Serum Tumor Markers in Skeletal Metastasis

Jpn J Clin Oncol 2006;36(7)439–444 doi:10.1093/jjco/hyl046

Serum Tumor Markers in Skeletal Metastasis Satoshi Tsukushi1, Hirohisa Katagiri2, Takae Kataoka3, Yoshihiro Nishida1 and Naoki Ishiguro1 1

Department of Orthopaedic Surgery, Nagoya University Graduate School and School of Medicine, Nagoya, Department of Orthopaedic Surgery, Shizuoka Cancer Center, Sunto-gun, Shizuoka and 3 Department of Clinical Oncology, Nagoya Memorial Hospital, Nagoya, Japan 2

Received October 31, 2005; accepted April 3, 2006; published online June 30, 2006

Key words: skeletal metastasis – tumor marker – diagnosis – carcinoembryonic antigen – carbohydrate antigen 19-9

INTRODUCTION Skeletal metastasis is the most common of all bone neoplasms. In patients older than 50 years with painful, osteolytic and poorly marginated skeletal lesions, the most frequent diagnosis is skeletal metastasis. Skeletal metastases are reportedly the first manifestation of malignancy in 5–20% of patients with cancer of unknown origin (1–5). The primary site at the time of presentation is unknown in 30% of patients with skeletal metastases (6). However, these skeletal lesions often include primary bone tumor or hematological malignancies. It is often difficult to distinguish skeletal metastasis of carcinoma from primary bone tumor and hematological malignancy with the imaging studies alone. Serum tumor markers may be useful in differentiating skeletal metastasis of carcinoma from primary bone

For reprints and all correspondence: Satoshi Tsukushi, Department of Orthopaedic Surgery, Nagoya University, 65 Tsurumaicho, Showa-ku, Nagoya City 466-8550, Japan; E-mail: [email protected]

tumor and hematological malignancy. In recent years, various tumor markers have been established as indicators of the presence of tumors and their volume (7–13). However, there have been no well-documented reports detailing the relationship between skeletal metastasis and tumor markers in a large series of patients. The purpose of our study was to assess the relationship between the clinical features of skeletal metastasis and serum tumor markers and to determine whether tumor markers are a useful modality in the differential diagnosis of skeletal metastasis of carcinoma.

PATIENTS AND METHODS We retrospectively reviewed consecutive 458 patients who underwent surgical or non-surgical treatment for skeletal metastasis from July 1992 to December 2002 at Nagoya Memorial Hospital and Nagoya University Hospital, Nagoya, Japan. Patients were prospectively identified and registered # 2006 Foundation for Promotion of Cancer Research

Downloaded from http://jjco.oxfordjournals.org/ by guest on June 1, 2013

Background: There have been no well-documented reports detailing the relationship between skeletal metastasis and tumor markers in a large series of patients. The purpose of our study was to assess the relationship between the clinical features of skeletal metastasis and serum tumor markers and to determine whether tumor markers are a useful modality in the differential diagnosis of skeletal metastasis. Methods: We retrospectively reviewed consecutive 458 patients with skeletal metastasis and divided the patients into two groups according to six clinical presenting factors. We assessed whether these groups influenced the level of the tumor markers in univariate and multivariate analysis. Results: Patients with skeletal metastasis of carcinoma had a higher level of markers CEA (P < 0.0001) and CA19-9 (P = 0.0008) than patients with primary bone tumors and hematological malignancies. Univariate analysis of clinical variables revealed that metastasis on axial skeleton, multiple skeletal metastases and visceral metastasis were associated with the serum CEA and CA19-9 levels. By multivariate analysis, metastasis on axial skeleton, multiple skeletal metastases and visceral metastasis were found to be associated with the serum CEA and CA19-9 levels. At least one of the tumor markers was elevated in 73% of all patients. Conclusions: The higher tumor marker level (CEA, CA19-9) is predictive of metastasis on the axial skeleton, multiple skeletal metastases and visceral metastasis. Tumor markers are useful as a screening test to distinguish skeletal metastases of carcinoma from primary bone tumors or hematological malignancy from primary bone tumor and hematological malignancy.

