Hyperthermic Isolated Limb Perfusion with Tumor Necrosis Factor-a and Melphalan in Patients with Locally Advanced Soft Tissue Sarcomas: Treatment Response and Clinical Outcome Related to Changes in Proliferation and Apoptosis

Hyperthermic isolated limb perfusion with tumor necrosis factor-a and melphalan (HILP-TM) with or without IFN-g is a promising local treatment in patients with locally advanced extremity soft tissue sarcomas (STSs), with response rates of up to 84%. The mechanisms of the treatment response are poorly understood. Here, we determined the HILP-TM-induced changes in mitotic activity, proliferation, and apoptosis in 37 STSs; the additional effect of IFN-g; and the association of HILP-TM with treatment response and clinical outcome. On archival material, obtained before and 6–8 weeks after HILP-TM with (n 5 15) or without (n 5 22) IFN-g, the number of mitoses was counted, and the proliferation fraction was determined by immunohistological staining for the proliferation associated Ki-67 antigen (MIB1). Apoptosis was visualized by enzymatic detection of DNA fragmentation (terminal deoxynucleotidyl transferase-mediated nick end labeling method). Clinical and histological response, follow-up status, and survival were recorded. The number of mitoses dropped 57% and proliferation rate decreased with 40% after HILP-TM, whereas the amount of apoptosis after HILP-TM more than doubled as before HILP-TM. The addition of IFN-g to HILP-TM did not influence the changes in tumor parameters and did not affect treatment response. A better clinical response to HILP-TM was correlated with high mitotic activity and low amount of apoptosis in tumor samples before HILP-TM. Patients with highly proliferative STS before and after HILP-TM had a relatively poor prognosis. Furthermore, patients who developed distant metastases after HILP-TM had a relatively high number of dividing cells in the tumor remnants after treatment. INTRODUCTION STSs constitute a heterogeneous group of malignant tumors that arise from tissue of mesenchymal origin and account for ;1% of all malignancies. STSs predominantly affect the limbs, and surgical resection in combination with high-dose adjuvant external beam radiotherapy, after marginal resection, is the treatment of choice (1–3). Because of the size and extent of the tumor or the proximity of vital structures, limb-saving surgery is not always possible, and mutilating surgical procedures or amputation are occasionally necessary (2, 4–8). Currently, one of the most promising limb-saving procedures is the combination of HILP with delayed surgical tumor excision (9, 10). The major advantages of HILP are that regional cytostatic concentrations can be 15–20 times as high as after systemic administration and that hyperthermia may enhance the cytotoxic effect (11, 12). In 1992, Lienard et al. (13) used HILP to administer high-dose recombinant TNF-a in combination with melphalan, an alkylating chemotherapeutic agent, resulting in encouraging outcomes. In HILP-TM with or without IFN-g, response rates of up to 84% have been reported in the treatment of STS (9, 10, 13–15). Because the mechanism of the antitumor effects of HILP-TM is still poorly understood and studies suggest that chemotherapeutic drugs in general inhibit proliferation by generating cell cycle arrest and subsequent apoptosis, these tumor parameters were investigated in HILP-TM (16–23). The objective of this study was to investigate whether the treatment response and clinical outcome of patients treated with HILP-TM is associated with tumor grade, mitotic activity, proliferation, and apoptosis in the tumor specimens of STS obtained before and after HILP-TM. Furthermore, the influence of HILP-TM on these tumor markers and additional effect of IFN-g was evaluated. Received 6/26/98; revised 3/9/99; accepted 3/25/99. 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 the Dutch Cancer Society (Koningin Wilhelmina Fonds) Grant 95-1085. 2 To whom requests for reprints should be addressed, at Department of Pathology, University Hospital Groningen, P. O. Box 30.001, 9700 RB Groningen, the Netherlands. Phone: (31) 50 3619530; Fax: (31) 50 3632510; E-mail: [email protected]. 3 The abbreviations used are: STS, soft tissue sarcoma; HILP, hyperthermic isolated limb perfusion; TNF, tumor necrosis factor; HILP-TM, HILP with TNF-a and melphalan; DIG, digoxigenin; LI, labeling index; CR, complete response; PR, partial response; NC, no change; OS, overall survival; DFS, disease-free survival. 1650 Vol. 5, 1650–1657, July 1999 Clinical Cancer Research Research. on October 24, 2017. © 1999 American Association for Cancer clincancerres.aacrjournals.org Downloaded from PATIENTS AND METHODS Patients. The inclusion criteria for this study were: (a) histological diagnosis of a malignant mesenchymal tumor, located in the soft tissues; (b) primary nonresectable tumor of lower or upper limb, which was treated with HILP-TM with or without IFN-g; and (c) the availability of paraffin-embedded tumor material obtained within 3 months before HILP-TM, with a maximum period between pretreatment and posttreatment tumor samples of 5 months. Between 1991 and 1997, 37 patients with primary nonresectable STS of the lower or upper limb underwent HILP-TM. HILP-TM was combined with IFN-g in 15 of the 37 analyzed patients. The mean age (6 SD) of the 18 men and 19 women, at the time of diagnosis, was 46 6 16.7 years (range, 18–80 years; median, 48 years). Four tumors were located in the upper limb (11%), and 33 were located in the lower limb (89%). At the time of HILP-TM, six patients (16%) had distant metastases, and three patients (8%) had regional lymph node metastases. The mean time (6 SD) between incisional biopsy and perfusion was 22 6 17 days (range, 0–83 days; median, 17 days). In one patient, an incisional biopsy of multiple superficial sarcomas was performed at the day of HILP-TM to obtain fresh tumor material. In 35 of the 37 patients, HILP-TM was