- Email: [email protected]
Low-grade chondrosarcomas: a difficult target for radionuclide imaging. Case report and review of the literature
European Journal of Radiology 43 (2002) 66 – 72 www.elsevier.com/locate/ejrad
Low-grade chondrosarcomas: a difficult target for radionuclide imaging. Case report and review of the literature Georgios Arsos a,*, Ioannis Venizelos b, Nikolaos Karatzas a, Apostolos Koukoulidis c, Constantinos Karakatsanis a a
Department of Nuclear Medicine, Aristotle Uni6ersity Medical School, Hippocration Hospital, Thessaloniki, Greece b Department of Pathology, Hippocration Hospital, Thessaloniki, Greece c Department of Orthopedics, Hippocration Hospital, Thessaloniki, Greece Received 14 May 2001; received in revised form 7 September 2001; accepted 10 September 2001
Abstract Bone scan with Tc-99m (technetium) diphosphonate is sensitive, but non-specific for musculoskeletal tumors. Tl-201 (thallium), Tc-99m-sestamibi, Tc-99m-tetrofosmin, and F-18-fluorodeoxyglucose (F-18-FDG) can visualize tumors more specifically and are therefore useful in orthopedic oncology. However, cartilaginous tumors are characterized by histological and biological features, which potentially impair specific radionuclide imaging. A case of a patient with a low-grade primary chondrosarcoma of the femur and a false negative Tl-201 scan is presented. Tc-99m-based tumor-localizing compounds (sestamibi, tetrofosmin), as well as metabolic and receptor-imaging radiopharmaceuticals have also been reported to fail in low-grade chondrosarcomas imaging. Low cellularity, mitochondrial specialization and the presence of an efflux membrane pump may contribute to poor imaging. A negative Tl-201 or Tc-99m-sestamibi scan should be interpreted with caution, when the possibility of a chondrosarcoma is not negligible. © 2002 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Radionuclide; Imaging; Thallium-201; Chondrosarcoma; Mitochondria
1. Introduction Chondrosarcomas are chondrogenic tumors, classified as primary and secondary. Eighty-five percent of primary chondrosarcomas, typically found in adults, are low-grade tumors [1]. Both normal cartilage and chondrogenic tumors are characterized by the presence of chondrocytes dispersed within and surrounded by cartilaginous matrix. The three-grade (I – III) system proposed by Evans et al. [2] for the classification of bone chondrosarcomas is based on histological features, such as nuclear size, mitotic rate, frequency of lacunae containing multiple nuclei, character of intercellular background and cellularity. Plain radiography is the initiating imaging modality. Computed tomography (CT) and magnetic resonance imaging (MRI) provide a wealth of information about location, shape, * Corresponding author. Present address: Zaka 19, Panorama, 552 36 Thessaloniki, Greece. Tel.: +30-31-992-808. E-mail address: [email protected] (G. Arsos).
size and composition, thus allowing an accurate locoregional staging of chondrogenic tumors [3]. Bone scan with Tc-99m (technetium) diphosphonate is the standard scintigraphic modality for the evaluation of both local tumor conditions and remote bone involvement. Tl-201 (thallium), Tc-99m-sestamibi, Tc99m-tetrofosmin and metabolic positron emission tomography (PET) radiopharmaceuticals are characterized as ‘tumor seeking’ agents and have been found useful in initial diagnosis, grading and post-therapy evaluation of primary bone tumors [4]. Tl-201 in particular, has been widely applied in musculoskeletal tumor imaging. Persistently high Tl-201 uptake indicates the presence of viable neoplasmatic tissue. Tracer uptake depends mainly on blood flow and cellularity of the tumor. However, differentiation between benign and malignant tumors is unsatisfactory as considerable uptake has been reported by some benign lesions [5]. Nevertheless, a negative predictive value of 94% has been reported for bone sarcomas, thus limiting the use of invasive diagnostic procedures [6]. The aims of the
0720-048X/02/$ - see front matter © 2002 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0 7 2 0 - 0 4 8 X ( 0 1 ) 0 0 4 2 7 - 2
G. Arsos et al. / European Journal of Radiology 43 (2002) 66–72
present study are to report on a false negative Tl-201 scan in a patient with a histologically proven low-grade chondrosarcoma and review the literature on scintigraphic diagnosis of chondrosarcoma with tumor-localizing radiopharmaceuticals.
2. Case report The patient was a 67-year-old lady with a short history of left hip swelling and pain. Hematological laboratory values and chest radiograph (CXR) were normal. A radiograph of the femurs showed an extensive osteolytic lesion occupying and deforming the proximal third of the left femur (Fig. 1(a)). CT revealed both intra and extra-osseus expansion of the lesion (Fig. 1(b)). Three-phase bone-scan with 740 MBq (20 mCi) Tc-99m-methylene diphosphonate (MDP) showed increased blood flow, hyperemia and intense osteoblastic activity in the upper third of the left femur (Fig. 2(a), (b) and (c)). Remote lesions were not detected by whole body scanning. A Tl-201 scan, subsequently performed with 111 MBq (3 mCi) of Tl-201-chloride, showed high early concentration of the tracer in the proximal third of the left femur, which substantially fainted on delayed imaging (Fig. 3, arrow). For each phase of both the bone and the Tl-201 scans, symmetric regions of interest (ROIs) were drawn over the lesion and the normal right femur. Average counts/pixel of lesion to contralateral normal tissue (L/N ratio) was calculated. Results were incorporated in Table 1 (present study). According to Tl-201 scan findings, the likelihood of malignancy was considered remote. Core
67
biopsy material consisted of reactive bone and irregularly shaped lobules of cartilaginous tumor, which were separated by narrow fibrous bands. There were a few chondrocytes within a mainly chondroid stroma. There were also a few myxoid areas. Most of chondrocytes had a dense, small nucleus, whilst a few binucleated cells were also present. On the basis of these histological findings, the diagnosis of grade I chondrosarcoma was made (Fig. 4).