440

Tumor markers in skeletal metastasis

Table 1. Primary site in patients with skeletal metastases Primary lesion

Number of patients Total

With primary lesion

With previously resected lesion

Lung

111

46

65

Breast

83

71

12

Prostate

37

16

21

Stomach

27

22

5

Liver

27

16

11

Myeloma

27

4

23

Colon

24

21

3

Kidney

23

10

13

Lymphoma

21

5

16

Primary bone tumor

17

7

10

Thyroid

11

8

3

Uterus cervical

7

7

0

Bile duct

6

6

0

Pancreas

6

3

3

Uterus endometrial

4

2

2

Unknown

14

0

14

Others

13

13

0

Total

458

257

201

(alpha-fetoprotein), PSA (prostate-specific antigen), CA15-3 (carbohydrate antigen 15-3) and SCC (squamous cell carcinoma antigen). We measured AFP in patients with liver dysfunction in which hepatocellular carcinoma was suspected, CA125 in women with a history of gynecological cancer or with gynecological symptoms, SCC in patients who were suspected to have squamous cell carcinoma from lung images and PSA in men with osteosclerotic lesions. The final choice of tumor markers to test for was made by the doctor in charge of each patient. Measurements were made by chemiluminescent enzyme immunoassay. All patients gave their informed consent for every examination. Carcinoembryonic antigen was measured in 342 patients, CA19-9 in 276 patients, AFP in 152 patients, CA125 in 116 patients, SCC in 44 patients, PSA in 30 patients and CA15-3 in 15 patients. Cut-off values for the respective tumor markers were as follows: CEA, 5 ng/ml; CA19-9, 37 U/ml; AFP, 20 ng/ml; CA125, 35 U/ml; SCC, 1.5 ng/ml; PSA, 4 ng/ml; and CA15-3, 30 U/ml. We analyzed the abnormal elevation rate beyond the cut-off value of each tumor marker. We divided the patients into two groups according to the six factors described below and assessed whether these groupings influenced the level of the tumor markers in univariate and multivariate analysis. Twelve potentially prognostic factors, including the condition of the primary lesion, disease-free interval from primary cancer to development of skeletal metastasis, the site of skeletal metastases, pathological fracture, metastatic load and visceral or brain metastasis, were investigated. Each of these six potentially significant factors was grouped in two categories for statistical analysis (Table 2). Condition of the primary lesion was divided into two categories (i) the primary lesion was cured and (ii) the primary lesion was untreated or recurrent. Disease-free interval from primary cancer to development of skeletal metastasis was divided into two categories: (i) patients with no disease-free interval in which skeletal metastasis was the first manifestation of malignancy or found concurrently with the primary lesion and (ii) patients with a disease-free interval in which skeletal metastasis developed after treatment of the primary lesion. Visceral or brain metastases were grouped into two categories: (i) no visceral or brain metastasis and (ii) any visceral or brain metastasis detectable with imaging study. Sites of skeletal metastases were divided into two categories: (i) limited to appendicular bone and (ii) axial bone affected. Metastatic load was subdivided into (i) solitary and (ii) multiple skeletal metastases. Pathological fractures were also divided into (i) with apparent fracture of spine or extremity and (ii) without apparent fracture. The SPSS 11.0 program for Windows was used for statistical analysis. Univariate analysis was performed with the t-test for categorical variables. Multivariate analysis was performed with multiple linear regression to assess the relationship between clinical features of skeletal metastasis and serum tumor markers. Differences were considered to be statistically significant at P < 0.05.


during that period. There were 258 men and 200 women, with a mean age at the time of diagnosis of 61.1 years (range 28–89 years). The lesions came from a variety of primary sites (Table 1). The diagnoses were made by independent radiologists and pathologists. The primary bone tumors were five osteosarcomas, five malignant fibrous histiocytomas, three liposarcomas, clear cell sarcoma and angiosarcoma. Multiple myeloma lesions were considered as skeletal metastases, consistent with other studies reporting skeletal metastasis (14–18). However, solitary plasmacytoma was excluded from our study. Similarly, malignant lymphoma of bone that did not have an extraosseous lesion was also excluded, but patients with primary lymph node lesion and secondary bone metastasis were included. The primary cancer could not be found in 14 patients despite thorough investigation. Of the 14 patients with undetermined primary cancer, a skeletal biopsy was performed in 10 patients, and the lesion was identified as adenocarcinoma in 9 patients and squamous cell carcinoma in 1 patient. Of 458 patients, 257 (56%) had a history of cancer, and the bone tumor appeared as a metastasis during follow-up. In 201 patients (44%) skeletal metastasis was the first manifestation of malignancy. The health insurance system in our region allows tests of up to three types of tumor markers in 1 month. Therefore, at initial examinations for skeletal metastasis we selected two or three tumor markers to test for from among the following: serum CEA (carcinoembryonic antigen), CA19-9 (carbohydrate antigen 19-9), CA125 (cancer antigen 125), AFP

