Departments of 1 Oncology, Radiology and Clinical Immunology, and 2 Medical Sciences, Uppsala University, Akademiska sjukhuset, Uppsala; 3 Division of Research and Development, AB Sangtec Medical, Bromma, Sweden
Received 9 November 2001; revised 4 February 2002; accepted 6 March 2002
![]() |
Abstract |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
To evaluate the reliability and validity of serum carcinoembryonic antigen (CEA), tissue polypeptide-specific antigen (TPS), vascular endothelial growth factor (VEGF) and basic fibroblast growth factor (bFGF) in monitoring palliative chemotherapy in advanced colorectal cancer (ACRC).
Methods:
Serum was prospectively collected from 87 patients with ACRC treated with first-line 5-fluorouracil and leucovorin before and 2, 4 and 10 weeks after induction.
Results:
Eight patients had normal baseline TPS levels, and these patients had a favourable outcome with prolonged survival and a higher rate of objective responses than patients with elevated TPS levels. At 10 weeks, all responders had a decreasing TPS value. The sensitivity for a decrease of >25% using TPS was 83% and 86% for objective and subjective responses, respectively, and the specificity was 65% and 72%, respectively. CEA had, in the same setting, a sensitivity of 45% and 46%, respectively, and the specificity was 88%. VEGF was elevated in 54% of the patients and bFGF in 15% of the patients. The VEGF values decreased during therapy in 94% of the patients, but the changes in serial VEGF values did not correlate with survival or response. Tumour markers used together did not enhance the predictive values of TPS alone.
Conclusions:
Repeated measurements of CEA, VEGF and bFGF in serum are of limited value in monitoring chemotherapy in ACRC. TPS seems to be of greater interest, but does not predict exactly which patients are going to have a positive outcome of palliative chemotherapy.
Key words: bFGF, CEA, colorectal carcinoma, palliative chemotherapy, TPS, VEGF
![]() |
Introduction |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
The tumour marker carcinoembryonic antigen (CEA) [5] has frequently been used for the last 30 years in a variety of clinical situations, including monitoring of chemotherapy treatment in ACRC, although a proper validation of the use of this marker for monitoring has never been performed [68].
Tissue polypeptide-specific antigen (TPS) has been reported to be a useful tumour marker in various malignant diseases [914] and as a predictor for response in monitoring chemotherapy in various advanced gastrointestinal carcinomas [11, 15]. TPS measures a specific epitope structure of soluble fragments of human cytokeratin 18 in the circulation [16].
Angiogenesis is a rapidly expanding research field, increasing the understanding of tumour growth and the metastatic process. Measurements of angiogenic peptides such as vascular endothelial growth factor (VEGF) and basic fibroblast growth factor (bFGF) in serum have been presented in several pilot studies, and indicate clinical impact as markers for staging, prognosis [1724] and for prediction of outcome in the palliative setting in gastrointestinal cancer [24].
In 1992, we began to prospectively collect serum prior to all chemotherapy courses in patients with ACRC. The following research questions were raised. (i) Can tumour markers prior to therapy predict outcome of palliative chemotherapy? (ii) Can tumour markers predict response early during therapy before any imaging technique may reveal a change in tumour size? (iii) Is it necessary to use imaging for response evaluation, if tumour markers turn out to have the same capability to predict prolonged survival and palliation of symptoms? Even if radiological assessment is the gold standard for response evaluation, it is still a surrogate marker for the main objectives of palliative chemotherapy, namely to prolong survival and to relieve/delay the occurrence of tumour-related symptoms. Here we report the results of measuring CEA, TPS, VEGF and bFGF according to tumour response and subjective outcome.
![]() |
Materials and methods |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
|
Assessment of tumour response
Objective and subjective response evaluations were performed according to study protocols [25, 26] every second month, i.e. after every fourth leucovorin treatment. Objective tumour response was assessed according to standard criteria. A complete response (CR) or partial response (PR) and stable disease (SD4) had to be present at two consecutive evaluations, i.e. with a minimum duration of 4 months. Patients who had a response or stable disease after 2 months and then progressed before the next evaluation were designated as stable disease for 2 months (SD2). In all other instances, progressive disease (PD) was recorded. All responses among patients in the trials have been reviewed independently. All assessments were carried out with computed tomography (CT) scan and/or other radiological examinations. In the following text, when the term objective response is used, it includes CR, PR and SD4. In previous studies from the Nordic group, patients with CR, PR or SD4 have, as a group, shown clinical benefit from the treatment, with prolonged survival and symptom improvement [27].