followed by surgical resection 6–8 weeks later, according to the study protocol. In two patients (5%), histological examination after HILP-TM was not possible: one patient had progressive lung metastases requiring chemotherapy, and in the other patient, the limb had to be amputated because of vascular occlusion 2 days after HILP-TM. Local progressive disease or treatment-related morbidity were reasons for performing the delayed local surgical resection earlier or later than planned in the protocol. Specimens for the assessment of histological reaction were obtained after a mean period of 61 6 16 days following HILP-TM (range, 12–103 days; median, 60 days). In three patients, in which HILP-TM was performed twice, histological tumor reaction was assessed after the first HILP-TM in two patients and after the second HILP-TM in one patient, of whom no tumor material was available after the first HILP-TM. The mean time between the pretreatment sample and the posttreatment sample was 83 6 22 days (range, 33–138 days; median, 80 days). No additional treatment was given between HILP-TM and the delayed resection. Patient characteristics, histology, and clinical data are presented in Table 1. Histology. In all cases, the histological diagnosis was made on H&E-stained paraffin sections of incisional biopsies with or without additional immunohistological stains. All cases were classified according to Enzinger and Weiss (1), who revealed 16 different histological types (Table 1). The STSs were graded according to the grading system of Coindre et al. (24), in which points are assigned to differentiation level, mitotic index, and necrosis. This resulted in 7 grade I (19%), 19 grade II (51%), and 11 grade III (30%) STSs (Table 1). In the resection specimens, the amount of necrosis was estimated on gross examination. For histology, at least one section per centimeter of the largest tumor diameter was taken. Care was taken to histologically document the presence of necrosis, viable tumor, or fibrosis. Tumor Parameters. The following tumor parameters were studied on adjacent slides of the macroscopically and microscopically most viable parts of the tumor before and after HILP-TM: mitotic index, proliferation, and apoptosis. The number of mitoses per 2 mm (mitotic index) in adjacent fields was counted on H&E-stained paraffin sections. For proliferation, the monoclonal antibody MIB1 (Immunotech S.A., Marseille, France) was used. This antibody recognizes an epitope of the Ki-67 antigen, which is present in the nucleus of cells in all phases of the cell cycle except G0 and early G1 (25). Immunohistochemistry was performed on paraffin sections (4 mm) according to a method modified from Shi et al. (26, 27). Briefly, after heating on a hot plate, slides were dewaxed in xylene and rehydrated in serial ethanol washes (100, 96, and 70%). After two incubations in an autoclave for 10 min at 110°C in a 20 mM blocking reagents (Boehringer Mannheim, Mannheim, Germany; pH 6.0), slides were incubated with a 1:400 dilution of the antibody in BSA buffer (pH 7.4). The primary antibody was detected with a biotinylated secondary antibody (multilink), followed by a streptavidin-alkaline phosphatase conjugate (Ready-to-Use Link and Label; Biogenex, San Ramon, CA). Final color was developed by the bromochloroindolyl-phosphate 4-nitroblue tetrazolium chloride method (Boehringer Mannheim). Apoptosis was studied in 4-mm sections of formaldehydefixed and paraffin-embedded tissue using the terminal deoxynucleotidyl transferase-mediated dUTP nick end labeling method. After deparaffination, sections were subjected to RNase (1 mg/ml in 50 mM Tris-HCl and 150 mM NaCl, pH 7.5; Aldrich, Milwaukee, WI) treatment for 30 min at room temperature to prevent nonspecific binding to DIG-11-dUTP. Slides were subsequently incubated with 15 ml of a 100-ml solution containing terminal deoxynucleotidyl transferase buffer (Pharmacia, Uppsala, Sweden) with 5 mM cobalt chloride and 1% BSA (Aldrich) together with 5 nM DIG-11-dUTP (Boehringer Mannheim) for 60 min at 37°C. After this reaction, slides were rinsed twice in a 0.13 SSC solution for 15 min at room temperature. Nonspecific background staining was prevented by preincubating slides with 5% blocking reagents (Boehringer Mannheim). Sections were incubated overnight at 6°C with alkaline phosphataselabeled sheep anti-DIG Fab fragments diluted 1:300 in 2% blocking reagents (Boehringer Mannheim). After extensive washing, alkaline-phosphatase was visualized with nitroblue tetrazolium 5-bromo-4-chromo-3-indolyl-phosphatase (Boehringer Mannheim) for 50 min at room temperature. Sections were counterstained with hematoxylin and coverslipped with a mounting medium soluble in xylene. Quantification of Ki-67 and Apoptosis. All sections stained for proliferation and apoptosis were quantified without knowing whether this section was before or after HILP-TM. For measuring the Ki-67 LI and the apoptosis LI, we used ocular micrometry on a Leitz microscope by using an eyepiece grid at 3400 magnification. Fifteen adjacent fields in histologically viable areas were quantified. The positive and negative nuclei were counted. Endothelial cells, inflammatory cells, and necrosis were excluded. The number of positive nuclei was then divided by the total number of nuclei in each of the 15 randomly selected fields to calculate an index per field. The Ki-67 LI (representing proliferative activity) and the apoptosis LI (representing apoptotic activity) were defined as the mean of the indices of the 15 randomly selected fields. 1651 Clinical Cancer Research Research. on October 24, 2017. © 1999 American Association for Cancer clincancerres.aacrjournals.org Downloaded from T ab le 1 Pa tie nt ch ar ac te ri st ic s, hi st ol og y, H IL PT M m od al ity , tr ea tm en t (c lin ic al , hi st ol og ic al , an d ov er al l) re sp on se , an d cl in ic al ou tc om e a Pa tie nt no . A ge (y r) Se x L im b D ia gn os is T um or gr ad e St ag e H IL PT M