Jpn J Clin Oncol 2006;36(7)

Table 2. Six clinical significant factors for statistical analysis Characteristics

Number of patients

Total number of patients

458

Male/female

258/200

Median age, years (range)

61.1 (28–89)

With primary lesion

286

With previously resected primary lesion

172

With a history of malignancy

257

Without a history of malignancy

201

With metastasis on axial skeleton

416

Without metastasis on axial skeleton

42 211

Without pathological fracture

247

With solitary skeletal metastasis

136

With multiple skeletal metastases

322

With visceral metastasis

230

Without visceral metastasis

228

RESULTS ABNORMAL ELEVATION OF EACH TUMOR MARKER CEA was abnormally elevated in 166 out of 342 (49%) patients: lung 64%, breast 52%, prostate 35%, stomach 48%, liver 17%, colon 80%, multiple myeloma 0%, malignant lymphoma 15%, kidney 0%, 5 of primary unknown 50%, solid carcinoma 51% and hematological malignancies 12%. CA19-9 was abnormally elevated in 78 out of 276 (28%) patients: lung 34%, breast 20%, prostate 21%, stomach 36%, liver 35%, colon 37%, multiple myeloma 0%, malignant lymphoma 0%, kidney 13%, primary unknown 20%, solid carcinoma 30% and hematological malignancies 0%. AFP was abnormally elevated in 29 out of 152 (19%) patients: lung 0%, breast 6%, prostate 0%, stomach 11%, liver 81%, solid carcinoma except liver 6% and hematological malignancies 0%. CA125 was abnormally elevated in 37 out of 116 (32%) patients: lung 63%, breast 27%, stomach 60%, solid carcinoma 38% and 7% (1/14) in hematological malignancies (Table 3). PSA was measured in 30 out of 35 patients with prostate carcinoma and was abnormally elevated in 27 patients (90%). At least one of the tumor markers of CEA, CA19-9, CA15-3, CA125 and PSA was elevated in 73% of all 458 cases, 76% of the patients with solid carcinoma, 0% of the patients with sarcoma and 9% of the patients with hematological malignancies. The serum levels of CEA were elevated in 51% of the patients with carcinoma but in 12% of the patients with hematological malignancy and in 0% of the patients with primary bone tumor. The median of the serum CEA levels of the patients with carcinoma was 840.8 ng/ml. The median of the serum CEA levels of the patients with primary bone tumor and hematological malignancy was 1.9 ng/ml. The serum levels of

CA19-9 were elevated in 30% of the patients with carcinoma but in 0% of the patients with hematological malignancy and in 0% of the patients with primary bone tumor. The median of the serum CA19-9 levels of the patients with carcinoma was 2314.4 U/ml. The median of the serum CA19-9 levels of the patients with primary bone tumor and hematological malignancy was 86.6 U/ml. Patients with skeletal metastasis of carcinoma had a higher (P < 0.0001) level of marker CEA than patients with primary bone tumors and hematological malignancies; sensitivity 51%, specificity 91%. Patients with skeletal metastasis of carcinoma had a higher (P = 0.0008) level of marker CA19-9 than patients with primary bone tumors and hematological malignancies; sensitivity 30%, specificity 100% (Fig. 1). Patients with metastasis on the axial skeleton had a higher (P < 0.0001) level of marker CEA than patients without metastasis on axial skeleton. Patients with metastasis on the axial skeleton had a higher (P = 0.004) level of marker CA19-9 than patients without metastasis on axial skeleton. Patients with multiple metastases had a higher (P = 0.002) level of marker CEA than patients with solitary metastasis. Patients with multiple metastases had a higher (P = 0.005) level of marker CA19-9 than patients with solitary metastasis. Patients with visceral metastasis had a higher (P < 0.0001) level of marker CEA than patients without visceral metastasis. Patients with visceral metastasis had a higher (P < 0.0001) level of marker CA19-9 than patients without visceral metastasis (Table 4). At least one of the tumor markers of CEA, CA19-9, CA15-3, CA125 and PSA was elevated in 43% of patients with solitary metastasis or without metastasis on axial skeleton. Univariate analysis of clinical variables revealed that metastasis on axial skeleton, multiple skeletal metastases and visceral metastasis were associated with the serum CEA and CA19-9 levels. By multivariate analysis, metastasis on axial skeleton, multiple skeletal metastases and visceral metastasis were found to be associated with the serum CEA and CA19-9 levels.

