|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
© 2002 American Society for Clinical Oncology Prospective Evaluation of the Peptide-Bound Collagen Type I Cross-Links N-Telopeptide and C-Telopeptide in Predicting Bone Metastases StatusByFrom the Hospital de Santa Maria, Faculdade de Medicina de Lisboa, Lisbon, Portugal; Milton S. Hershey Medical Center, Penn State University, Hershey, PA; and Department of Biometry and Epidemiology, Medical University of South Carolina, Charleston, SC. Address reprint requests to Luis Costa, Unidade de Oncologia, Hospital de Santa Maria, Av Professor Egas Moniz, 1649-035 Lisbon, Portugal; email: luiscosta.p{at}mail.telepac.pt
PURPOSE: The objective assessment of bone metastases is currently based on serial changes in skeletal survey. We performed a prospective study to determine whether a correlation exists between the biochemical markers of bone turnover and x-ray evaluation of bone metastases in patients with or without bisphosphonate therapy, and whether bone markers are influenced by extraskeletal disease. PATIENTS AND METHODS: Patients with either bone or extraskeletal metastases were consecutively enrolled and World Health Organization response criteria were applied for both bone and extraosseous disease every 3 to 4 months. Serum levels of bone-specific alkaline phosphatase (B-AP) and C-telopeptide (ICTP) and urine levels of N-telopeptide (NTX) were measured monthly. The data were analyzed by generalized estimation equation regression. RESULTS: We studied 97 patients with bone metastases (52 also with extraskeletal metastases) and 26 with extraosseous disease only. Median time on study was 153 days, and 281 objective evaluations (171 in bone) were performed. With bisphosphonates (49 patients receiving pamidronate and three receiving clodronate), percent change from levels without therapy was 47% for NTX (P < .001) and 69% for B-AP (P = .008). With disease progression in bone, percent change from mean levels during stable disease was 152% for NTX (P < .001) and 144% for ICTP (P < .001) regardless of bisphosphonate therapy. NTX had the highest positive predictive value (71%) for the diagnosis of bone metastases progression. Extraskeletal disease had no significant effect on bone markers. CONCLUSION: Urinary NTX may be a valuable bone marker to assess the antiresorptive effect of bisphosphonate therapy and to evaluate the progression of bone metastases.
METASTATIC SPREAD of cancer to bone is a common occurrence in many malignancies. Necropsy studies1 have shown a high incidence of bone metastases in the most frequent cancers: breast, 73%; prostate, 68%; lung, 36%; uterine cervix, 50%; thyroid, 42%, and vesical, 40%. In a prospective study2 of breast cancer patients, bone was the organ most frequently affected by metastases (69% of 485 patients). In many cases, bone is the first and only site of disease for a considerable period of time. Breast cancer, prostate cancer, and multiple myeloma are good examples of this. In patients with these diseases, an accurate assessment of how bone metastases respond to treatment is particularly important. However, the objective assessment of bone metastases is difficult to achieve, because radiologic changes observed during the clinical course are generally slow and sometimes equivocal. Nevertheless, changes seen on serial radiographs remain the standard method of evaluating bone metastases status (World Health Organization [WHO] and International Union Against Cancer criteria). The radiologic presentation of bone metastases is not uniform and can be classified into different x-ray patterns (blastic, lytic, or mixed). This observation is an important factor contributing to the difficulty in assessing the status of bone metastases. In recent years, new biochemical markers of the remodeling process in bone have become available. Among the new biochemical markers of bone turnover, bone-specific alkaline phosphatase (B-AP) reflects bone formation, while the peptide-bound collagen type I cross-links N-telopeptide (NTX) and C-telopeptide (ICTP) represent degradative products of mature collagen and reflect bone resorption. These biochemical markers of bone turnover seem to correlate with the presence3 and the extent4 of skeletal metastases and have been used, in clinical trials, to assess the effects of antiresorptive bisphosphonate therapy in patients with tumor-induced hypercalcemia.5 Since bisphosphonates have achieved an established role in the treatment of osteolytic bone metastases in breast cancer6 and in multiple myeloma,7 other studies analyzed the utility of bone markers to assess the response to the antiresorptive effect of bisphosphonates.8,9 Since conventional tumor markers are not bone-specific and in many cancer patients are not sensitive enough to predict metastatic disease evolution, we performed a prospective study to determine whether a correlation exists between the bone markers and bone metastases status as determined by standard radiologic criteria and to evaluate the effects of bisphosphonate therapy on the bone markers. The secondary objectives of the study were to analyze the possible influence of extraskeletal disease on the bone markers and to perform a comparison between tumor markers and bone markers in predicting bone metastases status.
