 14 (78) | NS |
| VEGFR1+ | 27 (38) | 0 (0) | 22 (41) | 4 (22) | ||
| VEGFR2− | 32 (43) | 2 (29) | NS | 23 (42) | 8 (44) | NS |
| VEGFR2+ | 42 (57) | 5 (71) | 32 (58) | 10 (56) | ||
| VEGFR3− | 50 (57) | 4 (57) | NS | 39 (65) | 10 (38) | 0.022 |
| VEGFR3+ | 37 (43) | 3 (43) | 21 (35) | 16 (61) |
Only primary tumor cases included in the analysis.
A negative expression is indicated by ‘−’.
A positive expression is indicated by ‘+’.
Protein expression of EGFR, KIT, MET, PDGFRB, VEGF, VEGFR1, VEGFR2, and VEGFR3 according to the RET mutation in primary MTCs
| Protein | C634 muta n (%) | Restb n (%) | P | M918T mutc n (%) | Restb n (%) | P | Wt RETd n (%) | Restb n (%) | P |
|---|---|---|---|---|---|---|---|---|---|
| EGFR−e | 22 (92) | 56 (90) | NS | 14 (93) | 64 (90) | NS | 33 (92) | 45 (90) | NS |
| EGFR+f | 2 (8) | 6 (10) | 1 (7) | 7 (10) | 3 (8) | 5 (10) | |||
| KIT− | 13 (54) | 48 (83) | 0.0069 | 12 (92) | 49 (71) | NS | 29 (83) | 32 (68) | NS |
| KIT+ | 11 (46) | 10 (17) | 1 (8) | 20 (29) | 6 (17) | 15 (32) | |||
| MET− | 13 (77) | 46 (84) | NS | 9 (75) | 50 (83) | NS | 31 (91) | 28 (74) | 0.054 |
| MET+ | 4 (22) | 9 (16) | 3 (25) | 10 (17) | 3 (9) | 10 (26) | |||
| PDGFRB− | 13 (54) | 43 (75) | 0.058 | 10 (71) | 46 (69) | NS | 27 (82) | 29 (60) | 0.040 |
| PDGFRB+ | 11 (46) | 14 (25) | 4 (29) | 21 (31) | 6 (18) | 19 (40) | |||
| VEGF− | 18 (75) | 41 (67) | NS | 10 (70) | 49 (67) | NS | 23 (64) | 36 (73) | NS |
| VEGF+ | 6 (25) | 20 (33) | 5 (30) | 21 (33) | 13 (36) | 13 (27) | |||
| VEGFR1− | 14 (82) | 28 (55) | 0.044 | 4 (44) | 38 (64) | NS | 20 (61) | 22 (63) | NS |
| VEGFR1+ | 3 (18) | 23 (45) | 5 (56) | 21 (36) | 13 (39) | 13 (37) | |||
| VEGFR2− | 7 (41) | 25 (47) | NS | 6 (54) | 26 (44) | NS | 15 (45) | 17 (46) | NS |
| VEGFR2+ | 10 (59) | 28 (53) | 5 (44) | 33 (56) | 18 (55) | 20 (54) | |||
| VEGFR3− | 8 (33) | 39 (67) | 0.0047 | 9 (64) | 38 (56) | NS | 24 (71) | 23 (48) | 0.041 |
| VEGFR3+ | 16 (67) | 19 (33) | 5 (36) | 30 (44) | 10 (29) | 25 (52) |
Tumors with an activating mutation in RET residue C634, either germline or somatic.
The rest of the tumors that do not have the analyzed genetic characteristic.
Tumors with the activating somatic RET mutation M918T.
Sporadic tumors with no mutations in RET exons 10, 11, 13, 14, 15, and 16.
A negative expression is indicated by ‘−’.
A positive expression is indicated by ‘+’.