3. Review of the literature Data on chondrosarcoma imaging with Tl-201, Tc99m-sestamibi, Tc-99m-tetrofosmin or F-18-fluorodeoxyglucose (F-18-FDG) were collected from literature [7–12]. Only base-line studies reporting individual L/N ratios were considered, allowing descriptive statistics (mean, standard deviation and ranges) to be calculated for each set of data (Table 1). Data on osteosarcomas, where available, are juxtaposed.
4. Discussion Chondrosarcomas, in general, showed a lower L/N ratio than osteosarcomas in both early and late phases of Tl-201 scan (Table 1). L/N ratio in a chondrosarcoma has been found lower than the lowest L/N ratio of three osteosarcomas (1.72 vs 2.01) [7]. Caluser et al. [8] were the first to report the unusually high rate of false negative Tl-201 scans (three out of five in their series) in chondrosarcomas, without commenting on
Fig. 1. (a) Radiograph of the left femur and (b) CT section of femurs at the level of the lesion.
68
G. Arsos et al. / European Journal of Radiology 43 (2002) 66–72
Fig. 2. Three phase Tc-99m-MDP bone scan, anterior views of hip and femur. (a) Dynamic phase (2 frames/s); (b) blood pool phase; and (c) delayed static image.
tumor grade. The authors attributed the reduced Tl-201 uptake to poor vascularity and cellularity within the tumor. In the series of Nishiyama et al. [9], the Tl-201 L/N ratio of a chondrosarcoma was lower, in both the early and late phases, than the corresponding average value of a group of four osterosarcomas; however, no significant difference was observed in blood flow and blood pool L/N ratios between the two types of tumor. Low late Tl-201 L/N ratio, despite the high blood flow and blood pool L/N ratios was also observed in our case (Table 1, present study). These data do not advocate poor vascularity as cause of reduced Tl-201 uptake. In the series of Nishiyama et al. [9], as well as in our case, late Tl-201 L/N ratio of chondrosarcomas was lower than that of the early phase. Early Tl-201 uptake
mainly reflects vascularity and vascular permeability whilst late phase depends more on specific, active intracellular tumor uptake [13]. Significant correlation (r= 0.84, PB 0.01) between the L/N ratio of the early phase of Tl-201 scan and that of the blood pool phase of the Tc-99m-MDP scan has been reported [7]. Late Tl-201 uptake is considered more specific for the diagnosis of malignancy, as some benign lesions (e.g. inflammations) may show increased early Tl-201 uptake. Accordingly, the false negative diagnosis of our case with a late L/N ratio very close to unity (1.08) was not unjustified. Low Tl-201 uptake by chondrosarcomas, resulting in false negative diagnoses, may be due to the low cellularity of the most common low-grade chondrosarcomas. Cellularity is an important feature for grading and is
G. Arsos et al. / European Journal of Radiology 43 (2002) 66–72
low in grade I chondrosarcomas. On the other end of the spectrum, cellular density may be so prominent and chondroid matrix so limited in grade III chondrosarcomas that the tumor could resemble spindle cell sarcoma [2]. Imaging intensity is proportional to the density of viable cells per volume unit and consequently, considerable specific uptake could be expected by the hypercellularised, high-grade chondrosarcomas [13]. Unfortunately, data on tumor grade were not available in some studies included in Table 1 [7– 9]. Evidence on the grade-uptake relation in cartilaginous tumors has been provided by Aoki et al. [12], who studied 11 patients, five with benign cartilaginous tumors and six with chondrosarcomas, with F-18-FDG PET. All patients had histologic characterization. Discrimination between benign lesions and chondrosarcomas was generally fea-
Fig. 3. Dual phase Tl-201 scan, anterior views (low-energy all-purpose collimator, 500 Kcts/ image, 256 × 256 matrix size). Upper, early phase at 15 min; lower, late phase at 3 h post-injection.