Patients and Study Design Consecutive cancer patients with documented advanced disease in bone (radiologic evidence) or assessable extraskeletal disease were included in this study. Since the objectives of the study were to examine whether a correlation exists between bone markers and bone metastases status and to analyze the possible influence of extraskeletal disease on the interpretative use of the bone markers, three times as many patients with bone metastases than patients with extraskeletal disease only were included. In each group, patients were enrolled consecutively. Patients were evaluated for the presence of osseous and extraosseous metastases at baseline and followed up prospectively with an objective evaluation of metastatic disease performed every 3 or 4 months. Patients with bone metastases underwent a bone survey (CT scan for vertebral disease) and bone scan (technetium-99) at baseline and every 3 or 4 months. The x-rays and computed tomography scans were evaluated and reported by an independent viewer, and WHO response criteria were applied. At baseline, all patients with documented bone metastases were classified according to x-ray pattern (lytic, blastic, or mixed) and extent of bone metastases (number of skeletal segments involved). The pelvis, spine, skull, long bone, and ribs plus scapula were considered as individual segments. In those patients who had extraosseous disease at baseline, an objective evaluation of metastatic sites was performed every 3 or 4 months and WHO response criteria were applied. When patients had only bone metastases at baseline, the objective assessment of extraskeletal sites was repeated at the end of the follow-up period. Similarly, those patients who had extraosseous disease only at baseline had a repeat bone scan and skeletal survey at the end of follow-up. Patients were assessable if they had at least two consecutive objective assessments of metastatic sites of disease. Patients were planned to be on study until they had a modification of the status of metastatic disease. During the study, a monthly clinical assessment included an objective clinical examination, assessment of performance status according to WHO criteria, an analgesic score for bone pain using the Radiation Therapy Oncology Group scale,10 and treatment evaluation (anticancer treatment and bisphosphonate therapy). Bisphosphonate therapy consisted of two possible regimens: pamidronate 90 mg intravenously monthly or oral clodronate 1,600 mg daily. Treatment modifications during the study were performed according to the usual medical oncology criteria, ie, progression of disease under previous treatment, serious toxicity, and acute medical complications requiring changes in oncology treatment. An evaluation of skeletal complications (pathologic fractures, need for radiation to bone, spinal cord compression, and episodes of hypercalcemia) was also performed during the study. The ethics committee of the Hospital Santa Maria at Lisbon approved patients and study design. A witnessed oral informed consent was obtained from each patient.
Biochemical Analysis Sera for B-AP and ICTP determinations were obtained from fasting morning blood samples. ICTP was measured in serum using immunoradiometric assay reagents obtained from DiaSorin (Stillwater, MN). The monoclonal antibody in this assay recognizes a large ICTP fragment of human bone collagen. The assay detection limit is 0.4 µg/L with an average imprecision of 8%. The reference range for ICTP is 2.5 to 4.0 µg/L. B-AP was measured in serum using immunoradiometric assay reagents from Hybritech Inc (San Diego, CA). The reference range is 8 to 12.5 µg/L. Tumor markers were determined monthly in those patients whose levels were initially elevated. We measured carcinoembryonic antigen (CEA) and CA 15.3 in breast cancer patients, prostate-specific antigen (PSA) in prostate cancer patients, and CEA and CA 19.9 in gastrointestinal cancer patients. Serum calcium and creatinine were measured monthly in all patients.