Results
Immunohistochemical study in MTC cases
The different TKI targets tested in the 103 MTC samples available for the study presented a variable expression (Supplementary Figure 1, see section on supplementary data given at the end of this article). Many MTCs showed positive staining with 57, 43, 34, 33, 32, 23, 20 and 10% positive cases for VEGFR2, VEGFR3, VEGFR1, PDGFRB, VEGF, KIT, MET and EGFR, respectively. The IHC intensity observed for each marker is summarized in detail in Supplementary Table 2.
The expression of several TKI targets was correlated. The strongest correlations corresponded to PDGFRB, which exhibited a positive correlation with VEGFR3 and a negative correlation with VEGFR1 (correlation coefficients of 0.51 and −0.34, P=9×10−8 and 0.002 respectively). A positive correlation was also detected between VEGF and VEGFR2 (correlation coefficient of 0.37, P=0.0006). The expression of EGFR correlated with that of MET and VEGFR3, and the expression of MET correlated with that of VEGFR2 and VEGF (see Supplementary Table 3, see section on supplementary data given at the end of this article). Similar results were observed when only primary tumor samples were analyzed.
Unsupervised clustering analysis of the primary MTCs based on the eight immunohistochemical markers evaluated clustered tumor samples into two main groups (Fig. 1). The left branch, which grouped 35% of the MTC cases, was characterized by positive PDGFRB and VEGFR3 staining and included most cases with EGFR overexpression. The right branch of the cluster included some of the VEGFR2-, VEGF-, MET-, KIT-, and VEGFR1-positive cases and cases displaying a negative expression for all the tested proteins. The left branch tended to be enriched in cases carrying the C634 RET mutation, either germline or somatic (35 vs 21%, in the left and right branches respectively), while the right branch was apparently enriched in tumors with the M918T RET somatic mutation (14 vs 6%).
Protein expression of TKI targets in different MTC types
Gender was associated with the expression of VEGFR1 (45% of female MTC cases exhibited positive staining vs 16% of the male cases, P=0.007) and VEGF (23 and 46% positive cases for females and males respectively, P=0.017), but age at diagnosis was not associated with IHC staining. We next determined whether the expression of the markers could be influenced by the type of tumor: primary tumors vs metastases and sporadic vs familial. MET displayed a higher expression in the metastatic cases than in the primary tumor cases (P=0.049; Table 2), similar to EGFR, PDGFRB, and VEGF, although differences for the latter proteins did not reach statistical significance (see Table 2). On the other hand, VEGFR1 displayed a higher expression in the primary tumor samples (P=0.046), similar to KIT, although this protein did not show any statistically significant difference. Some of the markers exhibited large differences in expression in relation to the inherited character of the disease. We found that positive KIT, PDGFRB, and VEGFR3 staining was more frequently associated with familial forms than with sporadic forms (P values of 0.0040, 0.010, and 0.022 respectively; Table 2; see also Supplementary Table 4). On the other hand, EGFR, MET, VEGF, VEGFR1, and VEGFR2 exhibited similar positive staining in familial and sporadic forms. Because familial cases are associated with specific mutations, these data suggest that different RET mutations can predict differences in the expression of some of these proteins.
Expression of TKI targets according to RET mutations
Because the metastatic samples exhibited altered expression of some of the proteins studied (see Table 2), we examined only primary tumor cases to determine the effect of specific RET mutations on protein levels. Cases with the C634 RET mutation included both germline and sporadic tumors, while cases with the M918T RET mutation and cases with no mutation in RET were all sporadic forms. We found that 67% of the MTC cases with the C634 RET mutation expressed VEGFR3, compared with 36 and 29% respectively of the MTC cases with the M918T RET mutation and the MTC cases without the RET mutation (P=0.0047; see Table 3; OR=4.1, 95% CI=1.5–11; Supplementary Table 4). The expression of KIT was also more frequently found in the C634 RET mutation cases than in the M918T RET mutation cases and non-mutated RET cases (P=0.0069; Table 3; OR=4.1, 95% CI=1.4–11; Supplementary Table 4). By contrast, VEGFR1 exhibited a lower expression in the C634 RET mutation cases than in the M918T or non-mutated RET cases (18, 56, and 39% respectively, P=0.044; Table 3). The expression of PDGFRB was higher in the C634 and M918T RET mutation cases than in the non-mutated RET cases (P=0.040; Table 3). Other proteins such as EGFR, VEGF, and VEGFR2 did not display differences in expression among the different types of RET-mutated tumors, while MET had a tendency toward a lower expression in MTC cases without RET mutations (Table 3).