A subjective response evaluation was performed by the treating doctor at the same time as the objective response evaluation. The subjective evaluation was based on a personal interview. A subjective response (designated improved) was present when the patients symptomatology had improved for at least 4 months with no signs of severe adverse treatment effects. If the patients did not have any symptoms of the disease prior to treatment and at the evaluation, they were designated symptom-free. Symptom-free and improved patients are grouped together under the category improved. Both the objective and the subjective response evaluations and their durations were identical to those used in previous trials [25, 2830].
Serum sampling
Serum was obtained from all patients before chemotherapy. Subsequent sampling was performed prior to all treatment courses, i.e. with an interval of 2 weeks. The serum samples were aliquoted and immediately frozen at 20°C and stored until assayed.
Tumour marker assessment
All tumour markers (see below) were retrospectively analysed at baseline, after 2 and 4 weeks of treatment and at the same time as the first radiographic examination prior to the fifth course (after 10 weeks). Among patients with a normal baseline tumour marker, no calculation of change for this specific marker was performed. Altogether, 923 samples have been analysed. A further 126 samples have not been analysed due to missing samples or insufficient quantity of serum.
CEA measurements
Serum CEA was analysed by a microparticle enzyme immunoassay (MEIA) on an IMx (Abbott Laboratories, Abbott Park, IL, USA). Normal reference interval <3.1 µg/l.
TPS measurements
TPS was measured using TPSTM IRMA (IDL Biotech, Stockholm, Sweden). All samples were run in duplicate. The upper limit of normal values of TPS in serum was defined as 80 U/l.
Angiogenesis factors
A pilot study was performed in 33 patients in this cohort with both VEGF and bFGF. VEGF indicated some correlation with response to therapy, and therefore this marker was enrolled. bFGF did not reveal any correlation to response or to survival. Elevated bFGF values were also only seen in a minority of the patients; therefore, no further determination of bFGF was performed.
VEGF measurements
A quantitative sandwich ELISA technique was used (human VEGF; Quantikine, R&D Systems, Minneapolis, MN, USA). The VEGF concentration in the samples was determined by comparing the optical density of the samples to the standard curve. The lowest detectable value for VEGF was 9 pg/ml. The upper normal limit of VEGF is 500 pg/ml (95th percentile) [17].
bFGF measurements
An ELISA technique was used to measure bFGF (human bFGF; Quantikine R&D). The lowest detectable value for bFGF was 0.25 pg/ml. The upper normal limit was determined to be 7.25 pg/ml [17]. The specificity of the assay has been verified by the manufacturer with many different recombinant human and mouse cytokines.
Statistical analyses
Comparisons of proportions between groups were performed by 2 analyses.
Survival curves were constructed according to the KaplanMeier product-limit method and log rank test was used to test differences. Correlations were studied with Spearmans rank correlation test. The proportional Coxs hazards model was used to evaluate the importance of prognostic variables one by one and in multivariate analyses. Comparison of changes in tumour marker levels in relation to objective and subjective responses were expressed in terms of sensitivity and specificity according to conventional definitions.
![]() |
Results |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
|
|
|
|
|
|
Baseline bFGF values
Only five (15%) of 33 patients had elevated bFGF values at baseline. bFGF values correlated significantly to VEGF but not to any of the other above-mentioned parameters (data not shown). Since only a minority of the patients had elevated bFGF values, further analyses were not performed.
Other markers for prediction of response and survival
There were no statistically significant differences in survival or response with respect to age, gender or primary tumour site (rectum or colon). Performance status correlated significantly to both response and survival, and baseline haemoglobin values to survival (Table 2).
Tumour markers used together
The sensitivity and specificity for response and prolonged survival did not increase when TPS and CEA were used together compared with those values calculated by TPS alone, either in univariate analyses or in multivariate analyses. Because patients with one or two normal baseline tumour marker levels were excluded, fewer patients were used in this group. When either of the tumour markers was used together with subjective response, the sensitivity and specificity for an objective response was not increased compared with subjective response only. As shown in Table 2, subjective and objective responses were highly correlated.
