976 Multidrug Resistance Phenotype in High Grade Soft Tissue Sarcoma Correlation of P-Glycoprotein Immunohistochemistry with Pathologic Response to Chemotherapy Rafael E. Jimenez, M.D.1 Mark M. Zalupski, M.D.2 John J. Frank, M.T.1 Wei Du, Ph.D.2 James R. Ryan, M.D.3 David R. Lucas, M.D.1 1 Department of Pathology, Harper Hospital, Wayne State University and the Karmanos Cancer Institute, Detroit, Michigan. 2 Department of Oncology, Harper Hospital, Wayne State University and the Karmanos Cancer Institute, Detroit, Michigan. 3 Department of Orthopedics, Hutzel Hospital, Wayne State University and the Karmanos Cancer Institute, Detroit, Michigan. BACKGROUND. P-glycoprotein–mediated drug efflux has been implicated as an important mechanism of multidrug resistance (MDR) in cancer. Its role in chemotherapy resistance in soft tissue sarcoma is unclear. METHODS. Tumor specimens prior to and following neoadjuvant chemotherapy from 29 cases of high grade soft tissue sarcoma were analyzed with 2 monoclonal antibodies (C494 and JSB-1) that recognize different epitopes of P-glycoprotein. Staining intensity was graded 0 5 negative, 1 5 equivocal, 2 5 moderate, 3 5 strong. Only cases with Grade 2 or 3 staining intensity with both antibodies were considered MDR positive. The resection specimens were evaluated for tumor necrosis postchemotherapy. Pathologic response was graded as good for ,15%, moderate for 15–50%, or poor for .50% posttreatment tumor viability. RESULTS. Of the 29 pretreatment specimens, 10 (34%) were MDR positive and 19 (66%) were MDR negative. Pathologic response to treatment was characterized as good in 6, moderate in 7, and poor in 16 patients. Of the MDR positive biopsies, 9 (90%) had poor response, compared with 7 (36%) in the MDR negative biopsy group (P 5 0.0078). None of the cases with MDR positive biopsies had a good response, compared with 6 cases in which biopsies were MDR negative (32%) (P 5 0.057). Only one MDR negative case became MDR positive posttreatment. CONCLUSIONS. Expression of MDR phenotype is found in approximately one-third of high grade soft tissue sarcomas. These preliminary data show a significant correlation between MDR phenotype and poor pathologic response to chemotherapy, and suggest that MDR induction by chemotherapy in soft tissue sarcoma is an uncommon event. Cancer 1999;86:976 – 81. © 1999 American Cancer Society. KEYWORDS: soft tissue sarcoma, chemotherapy, chemoresistance, MDR, P-glycoprotein, JSB-1, C494. P Presented at the 87th Annual Meeting of the United States and Canadian Academy of Pathology, Boston, Massachusetts, February 28 to March 6, 1998. Address for reprints: David R. Lucas, M.D., Department of Pathology, Harper Hospital, 3990 John R, Detroit, MI 48201. Received October 15, 1998; revision received February 10, 1999; accepted April 7, 1999. © 1999 American Cancer Society -glycoprotein (Pgp) expression has been implicated as a main cause for multidrug resistance (MDR) in several human neoplasms.1 This phenomenon is defined as an intrinsic or acquired resistance to multiple anticancer drugs that appear to be structurally and functionally unrelated. Pgp, the product of the mdr1 gene, localized in the long arm of chromosome 7, is a 170 kD cell membrane protein that actively pumps out drugs through a transmembrane channel. It is expressed in normal tissues, particularly in specialized epithelial cells with secretory functions, such as hepatocytes, colonic mucosa, biliary and pancreatic ductules, and brush border of the proximal tubules of the kidney, among others. The main characteristic of this protein is its broad substrate specificity, which encompasses a large list of anticancer drugs, including doxorubicin, an agent MDR in Soft Tissue