The expression of KIT, PDGFRB, VEGFR1, and VEGFR3 was also examined in stroma: blood vessels, inflammatory cells, and fibroblasts (data not shown). KIT was only present in the inflammatory population in scattered cells; PDGFRB was expressed in blood vessels and fibroblasts, with an expression level ranging from moderate to high; and VEGFR1 and VEGFR3 were expressed in most blood vessels and occasionally in inflammatory cells. In addition, the expression of KIT was significantly lower in MTC cases with the C634 mutation (P=0.0018); no statistically significant differences for the expression of PDGFRB, VEGFR1, and VEGFR3 were found.
Discussion
The lack of effective therapies for MTC may be changed drastically by the use of already available TKIs and the development of novel targeted drugs. Some of the TKIs have demonstrated remarkable clinical responses in MTC patients, and several new trials with these molecules are currently being conducted (Cohen et al. 2008, Schlumberger et al. 2009, Carr et al. 2010, Lam et al. 2010, Robinson et al. 2010, Wells et al. 2010, 2012, Ahmed et al. 2011, Hong et al. 2011, Kurzrock et al. 2011). However, there is a large inter-patient variability in TKI responses. The molecular basis for this variability is unknown, and the identification of biomarkers that could identify MTC patients who will probably benefit from these different drugs will improve clinical trial outcomes and ultimately progression-free survival and overall survival of patients with a disease that not long ago was untreatable. In this study, we used a large retrospective series of well-characterized MTC patients to define the expression of TKI target proteins in relation to specific clinical and tumor characteristics to provide data that could be used for rational selection of patients for TKI treatment.
So far, vandetanib and cabozantinib, which have markedly improved the progression-free survival of MTC patients (Schoffski et al. 2013, Thornton et al. 2012), are the only agents that have been approved for the treatment of this disease, but a wide variety of additional multi-targeted kinase inhibitors have entered clinical trials and several have shown clinical benefit in MTC patients. Promising results have been obtained with agents that primarily target angiogenesis and inhibit VEGF receptors at nanomolar concentrations. Because of the structural similarity between kinases, many of these molecules also have an effect on RET and other kinases, such as PDGFR, MET, and KIT, which could be all-important for the clinical responses observed. However, axitinib does not have an anti-RET activity and has displayed objective responses in MTC patients, suggesting that VEGFR might be as important as RET for targeted therapy (Cohen et al. 2008). Comparison of outcomes among the various phase II trials carried out so far is limited by variations in patient eligibility and response assessment. In addition, in most cases, genetic characteristics of the different MTC patients included (e.g. sporadic or familial and type of RET mutation) have not been reported.
