Article Text

Download PDFPDF

mTOR in squamous cell carcinoma of the oesophagus: a potential target for molecular therapy?
  1. J Boone1,
  2. F J W Ten Kate2,
  3. G J A Offerhaus2,
  4. P J van Diest2,
  5. I H M Borel Rinkes1,
  6. R van Hillegersberg1
  1. 1
    Department of Surgery, University Medical Center Utrecht, Heidelberglaan 100, Utrecht, The Netherlands
  2. 2
    Department of Pathology, University Medical Center Utrecht, Heidelberglaan 100, Utrecht, The Netherlands
  1. Richard van Hillegersberg, Department of Surgery (G04.228), University Medical Center Utrecht, Heidelberglaan 100, 3584 CX Utrecht, The Netherlands; r.vanhillegersberg{at}


Aims: The mammalian target of rapamycin (mTOR), an important regulator of protein translation and cell proliferation, is activated in various malignancies. In a randomised controlled trial of advanced renal cell carcinoma patients, targeted therapy to mTOR by means of rapamycin analogues has been shown to significantly improve survival. An in vitro study has revealed that mTOR is activated in oesophageal squamous cell carcinoma (OSCC) cell lines and that mTOR expression is inhibited by rapamycin. The objectives of this histological study were to determine the proportion of OSCC tissues with activated mTOR (p-mTOR) expression, thereby assessing the percentage of patients with OSCC that would possibly benefit from neoadjuvant rapamycin therapy, and to identify the clinicopathological features of these potentially rapamycin-sensitive tumours.

Methods: The expression of p-mTOR (Ser2448) was immunohistochemically assessed in a validated tissue microarray comprising triplicate tissue biopsy cores of 108 formalin-fixed, paraffin-embedded OSCCs. Staining results were correlated with clinicopathological data.

Results: Normal oesophageal epithelium was negative for p-mTOR. Activated mTOR expression was located in the cytoplasm of oesophageal tumour cells. 26 (25%) of 105 assessable OSCCs showed tumour cells with positive staining for activated mTOR. Activated mTOR expression was associated with a lesser degree of differentiation only (p = 0.024). No correlation was detected between p-mTOR and the proliferation marker Ki-67.

Conclusions: Activated mTOR can be detected in one-quarter of OSCCs. Since this subset of patients may potentially benefit from mTOR inhibiting therapy, a phase II clinical trial of neoadjuvant mTOR-inhibiting therapy in patients with OSCC may be considered.

Statistics from

Oesophageal cancer is the eighth most common malignancy in the world, with an estimated 462 000 new cases diagnosed in 2002.1 Radical surgical resection of the oesophagus with an extensive lymph node dissection offers the best chance for cure.2 Notwithstanding recent advances in surgical strategies and postoperative care, the prognosis of patients is limited with an overall 5-year survival rate after oesophagectomy of around 30%.3 4 A meta-analysis of randomised controlled trials has shown that neoadjuvant chemoradiotherapy, compared with surgery alone, improves the 2-year survival rate in oesophageal cancer patients by approximately 10% by enhancing locoregional control and opposing early metastatic spread.5 Although at present the most commonly used neoadjuvant agents are conventional chemotherapeutics, the expanding knowledge on molecular carcinogenetic pathways has ushered in a new era of neoadjuvant therapy: molecular targeted therapy.6 The addition of molecular targeted agents to current neoadjuvant treatment regimens might further improve the survival after surgery for oesophageal cancer.

An innovative target for molecular therapy is the mammalian target of rapamycin (mTOR). This 289 kDa serine/threonine protein kinase is activated by the phosphatidylinositol-3-kinase (PI3K)/Akt signalling pathway and functions as a key regulator of protein translation and cell proliferation by phosphorylating its downstream markers p70s6k and 4E-binding protein1.79 mTOR is upregulated in various adenocarcinomas, such as prostate cancer, renal cell carcinoma and breast cancer.1013 This is of particular clinical interest as mTOR inhibitors (ie, rapamycin analogues such as temsirolimus and everolimus) are commercially available.14 A recent randomised controlled trial has revealed that temsirolimus significantly improves the overall survival of advanced renal cell carcinoma patients.15

Regarding squamous cell carcinomas, analysis of the protein expression of mTOR is scarce and restricted to oral cancer and head and neck carcinomas.16 17 To date, in oesophageal squamous cell carcinoma (OSCC), only an in vitro study has been performed, and this has shown an activated state of mTOR in cell lines, with rapamycin reducing mTOR expression.18 However, no studies are available that have assessed the activation status of mTOR protein in OSCC specimens.