DISCUSSION In patients older than 50 years with osteolytic and poorly marginated skeletal lesions, the most frequent diagnosis is skeletal metastasis of carcinoma. However, these skeletal lesions often include primary bone tumors or hematological malignancies. In skeletal metastasis of carcinoma, main treatment is palliative therapy including radiotherapy, chemotherapy and surgery to maintain quality of life. On the other hand, in primary bone tumor or hematological malignancy radical surgery and intensive chemotherapy are required. In our series, 44% of patients presented without a known primary site and 30% of patients presented with solitary bone lesions. It is sometimes difficult to distinguish skeletal metastasis of carcinoma from a primary bone tumor or hematological malignancy in these patients with the imaging studies alone including plain radiography, computed tomography (CT),


With pathological fracture

441

442


Table 3. Abnormal elevation of each tumor marker Primary lesion

CEA (percent)

Primary Known

CA19-9 (percent)

Unknown

Primary Known

Unknown

AFP (percent)

CA125 (percent)

Lung

64 (56/87)*

73

62

34 (21/62)

55

32

0 (0/27)

63 (15/24)

Breast

52 (33/63)

52

72

20 (9/40)

23

0

6 (1/17)

27 (8/30)

Prostate

35 (7/20)

43

21

21 (3/14)

0

17

Stomach

48 (10/21)

53

33

36 (8/22)

30

75

11 (1/9)

Liver

81 (17/21)

–

– 60 (6/10)

17 (3/18)

0

18

35 (6/17)

25

44

Kidney

0 (0/9)

0

0

13 (1/8)

25

0

–

Colon

80 (16/20)

82

100

37 (7/19)

40

33

–

–

0 (0/13)

–

0

0 (0/13)

–

0

– (0/5)

– (0/5)

Lymphoma

15 (2/13)

–

18

0 (0/13)

–

0

– (0/3)

– (0/3)

Unknown

50 (5/10)

–

42

20 (2/10)

–

25

–

–

Myeloma

51 (138/269)

55

47

30 (64/211)

34

29

6 (5/85)

Hematological malignancies

12 (3/26)

–

8

0 (0/26)

–

0

0 (0/23)

Primary bone tumor

0 (0/10)

Total

49 (166/342)

0

0

55

41

0 (0/8) 28 (78/276)

0

0

34

24

0 (0/8) 19 (29/152)

–

38 (138/269) 7 (1/14) 0 (0/8) 32 (37/116)

CEA, carcinoembryonic antigen; CA19-9, carbohydrate antigen 19-9; AFP, alpha-fetoprotein; CA125, cancer antigen 125. *Parenthesis represents ‘the number of patients with abnormal elevation of tumor marker/the total number of the patients’.

CEA

CA19-9

105

104

103

102

101

100

10-1 Skeletal metastasis Primary bone tumors and hematological malignancies of carcinoma

106 serum carbohydrate antigen 19-9 level (U/mL)

serum carcinoembryonic antigen level (ng/mL)

106

105

104

103

102

101

100

10-1 Skeletal metastasis Primary bone tumors and hematological malignancies of carcinoma

Figure 1. Charts show that the levels of CEA and CA19-9 were significantly higher in patients with skeletal metastasis of carcinoma than patients with primary bone tumors and hematological malignancies (CEA, P < 0.0001; CA19-9, P = 0.0008); dashed line: upper limit of the marker; continuous line: median of the marker.