69
sible on the basis of standardized uptake value (SUV), a measure of glucose metabolism. SUV value in the six chondrosarcomas paralleled the increasing histologic grade. SUV of the single grade I chondrosarcoma and one grade II tumor was as low as 1.3, the reference cut-off value for discrimination between benign and malignant lesions being 2.0. In another PET study [14] of malignant tumors with L-[Carbon-11]-tyrosine, a 5 cm low-grade hondrosarcoma of the sacrum was the only out of 16 tumors, which was not visualized. Imaging failure was attributed to low metabolic activity of the tumor. Imaging of chondrosarcomas with the newer tumorseeking agents Tc-99m-sestamibi and Tc-99m-tetrofosmin is of particular interest. Reduced uptake of both agents by chondrosarcomas as compared to osteosarcomas has been reported [10]. There is strong evidence that mitochondria are the intracellular site of sestamibi accumulation, the negative potential of the inner mitochondrial membrane being the driving force [15]. Mitochondria of chondrocytes are different from those of other, non-mineralizing tissues. Analysis of the enzymatic content of mitochondria in chondrocytes within the growth plate supports the idea that they are specialized in the process of matrix calcification [16]. It is possible that in relation to matrix mineralization process, Ca2 + content of mitochondria in chondrocytes is high being maximum in locations of active mineralization [17]. High Ca2 + concentration could diminish (or even revert) the negative inner potential, thus abolishing the drive for intramitochondrial accumulation of cationic molecules like sestamibi. Conventional lowgrade chondrosarcoma is an actively mineralized tissue showing both a characteristic radiographic pattern of calcification and intense Tc-99m-MDP accumulation (Figs. 1(a), 2(c) and Table 1). Collectively, these data may support a ‘mitochondrial’ explanation of poor sestamibi accumulation in chondrosarcomas. Low (as compared to osteosarcomas) sestamibi uptake by even highly cellularized, high-grade (III) chondrosarcomas could be similarly explained [2,10]. The above comments for sestamibi could also account for tetrofosmin, a newer cationic compound with intracellular distribution similar to that of sestamibi. The L/N ratio has been found lower in chondrosarcomas than in osteosarcomas, when tetrofosmin was used for imaging [10]. Tetrofosmin uptake and retention was so poor in a case of a chondrosarcoma of the scapula that a L/N ratio less than unity was calculated [11]. An alternative or even additional mechanism that could account for the low uptake of both sestamibi and tetrofosmin by chondrosarcomas, could be the pumping of these radiopharmaceuticals out of the cell by P-glycoprotein (Pgp), which is a cation-pumping membrane molecule also responsible for multi-drug resistance to cytotoxic agents [18]. Expression of Pgp has been re-
70
G. Arsos et al. / European Journal of Radiology 43 (2002) 66–72
Table 1 Lesion to normal (L/N) ratios of various radiopharmaceuticals and SUV of F-18-FDG for chondrosarcomas and osteosarcomas Radiopharmaceutical
L/N ratio or SUVa
Reference
Chondrosarcoma
Osteosarcoma
1/nr (N/gr)
3/nr (N/gr)
Tl-201 Early Late
1.72 –
2.58 9 0.66 (2.01–3.31) –
Tc-99m-MDP BF BP DB
2.70 2.80 4.21
3.11 91.26 (1.70–4.12) 3.09 9 1.87 (1.70–5.22) 5.94 9 5.43 (2.60–12.20)
5/nr (N/gr)
9/nr (N/gr)
1.239 0.36 (0.94–1.82) –
4.52 9 2.98 (1.27–10.30) –
1.469 0.87 (1.01–3.00) 1.399 0.65 (0.91–2.49) 5.68 9 2.64 (3.22–10.10)
7.35 96.68 (1.80–19.10) 4.80 93.02 (2.30–11.80) 8.12 9 2.94 (2.80–12.70)
1/nr (N/gr)
4/nr (N/gr)
2.36 1.66
4.25 9 0.61 (3.63–5.12) 3.85 90.51 (3.23–4.29)
3.33 2.82 8.28
3.77 91.54 (2.76–5.71) 3.17 9 0.42 (2.85–3.75) 4.39 9 1.04 (3.05–5.50)
Tl-201 Early Late Tc-99m-MDP BF BP DB Tl-201 Early Late Tc-99m-MDP BF BP DB Tl-201 Early Late
1/I (N/gr)
[7]
[8]
[9]
Present study
2.52 1.08
Tc-99m-MDP BF BP DB
4.73 2.26 7.17
Tc-99m-sestamibi Tc-99m-tetrofosmin
2/III (N/gr) 1.199 0.12 (1.07–1.30) 1.45 9 0.27 (1.18–1.72) 1/nr (N/gr) 0.58
[11]
Tc-99m-tetrofosmin
1/I (N/gr) 1.30a 4/II (N/gr) 2.309 0.78 (1.30–3.10)a 1/III (N/gr) 3.30a
[12]
F-18 -FDG F-18 -FDG F-18 -FDG
9/III–IV (N/gr) 2.79 9 1.27 (1.81–5.85) 2.55 90.94 (1.66–4.78)
[10]
N/grade, number of cases/tumor grade. Figures represent mean 9 standard deviation (when N\1). nr, not reported; MDP, methylene diphosphonate; BF, blood flow; BP, blood pool; DB, delayed bone; and F-18-FDG, F-18-fluorodeoxyglucose. a Standardized uptake value.
ported in both benign and malignant cartilaginous tumors, being greatest in high-grade chondrosarcomas [19]. Successful imaging of sarcomatous tumors with In-
111-DTPA-octreotide, a somatostatin labeled analogue, has been recently reported. Interestingly, the only one out of 13 tumors, which failed to be visualized was again a chondrosarcoma [20].
G. Arsos et al. / European Journal of Radiology 43 (2002) 66–72
71
Fig. 4. Histological section of the biopsy specimen. A few, widely spaced mononucleated and binucleated chondrocytes (arrow) are surrounded by abundant chondroid stroma (hematoxylin and eosin, 400X).