In this study, we examined the expression of eight key proteins for TKI action (EGFR, KIT, MET, PDGFRB, VEGF, VEGFR1, VEGFR2, and VEGFR3) in 103 paraffin-embedded MTC samples characterized for RET mutations. We observed that these proteins were highly expressed in a subset of MTCs in a coordinated manner (Fig. 1, Supplementary Table 3). A previous publication that examined VEGF, VEGFR1, and VEGFR2 in 38 MTC samples has reported 95, 96, and 91% positive expression for these proteins respectively (Capp et al. 2010). In the present study, we adopted more strict criteria to define immunopositivity and, thus, the proportions of MTC samples exhibiting overexpression of these proteins were lower (32, 34, and 57% respectively). Applying criteria the same as those used by Capp et al., we obtained similar results for VEGF and VEGFR2, but observed a lower expression of VEGFR1 (see Supplementary Table 2). This difference could be due to the different VEGFR1 antibody used. The expression of some proteins was significantly different between primary tumor and metastatic samples (Table 2), but the number of metastatic samples included in this study was small, and we cannot rule out the possibility that additional changes would be detected in a larger series. In fact, a previous study by our group, including more metastatic cases and paired primary tumor/metastatic samples, had found a significant overexpression of VEGFR2 in the metastatic samples (Rodriguez-Antona et al. 2010). In the present study, we also found a higher expression of VEGFR2 in the metastatic samples than in the primary tumor samples (71 vs 55%), but it did not reach statistical significance. Additional histopathological information and clinical outcome data were not available to conduct additional analyses. Differences in the expression of the proteins were evident when comparing sporadic and familial MTC cases (Table 2), with a higher expression of KIT, PDGFRB, and VEGFR3 being observed in the familial cases. Most hereditary MTCs are caused by RET mutations affecting residue C634 and most sporadic cases have the M918T RET mutation (see Table 1), suggesting that the expression of TKI targets and RET mutations may be associated. In agreement with this, we found that MTC cases exhibiting VEGFR3 and KIT overexpression were mainly C634 RET-mutated cases; VEGFR1-positive cases were mainly M918T RET-mutated cases and cases without RET mutations; and MTC cases exhibiting PDGFRB overexpression were mainly tumors without RET mutations. VEGFR2, which is a target of several TKIs, exhibited similar staining among the different RET-mutated tumor cases, with the number of positive cases exceeding 40% in all cases (Table 3). When we examined the expression of KIT, PDGFRB, VEGFR1, and VEGFR3 in the tumor stroma, only KIT displayed a statistically significant difference in the C634 RET-mutated tumor cases, suggesting that for the expression of the receptors, RET mutation is only relevant in the tumor cells. In general, and although a validation with an independent series is required to confirm these data, these results indicate that different TKI treatments could be more effective according to the specific RET mutation present in the MTC. It could be suggested that C634 RET-mutated cases (mainly hereditary), with a higher expression of VEGFR3, PDGFRB, and KIT, might benefit from drugs with a high affinity for these targets, while for the M918T RET-mutated cases (mostly sporadic), TKIs targeting RET and VEGFR2 might be more appropriate. In fact, a phase III clinical trial comparing vandetanib vs placebo has found that the response rate to this drug is greater in patients with sporadic tumors who had a M918T RET mutation (Wells et al. 2012). However, it is also important to note that intratumoral heterogeneity and changes in tumor molecular profile through the acquisition of new somatic mutations suggest that combinations of more than one TKI may be more effective than single-agent treatments and that changing to different TKIs over time might also be needed.
In conclusion, this study shows for the first time to our knowledge that a substantial number of MTCs exhibit high expression levels of kinases targeted by TKIs for which promising results have been obtained in recent clinical trials. Furthermore, the expression of these targets is associated with clinical and molecular characteristics of the MTCs, supporting the notion that these data could be used for the identification of patients most likely to benefit from specific TKIs, thus helping to design rational clinical trials and perform a molecular selection of treatments to ultimately improve the clinical response of MTC patients.
Supplementary data
This is linked to the online version of the paper at http://dx.doi.org/10.1530/ERC-12-0316.
Declaration of interest
The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of the research reported.
Funding
This work was supported by Fondo de Investigaciones Sanitarias – FIS project PI11/01359 (to M Robledo), the Comunidad de Madrid S2011/BMD-2328 TIRONET, and SAF2012-35779. L Inglada-Pérez is supported by CIBERER. A Maliszewska, V Mancikova, and A A de Cubas are predoctoral fellows of the ‘la Caixa’/CNIO international PhD program. L J Leandro-García is a predoctoral fellow of the Fondo de Investigaciones Sanitarias. L Sanchez is supported by a project from the Spanish Ministry of Science and Innovation (SAF2009-08307).