Sarcoma/Jimenez et al. widely used in sarcoma treatment.2,3 Overexpression of Pgp has been shown to correlate with chemoresistance in a variety of cancers.4 –13 In addition, induction of Pgp expression by chemotherapeutic agents has also been reported,7,14,15 suggesting an explanation for failure of treatment in tumors initially MDR negative. Although its effectiveness is still somewhat controversial, neoadjuvant chemotherapy is widely used in the treatment of adult soft tissue sarcoma (STS).16,17 Multiple clinical trials have suggested a better outcome for treated patients,2,3 although the success obtained with osteosarcoma18,19 or pediatric soft tissue tumors13,14 has not been achieved. To what extent MDR expression determines the response to chemotherapy in various soft tissue neoplasms has been a subject of investigation in multiple recent studies.13,14,20 –26 The role that it plays specifically in adults STS is still not completely understood, and contradictory data have emerged recently. In this study we investigated expression of MDR by immunohistochemical methods, utilizing two monoclonal antibodies that identify different epitopes of Pgp. We studied pretreatment and posttreatment specimens from 29 cases of a variety of high grade STS, and the data were correlated with pathologic response to chemotherapy. We found expression of MDR phenotype in approximately one-third of cases, as well as significant correlation between MDR phenotype and poor pathologic response to therapy. 977 case containing diagnostic tissue was chosen to perform additional stains. Immunohistochemistry Slides for immunohistochemical studies were stained with C494 (Signet Laboratories, Dedham, MA; 1:100 dilution of a 0.100 mg/mL concentrate mouse monoclonal, 20-minute incubation) and with JSB-1 (Signet Laboratories; 1:10 dilution of a 0.20 mg/mL concentrate mouse monoclonal, overnight incubation). Vimentin staining was performed on the Ventana Medical Systems ES automated slide stainer using prediluted antibody (clone 3B4, Ventana Medical Systems, Tucson, AZ). Sections of kidney, liver, adrenal gland, and colonic mucosa were used as positive controls for C494 and JSB-1. Negative controls were run on every sample. Scoring of Immunostaining Tissue immunoreactivity was evaluated with vimentin immunostaining. Staining intensity for C494 and JSB-1 was scored as 0 (negative), 1 (equivocal), 2 (moderate), or 3 (strong). Cytoplasmic (Fig. 1), membranous (Fig. 2), or Golgi region (Fig. 3) staining were considered positive. To avoid false-positive results due to ambiguous and/or excessive background staining, only moderate or strong staining (Grades 2 and 3) were considered positive. Only cases positive for both antibodies were considered MDR positive. Statistical Analysis METHODS General Paraffin embedded tissue blocks from 29 patients with adult STS, including pretreatment and posttreatment specimens (58 cases), were retrieved. Clinical and pathologic information was obtained from medical records and surgical pathology reports. Pathology Hematoxylin and eosin (H&E) diagnostic materials were reviewed by two pathologists (R.E.J., D.R.L.) with a double-headed microscope. The cases were evaluated to corroborate histologic type and grade of sarcoma and to assess pathologic response to chemotherapy. Histologic grade was determined in the pretreatment specimen according to cellularity, mitotic activity, nuclear hyperchromasia, pleomorphism, and necrosis. A four-tier grading system was used. Pathologic response was evaluated according to the degree of necrosis present in the posttreatment resection specimens, and was considered good for ,15%, moderate for 15–50%, and poor for .50% of tumor viability. At least one representative