The expression of molecular markers in formalin-fixed, paraffin-embedded tissues is usually assessed by immunohistochemistry. In 1998, Kononen et al introduced the tissue microarray (TMA) technology to enhance throughput.19 By inserting small (diameter 0.6 mm) cores of paraffin-embedded tissues into a single recipient block, this technique allows for rapid immunohistochemical analysis of hundreds of tissues concurrently under identical laboratory and evaluation conditions, without significantly damaging the patient’s tissue.20

The objectives of the present study were to determine the proportion of OSCC tissues showing expression of activated mTOR (p-mTOR), thereby assessing the percentage of patients with OSCC that would possibly benefit from neoadjuvant rapamycin therapy and identifying the clinicopathological features of these potentially rapamycin-sensitive tumours.


Patients and specimens

One-hundred and eight consecutive patients with OSCC who had undergone transhiatal or transthoracic oesophagolymphadenectomy between 1989 and 2006 at the University Medical Center Utrecht were included in this study. Patients who had received neoadjuvant therapy were excluded.

Paraffin-embedded tissue specimens of all patients were retrieved from the archives of the Department of Pathology. The study was carried out in accordance with the ethical guidelines of our institution concerning allowing anonymous or coded use of left over tissue from surgical procedures.21 From all paraffin blocks, a 4 μm slide was stained with H&E for histopathological diagnosis. All tumours were graded for differentiation (G1, well differentiated; G2, moderately differentiated; G3, poorly differentiated),22 infiltration depth (T1–T4),23 number of dissected lymph nodes, metastatic lymph node involvement and presence of extracapsular growth by an experienced pathologist (FJWTK). Tumours were staged according to the most recent tumour, node, metastasis (TNM) staging system.23

TMA construction

On one selected H&E-stained section of each tumour, three representative tumour regions were marked, avoiding areas of necrosis. From these regions, a tissue cylinder with a diameter of 0.6 mm was punched out of the corresponding paraffin block and placed into the TMA paraffin block using a custom-made precision instrument (MTA-I; Beecher Instruments, Sun Prairie, Wiconsin, USA), which was guided by the MTA Booster (Alphelys, Plaisir, France). The distribution and position of the cores was determined in advance with the TMA-Designer Software (V.1.6.8; Alphelys). Cores of normal oesophageal mucosa, lymph node, kidney, liver, spleen and prostate were incorporated in the TMA as internal controls. In a previous study, we have validated this TMA approach using established molecular markers with various expression patterns.24


A 4 μm section of the TMA was deparaffinised in xylene for 10 min, followed by dehydration in serial ethanol dilutions. Antigen retrieval was carried out by boiling the slides in sodium citrate (pH 6.0) for 20 min. After a cooling-off period in the citrate buffer solution for 30 min, endogenous peroxidase activity was blocked for 15 min. To avoid aspecific staining, endogenous avidin and biotin were blocked with Avidin Biotin blocking solution (Dako, Glostrup, Denmark; catalogue no. X0590) for 10 min. Subsequently, incubation with normal swine serum (Dako; catalogue no. X0901; dilution 1:5) was carried out for 10 min. Then, slides were incubated with the primary antibody against p-mTOR (Ser2448; Cell Signaling Technology, Danvers, Massachusetts, USA; catalogue no. 2976; dilution 1:50) in normal swine serum overnight in 4°C. The next day, incubation with the secondary antibody (swine-anti-rabbit; Dako; catalogue no. E0353; dilution 1:300 in 10% human serum) was applied for 30 min, followed by the Streptavidin–Biotin Complex (Dako; catalogue no. K0377) for 60 min. The peroxidase reactivity was developed by 3,3′-diaminobenzidine (10 min) and slides were counterstained with Mayer’s haematoxylin. Between each step, slides were washed in phosphate-buffered saline (pH 7.4). Breast carcinoma known for its p-mTOR positivity was used as positive control tissue (fig 1).11 13 25 A negative control was obtained by omitting the primary antibody.

Figure 1 Positive activated mammalian target of rapamycin (p-mTOR) staining in control breast cancer tissue (A, original magnification ×100; B, original magnification ×200; C, original magnification ×400). A strong, brown, cytoplasmic staining is seen.

Since mTOR is involved in cell proliferation, a second TMA-slide was stained for the proliferation marker Ki-67 (MIB-1) to assess whether a correlation existed. After deparaffinisation, rehydration, endogenous peroxide blocking and boiling in citrate buffer, as described above, slides were incubated with the primary antibody against MIB-1 (Dako; catalogue no. M7240; dilution 1:100) on an autostaining machine for 60 min. Subsequently, slides were incubated with the secondary antibody followed by the Streptavidin–Biotin Complex. Then, the peroxidase reactivity was developed and slides were counterstained with haematoxylin. Tonsil was used as positive control tissue.


