Lack of Correlation of Functional Scintigraphy with Technetium- Methoxyisobutylisonitrile with Histological Necrosis following Induction Chemotherapy or Measures of P-Glycoprotein Expression in High-Grade Osteosarcoma

In osteosarcoma, some studies have suggested Pglycoprotein expression is a prognostic factor. The clearance of technetium hexakis-2-methoxyisobutylisonitrile (Tc-MIBI) has been used in some tumor systems as an in vivo measure of P-glycoprotein-mediated efflux. In this study we explored the correlation between TcMIBI clearance and histological necrosis following induction chemotherapy and P-glycoprotein expression in osteosarcoma. The primary tumors of 20 patients with high-grade osteosarcoma were imaged at diagnosis with Tc-MIBI, and the uptake ratios and biological halflives were calculated. P-Glycoprotein expression in the tumor tissue was determined immunohistochemically and by measuring mRNA expression of the multidrug resistance-1 gene. The histological necrosis following induction chemotherapy was assessed by the Huvos grading system. The biological half-life of Tc-MIBI ranged from 1.4 to 52.5 h. Seven of the 20 tumor samples had a favorable extent of necrosis following induction chemotherapy. The Tc-MIBI half-life and uptake ratio showed no correlation with histological necrosis following induction chemotherapy. The Tc-MIBI half-life and uptake ratio did not correlate with either measure of P-glycoprotein expression. The results of this pilot study indicate that Tc-MIBI imaging is not an effective predictor of histological necrosis following induction chemotherapy in high-grade osteosarcoma. Tc-MIBI imaging did not correlate with measures of P-glycoprotein expression in the tumor tissue. INTRODUCTION OS, the most common primary malignant bone tumor, occurs mainly in children and adolescents. In OS, prognostic factors at diagnosis other than clinical stage have not been clearly identified (1). The extent of necrosis following induction chemotherapy (Huvos grade) is a strong predictor of patient outcome, but attempts to improve outcome by intensifying therapy at that time have largely been unsuccessful (2). Because 60–70% of patients with localized OS are cured with standard approaches, a need exists to be able to identify a “poor risk” subgroup of patients at diagnosis for stratification of therapy (1, 3). Some retrospective studies have suggested that pgp expression at diagnosis may be a prognostic factor for OS (4, 5), whereas other studies have failed to confirm this relationship (6–8). Several potential explanations for these conflicting results exist, but one concern is that immunohistochemistry, which has been relied on in prior studies, may not be an appropriate method to evaluate pgp expression (4–6, 8). It has been reported that immunohistochemistry for pgp does not always correlate with multidrug resistance-1 mRNA levels in OS (9). In other tumor systems, the simultaneous use of multiple methodologies to detect pgp expression and function has been advocated (10). An imaging procedure that can detect functional pgp expression and thus potentially serve as a prognostic factor provides a number of theoretical advantages, particularly in OS. Frequently the only tumor tissue available from the diagnostic biopsy is paraffin-embedded material, which precludes analysis by any method other than immunohistochemistry. The biopsy represents a small portion of a tumor, which frequently is large and heterogeneous. A series of lipophilic cationic radiopharmaceuticals, including Tc-MIBI (sestamibi), have been identified and validated as transport substrates for pgp, thus enabling functional imaging (11–13). In many tumor systems, including, among others, breast cancer, lung cancer, hepatocellular carcinoma, malignant lymphomas, soft tissue sarcoma, and bone Received 12/4/00; revised 5/25/01; accepted 6/21/01. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1 Supported by Grant CA-83132 from the National Cancer Institute. R. G. is the recipient of an ASCO Career Development Award. 2 To whom requests for reprints should addressed, at Department of Pediatrics, Memorial Sloan-Kettering Cancer Center, 1275 York Avenue, Mailbox #376, New York, NY 10021. Phone: (212) 639-8392; Fax: (212) 639-2767; E-mail: [email protected]. 3 The abbreviations used are: OS, osteosarcoma; pgp, P-glycoprotein; Tc-MIBI, technetium hexakis-2-methoxyisobutylisonitrile; RTPCR, reverse transcription-PCR. 3065 Vol. 7, 3065–3070, October 2001 Clinical Cancer Research Research. on April 20, 2017. © 2001 American Association for Cancer clincancerres.aacrjournals.org Downloaded from sarcomas, significant correlations have been reported between Tc-MIBI scintigraphy and pgp immunohistochemistry, in vitro cytotoxicity assays, chemotherapy response, and/or patient outcome (14–20). In some of the same histologies as well as others, including parathyroid adenomas or head and neck cancer, some published reports have failed to identify a significant correlation between Tc-MIBI scintigraphy and measures of pgp expression or patient outcome (21–24). In this pilot study, the uptake and clearance of Tc-MIBI in patients with newly diagnosed high-grade extremity OS were measured. The scintigraphy results were related to the extent of necrosis (Huvos grade) in the tumor following induction chemotherapy as well as measures of pgp expression, determined both immunohistochemically and by quantitative RT-PCR for multidrug resistance-1 expression in the tumor tissue. This may indicate the potential for Tc-MIBI