Author contribution statement
C Rodríguez-Antona was responsible for study design, data analysis, data interpretation, manuscript writing, and final approval of the manuscript. I Muñoz-Repeto, L Inglada-Pérez, A A de Cubas, and M Cañamero were involved in data analysis, data interpretation, and final approval of the manuscript. V Mancikova and A Cascón were responsible for data collection, data interpretation, and final approval of the manuscript. A Maliszewska, L Sanchez, A Gómez, R Letón, C Álvarez-Escolá, and J Aller were responsible for data collection and final approval of the manuscript. L J Leandro-García and I Comino-Méndez were responsible for data interpretation and final approval of the manuscript. M Robledo was responsible for study design, data interpretation, manuscript writing, and final approval of the manuscript.
Acknowledgements
The authors thank Tissue Bank Network coordinated by the Molecular Pathology Program of the Spanish National Cancer Centre (CNIO) for providing tissue samples.
References
Ahmed M, Barbachano Y, Riddell A, Hickey J, Newbold KL, Viros A, Harrington KJ, Marais R & Nutting CM 2011 Analysis of the efficacy and toxicity of sorafenib in thyroid cancer: a phase II study in a UK based population. European Journal of Endocrinology 165 315–322. (doi:10.1530/EJE-11-0129).
Brandi ML, Gagel RF, Angeli A, Bilezikian JP, Beck-Peccoz P, Bordi C, Conte-Devolx B, Falchetti A, Gheri RG & Libroia A et al. 2001 Guidelines for diagnosis and therapy of MEN type 1 and type 2. Journal of Clinical Endocrinology and Metabolism 86 5658–5671. (doi:10.1210/jc.86.12.5658).
Capp C, Wajner SM, Siqueira DR, Brasil BA, Meurer L & Maia AL 2010 Increased expression of vascular endothelial growth factor and its receptors, VEGFR-1 and VEGFR-2, in medullary thyroid carcinoma. Thyroid 20 863–871. (doi:10.1089/thy.2009.0417).
Carr LL, Mankoff DA, Goulart BH, Eaton KD, Capell PT, Kell EM, Bauman JE & Martins RG 2010 Phase II study of daily sunitinib in FDG–PET-positive, iodine-refractory differentiated thyroid cancer and metastatic medullary carcinoma of the thyroid with functional imaging correlation. Clinical Cancer Research 16 5260–5268. (doi:10.1158/1078-0432.CCR-10-0994).
Cascon A, Ruiz-Llorente S, Rodriguez-Perales S, Honrado E, Martinez-Ramirez A, Leton R, Montero-Conde C, Benitez J, Dopazo J & Cigudosa JC et al. 2005 A novel candidate region linked to development of both pheochromocytoma and head/neck paraganglioma. Genes, Chromosomes & Cancer 42 260–268. (doi:10.1002/gcc.20139).
Ceccherini I, Bocciardi R, Luo Y, Pasini B, Hofstra R, Takahashi M & Romeo G 1993 Exon structure and flanking intronic sequences of the human RET proto-oncogene. Biochemical and Biophysical Research Communications 196 1288–1295. (doi:10.1006/bbrc.1993.2392).
Cohen EE, Rosen LS, Vokes EE, Kies MS, Forastiere AA, Worden FP, Kane MA, Sherman E, Kim S & Bycott P et al. 2008 Axitinib is an active treatment for all histologic subtypes of advanced thyroid cancer: results from a phase II study. Journal of Clinical Oncology 26 4708–4713. (doi:10.1200/JCO.2007.15.9566).
Cote GJ & Gagel RF 2003 Lessons learned from the management of a rare genetic cancer. New England Journal of Medicine 349 1566–1568. (doi:10.1056/NEJMe038148).