In conclusion, visualization of low-grade chondrosarcomas with a variety of radiopharmaceuticals is problematic. Low cellularity may be responsible for false negative Tl-201 scans and metabolic PET imaging. Evaluation of malignancy on the basis of Tl-201 scan findings is not prudent for low-grade chondrosarcomas. Poor performance of the newer Tc-99m-based tumorseeking compounds in chondrosarcoma may be associated with chondrocyte-specific features, such as specialized mitochondrial function or the presence of a membrane efflux pump. As chondrosarcomas are the second most frequent malignant bone tumor and the vast majority of them are low-grade tumors, exclusion of malignancy on the basis of a negative Tl-201 or Tc-99m-sestamibi scan should be cautious, when the likelihood of chondrosarcoma is suggested by clinical, demographic and imaging data. A negative base-line scan excludes these agents from further specific, clinically useful tasks, such as the evaluation of response to treatment or characterization of post-therapeutic tissue changes.
References [1] Johnston JO. Tumors in orthopedics. In: Skinner HB, editor. Current diagnosis and treatment in orthopedics. Englewood Cliffs, NJ: Prentice-Hall, 1995:236 –314. [2] Evans HL, Ayala AC, Romsdahl MM. Prognostic factors in chondrosarcoma of bone. A clinicopathological analysis with emphasis on histologic grading. Cancer 1977;40:818 – 31. [3] Masciocchi C, Sparvoli L, Barile A. Diagnostic imaging of malignant cartilage tumors. Eur J Radiol 1998;27:S86 –90.
[4] Pneumaticos SG, Chatziioannou SN, Moore WH, Johnson M. The role of radionuclides in primary musculosceletal tumors beyond the ‘bone scan’. Crit Rev Oncol Hematol 2001;37:217 – 26. [5] Abdel-Dayem HM, Scott AM, Macapinlac HA, El-Gazzar AH, Larson SM. Role of thallium-201 chloride and Tc-99m sestamibi in tumor imaging. In: Freeman LM, editor. Nuclear medicine annual 1994. New York: Raven Press, 1994:181 – 234. [6] Van der Wall H, Murray PC, Huckstep RL, Philips RL. The role of thallium scintigraphy in excluding malignancy in bone. Clin Nucl Med 1993;18:551 – 7. [7] Caluser C, Macapinlac H, Healey J, Ghavimi F, Meyers P, Wollner N, et al. The relationship between thallium uptake, blood flow, and blood pool activity in bone and soft tissue tumors. Clin Nucl Med 1992;17:565 – 72. [8] Caluser CI, Abdel-Dayem HM, Macapinlac HA, Scott A, Healy JH, Huvos A, et al. The value of thallium and three-phase bone scans in the evaluation of bone and soft tissue sarcomas. Eur J Nucl Med 1994;21:1198 – 205. [9] Nishiyama Y, Yamamoto Y, Toyoma Y, Satoh K, Ohkawa M, Tanabe M. Diagnostic value of Tl-201 and three-phase bone scintigraphy for bone and soft tissue tumors. Clin Nucl Med 2000;3:200 – 5. [10] So¨ derlund V, Jonsson C, Bauer HCF, Brosjo O, Jacobsson H. Comparison of technetium-99m-MIBI and technetium-99mtetrofosmin uptake by musculosceletal sarcomas. J Nucl Med 1997;38:682 – 5. [11] Bu¨ yu¨ kdereli G, Sargin O8 , O8 zbarlas S. Tc-99m tetrofosmin imaging in chondrosarcoma. Clin Nucl Med 2000;1:64 – 5. [12] Aoki J, Watanabe H, Shinozaki T, Tokunaga M, Inoue T, Endo K. FDG-PET in differential diagnosis and grading of chondrosarcoma. J Comput Assist Tomog 1999;23:603 – 8. [13] Abdel-Dayem HM. The role of nuclear medicine in primary bone and soft tissue tumors. Sem Nucl Med 1997;24:355 –63. [14] Kole AC, Pruim J, Nieweg OE, Van Ginkel RJ, Hoekstra HJ, Koops HS, et al. PET with L-[Carbon-11]-tyrosine to visualise tumors and measure protein synthesis rates. J Nucl Med 1997;38:191 – 5. [15] Piwnica-Worms D, Kronauge JF, Chiu ML. Uptake and reten-
72
G. Arsos et al. / European Journal of Radiology 43 (2002) 66–72
tion of hexakis (2-methoxyisobutyl isonitrile) technetium(I) in cultured chick myocardial cells. Mitochondrial and plasma membrane potential dependence. Circulation 1990;82:1826 – 38. [16] Shapiro LM, Lee NH. Heterogeneity of chondrocyte mitochondria. A study of Ca2 + concentration and density banding characteristics of normal and rachitic cartilage. Biochim Biophys Acta 1976;451:583 – 91. [17] Spencer CA, Palmer TN, Mansosn RM. Intermediary metabolism of the Swarm rat chondrosarcoma chondrocyte. Biochem J 1990;265:911 –4. [18] Wyman JJ, Hornstein AM, Meinter PA, Mak S, Verdier P,
Block JA, et al. Multidrug resistance-1 and P-glycoprotein in human chondrosarcoma cell lines: expression correlates with decreased intracellular doxorubicin and in vitro chemoresistance. J Orthop Res 1999;17:935 – 40. [19] Rosier RN, O’Keefe RJ, Teot LA, Fox EJ, Nester TA, Pujas JE, et al. P-Glycoprotein expression in cartilaginous tumours. Surg Oncol 1997;65:95 – 105. [20] Giannakenas C, Kalofonos HP, Apostolopoulos D, Petsas T, Kalogeropoulou C, Tzorakelefterakis E, et al. Scintigraphic imaging of sarcomatous tumors with [(111)In-DTPA-phe-1]-octreotide. Oncology 2000;58:18 – 24.