Immunohistochemical staining of p-mTOR was scored conjointly by two observers (FJWTK and JB) using a scoring system that incorporates staining intensity and percentage of positive tumour cells. For each core, the intensity of p-mTOR staining (0, absent; 1, weak; or 2, strong) was multiplied by the percentage of positive-staining tumour cells, resulting in a score ranging from 0 to 200.26 Cases were marked as follows: −, cases in which all cores had a score of 0; +, when the highest score of the three cores was 1–19; ++, when the highest score was 20 or higher (fig 2).

Figure 2 No activated mammalian target of rapamycin (p-mTOR) staining was detected in normal oesophageal squamous epithelium (A, original magnification ×100). A strong cytoplasmic staining (score ++) was detected in oesophageal squamous carcinoma cells (B, original magnification ×100; C, original magnification ×200; D, original magnification ×400).

In each core, the percentage of tumour cells expressing Ki-67 was assessed. For determining the Ki-67 score of a tumour, the mean Ki-67 score of the corresponding tumour cores was calculated. Then, tumours were divided into three groups according to the Ki-67 score: negative, <10% of tumour nuclei stained; weakly positive, 10–50%; strongly positive, >50% (fig 3).27

Figure 3 Ki-67 staining of an oesophageal squamous cell carcinoma core showing a strong, nuclear staining in more than 50% of tumour cells (A, original magnification ×100; B, original magnification ×400).

TMA cores were considered lost if less than 10% of cells contained tumour (“sampling error”) or when less than 10% of tissue was present (“absent core”). Patient cases were excluded if two out of three cores were lost.

Statistical analysis

Statistical analysis was done using SPSS for Windows (V.12.0). Percentages were rounded to the nearest whole integer. Tumours with p-mTOR expression (score + and ++) and those without p-mTOR expression (score −) were compared with the Pearson’s χ2 test or Fisher’s exact test. The Spearman’s rank correlation coefficient (ρ) was computed to determine the correlation between the expression of p-mTOR and the expression of Ki-67. Two-tailed p values <0.050 were considered statistically significant.


Three (3%) of 108 cases were excluded, because only a single tumour biopsy core was present on the TMA slide stained for p-mTOR. The study population therefore consisted of 105 patients (56 men (53%) and 49 women) with a mean age of 62 years (range 36–79 years). The average tumour size was 4.7 cm (range 0.8–11.0 cm). Tumours were mainly located in the middle third (52%) or lower third (39%) of the oesophagus. The grade of differentiation was predominantly moderate (62%) or poor (32%). A mean of 15 lymph nodes were dissected during oesophagectomy. Sixty-two (59%) patients had lymph node metastases. In 55% of these lymph-node-positive patients, extracapsular lymph node involvement was detected. An overview of clinicopathological features of the study population is given in table 1.

Table 1 Activated mammalian target of rapamycin (p-mTOR) expression and clinicopathological characteristics of 105 patients with oesophageal squamous cell carcinoma

In the control tissue, a brown perinuclear staining of p-mTOR was seen (fig 1). As depicted in fig 2A, normal oesophageal epithelium was negative for p-mTOR. In OSCC, p-mTOR staining was noticed in the cytoplasm of tumour cells (fig 2B–D). Tumour cells expressing p-mTOR were noticed in 26/105 (25%) OSCCs. Eighteen (69%) of the p-mTOR-expressing tumours had a score between 1 and 19 (+). A score of 20 or higher (++) was given to eight (8%) OSCCs (table 1).

As shown in table 1, p-mTOR expression was associated with a lesser degree of differentiation only (p = 0.024). No statistical significant difference was detected between p-mTOR-positive and p-mTOR-negative tumours with regard to tumour location, TNM stage or the presence of lymph node metastasis. In addition, no correlation was detected between the expression of p-mTOR and Ki-67 (Spearman’s ρ 0.058, p = 0.56).


mTOR plays an important role in protein translation, cell growth and cell proliferation by integrating both environmental and internal factors, including nutrients, growth factors and energy levels.28 Subsequent to preclinical studies revealing the expression of mTOR protein in several cancer tissues, and showing mTOR inhibitors to inhibit cell proliferation in cancer cell lines, rapamycin and its analogues are increasingly being tested in oncological clinical trials.15 29 In a phase II trial of patients with locally advanced or metastatic breast carcinomas, weekly intravenous infusion of the rapalog temsirolimus has resulted in a response rate of 9.2%.29 In addition, a recent randomised controlled trial showed that temsirolimus significantly improved overall survival (p = 0.008) and progression-free survival (p<0.001) in advanced renal cell carcinoma patients.15