![]() |
Discussion |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
TPS measurements appear to be useful for these purposes and they are clearly superior to CEA measurements. TPS baseline values were not correlated with whether a response occurred or not, but were with better survival, and survival was particularly favourable among patients with normal TPS value. An early TPS evaluation (after 2 and 4 weeks, i.e. after the first and second treatment course) indicated clinical outcome, with sensitivities 75% but with lower specificity. At 10 weeks a TPS value decrease of >25% correlated significantly, in univariate and multivariate analyses, with both objective and subjective responses with sensitivity and specificity of
85% and 70%, respectively, and with survival. The results are in accordance with other studies that have explored any of the above research questions when monitoring patients with breast, lung and various gastrointestinal cancers [11, 15, 31, 32].
Baseline CEA levels, in contrast, did not give any clinically meaningful prognostic or predictive information about response. After 2 and 4 weeks, very few patients showed a decrease of >25%. At 10 weeks, the sensitivity for response was only 45%, although the specificity was higher, at 88%. Thus, a decrease in responding patients was seen considerably more often using TPS than CEA, whereas a decrease in non-responding patients was less frequent using CEA. Although the CEA changes after 10 weeks correlated significantly with response and survival in univariate analyses but not in multivariate analyses, the low sensitivity limits its use clinically. Repeated CEA measurements during therapy have been indicated in several reports as valuable in predicting response. The conclusions about the value of repeated CEA measurements, however, differ considerably between the reports, which often is the case when studies are mostly retrospective and based on small numbers of patients [3341]. Hamm and Cripps [6] could not find any benefit of monitoring therapy with CEA in their own study or in an overview of the collected literature. Two other reviews came to somewhat contradictory conclusions [7, 8]. Similar to our study, Hanke et al. [41] could not show that a CEA or a CA 19-9 rise after 6 weeks of 5-FUleucovorin treatment were appropriate for recording disease progression. PD was rarely seen among 43 patients if CEA levels were falling (specificity 99% and 88% in this study). Several authors have suggested usefulness in indicating PD after a period of treatment [8, 37, 39, 41]. The American Society of Clinical Oncology (ASCO) guidelines [42] from 1996 stated that existing data were not sufficient to recommend routine use of CEA alone for monitoring response to chemotherapy. However, if no alternative test was available to detect a response, it was suggested that CEA should be assayed at the commencement of treatment for metastatic disease and every 23 months during active therapy.
This study of homogeneously treated patients according to prospective protocols could not detect any clinically meaningful information from monitoring with serial CEA measurements. Combining CEA and other tumour markers or clinical parameters did not add further information. Thus, the ASCO guidelines are still valid and efforts should be made to objectively verify tumour extent prior to therapy, and not to rely solely upon an elevated CEA level. TPS, rather than CEA, may better separate responding from non-responding patients if therapy is initiated without measurable or otherwise evaluable disease.
Only about half of our patients had elevated VEGF values and 15% had elevated bFGF values. There was no correlation between the levels of these angiogenetic peptides and survival or response, either at baseline or at an early evaluation. Therefore, this study has not been able to confirm the results from small pilot studies claiming potentially valuable information from VEGF and bFGF in the palliative setting among patients with ACRC [1821, 23, 24].
Of interest is the striking finding that several patients, regardless of outcome, have a rapid and clear decrease in TPS and an even more definite decrease in VEGF after initiation of chemotherapy. The explanation of this is not clear, but needs to be explored further. We have previously speculated that the decrease in TPS levels indicates some tumour cell death in most patients (75%) with gastrointestinal cancer, but this cell death is not sufficient to give either a subjective response seen in 40% to 50% of the patients or an objective response in 10% to 25%, depending upon primary tumour site [15]. In order to qualify as a response, it must also have a minimum duration, here 4 months. Concerning VEGF, Verheul and Pinedo [43] found that the level correlated highly with platelet count. They did not find any direct relationship with tumour burden. Most cytotoxic agents, and 5-FU, influence platelet counts. The mechanisms behind the decrease of TPS and VEGF serum levels may thus be different.
This study was prospectively designed to be able to address the question of the practical value of measuring tumour markers. Aliquoted serum was analysed retrospectively to test the most promising markers when complete follow-up of all patients was available. The treatment given was one of many 5-FUleucovorin schedules that were the reference treatment up until the past year, when combinations with either irinotecan or oxaliplatin have shown superior results, although at the expense of higher toxicity and costs (see Ragnhammar et al. [1] for a review). However, we have no reason to believe that the tumour marker changes seen are not relevant also when using other 5-FUleucovorin regimens, with or without new drugs.