Low-grade chondrosarcomas: a difficult target for radionuclide imaging. Case report and review of the literature Georgios Arsos a,*, Ioannis Venizelos b, Nikolaos Karatzas a, Apostolos Koukoulidis c, Constantinos Karakatsanis a a
Department of Nuclear Medicine, Aristotle Uni6ersity Medical School, Hippocration Hospital, Thessaloniki, Greece b Department of Pathology, Hippocration Hospital, Thessaloniki, Greece c Department of Orthopedics, Hippocration Hospital, Thessaloniki, Greece Received 14 May 2001; received in revised form 7 September 2001; accepted 10 September 2001
Abstract Bone scan with Tc-99m (technetium) diphosphonate is sensitive, but non-specific for musculoskeletal tumors. Tl-201 (thallium), Tc-99m-sestamibi, Tc-99m-tetrofosmin, and F-18-fluorodeoxyglucose (F-18-FDG) can visualize tumors more specifically and are therefore useful in orthopedic oncology. However, cartilaginous tumors are characterized by histological and biological features, which potentially impair specific radionuclide imaging. A case of a patient with a low-grade primary chondrosarcoma of the femur and a false negative Tl-201 scan is presented. Tc-99m-based tumor-localizing compounds (sestamibi, tetrofosmin), as well as metabolic and receptor-imaging radiopharmaceuticals have also been reported to fail in low-grade chondrosarcomas imaging. Low cellularity, mitochondrial specialization and the presence of an efflux membrane pump may contribute to poor imaging. A negative Tl-201 or Tc-99m-sestamibi scan should be interpreted with caution, when the possibility of a chondrosarcoma is not negligible. © 2002 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Radionuclide; Imaging; Thallium-201; Chondrosarcoma; Mitochondria
1. Introduction Chondrosarcomas are chondrogenic tumors, classified as primary and secondary. Eighty-five percent of primary chondrosarcomas, typically found in adults, are low-grade tumors [1]. Both normal cartilage and chondrogenic tumors are characterized by the presence of chondrocytes dispersed within and surrounded by cartilaginous matrix. The three-grade (I – III) system proposed by Evans et al. [2] for the classification of bone chondrosarcomas is based on histological features, such as nuclear size, mitotic rate, frequency of lacunae containing multiple nuclei, character of intercellular background and cellularity. Plain radiography is the initiating imaging modality. Computed tomography (CT) and magnetic resonance imaging (MRI) provide a wealth of information about location, shape, * Corresponding author. Present address: Zaka 19, Panorama, 552 36 Thessaloniki, Greece. Tel.: +30-31-992-808. E-mail address: [email protected] (G. Arsos).
size and composition, thus allowing an accurate locoregional staging of chondrogenic tumors [3]. Bone scan with Tc-99m (technetium) diphosphonate is the standard scintigraphic modality for the evaluation of both local tumor conditions and remote bone involvement. Tl-201 (thallium), Tc-99m-sestamibi, Tc99m-tetrofosmin and metabolic positron emission tomography (PET) radiopharmaceuticals are characterized as ‘tumor seeking’ agents and have been found useful in initial diagnosis, grading and post-therapy evaluation of primary bone tumors [4]. Tl-201 in particular, has been widely applied in musculoskeletal tumor imaging. Persistently high Tl-201 uptake indicates the presence of viable neoplasmatic tissue. Tracer uptake depends mainly on blood flow and cellularity of the tumor. However, differentiation between benign and malignant tumors is unsatisfactory as considerable uptake has been reported by some benign lesions [5]. Nevertheless, a negative predictive value of 94% has been reported for bone sarcomas, thus limiting the use of invasive diagnostic procedures [6]. The aims of the
0720-048X/02/$ - see front matter © 2002 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0 7 2 0 - 0 4 8 X ( 0 1 ) 0 0 4 2 7 - 2
G. Arsos et al. / European Journal of Radiology 43 (2002) 66–72
present study are to report on a false negative Tl-201 scan in a patient with a histologically proven low-grade chondrosarcoma and review the literature on scintigraphic diagnosis of chondrosarcoma with tumor-localizing radiopharmaceuticals.
2. Case report The patient was a 67-year-old lady with a short history of left hip swelling and pain. Hematological laboratory values and chest radiograph (CXR) were normal. A radiograph of the femurs showed an extensive osteolytic lesion occupying and deforming the proximal third of the left femur (Fig. 1(a)). CT revealed both intra and extra-osseus expansion of the lesion (Fig. 1(b)). Three-phase bone-scan with 740 MBq (20 mCi) Tc-99m-methylene diphosphonate (MDP) showed increased blood flow, hyperemia and intense osteoblastic activity in the upper third of the left femur (Fig. 2(a), (b) and (c)). Remote lesions were not detected by whole body scanning. A Tl-201 scan, subsequently performed with 111 MBq (3 mCi) of Tl-201-chloride, showed high early concentration of the tracer in the proximal third of the left femur, which substantially fainted on delayed imaging (Fig. 3, arrow). For each phase of both the bone and the Tl-201 scans, symmetric regions of interest (ROIs) were drawn over the lesion and the normal right femur. Average counts/pixel of lesion to contralateral normal tissue (L/N ratio) was calculated. Results were incorporated in Table 1 (present study). According to Tl-201 scan findings, the likelihood of malignancy was considered remote. Core
67
biopsy material consisted of reactive bone and irregularly shaped lobules of cartilaginous tumor, which were separated by narrow fibrous bands. There were a few chondrocytes within a mainly chondroid stroma. There were also a few myxoid areas. Most of chondrocytes had a dense, small nucleus, whilst a few binucleated cells were also present. On the basis of these histological findings, the diagnosis of grade I chondrosarcoma was made (Fig. 4).