Duncan TJ, Al-Attar A, Rolland P, Scott IV, Deen S, Liu DT, Spendlove I & Durrant LG 2008 Vascular endothelial growth factor expression in ovarian cancer: a model for targeted use of novel therapies? Clinical Cancer Research 14 3030–3035. (doi:10.1158/1078-0432.CCR-07-1888).
Giraudet AL, Vanel D, Leboulleux S, Auperin A, Dromain C, Chami L, Ny Tovo N, Lumbroso J, Lassau N & Bonniaud G et al. 2007 Imaging medullary thyroid carcinoma with persistent elevated calcitonin levels. Journal of Clinical Endocrinology and Metabolism 92 4185–4190. (doi:10.1210/jc.2007-1211).
Hong DS, Cabanillas ME, Wheler J, Naing A, Tsimberidou AM, Ye L, Busaidy NL, Waguespack SG, Hernandez M & El Naggar AK et al. 2011 Inhibition of the Ras/Raf/MEK/ERK and RET kinase pathways with the combination of the multikinase inhibitor sorafenib and the farnesyltransferase inhibitor tipifarnib in medullary and differentiated thyroid malignancies. Journal of Clinical Endocrinology and Metabolism 96 997–1005. (doi:10.1210/jc.2010-1899).
de Hoon MJ, Imoto S, Nolan J & Miyano S 2004 Open source clustering software. Bioinformatics 20 1453–1454. (doi:10.1093/bioinformatics/bth078).
Kurzrock R, Sherman SI, Ball DW, Forastiere AA, Cohen RB, Mehra R, Pfister DG, Cohen EE, Janisch L & Nauling F et al. 2011 Activity of XL184 (cabozantinib), an oral tyrosine kinase inhibitor, in patients with medullary thyroid cancer. Journal of Clinical Oncology 29 2660–2666. (doi:10.1200/JCO.2010.32.4145).
Lam ET, Ringel MD, Kloos RT, Prior TW, Knopp MV, Liang J, Sammet S, Hall NC, Wakely PE Jr & Vasko VV et al. 2010 Phase II clinical trial of sorafenib in metastatic medullary thyroid cancer. Journal of Clinical Oncology 28 2323–2330. (doi:10.1200/JCO.2009.25.0068).
Lee JC, Li CF, Fang FM, Wang JW, Jeng YM, Yu SC, Lin YT, Wu JM, Tsai JW & Li SH et al. 2010 Prognostic implication of MET overexpression in myxofibrosarcomas: an integrative array comparative genomic hybridization, real-time quantitative PCR, immunoblotting, and immunohistochemical analysis. Modern Pathology 23 1379–1392. (doi:10.1038/modpathol.2010.128).
Maderna E, Salmaggi A, Calatozzolo C, Limido L & Pollo B 2007 Nestin, PDGFRβ, CXCL12 and VEGF in glioma patients: different profiles of (pro-angiogenic) molecule expression are related with tumor grade and may provide prognostic information. Cancer Biology & Therapy 6 1018–1024. (doi:10.4161/cbt.6.7.4362).
Miliaras D, Karasavvidou F, Papanikolaou A & Sioutopoulou D 2004 KIT expression in fetal, normal adult, and neoplastic renal tissues. Journal of Pathology 57 463–466. (doi:10.1136/jcp.2003.013532).
Nagy R, Sweet K & Eng C 2004 Highly penetrant hereditary cancer syndromes. Oncogene 23 6445–6470. (doi:10.1038/sj.onc.1207714).
Randolph GW & Maniar D 2000 Medullary carcinoma of the thyroid. Cancer Control 7 253–261.
Robinson BG, Paz-Ares L, Krebs A, Vasselli J & Haddad R 2010 Vandetanib (100 mg) in patients with locally advanced or metastatic hereditary medullary thyroid cancer. Journal of Clinical Endocrinology and Metabolism 95 2664–2671. (doi:10.1210/jc.2009-2461).