If the TPS results we obtained hold true in a larger trial, what is then the practical use of measuring TPS before and early during palliative chemotherapy for ACRC? Even if the baseline level has prognostic information, a high value does not preclude a reasonable chance of a favourable response. A normal or a very low baseline value indicates a very long median survival, and treatment initiation could probably safely be postponed until increasing levels, or other signs of PD, are seen, if this is agreed upon.
If toxicity to the first one or two courses is considered high by the patient, a clear decrease in TPS values after 2 or 4 weeks may indicate that a response is reasonably likely, and the patient may more easily accept the toxicity. Conversely, if no decrease is seen, the probability of a response is very low, and treatment could be terminated. If, on the other hand, toxicity is low, the sensitivity and specificity are not high enough to allow any recommendation of therapy change at this early time point. After 2 months of treatment, a lack of a decrease or an increase of TPS levels is a strong indicator of no response, and a confirming imaging investigation could be avoided. Similarly, with a clear decrease in values together with unequivocal symptomatic benefit and acceptable toxicity, the costs of imaging could be saved, and therapy continued. The cost of TPS measurement is low (Swedish crowns, SEK 125) compared with, for example, a CT scan of the liver (SEK 2500). Finally, it must be stressed that a complete concordance between imaging and changes in serum markers levels can never be reached. When the above discussions were retrospectively applied to the patient histories, we noticed several patients where knowledge of the TPS values would have influenced the practical care of the patients or being cost saving. The same rarely happened using CEA.
Conclusions
The results in this study indicate that repeated measurements of TPS levels during therapy can be of clinical relevance, primarily as a marker of lack of response. In this context, it is clearly superior to the commonly used CEA. We strongly question the wide use of CEA, both in clinical trials and for routine use, when monitoring 5-FU-based chemotherapy in patients with ACRC. VEGF and bFGF did not provide any relevant clinical information. However, sensitivity and specificity figures even for TPS are probably not high enough to be of definite relevance in an individual patient. In this context, so-called objective response assessments are not perfect reflectors of true patient benefit, symptom influence and survival prolongation [27, 44, 45]. Thus, there continues to be a desire to develop new methods for early measurement of anti-tumour effects of chemotherapy. Many asymptomatic patients do not have easily measurable disease, but still desire to be actively treated due to the generally poor prognosis. In these patients, changes in TPS levels, if elevated, appear to be of value to guide the length of treatment. If TPS is to be used in clinical practice, studies to confirm our results are warranted.
![]() |
Acknowledgements |
---|
![]() |
Footnotes |
---|
![]() |
References |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
2. Assersohn L, Norman A, Cunningham D et al. Influence of metastatic site as an additional predictor for response and outcome in advanced colorectal carcinoma. Br J Cancer 1999; 79: 18001805.[ISI][Medline]
3. Graf W, Bergstrom R, Pahlman L, Glimelius B. Appraisal of a model for prediction of prognosis in advanced colorectal cancer. Eur J Cancer 1994; 4: 453457.
4. Köhne C-H, Cunningham D, Di Costanzo F et al. Clinical determinants of survival in patients with 5-fluorouracil-based treatment for metastatic colorectal cancer: results of a multivariate analysis of 3825 patients. Ann Oncol 2002; 13: 308317.
5. Gold PF. Specific carcinoembryonic antigens of the human digestive tract. J Exp Med 1965; 122: 467481.[ISI][Medline]
6. Hamm CM, Cripps C. Carcinoembryonic antigen in metastatic colorectal cancer. Clin Invest Med 1998; 21: 186191.[ISI][Medline]
7. Mitchell EP. Role of carcinoembryonic antigen in the management of advanced colorectal cancer. Semin Oncol 1998; 25: 1220.