magnetic resonance imaging (MRI) and bone scan. Skeletal biopsy can often provide pathological diagnosis. In primary bone tumor, a well-planned bone biopsy that does not compromise a limb-salvage procedure is required, and in hematological malignancy it can be diagnosed with immunoelectrophoresis and bone marrow aspiration biopsy. Thus from

the clinical point of view it is necessary to differentiate skeletal metastasis of carcinoma from primary bone tumor and hematological malignancy. Patients with skeletal metastasis of carcinomas had a higher (P < 0.0001) level of marker CEA than patients with primary bone tumors and hematological malignancies. Patients with


Solid carcinoma

–

Jpn J Clin Oncol 2006;36(7)

443

Table 4. Relationship between clinical factors in skeletal metastasis and the tumor marker level (univariate and multivariate analysis) Variable

With primary lesion

CEA

CA19-9

0.528

0.597

0.222

0.099

0.393

0.597

0.0001>

0.004

0.0001>

0.018

0.195

0.607

0.157

0.654

0.002

0.005

0.005

0.013

0.0001>

0.0001>

0.0001>

0.001

195

2130.1

1690.3

1658.3

With a history of malignancy

1250.5

3858.1

Without a history of malignancy

167.5

106.8

With metastasis on axial skeleton

825.9

2186.8

23.9

44.5

With pathological fracture

97.4

190.5

Without pathological fracture

1316.9

3504.5

With multiple skeletal metastases

1025.2

2709.8

33.7

39.1

1315.9

3520.3

With solitary skeletal metastasis With visceral metastasis

98

CEA 76

CA19-9 0.82

63.9

skeletal metastasis of carcinomas had a higher (P = 0.0008) level of marker CA19-9 than patients with primary bone tumors and hematological malignancies. In patients suspected to have skeletal metastasis, measurement of tumor markers can be useful as a screening test to distinguish skeletal metastasis of carcinoma from primary bone tumor or hematological malignancy. The abnormal elevation rate of tumor markers varies greatly depending on the cancer stage. It is reported that CEA is elevated in about 20% of patients with lung, breast and colon carcinomas without metastasis (19–22). In present study it showed an especially high abnormal elevation rate with lung carcinoma (64%), breast carcinoma (52%) and colon carcinoma (80%). Our study included only patients with skeletal metastases, but the abnormal elevation rates of CEA and CA19-9 were higher than primary carcinomas without metastasis and almost the same as advanced cancers (23–27). There are only a few reports about the relation between skeletal metastasis and tumor marker. Shinozaki et al. reported that the tumor marker abnormal elevation rates were 34.6% for CEA, 30.8% for CA19-9 and 30.8% for CA125 in cases of skeletal metastasis (n = 26) (28). Despite the small number of patients in their report, the results were nearly the same as those from our analysis of a large series of patients. At least one of the tumor markers of CEA, CA19-9, CA15-3, CA125 and PSA was elevated in 73% of all 458 patients. When a cancer metastasis appears in bone, the cancer has metastasized hematogenously and abnormal elevation rate of tumor markers is so high that it is very useful as a screening test. It is reported that AFP is elevated in 90% of patients with hepatocellular carcinoma (29). In our study, the abnormal elevation rate was only 14%, but a high rate of 81% (17 of 21 patients) was seen in patients with hepatocellular carcinoma. Prostate-specific antigen was measured in 28 of 35 patients with prostate carcinoma, and a high specificity

of 89% (25 of 28 patients) was seen. When an initial diagnosis is skeletal metastasis of unknown primary cancer, use of organ-specific tumor markers such as PSA and AFP is beneficial in determining primary lesions of prostate and hepatocellular carcinoma. Other tumor markers with lower organ specificity are not particularly helpful in identifying the primary site but are useful in differential diagnosis. Tumor markers showed a lower abnormal elevation rate with renal cell carcinoma (CEA, 0%; CA 19-9, 13%) and thyroid carcinoma (CEA, 0%; CA19-9, 20%) than with other carcinomas in our study. Therefore, if all tumor marker levels are normal, primary bone tumor, hematological malignancy, renal cell carcinoma and thyroid carcinoma would be suspected rather than metastases from other carcinomas such as lung or colorectal carcinoma. Reportedly PET and PET-CT are helpful to determine the primary site in a bone metastasis patient with unknown origin (30). Unless the initial examinations such as serum tumor markers, chest CT and abdominal CT are able to detect the primary site, we recommend performing PET and PET-CT examination. By multivariate analysis, metastasis on axial skeleton, multiple skeletal metastases and visceral metastasis were found to be associated with the serum CEA and CA19-9 levels. However, there was no difference associated with the presence of primary lesions, pathological fracture or previous cancer history. The higher tumor marker level (CEA, CA19-9) is predictive of metastasis on the axial skeleton, multiple skeletal metastases and visceral metastasis. Carcinomas of the lung, breast and prostate are malignant tumors that frequently develop skeletal metastases. In advanced cases skeletal metastases occur in 69% of breast carcinomas (31), 33–85% of prostate carcinomas, 24–55% of lung carcinomas and 35% of renal cell carcinomas (32). In these patients, evaluation for the development of skeletal metastasis with bone scan is necessary during the follow-up. However, the high expense of bone scan makes it difficult to