Statistical Analysis We first analyzed the relationship between bone markers, disease progression in bone, and bisphosphonate administration. The factors of interest were disease progression, bisphosphonate administration, and the interaction of the two. Using GEE regression with the logarithm of each bone marker as dependent variables, we first tested the interaction of progression and bisphosphonate administration on marker levels. If this test was nonsignificant (P > .10), the interaction term was dropped from the model and the independent effects of disease progression in bone and administration of bisphosphonates were tested. In order to report the difference of bone marker level between stable disease and progression in bone, and between no administration and administration of bisphosphonates, the regression coefficients and 95% confidence limits were exponentiated. The resulting values represent the percent change in the marker level between levels of the explanatory variables. In the analysis of tumor markers, a nominal variable with the type of marker (CEA, CA 15.3, CA 19.9, and PSA) was included in the model as covariate. The exponentiated regression coefficients represent the percent change in tumor marker levels, averaged over all types of tumor markers. To evaluate whether the results of the previous analysis had clinical application, we determined the sensitivity, specificity, and negative and positive predictive values of disease progression in bone for each bone marker. In order to compare the diagnostic efficiency between the markers, receiver operating characteristic (ROC) curves were constructed for each biochemical marker. In these analyses, an allowance was made for the effect of bisphosphonate therapy on bone marker levels using the percent decrease in marker level under bisphosphonate therapy, as estimated by the primary analysis. To investigate the relationship between bone markers and extraosseous disease progression, a GEE regression was conducted in the total sample of 123 patients. The model included terms for disease progression in bone, extraosseous progression, and bisphosphonate administration. Confirmatory subanalyses were conducted in the group of 26 patients without bone metastases and in the group of 53 patients with both osseous and extraosseous metastases. Statistical calculations were performed with Stata 6.0 software (Stata Corp, College Station, TX).
One hundred forty-six cancer patients were included in this study, but only 123 patients had at least two consecutive objective evaluations of metastatic disease and were therefore assessable. The median age was 64 years (range, 30 to 92 years); 82 patients were female. Sixty-two patients had breast cancer, 22 had gastrointestinal cancer, 15 had prostate cancer, six had lung cancer, five had renal cancer, five had unknown primary carcinomas, two had urothelial carcinomas, two had head and neck cancer, two had neuroendocrine tumors, and two had multiple myeloma. Ninety-seven patients had documented bone metastases by x-ray (52 had extraskeletal metastases also and 45 had bone metastases only). The x-ray pattern of bone metastases was shown to be lytic in 44 patients, mixed in 31, and blastic in 22. Twenty-six patients had extraosseous disease only. The baseline characteristics of the patients with bone metastases and with extraosseous disease only are presented in Tables 1 and2.
The 123 assessable patients were observed for a median time of 153 days (range, 22 to 589 days). During the study, an objective evaluation of tumor response (metastases status) was done on 281 occasions (171 in bone, 110 in extraosseous metastases). On the basis of the site of metastases, three groups could be identified. (A) Forty-five patients with bone metastases only (27 women and 18 men; median age, 65.9 years) had 79 objective evaluations of bone metastases status after baseline (complete response, n = 1; partial response, n = 5; stable disease, n = 40; and disease progression, n = 33). (B) Fifty-two patients with both osseous and extraosseous metastases (39 women and 13 men; median age, 59.3 years) had 92 objective evaluations of bone metastases status after baseline (partial response, n = 3; stable disease, n = 51; disease progression, n = 38) and 77 objective evaluations of extraosseous disease after baseline (complete response, n = 3; partial response, n = 11; stable disease, n = 22; and disease progression, n = 41). (C) Twenty-six patients with extraosseous metastases only (16 women and 10 men; median age, 61 years) had 33 objective evaluations of extraosseous disease response after baseline (complete response, n = 1; partial response, n = 3; stable disease, n = 12; disease progression, n = 17). Two patients in this group developed bone metastases while on study. The differences in the response rate (both for partial and complete response) in group B should be attributed to the insensitivity of radiographic determination of bone response, since there was no difference in systemic treatment. Ninety patients received chemotherapy (66 patients with bone metastases and 24 patients with extraosseous disease only), 49 received hormone therapy (46 of the bone metastases group of patients), and 27 patients with bone metastases required radiation therapy to bone. During the study, 52 patients with bone metastases received bisphosphonate therapy. Forty-nine were treated with pamidronate 90 mg intravenously on a monthly basis, and three patients were treated with oral clodronate 1,600 mg daily. Table 3 gives the results of the analysis of the association of bone markers and tumor markers with disease progression in bone and with bisphosphonate administration in 97 patients with bone metastases at baseline. NTX and ICTP were strongly associated with disease progression in bone. The average percent change from the level of the bone marker during stable disease was 152% for NTX (P < .001) and 144% for ICTP (P < .001). Tumor markers were also associated with disease progression in bone (P = .03), with an average percent change from levels in stable disease of 156%. B-AP was not shown to be associated with disease progression in bone (P = .20).