block from each Comparisons between groups were made with the Fisher exact test. RESULTS There were 12 women and 17 men in the study population. The mean age was 52.2 6 16.2 years; the median age was 56 years. All patients were treated between 1990 and 1998 and received 3 cycles of neoadjuvant chemotherapy consisting of doxorubicin, dacarbazine, and ifosfamide, delivered by infusion over 72–96 hours. Cycles were generally repeated at 21-day intervals, with the last treatment completed 3–5 weeks prior to surgical resection. The histologic type of sarcoma included 12 malignant fibrous histiocytomas, 5 synovial sarcomas, 3 liposarcomas, 3 malignant peripheral nerve sheath tumors, 1 rhabdomyosarcoma, 2 leiomyosarcomas, 1 fibrosarcoma, 1 myxoid chondrosarcoma, and 1 undifferentiated sarcoma. All cases had a histologic Grade 3 (n 5 8) or 4 (n 5 21) out of 4. Twenty-two tumors were localized in the extremities and 7 in the trunk. The mean size of the tumors was 11.9 6 7.4 cm. Among the 29 pretreatment specimens, 10 (34%) 978 CANCER September 15, 1999 / Volume 86 / Number 6 FIGURE 1. (A) Intracytoplasmic immunopositivity with JSB-1 antibody is shown in a malignant peripheral nerve sheath tumor (original magnification 3100). (B) Diffuse positivity with C494 antibody is noted in this monophasic synovial sarcoma. The pattern of staining is cytoplasmic (original magnification 350). FIGURE 2. (A) This myxoid chondrosarcoma (chordoid sarcoma) displays prominent membranous staining with C494 antibody (original magnification 3100). (B) Prominent membranous and cytoplasmic pattern of immunostaining is shown in a rhabdomyosarcoma with C494 antibody (original magnification 3100). were positive for both C494 and JSB-1, and thus were considered MDR positive (95% confidence interval, 0.184 – 0.549). Twelve cases were negative for both markers, and 7 cases were positive with only 1 marker (Table 1). Thus, a total of 19 cases (66%) were considered MDR negative (95% confidence interval, 0.451– 0.816). Sixteen patients had a poor pathologic response, 7 had a moderate response, and 6 had a good response to preoperative chemotherapy. Of the cases with MDR positive biopsies, 9 (90%) had a poor response (95% confidence interval, 0.603– 0.995) and 1 (10%) had a moderate response (Table 2). In the MDR negative biopsy group, 7 (36%) had a poor response (95% confidence interval, 0.150 – 0.635) (P 5 0.0078) and 13 (68%) had a poor or moderate response (P 5 0.057) (i.e., none of MDR positive cases had a good response, compared with 32% MDR negative cases). No signifi- FIGURE 3. Golgi staining pattern with JSB-1 is seen in this malignant peripheral nerve sheath tumor (original magnification 3250). MDR in Soft Tissue Sarcoma/Jimenez et al. 979 TABLE 1 Summary of Clinical Data and Pathologic Findings Case Age (yrs) Gender Diagnosis Location JSB-1a C494a Response 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 74 65 40 55 56 56 47 23 39 36 70 68 52 29 40 42 70 65 75 21 62 56 18 62 39 64 70 61 68 M M M M M M M F M F M F F F M M F F M F F M M F M F M F M Leiomyosarcoma MFH MPNST MPNST Myxoid chondrosarcoma Myxoid liposarcoma Synovial sarcoma Synovial sarcoma Synovial sarcoma MFH MFH MFH MPNST Synovial sarcoma Synovial sarcoma MFH MFH MFH MFH Undifferentiated sarcoma Pleomorphic liposarcoma Fibrosarcoma Alveolar rhabdomyosarcoma Epithelioid leiomyosarcoma Round cell liposarcoma MFH MFH MFH MFH R thigh L thigh R arm L axilla Paraspinal R thigh L thigh L thigh L thigh L thigh R thigh L thigh R forearm R shoulder R groin R forearm L buttock R thigh L forearm L thigh L thigh L buttock L thigh L thigh R thigh R thigh R groin L thigh L forearm 1 1 1 1 1 1 1 1 1 2 2 2 2 2 2 2 1 2 2 2 2 2 2 2 2 1 2 2 2 1 1 1 1 1 1 1 1 1 