Without visceral metastasis

Multivariate P-value

CA19-9 (U/ml)

With previously resected primary lesion

Without metastasis on axial skeleton

Univariate P-value

Median level CEA (ng/ml)

444


perform repeatedly. When the tumor marker is elevating, one should suspect that skeletal metastasis is appeared or progressed. Several bone metabolic markers, bone resorption and bone formation markers, are available in recent years, and these markers are reported to be useful to diagnose and monitor the clinical course (33). In future work we would like to determine these bone metabolic markers in addition to tumor markers. Measuring these markers during follow-up is important as they are indicators for the timing of imaging studies for skeletal metastasis and may give a chance to detect skeletal metastases in its early stage. In patients suspected to have skeletal metastasis, tumor markers should be measured in the initial examination. This will shorten the time required for a definite diagnosis of skeletal metastasis; will be helpful in searching for the primary lesion; and will reduce unnecessary, painful and expensive examinations.

This work was supported in part by the Grant-in-Aid for Cancer Research (14–19) from the Ministry of Health, Labour and Welfare, Japan. No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. One or more of the authors have received funding from a Grant-in-Aid for Cancer Research (14–19) from the Ministry of Health, Labour and Welfare, Japan. Each author certifies that his or her institution has approved the human protocol for this investigation and that all investigations were conducted in conformity with ethical principles of research, and that informed consent was obtained.

References 1. Didolkar MS, Fanous N, Elias EG, Moore RH. Metastatic carcinomas from occult primary tumors: a study of 254 patients. Ann Surg 1977;186:625–30. 2. Holmes FF, Fouts TL. Metastatic cancer of unknown primary site. Cancer 1970;26:816–20. 3. Moertel CG, Reitemeier RJ, Schutt AJ, Hahn RG. Treatment of the patient with adenocarcinoma of unknown origin. Cancer 1972;30:1469–72. 4. Simon MA, Karluk MB. Diagnostic strategy for orthopedic surgeons. Skeletal metastases of unknown origin. Clin Orthop 1982;166:96–103. 5. Stewart JF, Tattersall MH, Woods RL, Fox RM. Unknown primary adenocarcinoma: Incidence of over investigation and natural history. Br Med J 1979;1:1530–3. 6. Katagiri H, Takahashi M, Inagaki J, Sugiura H, Ito S, Iwata H. Determining the site of the primary cancer in patients with skeletal metastasis of unknown origin: a retrospective study. Cancer 1999;86:533–537. 7. Chu TM, Nemoto TE. Evaluation of carcinoembryonic antigen in human mammary carcinoma. J Natl Cancer Inst 1973;51:1119–22. 8. Gold P, Freedman SO. Specific carcinoembryonic antigens of the human digestive system. J Exp Med 1965;122:467–81. 9. Gold P, Freedman SO. Demonstration of tumor specific antigen in human colonic carcinomata by immunological tolerance and absorption technique. J Exp Med 1964;121:439–62.