3. Review of the literature Data on chondrosarcoma imaging with Tl-201, Tc99m-sestamibi, Tc-99m-tetrofosmin or F-18-fluorodeoxyglucose (F-18-FDG) were collected from literature [7–12]. Only base-line studies reporting individual L/N ratios were considered, allowing descriptive statistics (mean, standard deviation and ranges) to be calculated for each set of data (Table 1). Data on osteosarcomas, where available, are juxtaposed.
4. Discussion Chondrosarcomas, in general, showed a lower L/N ratio than osteosarcomas in both early and late phases of Tl-201 scan (Table 1). L/N ratio in a chondrosarcoma has been found lower than the lowest L/N ratio of three osteosarcomas (1.72 vs 2.01) [7]. Caluser et al. [8] were the first to report the unusually high rate of false negative Tl-201 scans (three out of five in their series) in chondrosarcomas, without commenting on
Fig. 1. (a) Radiograph of the left femur and (b) CT section of femurs at the level of the lesion.
68
G. Arsos et al. / European Journal of Radiology 43 (2002) 66–72
Fig. 2. Three phase Tc-99m-MDP bone scan, anterior views of hip and femur. (a) Dynamic phase (2 frames/s); (b) blood pool phase; and (c) delayed static image.
tumor grade. The authors attributed the reduced Tl-201 uptake to poor vascularity and cellularity within the tumor. In the series of Nishiyama et al. [9], the Tl-201 L/N ratio of a chondrosarcoma was lower, in both the early and late phases, than the corresponding average value of a group of four osterosarcomas; however, no significant difference was observed in blood flow and blood pool L/N ratios between the two types of tumor. Low late Tl-201 L/N ratio, despite the high blood flow and blood pool L/N ratios was also observed in our case (Table 1, present study). These data do not advocate poor vascularity as cause of reduced Tl-201 uptake. In the series of Nishiyama et al. [9], as well as in our case, late Tl-201 L/N ratio of chondrosarcomas was lower than that of the early phase. Early Tl-201 uptake
mainly reflects vascularity and vascular permeability whilst late phase depends more on specific, active intracellular tumor uptake [13]. Significant correlation (r= 0.84, PB 0.01) between the L/N ratio of the early phase of Tl-201 scan and that of the blood pool phase of the Tc-99m-MDP scan has been reported [7]. Late Tl-201 uptake is considered more specific for the diagnosis of malignancy, as some benign lesions (e.g. inflammations) may show increased early Tl-201 uptake. Accordingly, the false negative diagnosis of our case with a late L/N ratio very close to unity (1.08) was not unjustified. Low Tl-201 uptake by chondrosarcomas, resulting in false negative diagnoses, may be due to the low cellularity of the most common low-grade chondrosarcomas. Cellularity is an important feature for grading and is
G. Arsos et al. / European Journal of Radiology 43 (2002) 66–72
low in grade I chondrosarcomas. On the other end of the spectrum, cellular density may be so prominent and chondroid matrix so limited in grade III chondrosarcomas that the tumor could resemble spindle cell sarcoma [2]. Imaging intensity is proportional to the density of viable cells per volume unit and consequently, considerable specific uptake could be expected by the hypercellularised, high-grade chondrosarcomas [13]. Unfortunately, data on tumor grade were not available in some studies included in Table 1 [7– 9]. Evidence on the grade-uptake relation in cartilaginous tumors has been provided by Aoki et al. [12], who studied 11 patients, five with benign cartilaginous tumors and six with chondrosarcomas, with F-18-FDG PET. All patients had histologic characterization. Discrimination between benign lesions and chondrosarcomas was generally fea-
Fig. 3. Dual phase Tl-201 scan, anterior views (low-energy all-purpose collimator, 500 Kcts/ image, 256 × 256 matrix size). Upper, early phase at 15 min; lower, late phase at 3 h post-injection.