Both NTX and B-AP levels were reduced in the presence of bisphosphonate therapy (70 objective assessments in 52 patients with bone metastases), with an average percent change of 47% (P < .001) and 69% (P = .008), respectively, from the levels observed before bisphosphonate administration. ICTP and tumor markers were not affected by bisphosphonate therapy (P = .88 and P = .54, respectively). There was no evidence of significant interaction between bisphosphonate administration and bone metastases progression on the level of the bone marker NTX. Therefore, the two effects may be taken as independent, in the sense that the expected change in bone marker levels after disease progression can be safely assumed to be constant, regardless of bisphosphonate administration. To assess the robustness of the previous analysis and the eventual effect of external factors in the estimates, we conducted a further analysis controlling for patient variables, since bone marker levels may be influenced by a number of factors. Using multiple linear regression with stepwise backward elimination, we tested the association between each bone marker and the following variables collected at baseline: age, sex, presence of extraosseous metastases, x-ray pattern of bone metastases, number of skeletal segments involved, performance status, analgesic score for bone pain, and breast cancer. The GEE regression was repeated but with the retained factors (sex, age, number of skeletal segments, and performance status) entered as covariates. In a second analysis, the same regression of the primary analysis was repeated in the subsample of patients with breast cancer. In both cases, the results were essentially the same as in the primary analysis (data not shown). In order to assess the clinical applicability of the previous results in the diagnosis of bone disease progression, the diagnostic indexes of each marker were calculated. The results are presented in Table 4, with cutoff values set for each marker at the average percent change with disease progression (comparing bone and tumor marker levels at the time that patient was classified to have disease progression in bone to the levels of the same markers in the previous objective evaluation performed). NTX had the highest sensitivity, specificity, and predictive values of all the markers evaluated. Specifically, the predictive value for bone disease progression of 152% or greater increase in NTX is 71%. This result is confirmed by the inspection of the ROC curves for the markers associated with disease progression (NTX, ICTP, and tumor markers), where NTX shows the largest area under the curve (Fig 1). According to these curves, the most likely cutoff point of NTX for disease progression in bone is an increase to 130% from previous levels, which has a sensitivity of 70%, a specificity of 80%, a positive predictive value of 72%, and a negative predictive value of 79%.
The results of the analysis of the association of tumor markers with disease progression, both in bone and extraosseous disease, and with bisphosphonate administration, in the entire sample of 123 assessable patients, are presented in Table 5. Only the tumor markers were significantly increased with extraosseous disease progression, with an average percent change from stable disease of 166% (P = .02). Progression of extraskeletal disease had no significant effect on NTX levels (average percent change from stable disease of 99%, P = .91) or B-AP levels (average percent change from stable disease of 94%, P = .61). For ICTP, the average percent change from stable disease was 118%, but this value did not reach statistical significance (P = .13). These results are the same when the analyses were conducted in the group of patients with extraosseous disease only at baseline and in the group of patients with extraosseous disease and bone disease at baseline.
Thirty-four skeletal related events (SREs) were reported during the study. Ten SREs occurred at the time of bone metastases diagnosis (first objective evaluation). The SREs observed were radiation therapy to bone (n = 20), pathologic fracture (n = 13), and spinal cord compression (n = 1). No episode of hypercalcemia was reported during the study.