Acknowledgments

10. Goslin RH, Skarin AT, Zamcheck N, Carcinoembryonic antigen: A useful monitor of therapy of small cell lung cancer. JAMA 1981;246:2173–6. 11. Hayakawa T, Kondo T, Shibata T, Hamano H, Kitagawa M, Sakai Y, et al. Sensitive serum markers for detecting pancreatic cancer. Cancer 1988;61:1827–31. 12. Vincent RG, Chu TM, Fergen TB, Ostrander M. Carcinoembryonic antigen in 228 patients with carcinoma of the lung. Cancer 1975;36:2069–76. 13. Yoshikawa T, Nishida K, Tanigawa M, Fukumoto K, Kondo M. Carbohydrate antigen determinant and other tumor markers in gastrointestinal malignancies. Digestion 1985;31:67–76. 14. Bauer HCF, Wedin R. Survival after surgery for spinal and extremity metastases. Acta Orthop Scand 1995;66:143–146. 15. DeWald RL, Bridwell KH, Prodromas C, Rodts M. Reconstructive spinal surgery as palliation for metastatic malignancies of the spine. Spine 1985l;10:21–6. 16. Fidler MW. Anterior decompression and stabilization of metastatic spinal fractures. J Bone Joint Surg 1986;68B:83–90. 17. Harrington KD. Anterior decompression and stabilization of the spine as a treatment for vertebral collapse and spinal cord compression from metastatic malignancy. Clin Orthop 1988;223:177–97. 18. Hosono N, Yonenobu K, Fuji T, Ebara S, Yamashita K, Ono K. Vertebral body replacement with a ceramic prosthesis for metastatic spinal tumors. Spine 1995;20:2454–62. 19. Guadagni F, Ferroni P, Carlini S, Mariotti S, Spila A, Aloe S, et al. A re-evaluation of carcinoembryonic antigen (CEA) as a serum marker for breast cancer: a prospective longitudinal study. Clin Cancer Res 2001;7:2357–62. 20. Ishigami S, Natsugoe S, Hokita S, Che X, Tokuda K, Nakajo A, et al. Clinical importance of preoperative carcinoembryonic antigen and carbohydrate antigen 19-9 levels in gastric cancer. J Clin Gastroenterol 2001;32:41–4. 21. Janssen CW Jr, Orjasaeter H. Carcinoembryonic antigen in patients with gastric carcinoma. Eur J Surg Oncol 1986;12:19–23. 22. Sawabata N, Ohta M, Takeda S, Hirano H, Okumura Y, Asada H, et al. Serum carcinoembryonic antigen level in surgically resected clinical stage I patients with non-small cell lung cancer. Ann Thorac Surg 2002;4:174–9. 23. Concannon JP, Dalbow MH, Frich JC Jr. Carcinoembryonic antigen (CEA) plasma levels in untreated cancer patients and patients with metastatic disease. Radiology 1973;108:191–193. 24. Kuusela P, Jalanko H, Roberts P, Sipponen P, Mecklin JP, Pitkanen R, et al. Comparison of CA 19-9 and carcinoembryonic antigen (CEA) levels in the serum of patients with colorectal diseases. Br J Cancer 1984;49:135–9. 25. Martin EW Jr, Kibbey WE, DiVecchia L, Anderson G, Catalano P, Minton JP. Carcinoembryonic antigen: Clinical and historical aspects. Cancer 1976;37:62–81. 26. Vincent RG, Chu TM, Lane WW. Related articles. The value of carcinoembryonic antigen in patients with carcinoma of the lung. Cancer 1979;44:685–91. 27. Waalkes TP, Gehrke CW, Tormey DC, Woo KB, Kuo KC, Synder J, et al. Biologic markers in breast carcinoma. IV. Serum fucose-protein ratio: Comparisons with carcinoembryonic antigen and human chorionic gonadotrophin. Cancer 41:1871–1882. 28. Shinozaki T, Chigira M, Kato K. Multivariate analysis of serum tumor markers for diagnosis of skeletal metastases. Cancer 1992;69:108–12. 29. Chen DS, Sung JL. Serum alphafetoprotein in hepatocellular carcinoma. Cancer 1977;40:779–83. 30. Delgado-Bolton RC, Frenandez-Perez C, Gonzalez-Mate A, Carreras JL. Meta-analysis of the performance of 18F-FDG PET in primary tumor detection in unknown primary tumors. J Nucl Med 2003;44:1301–14. 31. Coleman RE, Rubens RD. The clinical course of bone metastases from breast cancer. Br J Cancer 1987;55:61–6. 32. Harrington KD. Orthopaedic Management of Metastatic Bone Disease. St Louis: The CV Mosby Company 1988;15–31. 33. Koizumi M, Takahashi S, Ogata E. Comparison of serum bone resorption markers in the diagnosis of skeletal metastasis. Anticancer Res 2003;23:4095–9.