69
sible on the basis of standardized uptake value (SUV), a measure of glucose metabolism. SUV value in the six chondrosarcomas paralleled the increasing histologic grade. SUV of the single grade I chondrosarcoma and one grade II tumor was as low as 1.3, the reference cut-off value for discrimination between benign and malignant lesions being 2.0. In another PET study [14] of malignant tumors with L-[Carbon-11]-tyrosine, a 5 cm low-grade hondrosarcoma of the sacrum was the only out of 16 tumors, which was not visualized. Imaging failure was attributed to low metabolic activity of the tumor. Imaging of chondrosarcomas with the newer tumorseeking agents Tc-99m-sestamibi and Tc-99m-tetrofosmin is of particular interest. Reduced uptake of both agents by chondrosarcomas as compared to osteosarcomas has been reported [10]. There is strong evidence that mitochondria are the intracellular site of sestamibi accumulation, the negative potential of the inner mitochondrial membrane being the driving force [15]. Mitochondria of chondrocytes are different from those of other, non-mineralizing tissues. Analysis of the enzymatic content of mitochondria in chondrocytes within the growth plate supports the idea that they are specialized in the process of matrix calcification [16]. It is possible that in relation to matrix mineralization process, Ca2 + content of mitochondria in chondrocytes is high being maximum in locations of active mineralization [17]. High Ca2 + concentration could diminish (or even revert) the negative inner potential, thus abolishing the drive for intramitochondrial accumulation of cationic molecules like sestamibi. Conventional lowgrade chondrosarcoma is an actively mineralized tissue showing both a characteristic radiographic pattern of calcification and intense Tc-99m-MDP accumulation (Figs. 1(a), 2(c) and Table 1). Collectively, these data may support a ‘mitochondrial’ explanation of poor sestamibi accumulation in chondrosarcomas. Low (as compared to osteosarcomas) sestamibi uptake by even highly cellularized, high-grade (III) chondrosarcomas could be similarly explained [2,10]. The above comments for sestamibi could also account for tetrofosmin, a newer cationic compound with intracellular distribution similar to that of sestamibi. The L/N ratio has been found lower in chondrosarcomas than in osteosarcomas, when tetrofosmin was used for imaging [10]. Tetrofosmin uptake and retention was so poor in a case of a chondrosarcoma of the scapula that a L/N ratio less than unity was calculated [11]. An alternative or even additional mechanism that could account for the low uptake of both sestamibi and tetrofosmin by chondrosarcomas, could be the pumping of these radiopharmaceuticals out of the cell by P-glycoprotein (Pgp), which is a cation-pumping membrane molecule also responsible for multi-drug resistance to cytotoxic agents [18]. Expression of Pgp has been re-
70
G. Arsos et al. / European Journal of Radiology 43 (2002) 66–72
Table 1 Lesion to normal (L/N) ratios of various radiopharmaceuticals and SUV of F-18-FDG for chondrosarcomas and osteosarcomas Radiopharmaceutical
L/N ratio or SUVa
Reference
Chondrosarcoma
Osteosarcoma
1/nr (N/gr)
3/nr (N/gr)
Tl-201 Early Late
1.72 –
2.58 9 0.66 (2.01–3.31) –
Tc-99m-MDP BF BP DB
2.70 2.80 4.21
3.11 91.26 (1.70–4.12) 3.09 9 1.87 (1.70–5.22) 5.94 9 5.43 (2.60–12.20)
5/nr (N/gr)
9/nr (N/gr)
1.239 0.36 (0.94–1.82) –
4.52 9 2.98 (1.27–10.30) –
1.469 0.87 (1.01–3.00) 1.399 0.65 (0.91–2.49) 5.68 9 2.64 (3.22–10.10)
7.35 96.68 (1.80–19.10) 4.80 93.02 (2.30–11.80) 8.12 9 2.94 (2.80–12.70)
1/nr (N/gr)
4/nr (N/gr)
2.36 1.66
4.25 9 0.61 (3.63–5.12) 3.85 90.51 (3.23–4.29)
3.33 2.82 8.28
3.77 91.54 (2.76–5.71) 3.17 9 0.42 (2.85–3.75) 4.39 9 1.04 (3.05–5.50)
Tl-201 Early Late Tc-99m-MDP BF BP DB Tl-201 Early Late Tc-99m-MDP BF BP DB Tl-201 Early Late
1/I (N/gr)
[7]
[8]
[9]
Present study
2.52 1.08
Tc-99m-MDP BF BP DB
4.73 2.26 7.17
Tc-99m-sestamibi Tc-99m-tetrofosmin
2/III (N/gr) 1.199 0.12 (1.07–1.30) 1.45 9 0.27 (1.18–1.72) 1/nr (N/gr) 0.58
[11]
Tc-99m-tetrofosmin
1/I (N/gr) 1.30a 4/II (N/gr) 2.309 0.78 (1.30–3.10)a 1/III (N/gr) 3.30a
[12]
F-18 -FDG F-18 -FDG F-18 -FDG
9/III–IV (N/gr) 2.79 9 1.27 (1.81–5.85) 2.55 90.94 (1.66–4.78)
[10]
N/grade, number of cases/tumor grade. Figures represent mean 9 standard deviation (when N\1). nr, not reported; MDP, methylene diphosphonate; BF, blood flow; BP, blood pool; DB, delayed bone; and F-18-FDG, F-18-fluorodeoxyglucose. a Standardized uptake value.
ported in both benign and malignant cartilaginous tumors, being greatest in high-grade chondrosarcomas [19]. Successful imaging of sarcomatous tumors with In-
111-DTPA-octreotide, a somatostatin labeled analogue, has been recently reported. Interestingly, the only one out of 13 tumors, which failed to be visualized was again a chondrosarcoma [20].
G. Arsos et al. / European Journal of Radiology 43 (2002) 66–72
71
Fig. 4. Histological section of the biopsy specimen. A few, widely spaced mononucleated and binucleated chondrocytes (arrow) are surrounded by abundant chondroid stroma (hematoxylin and eosin, 400X).