8. Macdonald JS. Carcinoembryonic antigen screening: pros and cons. Semin Oncol 1999; 26: 556560.[ISI][Medline]
9. Polito M, Minardi D, Recchioni A et al. Serum markers for monitoring of prostatic carcinoma. Prostate 1997; 33: 208216.[ISI][Medline]
10. Rebhandl W, Paya K, Felberbauer FX et al. Tissue polypeptide-specific antigen (TPS) in pediatric malignancies. Anticancer Res 1997; 17: 28652868.[ISI][Medline]
11. Kornek G, Schenk T, Raderer M et al. Tissue polypeptide-specific antigen (TPS) in monitoring palliative treatment response of patients with gastrointestinal tumours. Br J Cancer 1995; 71: 182185.[ISI][Medline]
12. Mishaeli M, Klein B, Sadikov E et al. Initial TPS serum level as an indicator of relapse and survival in colorectal cancer. Anticancer Res 1998; 18: 21012105.[ISI][Medline]
13. Zygmunt A, Madry R, Markowska J, Fischer Z. Estimation of the usefulness of neoplastic markers TPS and CA 125 in diagnosis and monitoring of ovarian cancer. Eur J Gynaecol Oncol 1999; 20: 298301.[ISI][Medline]
14. Abbasciano V, Sartori S, Trevisani L et al. Neuron-specific enolase, thymidine kinase, and tissue polypeptide-specific antigen in diagnosis and response to chemotherapy of small-cell lung cancer. Cancer Detect Prev 1999; 23: 309315.[ISI][Medline]
15. Glimelius B, Hoffman K, Einarsson R et al. Monitoring palliative chemotherapy in advanced gastrointestinal cancer using serial tissue polypeptide specific antigen (TPS) measurements. Acta Oncol 1996; 35: 141148.
16. Rydlander L, Ziegler E, Bergman T et al. Molecular characterization of a tissue-polypeptide-specific-antigen epitope and its relationship to human cytokeratin 18. Eur J Biochem 1996; 241: 309314.[Abstract]
17. Brattstrom D, Bergqvist M, Larsson A et al. Basic fibroblast growth factor and vascular endothelial growth factor in sera from non-small cell lung cancer patients. Anticancer Res 1998; 18: 11231127.[ISI][Medline]
18. Dirix LY, Vermeulen PB, Hubens G et al. Serum basic fibroblast growth factor and vascular endothelial growth factor and tumour growth kinetics in advanced colorectal cancer. Ann Oncol 1996; 7: 843848.[Abstract]
19. Davies MM, Jonas SK, Kaur S, Allen-Mersh TG. Plasma vascular endothelial but not fibroblast growth factor levels correlate with colorectal liver mestastasis vascularity and volume. Br J Cancer 2000; 82: 10041008.[ISI][Medline]
20. Dirix LY, Vermeulen PB, Pawinski A et al. Elevated levels of the angiogenic cytokines basic fibroblast growth factor and vascular endothelial growth factor in sera of cancer patients. Br J Cancer 1997; 76: 238243.[ISI][Medline]
21. Fujisaki K, Mitsuyama K, Toyonaga A et al. Circulating vascular endothelial growth factor in patients with colorectal cancer. Am J Gastroenterol 1998; 93: 249252.[ISI][Medline]
22. Kumar H, Heer K, Lee PW et al. Preoperative serum vascular endothelial growth factor can predict stage in colorectal cancer. Clin Cancer Res 1998; 4: 12791285.[Abstract]
23. Takeda A, Shimada H, Imaseki H et al. Clinical significance of serum vascular endothelial growth factor in colorectal cancer patients: correlation with clinicopathological factors and tumor markers. Oncol Rep 2000; 7: 333338.[ISI][Medline]
24. Hyodo I, Doi T, Endo H et al. Clinical significance of plasma vascular endothelial growth factor in gastrointestinal cancer. Eur J Cancer 1998; 34: 20412045.[ISI][Medline]
25. Glimelius B. Biochemical modulation of 5-fluorouracil: a randomized comparison of sequential methotrexate, 5-fluorouracil and leucovorin versus sequential 5-fluorouracil and leucovorin in patients with advanced symptomatic colorectal cancer. The Nordic Gastrointestinal Tumor Adjuvant Therapy Group. Ann Oncol 1993; 4: 235240.[Abstract]
26. Glimelius B, Jakobsen A, Graf W et al. Bolus injection (24 min) versus short-term (1020 min) infusion of 5-fluorouracil in patients with advanced colorectal cancer: a prospective randomised trial. Nordic Gastrointestinal Tumour Adjuvant Therapy Group. Eur J Cancer 1998; 34: 674678.[ISI][Medline]