During the observation period, there was no statistically significant association between the value of the bone markers and the occurrence of SREs. There was a statistically significant association between SREs and the following baseline characteristics: worse performance status (P = .002) and analgesic bone pain score
Metastatic disease to bone disrupts the resorption/formation balance and in many circumstances favors the increase of bone resorption. Bone destruction in the presence of cancer cells (tumor osteolysis) is believed to be mediated by the activation of osteoclasts.11 Bone degradation is a very important phenomenon, even when there is osteoblast reaction and bone metastases present on x-ray as blastic lesions. Several new biochemical markers of bone turnover have been developed. Pyridinium cross-links and associated telopeptides are among the new bone resorption markers and represent fragments of cross-linking amino acid derivatives that stabilize the collagen type I fibrils in bone. These cross-linking compounds, known as the pyridinium cross-links, result from posttranslational modification of lysine and hydroxylysine in bone. Two nonreducible pyridinium compounds are present in bone collagen, namely, pyridinoline (PYD) and deoxypyridinoline (DPD). In the process of collagen breakdown, PYD and DPD are released into the circulation both in a free state and bound to peptide. Approximately 60% to 65% of the cross-links present in urine are bound to peptide. Using monoclonal antibodies, it is now possible to recognize and measure by immunoassay the telopeptide fragments to which the pyridinium cross-links are attached in bone collagen and to measure the specific peptide fragment from either the amino-terminal telopeptide (NTX) or carboxy-terminal telopeptide of the bone collagen fibril. In the case of the carboxy-terminal peptide, two monoclonal antibodies have been developed as markers of collagen breakdown (ICTP and cross-laps). The standard method of bone metastases evaluation (skeletal survey) is less sensitive than other radiologic methods for assessing extraskeletal disease, as it was confirmed on this study (patients with both bone and extraskeletal disease had differences in the response rate to the same treatment). When bone metastases are the only site of clinically detectable disease, radiographic evaluation of bone metastases is the only objective criterion to support treatment decisions. Our study addressed the question of whether the newer biochemical markers of bone turnover are of value in predicting the status of bone metastases. We performed a prospective study with independent objective evaluation of bone metastases and extraskeletal disease in patients with a variety of cancers and different radiologic patterns of bone metastases. Since bisphosphonate therapy can significantly decrease the bone markers, we tested whether the value of bone markers in predicting bone metastases progression could be affected by bisphosphonate therapy. We found that urine NTX level is a specific marker for bone disease and better than conventional tumor markers in predicting progression of metastatic disease in bone. NTX was increased on average by 50% (95% confidence interval, 30% to 70%) from the previous objective evaluation at the time of disease progression in bone. As observed by other investigators, we found that urine NTX levels decreased with bisphosphonate therapy. However, when bone metastases progressed while the patient was receiving bisphosphonate therapy, NTX still increased by the same amount (mean 50%) between two consecutive objective evaluations. We conclude that NTX seems to be a valid bone resorption marker to predict bone metastases progression in patients with or without bisphosphonate therapy. Inspection of the ROC curve for NTX (Fig 1) shows that if we use a 30% increase in NTX levels as the cutoff instead of 50% to determine disease progression in bone, we increase the sensitivity of the test to 70%, while the specificity only decreases from 85% to 80%, which suggests that an increase over 30% in the urine level of NTX could be indicative of bone metastases progression. The application of these results can be clinically useful. We believe that serial determinations of urine NTX should not replace the skeletal survey as the standard method of evaluating bone metastases status yet, but they could be used to indicate the need to repeat a new radiographic evaluation and to help clinicians overcome some of the difficulties raised by the equivocal interpretation of radiologic changes in bone metastases. In this study, ICTP also increased significantly (mean 44%) with disease progression in bone. However, we did not find a significant decrease in ICTP levels with bisphosphonate therapy. It is possible that ICTP represents a bone collagen product derived from an osteoclast-independent mechanism of bone degradation and therefore is not influenced by bisphosphonate therapy. On the other hand, it is possible that serum ICTP represents two different sources of collagen degradation: bone and soft tissue. In fact, our results show that ICTP levels increased significantly with bone metastases progression, and there was a trend to increase with extraskeletal disease progression as well. These results suggests that future studies of new biochemical indicators of bone turnover to monitor bone metastases should take into consideration the effect of bisphosphonate therapy and the possible influence of extraskeletal disease on the markers. NTX and C-telopeptides have been used to assess patient response to bisphosphonates in tumor-induced hypercalcemia5 and to evaluate the analgesic effect in patients with bone metastases.12 There is preliminary evidence that breast cancer patients with bone metastases who achieve normalization of NTX levels after pamidronate therapy have fewer pathologic fractures.13 In the present study, there was no significant association between decrease in NTX and a reduction in SREs during the time of study. This could be due to the shorter time of observation of this study or to the fact that patients were not randomized to receive or not bisphosphonate therapy. In a small cohort of breast cancer patients with bone metastases (n = 37), investigators found that urinary NTX levels were a valid marker to predict response to anticancer treatment at 4 months of follow-up.14 In this study, patients were receiving oral pamidronate, a regimen that it is now not considered effective. In another study with breast cancer patients receiving hormone therapy, urinary PYD and DPD increased significantly in nonresponders (P = .03).15 ICTP was found to be significantly elevated in the serum of prostate cancer patients with bone metastases,16 and in another report, the level of ICTP at baseline correlated with a poor response to chemotherapy in multiple myeloma patients.17 However, other investigators more recently reported that ICTP was significantly elevated, and an important prognostic factor, in ovarian cancer patients without bone metastases.18 ICTP can be of use as a marker of collagen breakdown in cancer patients, but it is not specific for bone. The bone formation marker B-AP, in our study, did not add any further information to the resorption markers regarding the interpretation of bone metastases evolution. In conclusion, a tissue-specific marker for bone collagen (NTX) can reflect bone metastases evolution and became a useful marker for monitoring the treatment of bone metastases. NTX had the highest diagnostic accuracy for bone metastases status when compared with other bone markers (B-AP and ICTP) and tumor markers.