In conclusion, visualization of low-grade chondrosarcomas with a variety of radiopharmaceuticals is problematic. Low cellularity may be responsible for false negative Tl-201 scans and metabolic PET imaging. Evaluation of malignancy on the basis of Tl-201 scan findings is not prudent for low-grade chondrosarcomas. Poor performance of the newer Tc-99m-based tumorseeking compounds in chondrosarcoma may be associated with chondrocyte-specific features, such as specialized mitochondrial function or the presence of a membrane efflux pump. As chondrosarcomas are the second most frequent malignant bone tumor and the vast majority of them are low-grade tumors, exclusion of malignancy on the basis of a negative Tl-201 or Tc-99m-sestamibi scan should be cautious, when the likelihood of chondrosarcoma is suggested by clinical, demographic and imaging data. A negative base-line scan excludes these agents from further specific, clinically useful tasks, such as the evaluation of response to treatment or characterization of post-therapeutic tissue changes.
References [1] Johnston JO. Tumors in orthopedics. In: Skinner HB, editor. Current diagnosis and treatment in orthopedics. Englewood Cliffs, NJ: Prentice-Hall, 1995:236 –314. [2] Evans HL, Ayala AC, Romsdahl MM. Prognostic factors in chondrosarcoma of bone. A clinicopathological analysis with emphasis on histologic grading. Cancer 1977;40:818 – 31. [3] Masciocchi C, Sparvoli L, Barile A. Diagnostic imaging of malignant cartilage tumors. Eur J Radiol 1998;27:S86 –90.
[4] Pneumaticos SG, Chatziioannou SN, Moore WH, Johnson M. The role of radionuclides in primary musculosceletal tumors beyond the ‘bone scan’. Crit Rev Oncol Hematol 2001;37:217 – 26. [5] Abdel-Dayem HM, Scott AM, Macapinlac HA, El-Gazzar AH, Larson SM. Role of thallium-201 chloride and Tc-99m sestamibi in tumor imaging. In: Freeman LM, editor. Nuclear medicine annual 1994. New York: Raven Press, 1994:181 – 234. [6] Van der Wall H, Murray PC, Huckstep RL, Philips RL. The role of thallium scintigraphy in excluding malignancy in bone. Clin Nucl Med 1993;18:551 – 7. [7] Caluser C, Macapinlac H, Healey J, Ghavimi F, Meyers P, Wollner N, et al. The relationship between thallium uptake, blood flow, and blood pool activity in bone and soft tissue tumors. Clin Nucl Med 1992;17:565 – 72. [8] Caluser CI, Abdel-Dayem HM, Macapinlac HA, Scott A, Healy JH, Huvos A, et al. The value of thallium and three-phase bone scans in the evaluation of bone and soft tissue sarcomas. Eur J Nucl Med 1994;21:1198 – 205. [9] Nishiyama Y, Yamamoto Y, Toyoma Y, Satoh K, Ohkawa M, Tanabe M. Diagnostic value of Tl-201 and three-phase bone scintigraphy for bone and soft tissue tumors. Clin Nucl Med 2000;3:200 – 5. [10] So¨ derlund V, Jonsson C, Bauer HCF, Brosjo O, Jacobsson H. Comparison of technetium-99m-MIBI and technetium-99mtetrofosmin uptake by musculosceletal sarcomas. J Nucl Med 1997;38:682 – 5. [11] Bu¨ yu¨ kdereli G, Sargin O8 , O8 zbarlas S. Tc-99m tetrofosmin imaging in chondrosarcoma. Clin Nucl Med 2000;1:64 – 5. [12] Aoki J, Watanabe H, Shinozaki T, Tokunaga M, Inoue T, Endo K. FDG-PET in differential diagnosis and grading of chondrosarcoma. J Comput Assist Tomog 1999;23:603 – 8. [13] Abdel-Dayem HM. The role of nuclear medicine in primary bone and soft tissue tumors. Sem Nucl Med 1997;24:355 –63. [14] Kole AC, Pruim J, Nieweg OE, Van Ginkel RJ, Hoekstra HJ, Koops HS, et al. PET with L-[Carbon-11]-tyrosine to visualise tumors and measure protein synthesis rates. J Nucl Med 1997;38:191 – 5. [15] Piwnica-Worms D, Kronauge JF, Chiu ML. Uptake and reten-
72
G. Arsos et al. / European Journal of Radiology 43 (2002) 66–72
tion of hexakis (2-methoxyisobutyl isonitrile) technetium(I) in cultured chick myocardial cells. Mitochondrial and plasma membrane potential dependence. Circulation 1990;82:1826 – 38. [16] Shapiro LM, Lee NH. Heterogeneity of chondrocyte mitochondria. A study of Ca2 + concentration and density banding characteristics of normal and rachitic cartilage. Biochim Biophys Acta 1976;451:583 – 91. [17] Spencer CA, Palmer TN, Mansosn RM. Intermediary metabolism of the Swarm rat chondrosarcoma chondrocyte. Biochem J 1990;265:911 –4. [18] Wyman JJ, Hornstein AM, Meinter PA, Mak S, Verdier P,
Block JA, et al. Multidrug resistance-1 and P-glycoprotein in human chondrosarcoma cell lines: expression correlates with decreased intracellular doxorubicin and in vitro chemoresistance. J Orthop Res 1999;17:935 – 40. [19] Rosier RN, O’Keefe RJ, Teot LA, Fox EJ, Nester TA, Pujas JE, et al. P-Glycoprotein expression in cartilaginous tumours. Surg Oncol 1997;65:95 – 105. [20] Giannakenas C, Kalofonos HP, Apostolopoulos D, Petsas T, Kalogeropoulou C, Tzorakelefterakis E, et al. Scintigraphic imaging of sarcomatous tumors with [(111)In-DTPA-phe-1]-octreotide. Oncology 2000;58:18 – 24.