27. Graf W, Pahlman L, Bergstrom R, Glimelius B. The relationship between an objective response to chemotherapy and survival in advanced colorectal cancer. Br J Cancer 1994; 70: 559563.[ISI][Medline]
28. Glimelius B, Hoffman K, Graf W et al. Quality of life during chemotherapy in patients with symptomatic advanced colorectal cancer. The Nordic Gastrointestinal Tumor Adjuvant Therapy Group. Cancer 1994; 73: 556562.[ISI][Medline]
29. Glimelius B, Graf W, Hoffman K et al. General condition of asymptomatic patients with advanced colorectal cancer receiving palliative chemotherapy. A longitudinal study. Acta Oncol 1992; 31: 645651.[ISI][Medline]
30. Superiority of sequential methotrexate, fluorouracil, and leucovorin to fluorouracil alone in advanced symptomatic colorectal carcinoma: a randomized trial. The Nordic Gastrointestinal Tumor Adjuvant Therapy Group. J Clin Oncol 1989; 7: 14371446.[Abstract]
31. Ray P, Quantin X, Grenier J et al. Predictive factors of tumor response and prognostic factors of survival during lung cancer chemotherapy. Cancer Detect Prev 1998; 22: 293304.[ISI][Medline]
32. van Dalen A. TPS monitoring in metastatic breast cancer. Int J Biol Markers 1999; 14: 4548.[ISI][Medline]
33. Al-Sarraf M, Baker L, Talley RW et al. The value of serial carcinoembryonic antigen (CEA) in predicting response rate and survival of patients with gastrointestinal cancer treated with chemotherapy. Cancer 1979; 44: 12221225.[ISI][Medline]
34. Allen-Mersh TG, Kemeny N, Niedzwiecki D et al. Significance of a fall in serum CEA concentration in patients treated with cytotoxic chemotherapy for disseminated colorectal cancer. Gut 1987; 28: 16251629.[Abstract]
35. Herrera MA, Chu TM, Holyoke ED, Mittelman A. CEA monitoring of palliative treatment for colorectal carcinoma. Ann Surg 1977; 185: 2330.[ISI][Medline]
36. Hine KR, Dykes PW. Prospective randomised trial of early cytotoxic therapy for recurrent colorectal carcinoma detected by serum CEA. Gut 1984; 25: 682688.[Abstract]
37. Mayer RJ, Garnick MB, Steele GD Jr, Zamcheck N. Carcinoembryonic antigen (CEA) as a monitor of chemotherapy in disseminated colorectal cancer. Cancer 1978; 42: 14281433.[ISI][Medline]
38. Ravry M, Moertel CG, Schutt AJ, Go VL. Usefulness of serial serum carcinoembryonic antigen (CEA) determinations during anticancer therapy or long-term follow-up of gastrointestinal carcinoma. Cancer 1974; 34: 12301234.[ISI][Medline]
39. Shani A, OConnell MJ, Moertel CG et al. Serial plasma carcinoembryonic antigen measurements in the management of metastatic colorectal carcinoma. Ann Intern Med 1978; 88: 627630.[ISI][Medline]
40. Staab HJ, Anderer FA, Hornung A et al. Doubling time of circulating CEA and its relation to survival of patients with recurrent colorectal cancer. Br J Cancer 1982; 46: 773781.[ISI][Medline]
41. Hanke B, Riedel C, Lampert S et al. CEA and CA 19-9 measurement as a monitoring parameter in metastatic colorectal cancer (CRC) under palliative first-line chemotherapy with weekly 24-hour infusion of high-dose 5-fluorouracil (5-FU) and folinic acid (FA). Ann Oncol 2001; 12: 221226.[Abstract]
42. ASCO. 1997 update of recommendations for the use of tumor markers in breast and colorectal cancer. Adopted on November 7, 1997 by the American Society of Clinical Oncology. J Clin Oncol 1998; 16: 793795.[Abstract]
43. Verheul HM, Pinedo HM. Tumor growth: a putative role for platelets? Oncologist 1998; 3: II.[Medline]
44. Buyse M, Thirion P, Carlson RW et al. Relation between tumour response to first-line chemotherapy and survival in advanced colorectal cancer: a meta-analysis. Meta-Analysis Group in Cancer. Lancet 2000; 356: 373378.[ISI][Medline]
45. Nygren P, Glimelius B. The Swedish Council on Technology Assessment in Health Care (SBU) report on Cancer ChemotherapyProject objectives, the working process, key definitions and general aspects on cancer trial methodology and interpretation. Acta Oncol 2001; 40: 155165.[ISI][Medline]