scintigraphy to predict response to chemotherapy, patient outcome, and pgp expression in high-grade OS. PATIENTS AND METHODS Twenty patients with newly diagnosed high-grade extremity OS participated in this study between May 1998 and May 2000. All newly diagnosed extremity OS patients were offered participation in the study. Tumor tissue was obtained from the patients after the patient or guardian provided informed written consent in accordance with a research protocol approved by the Memorial Hospital Institutional Review Board. All patients received standard OS induction chemotherapy comprising four courses of high-dose methotrexate with leucovorin rescue and two cycles of cisplatin and doxorubicin, as has been described previously in the Memorial Sloan-Kettering Cancer Center T10 and T12 protocols (2, 25). Histological necrosis following induction chemotherapy was assessed at the time of definitive surgery based on the Huvos grading system as described previously (25). Thallium scans were performed and analyzed as a routine clinical study as has been described previously (26). Tc-MIBI Scintigraphic Imaging. The interval from initial biopsy to Tc-MIBI scan averaged 8 days (range, 1–15 days). Despite the patient’s age, Clark’s rule was applied to the radioactivity dosage, which was based on proportional body weights as related to the standard weight mean of 150 pounds, to give 740 MBq (20 mCi) Tc-MIBI (Cardiolite; DuPont Pharma, Wilmington, DE). Serial planar images were acquired for 5 min each at four time points at 20 min and 1, 2, and 4 h after i.v. administration of radiotracer. These spot images were centered on the lesion and included the opposite limb. A dualhead ADAC Genesys camera with a LEHR collimator, equipped with SUN Spark series computer and Pegasys Image Processing System (ADAC laboratories, Milpitas, CA) was used for image acquisition. Whole-body images were acquired after each planar spot image. All calculations were performed using previously described methods (20). To calculate the biological half-life, we drew a manual region of interest over the entire area of tumor uptake. The count density in this region was determined for each of the time points. These data were adjusted for decay correction and were used to produce a fitted monoexponential curve to derive the tumor effective biological half-life. The uptake ratio was determined by relating the count density in the same region of interest on the 20 min image to the identical area on the contralateral leg, which did not include tumor. For the uptake ratio, the mean of the anterior and posterior images was calculated and presented. Immunohistochemistry. All paraffin-embedded material was retrieved from the Department of Pathology at Memorial Sloan-Kettering Cancer Center. Immunohistochemistry was performed as has been described previously (4, 8, 27). Briefly, sections from decalcified bone tumor specimens were cut at 4–5 m, deparaffinized, and rehydrated. Pretreatment comprised digestion with 0.05% trypsin followed by microwave treatment for 10 min. JSB-1 (Signet Laboratories, Dedham, MA) at a dilution of 1:20 and C494 at a dilution of 1:20 (Signet Laboratories, Dedham, MA) were used for pgp. Incubations were at 4°C for 16 h. Color reactions were obtained using a standard avidin and peroxidase-conjugated streptavidin (Dako Corporation, Carpinteria, CA) technique. Slides were counterstained with Harris modified hematoxylin (Fisher, Pittsburgh, PA). Positive controls, constituting normal adrenal gland, as well as negative controls, in which the primary antibody was omitted, were included with each run. A pathologist blinded to patient identity scored each case. The cases were scored as 0 (no staining), 1 (1–25% of the cells staining positive), 2 (26– 50% positive), 3 (51–75% positive), and 4 (76–100% positive). The staining was considered positive only if it localized predominantly to the membrane. Quantitative RT-PCR. RNA was prepared from frozen tumor tissue by use of RNAzol as per manufacturer’s instructions (BiotecX Laboratory). Quantitative RT-PCR was performed using methodologies described previously (28–31). Total RNA (5–10 g) was treated with RNase-free DNase (Boehringer-Mannheim, Indianapolis, IN), reextracted with phenol-chloroform, and reverse transcribed using random hexamers with a cDNA cycle kit, according to the manufacturer’s instructions (Invitrogen, Carlsbad, CA). Relative gene expression was calculated by determining the ratio between the amount of the radiolabeled PCR product within the linear range of the multidrug resistance-1 gene and the -actin gene. The PCR conditions and gel electrophoresis have all been described previously (28–31). Radioactivity was quantitated on a Fuji BAS 2500 phosphorimager (Fuji Photo Film, Tokyo, Japan). The primers used were 5 -CCCATCATTGCAATAGCAGG-3 and 5 -GTTCAAACTTCTGCTCCTGA-3 for MDR-1, and BA67 and BA68 for -actin; all primers have been described previously (27–30). The CCRF-CEM cell line served as a negative control, and CEM-VBL, which is known to overexpress multidrug resistance-1 (obtained from Dr. William Beck, University of Illinois at Chicago, Chicago, IL), served as a positive control. Any detectable PCR product of the appropriate size was considered as positive for MDR-1 gene expression. Statistical Methods. The results of the Tc-MIBI scans were related to the Huvos grade and measures of pgp expression by a t test. Measures of pgp expression and the thallium scan results were related to Huvos grade and each other by a Fisher exact test.