2 2 2 2 2 2 1 1 2 2 2 1 1 1 1 1 2 2 2 2 Poor Poor Poor Poor Poor Poor Poor Poor Poor Poor Poor Poor Poor Poor Poor Poor Moderate Moderate Moderate Moderate Moderate Moderate Moderate Good Good Good Good Good Good MFH: malignant fibrous histiocytoma; MPNST: malignant peripheral nerve sheath tumor. a Positivity in pretreatment specimen. TABLE 2 MDR Positivity vs. Pathologic Response MDR positive MDR negative Total Good Moderate Poor Total 0 6 (32%) 6 (21%) 1 (10%) 6 (32%) 7 (24%) 9 (90%) 7 (36%) 16 (55%) 10 19 29 MDR: multidrug resistance. cant correlation was observed between histologic type, grade, or any histologic feature and response. Only in one case with an MDR negative pretreatment specimen was Pgp expressed in the resection specimen, suggesting a low incidence of MDR phenotype induction by chemotherapy. DISCUSSION MDR, as defined by Pgp overexpression, has been extensively implicated in chemotherapy resistance in numerous human cancers, particularly in neoplasms derived from tissues that normally express Pgp (i.e., colonic and renal carcinomas).1 However, the list of neoplasms in which MDR is believed to play a significant role in chemoresistance is increasing. In addition, MDR phenotype induction has been documented in neoplasms that did not express it prior to chemotherapy.7,14,15 Recently, the question of whether chemoresistance in adult STS could be explained by Pgp expression has been raised.14,20 –26 Recent data have suggested a strong correlation between Pgp expression and outcome following therapy in pediatric STS13,14 and osteosarcoma.19,27 However, few studies have looked into prediction of response of adult STS to chemotherapy. The reported incidence of Pgp overexpression in STS ranges from 0% to 100%.20 This large range of data is probably best explained by the multiple techniques utilized in the evaluation of either mdr1 expression (mRNA) or Pgp detection, including slot or dot blot, 980 CANCER September 15, 1999 / Volume 86 / Number 6 Northern blot, reverse transcriptase–polymerase chain reaction, RNase protection assay, in situ hybridization, immunoblot, and immunohistochemistry. With so many techniques and criteria used to assess MDR phenotype, results are difficult to compare. Immunohistochemical studies such as ours also report variable incidences, which is probably a reflection of the different numbers and clones of monoclonal antibodies and the variety of criteria utilized for interpretation. For example, Nakanishi et al.22 used staining with either of two antibodies as criterion for MDR positivity and reported an incidence of 62% of MDR phenotype expression in a variety of STS. On the other hand, Vergier et al.23 considered only 1 of 22 cases of STS to be MDR positive, using immunostaining with all of 3 different antibodies as the criterion for MDR positivity. Advantages of immunohistochemical assessment of Pgp expression over other techniques include accessibility to archival fixed tissue, the ability to distinguish between tumor cell expression and normal cell expression, assessment of Pgp directly, and relatively good sensitivity. Limitations of the technique, however, include control of fixation of the tissue and cross-reactivity of monoclonal antibodies with other non-Pgp-related proteins.14,23,28 In our study, adequate immunoreactivity was assessed by staining of the tissue with vimentin, indirectly reflecting an acceptable degree of fixation of the tissue. To address the issue of cross-reactivity, several authors, including a consensus panel recommendation,28 suggest that at least two different monoclonal antibodies directed toward two different epitopes of Pgp should be used to determine MDR positivity by immunohistochemistry. Our criterion for MDR