1. Galasko CSB: The anatomy and pathways of skeletal metastases, in Weiss L, Gilbert AH (eds): Bone Metastasis. Boston, MA, Hall, 1981, pp 49-63 2. Coleman RE, Rubens RD: The clinical course of bone metastases from breast cancer. Br J Cancer 55: 130-135, 1987
3.
Demers LM, Costa L, Chinchilli V, et al: Biochemical markers of bone turnover in patients with metastatic bone disease. Clin Chem 41: 1489-1494, 1995 4. Demers LM, Costa L, Lipton A: Biochemical markers and skeletal metastases. Cancer 88: 2919-2926, 2000 (suppl)[CrossRef][Medline]
5.
Vinholes J, Guo CY, Purohit OP, et al: Evaluation of new bone resorption markers in a randomized comparison of pamidronate or clodronate for hypercalcemia of malignancy. J Clin Oncol 15: 131-138, 1997
6.
Hortobagyi GN, Theriault R, Poter L, et al: Efficacy of pamidronate in reducing skeletal complications in patients with breast cancer and lytic bone metastases. N Engl J Med 335: 1785-1791, 1996
7.
Berenson JR, Lichtenstein A, Poter L, et al: Efficacy of pamidronate in reducing skeletal events in patients with advanced multiple myeloma. N Engl J Med 334: 488-493, 1996 8. Body JJ, Dumon JC, Gineyts E, et al: Markers of bone resorption in breast cancer-induced osteolysis before and after bisphosphonates. Br J Cancer 75: 408-412, 1997[Medline] 9. Withold W, Arning M, Schwarz M, et al: Monitoring of multiple myeloma patients by simultaneously measuring markers substances of bone resorption and formation. Clin Chim Acta 296: 21-30, 1998 10. Tong D, Gillick L, Hendrickson FR: The palliation of symptomatic osseous metastases: Final results of the study by the Radiation Therapy Oncology Group. Cancer 50: 839-839, 1982 11. Taube T, Elomaa I, Blomqvist C, et al: Histomorphometric evidence for osteoclast-mediated bone resorption in metastatic breast cancer. Bone 15: 161-166, 1994[Medline]
12.
Vinholes J, Purohit OP, Abbey ME, et al: Relationships between biochemical and symptomatic response in a double-blind randomised trial of pamidronate for metastatic bone disease. Ann Oncol 8: 1243-1250, 1997 13. Lipton A, Demers LM, Curley E, et al: Markers of bone resorption in patients treated with pamidronate. Eur J Cancer 34: 2021-2026, 1998 14. Vinholes J, Coleman RE, Lacombe D, et al: Assessment of bone response to systemic therapy in a EORTC trial: Preliminary experience with the use of collagen cross-link excretion. Br J Cancer 80: 221-228, 1999[CrossRef][Medline] 15. Walls J, Assiri A, Howell A, et al: Measurement of urinary collagen cross-links indicate response to therapy inpatients with breast cancer and bone metastases. Br J Cancer 80: 1265-1270, 1999[CrossRef][Medline] 16. Koga H, Naito S, Koto S, et al: Use of bone turnover marker, pyridinoline cross-linked carboxyterminal telopeptide of type I collagen (ICTP), in the assessment and monitoring of bone metastasis in prostate cancer. Prostate 39: 1-7, 1999[CrossRef][Medline] 17. Elomaa I, Virkkuunen P, Risteli L, et al: Serum concentration of the cross-linked carboxyterminal telopeptide of type I collagen (ICTP) is a useful prognostic indicator in multiple myeloma. Br J Cancer 66: 337-341, 1992[Medline]
18.
Santala M, Simojoni M, Risteli J, et al: Type I and type III collagen metabolites as predictors of clinical outcome in epithelial ovarian cancer. Clin Cancer Res 5: 4091-4096, 1999 Submitted January 22, 2001; accepted September 18, 2001.
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2002 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
3 (P = .03). 