As far as we know, no studies have yet investigated p-mTOR expression in OSCC, or the application of mTOR inhibitors in patients with OSCC. However, Hou et al have shown expression of mTOR in two OSCC cell lines, with a higher expression in the poorly differentiated cell line.18 This is in line with the statistical significant association that was found in the current study between the expression of p-mTOR in OSCC and a poorer degree of tumour differentiation. Moreover, they detected a significant decrease in mTOR mRNA levels in OSCC cells treated with rapamycin.18 The goal of our study was to investigate the frequency of activated mTOR in human OSCC tissue samples, thereby assessing the amount of patients that would possibly benefit from mTOR inhibiting therapy.

To determine the expression of activated mTOR in a large study population under identical laboratory and evaluation conditions, we used a TMA slide containing triplicate core biopsies of 108 OSCCs. The expression of mTOR has previously been determined by means of TMA technology in malignancies such as prostate, breast and renal cell cancer.11 12 3032 Since a potential disadvantage of TMA technology might be that the small (diameter 0.6 mm) tissue cores are not representative for the full section, we have recently validated our TMA using well-known molecular markers.24 Although our TMA has proven valid, one should keep in mind when interpreting the results of the current study that false-negative observations may have occurred. Yet, this would only lead to a higher number of OSCCs that are p-mTOR positive, and which therefore might benefit from rapamycin therapy. Moreover, when performing immunohistochemical analysis on full sections instead of a TMA slide, it might also be questioned whether a single slide is representative of the entire tumour, and false-negative results may also occur. As with all immunohistochemical studies, it would be worthwhile to perform this study in a different OSCC population to confirm the current data.

The reported percentages of malignancies expressing activated mTOR vary widely, from 15% in hepatocellular carcinomas to around 60% in gastric adenocarcinoma and biliary tract carcinoma.26 33 34 Although this could be explained by differences in tumour biology, it should be taken into account that the percentage of carcinomas expressing p-mTOR also depends to a great extent on the scoring system used. In several studies that have assessed the expression of mTOR in cancer tissues by immunohistochemistry, tumours were marked as positive in cases in which any positivity was noticed.35 36 According to this scoring system, 25% of our tumours would be positive for p-mTOR. However, in a recently published scoring system of p-mTOR applied in biliary tract adenocarcinomas, tumours were considered positive when the multiplication of percentage of staining tumour cells and staining intensity (on a scale of 0–2) was 20 or higher.26 When we apply this scoring system, only 8% of our OSCCs would be classed as expressing p-mTOR. From a clinical point of view, it remains to be elucidated which degree of p-mTOR expression would benefit from mTOR inhibitors. One could imagine that in tumours in which only 25% of tumour cells express activated mTOR, mTOR inhibitors would destroy only 25% of tumour cells and leave the rest unaffected, and leading to outgrowth of resistant clones. Nevertheless, in these tumours, mTOR inhibitors could still be effective when given in combination with chemotherapy.

Since p-mTOR is involved in the regulation of cellular proliferation,79 37 we assessed the correlation between the expression of p-mTOR and that of the proliferation marker Ki-67. However, comparable to the results reported in primary liver neoplasms,33 no correlation was found. This could be explained by the fact that the cell cycle is controlled by numerous other molecular pathways (eg, the retinoblastoma protein, E2F-family proteins, cyclins and p53 protein) that are also commonly mutated in malignancies.3841

Although some studies have shown p-mTOR to be an adverse prognostic marker in breast carcinoma and biliary tract cancer,13 26 we did not assess the prognostic significance of p-mTOR in OSCC since the current study population consisted mainly of patients with advanced disease. As the presence of (distant) lymph node metastases would probably affect disease-free survival and overall-survival more than would the presence of p-mTOR expression, it should be recommended to study the prognostic significance of p-mTOR in a larger population of early-staged OSCCs (ie, stage I–IIa).

In summary, activated mTOR was detected in 25% of patients with OSCC, and predominantly in poorly differentiated tumours. Since patients with such tumours may benefit from mTOR inhibiting therapy, further clinical studies are warranted.

Take-home messages

  • Activated mammalian target of rapamycin (mTOR) expression was detected in one-quarter of oesophageal squamous cell carcinomas and was significantly associated with a poorer degree of differentiation.

  • No correlation was detected between activated mTOR and the proliferation marker Ki-67.

  • On the basis of these results, a phase II clinical trial of neoadjuvant mTOR-inhibiting therapy in patients with oesophageal squamous cell carcinoma may be considered.



  • Competing interests: None.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.