Osteosarcoma is a disease, which predominantly targets the adolescent age group and the most common primary malignant bone tumor in general. In spite of advances in imaging techniques, surgery and especially combination chemotherapy, the overall survival rate over 5 years is still approximately 65%,1 and 25–50% of patients with initial diffusion succumb to lung metastasis. The rate of mortality for osteosarcoma is basically the same as that some decades ago.2, 3 Prognostication in individual cases remains a problem. It would be helpful if objective instruments were available for predicting the chance of survival or chemotherapy response, especially early in treatment, preferably even before surgery.
The reversion-inducing cysteine-rich protein with Kazal motifs (RECK) gene was originally isolated as a novel transformation suppressor gene by cDNA expression cloning.4 The RECK gene is extensively expressed in normal tissues and various tumor-derived cell lines, but down-regulated in tumor tissues.5, 6 It suggest a potential importance of RECK in the diagnosis and therapy of malignant tumors. The RECK gene encodes a glycosylphosphatidylinositol (GPI)-anchored glycoprotein (molecular weight, 110,000) as a negative regulator of matrix metalloproteinases (MMPs), including at least MMP-2, MMP-9, and MT1-MMP. Restoration of RECK in tumor-derived cell lines strongly suppresses their abilities to invade, metastasize, and angiogenesis.5 Positive correlation between residual RECK expression in tumor tissues and survival of the patients have been documented in a variety of cancers.7–10In this study, we aimed to determine whether the expression of RECK is a predictor of the clinical behavior of osteosarcoma and predict a high-risk group of patients who would benefit from new treatment strategies.
MATERIALS AND METHODS
Patients and Tumor Samples
We retrospectively identified 49 patients with primary osteosarcoma occurring between 2001 and 2004 who met the following criteria: (1) no metastasis at presentation; (2) no history of previous treatment; (3) adequate preoperative and postoperative chemotherapy; (4) tumor excision with a wide or radical margin; and (5) presence of incisional biopsy specimens available for immunohistochemical staining (IHS). Among the 49 selected patients, there were 35 males (71.4%) and 14 females (28.6%) with an average age of 20.2 years (range, 4–50 years). Following initial biopsy, all patients underwent multiagent, neoadjuvant chemotherapy, and definitive surgery. The locations of the tumor were the femur (25 cases), tibia (21 cases), proximal humerus (2 cases), and pelvic bone (1 case). The histologic subtypes were osteoblastic in 40 patients, and other in 9 patients. The tumors were classified as intracompartmental (Stage IIA) in 13 patients, extracompartmental (Stage IIB) in 34 patients and Stage I in 2 patients. Chemotherapy was given before and after surgery. The chemotherapy protocol was the modified T12 protocol.11 The surgical margins of the tumor specimens were histologically defined according to the system of Enneking.12 The margins were radical in eight patients and wide in 41 patients. During option of surgical procedures, the location and extent of the tumor and the life expectancy of the patient took into account. Limb-salvage procedures were performed in 42 patients (85.7%), whereas seven patients (14.3%) underwent amputation. All tumors specimens were immediately fixed in 10% formalin and then embedded in paraffin. The specimens we analyzed were from incisions of biopsy before chemotherapy. The paraffin blocks were evaluated again to select a representative area. The extent of tumor necrosis was evaluated according to a previously described semiquantitative method. The response of chemotherapy was considered good if the extent of tumor necrosis was >90% and poor if it was <90%.13
Follow-up duration was defined as the date of initial presentation to the date of death or last visit. The minimum follow-up for survivors was 12 months (mean, 64.1 months; range, 12–104 months). During the follow-up, 21 patients (42.8%) were continuously disease-free, 19 patients (38.8%) died of disease, and 2 patients died during chemotherapy. Distant metastasis developed in 13 patients at a mean of 14.7 months (range, 3–41 months) after the original diagnosis. Three patients had bone metastasis and 10 patients had lung metastasis. The onset of local recurrence was as early as 1 month (range, 1–13 months).
RECK expression was determined by using Streptavidin-perosidase immunohistochemistrical method. Serial 4-µm sections were cut from formalin-fixed, paraffin-embedded blocks and placed on silane-coated slides. After deparaffinization, sections were incubated in 3% H2O2 for 20 min to inactivate endogenous peroxidase. Deparrafinized and rehydrated specimens were heated in 10 mM citrate buffer (pH 6.0) for 10 min in an autoclave at 105°C. After being cooled to room temperature for 30 min, specimens were incubated with normal horse serum for 20 min at room temperature followed by incubation with polyclonal antibodies against RECK (H-300, rabbit polyclonal antibody; 1:50, Santa Cruz Biotechnology, Inc., Santa Cruz, CA sc-28918) for 16 h at 4°C. The sections were washed three times for 5 min in PBS and incubated for 1 h with biotinylated anti-rabbit IgG secondary antibodies. After rinsing, immune complexes were visualized by the standard avidin–biotin–peroxidase complex (ABC) method. After counterstaining with Meyer's hematoxylin, the slides were dehydrated and mounted.
The staining was evaluated independently by two pathologists, and the data from the two investigators were averaged. We interpreted immunoreactivity in a semiquantitative manner by an intensity proportion scoring system similar to that previously described for RECK expression scoring in non-small-cell lung cancer and Ezrin expression in osteossarcomas.14, 15 The score was calculated by summing intensity score and proportion score providing a score between 0 and 6. The proportion score was as follows: 0 = no positivity; +1 ≤25% tumor cell positivity; +2 = 25–50% tumor cell positivity; and +3 = more than 50% tumor cell positivity. The intensity score was as follows: 0 = no staining; +1 = weak staining; +2 = intermediate staining; and +3 = strong staining. We further divided RECK-positive specimens into low and high expression according to the score; specimens with scores 1–4 were regarded as low expression, whereas specimens with scores 5 or 6 were regarded as high expression.