positivity was unequivocal staining of tumor cells with both monoclonal antibodies utilized, thus diminishing the likelihood of falsepositive staining. The clinical significance of MDR positivity in STS is uncertain. Data from published studies are difficult to compare, due to the high variability of clinical outcome parameters (clinical response, pathologic response, and relapse free and overall survival) and modalities of treatment employed (mono/polychemotherapy, surgery, and/or radiation).20 Chan et al.13 found that positive staining for Pgp was a strong predictor of adverse outcome and poor response to chemotherapy in pediatric STS. Nakanishi et al.22 found a significant difference in the 5-year survival of patients with Pgp positive tumors (51%) versus Pgp negative tumors (84%). Conversely, Leuschner et al.21 and Stein et al.24 could not demonstrate any correlation between immunohistochemistry and treatment response. Our data show a significant correlation between MDR positivity, as defined by immunohistochemical detection of Pgp, and pathologic response to chemotherapy. Ninety percent of the patients with MDR positive tumors had a poor pathologic response, compared with only 36% of MDR negative cases. Although based on a relatively small number of cases with heterogeneous histologic types, these results suggest a role for MDR in resistance to chemotherapy in high grade STS. However, 42% of cases with poor response (7 of 16) were MDR negative, which suggests that other factors besides MDR phenotype may contribute to STS chemoresistance. Induction of MDR in neoplastic cells by chemotherapeutic agents has been a matter of concern expressed in several recent publications.7,14,15 Our data show that only one pretreatment MDR negative case expressed MDR phenotype after chemotherapy in the resection specimen. This information suggests that MDR phenotype induction by chemotherapeutic agents in the setting of adult STS is an infrequent event. CONCLUSIONS MDR phenotype, as determined by immunohistochemical detection of Pgp, is expressed in approximately one-third of high grade adult STS. These preliminary data show a significant correlation between MDR phenotype and poor pathologic response to chemotherapy. Further studies are needed to confirm the role of MDR phenotype and other possible mechanisms in the chemoresistance of soft tissue sarcoma. MDR phenotype induction by chemotherapeutic agents in STS appears to be an uncommon event. REFERENCES 1. 2. 3. 4. 5. 6. 7. Nooter K, Stoter G. Molecular mechanisms of multidrug resistance in cancer chemotherapy. Pathol Res Pract 1996; 192:768 – 80. Mertens WC, Bramwell VHC. Adjuvant chemotherapy for soft tissue sarcomas. Hematol Oncol Clin North Am 1995;9: 801–15. Zalupski MM, Baker LH. Systemic adjuvant chemotherapy for soft tissue sarcomas. Hematol Oncol Clin North Am 1995;9:787– 800. Fojo AT, Ueda K, Slamon DJ, Ppoplack DG, Gottesman MM, Pastan I. Expression of a multidrug-resistance gene in human tumors and tissues. Proc Natl Acad Sci U S A 1987;84:265–9. Goldstein LJ, Galski H, Fojo A, Willingham M, Lai SL, Gazdar A, et al. Expression of a multidrug resistance gene in human cancers. J Natl Cancer Inst 1989;81:116 –24. Salmon SE, Grogan TM, Miller T, Scheper R, Dalton WAS. Prediction of doxorubicin resistance in vitro in myeloma, lymphoma, and breast cancer by P-glycoprotein staining. J Natl Cancer Inst 1989;81:696 –701. Grogan TM, Spier CM, Salmon SE, Matzner M, Rybski J, Weinstein RS, et al. P-glycoprotein expression in human plasma cell myeloma: correlation with prior chemotherapy. Blood 1993;81:490 –5. MDR in Soft Tissue Sarcoma/Jimenez et al. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. Klein EA. The multidrug resistance gene in renal cell carcinoma. Semin Urol 1989;7:207–14. Satta T, Isobe K, Yamauchi M, Nakashima I, Takagi H. Expression of MDR1 and glutatione S transferase-p; genes and chemosensitivities in human gastrointestinal cancer. Cancer 1992;69:941– 6. Huang C, Wu M, Xu G, Li D, Cheng H, Tu Z, et al. Overexpression of the MDR1 gene and P-glycoprotein in human hepatocellular carcinoma. J Natl Cancer Inst 1992;84:262–3. Keith WN, Stallard S, Brown R. Expression of mdr1 and gst-p in human breast tumors: comparison to in vitro chemosensitivity. Br J Cancer 1990;61:712– 6. Volm M, Mattern J, Samsel B. Overexpression of P-glycoprotein and glutathione S-transferase-p; in resistant nonsmall cell lung carcinomas of smokers. Br J Cancer 1991;64: 700 – 4. Chan HSL, Thorner PS, Haddad G, Ling V. Immunohistochemical detection of P-glycoprotein: prognostic correlation in soft tissue sarcoma of childhood. J Clin Oncol 1990;8:689 – 704. Chan HSL, DeBoer G, Haddad G, Ling V. Multidrug drug resistance in pediatric sarcomas. Hematol Oncol Clin North Am 1995;9:889 –908. Chaudhary PM, Roninson IB. Induction of multidrug resistance in human cells by transient exposure to different chemotherapeutic drugs. J Natl Cancer Inst 1993;85:632–9. Blair SC, Zalupski MM, Baker LH. Ifosfamide and etoposide in the treatment of advanced soft tissue sarcomas. Am J Clin Oncol 1994;17:480 – 4. Zalupski M, Metch B, Balcerzak S, Fletcher WS, Chapman R, Bonnet JD, et al. A phase III comparison of doxorubicin and DTIC given by bolus vs. infusion in patients with soft tissue sarcomas: a Southwest Oncology Group study. J Natl Cancer Inst 1991;83:926 –32. Glasser DB, Lan JM, Huvos AG, Marcove RC, Rosen G. Survival, prognosis, and therapeutic response in osteogenic sarcoma: the Memorial Hospital experience. Cancer 1992; 69:698 –708. 981 19. Baldini N, Scotlandi K, Barbanti-Brodano G, Manara MC, Maurici D, Bacci G, et al. Expression of P-glycoprotein in high-grade osteosarcomas in relation to clinical outcome. N Engl J Med 1995;333:1380 –5. 20. Stein U, Shoemaker RH, Schlag PM. MDR1 gene expression: evaluation of its use as a molecular marker for prognosis and chemotherapy of bone and soft tissue sarcomas. Eur J Cancer 1996;32A:86 –92. 21. Leuschner I, Schmidt D, Hoffman H, Dietel M, Harms D. Expression of P-glycoprotein in soft tissue tumors of childhood: correlation to treatment response [abstract]. J Cancer Res Clin Oncol 1991;117(Suppl III):97. 22. Nakanishi H, Myoui A, Ochi T, Aozasa K. P-glycoprotein expression in soft-tissue sarcomas. J Cancer Res Clin Oncol 1997;123:352– 6. 23. Vergier B, Cany L, Bonnet F, Robert J, de Mascare A, Coindre JM. Expression of MDR1/P-glycoprotein in human sarcomas. Br J Cancer 1993;68:1221– 6. 24. Stein U, Wunderlich V, Haensch W, Schmidt-Peter P. Expression of the mdr1 gene in bone and soft tissue sarcomas of adult patients. Eur J Cancer 1993;29A:1979 – 81. 25. Noonan KE, Beck C, Hozmayer TA, Chin JE, Wunder JS, Andrulis IL, et al. Quantitative analysis of MDR1 (multidrug resistance) gene expression in human tumors by polymerase chain reaction. Proc Natl Acad Sci U S A 1990;87:7160 – 4. 26. Gerlach JH, Bell DR, Karakousis C, Slocum HK, Kartner N, Rustum YM, et al. P-glycoprotein in human sarcoma: evidence for multidrug resistance. J Clin Oncol 1987;5:1452– 60. 27. Takeshita H, Gebhardt MC, Springfield DS, Kusuzaki K, Mankin HJ. Experimental models for the study of drug resistance in osteosarcoma: P-glycoprotein-positive, murine osteosarcoma cell lines. J Bone Joint Surg 1996;78-A: 366 –75. 28. Beck WT, Grogan TM, Willman CL, Cordon-Cardo C, Parham DM, Kuttesch JF, et al. Methods to detect P-glycoprotein-associated multidrug resistance in patients’ tumors: consensus recommendations. Cancer Res 1996;56:3010 –20.