Rodriguez-Antona C, Pallares J, Montero-Conde C, Inglada-Perez L, Castelblanco E, Landa I, Leskela S, Leandro-Garcia LJ, Lopez-Jimenez E & Leton R et al. 2010 Overexpression and activation of EGFR and VEGFR2 in medullary thyroid carcinomas is related to metastasis. Endocrine-Related Cancer 17 7–16. (doi:10.1677/ERC-08-0304).
Schlumberger MJ, Elisei R, Bastholt L, Wirth LJ, Martins RG, Locati LD, Jarzab B, Pacini F, Daumerie C & Droz JP et al. 2009 Phase II study of safety and efficacy of motesanib in patients with progressive or symptomatic, advanced or metastatic medullary thyroid cancer. Journal of Clinical Oncology 27 3794–3801. (doi:10.1200/JCO.2008.18.7815).
Schoffski P, Elisei R, Müller S, Brose MS, Shah MH, Licitra LF, Jarzab B, Medvedev V, Kreissl M & Niederle B et al. 2013 An international, double-blind, randomized, placebo-controlled phase III trial (EXAM) of cabozantinib (XL184) in medullary thyroid carcinoma (MTC) patients (pts) with documented RECIST progression at baseline. Journal of Clinical Oncology 31 412–419. (doi:10.1200/JCO.2012.45.0494).
Schuffenecker I, Virally-Monod M, Brohet R, Goldgar D, Conte-Devolx B, Leclerc L, Chabre O, Boneu A, Caron J & Houdent C et al. 1998 Risk and penetrance of primary hyperparathyroidism in multiple endocrine neoplasia type 2A families with mutations at codon 634 of the RET proto-oncogene. Groupe D'etude des Tumeurs a Calcitonine. Journal of Clinical Endocrinology and Metabolism 83 487–491. (doi:10.1210/jc.83.2.487).
Thornton K, Kim G, Maher VE, Chattopadhyay S, Tang S, Moon YJ, Song P, Marathe A, Balakrishnan S & Zhu H et al. 2012 Vandetanib for the treatment of symptomatic or progressive medullary thyroid cancer in patients with unresectable locally advanced or metastatic disease: U.S. Food and drug administration drug approval summary. Clinical Cancer Research 18 3722–3730. (doi:10.1158/1078-0432.CCR-12-0411).
Uchino S, Noguchi S, Yamashita H, Sato M, Adachi M, Yamashita H, Watanabe S, Ohshima A, Mitsuyama S & Iwashita T et al. 1999 Somatic mutations in RET exons 12 and 15 in sporadic medullary thyroid carcinomas: different spectrum of mutations in sporadic type from hereditary type. Japanese Journal of Cancer Research 90 1231–1237. (doi:10.1111/j.1349-7006.1999.tb00701.x).
Wells SA Jr, Baylin SB, Johnsrude IS, Harrington DP, Mendelsohn G, Ontjes DJ & Cooper CW 1982 Thyroid venous catheterization in the early diagnosis of familial medullary thyroid carcinoma. Annals of Surgery 196 505–511. (doi:10.1097/00000658-198211000-00001).
Wells SA Jr, Gosnell JE, Gagel RF, Moley J, Pfister D, Sosa JA, Skinner M, Krebs A, Vasselli J & Schlumberger M 2010 Vandetanib for the treatment of patients with locally advanced or metastatic hereditary medullary thyroid cancer. Journal of Clinical Oncology 28 767–772. (doi:10.1200/JCO.2009.23.6604).
Wells SA Jr, Robinson BG, Gagel RF, Dralle H, Fagin JA, Santoro M, Baudin E, Elisei R, Jarzab B & Vasselli JR et al. 2012 Vandetanib in patients with locally advanced or metastatic medullary thyroid cancer: a randomized, double-blind phase III trial. Journal of Clinical Oncology 30 134–141. (doi:10.1200/JCO.2011.35.5040).