Data were compared by the χ2 test. Kaplan–Meier method was used to analyze the postoperative survival rate, and the differences in survival rates were assessed by the log-rank test. Any factor influencing prognosis in univariate analysis was analyzed in a multivariate analysis using Cox's proportional hazard regression model with a forward conditional stepwise procedure to determine whether the factor was acting independently. All calculations were performed with SPSS version 13.0 software, and p < 0.05 was considered significant.
RECK Expression and Clinical Characteristics
RECK immunoreactivity was found in the cytoplasm of tumor cells. On the basis of summing intensity score and proportion score of RECK in the primary tumors, 22 patients (44.9%) were classified as RECK-weak patients and 27 patients (55.1%) as RECK-strong patients. No correlation was revealed between the RECK status and age, sex, tumor location, or subtype (Table 1). RECK expression was inversely correlated with metastasis (r = −0.387, p = 0.006) and recurrence (r = −0.420, p = 0.003). Furthermore, RECK expression was correlated with histological response (r = −0.347, p = 0.015) and Ennecking stage (r = −0.323, p = 0.024; Table 1).1
Table 1. Characteristics of Patients with OS and RECK Expression
Prognostic Factors and Postoperative Survival
Univariate analyses of prognostic factors demonstrated that the RECK expression, metastasis, histological response, Ennecking stage, and recurrence were significant prognostic factors (Table 2). No correlations were found between prognosis and age, sex, tumor location, or subtype. The 5-year survival rate of RECK-strong patients was 81.5%, which was significantly higher than that of RECK-weak patients (36.4%; p = 0.003).2
Table 2. Univariate Analysis of Prognostic Factors in Osteosarcoma
5-Year Survival (%)
Age, years (≤18/>18)
Tumor location (femur/tibia/others)
Histological response (good/poor)
Ennecking stage (I/IIA/IIB)
Multivariate Analysis of Prognostic Factors
To assess whether the prognostic value of RECK was attributable to prognosis, COX regression analysis was performed with all variables. Reduced RECK expression was an independent and significant factor to predict a poor prognosis (p = 0.017; hazard ratio, 0.278; 95% confidence interval, 0.097–0.793; Table 3).
Table 3. Multivariate Survival Analysis
95% Confidence Interval
95% Confidence Interval
Osteosarcoma is a rare, highly malignant tumor that has a predilection for adolescents and continues to confer a generally poor prognosis in spite of newly devised chemotherapy regimes combined with wide-margin, limb-sparing surgery. Previous studies have shown that response to chemotherapy, tumor size, site and the presence of metastases as being key determinants of prognosis, but only apparent at a late stage in the course of disease. Therefore, elucidation of factors affecting behavior of osteosarcoma is indispensable to improve prognosis. Some molecular markers (in particular, P-glycoprotein,16 CXCR4,17 uPA/uPAR,18 and survivin19) have been uncovered as being useful both in predicting response to chemotherapy, overall prognosis, the likelihood of metastases at diagnosis, and at the same time providing targets for developing new therapeutic agents.
In this study, we focused on RECK expression in biopsy specimen and its prognostic significance in patients with osteosarcoma and demonstrated that reduced RECK expression was a significant factor to predict a poor prognosis. We also demonstrated that enhanced expression of RECK were correlated with reduced tumor recurrence and metastasis. Patients characteristics found in the present series such as gender (male, 71.4%), age (≤18, 53.1%), tumor location (femur, 51%; tibia, 42.9%), and subtype (osteoblastic, 81.6%) were similar to those in the previous reports in osteosarcoma.20, 15 In addition, uni- and multi-variate analyses in this study revealed clinicopathologic factors such as RECK, metastasis, histological response, Ennecking stage, and recurrence were useful prognostic determinants. This study also confirmed that the expression of RECK were correlated with metastasis and recurrence. Multivariate analysis revealed that positive staining of RECK persisted as an independent prognostic factor for favorable outcome. Kang et al.21 confirmed RECK over expression had been correlated with reduced cell invasion in vitro and also identified down regulation of RECK in a number of osteosarcoma cell lines and human osteosarcoma tissues.
The process of cancer metastasis consists of sequential steps, including invasion, detachment, intravasation, adhesion, and extravasation.22 Experimental studies have revealed that RECK can inhibit tumor progression (tumor invasion, metastasis, and prognosis) through controlling tumor angiogenesis, because RECK can regulate a variety of MMPs (MMP-2, MMP-9, and MT1-MMP).23–25 Therefore, a poor prognosis in a primary osteosarcoma with reduced RECK expression may be explained by the molecular mechanism of RECK demonstrated in these experimental studies. RECK has been already identified as an independent prognostic factor in several human malignancies. Masui et al.6 revealed an inverse correlation between RECK expression and MMP-2 activation in pancreas cancer, and this finding was consistent with experimental results. This study revealed 10 in 13 patients with metastasis showed RECK negative, while eight in nine patients with local recurrence showed RECK negative.
However, we have noted limitation in our study design. Our study just conducted an initial retrospective analysis of patients due to a relatively small sample size, and the relatively small sample size left the study with limited statistical power.
The expression of RECK in ostosarcoma biopsy specimens appears a promising prognostic maker in predicting the outcome of patients. Determination of pre-treatment RECK levels may help to improve the selection of the most appropriate therapy regime for an individual patient with osteosarcoma. Moreover, its clinical use as a prognostic indicator needs further evaluation.
We thank Honglin Yin and Zhenfeng Lu, Department of Pathology, Jinling Hospital, for their technical assistance.