Laboratory variables to stratify metastatic colorectal cancer patients

J. Watine

Member of the Committee on Evidence-Based Laboratory Medicine of the International Federation of Clinical Chemistry and Laboratory Medicine, Laboratoire de biologie polyvalente, Hôpital Général, Rodez, France (E-mail: watine61@hotmail.com)

In an article published in the February 2002 issue of Annals of Oncology by Koehne et al. [1] it was demonstrated, in a series of 3825 metastatic colorectal cancer patients who participated in 19 prospective randomised chemotherapy trials, that four baseline variables, i.e. performance status, alkaline phosphatase (ALP), white blood cell count (WBC) and number of metastatic sites, can define at least three risk groups of patients. Two additional laboratory variables, i.e. platelet count and haemoglobin, were also of independent prognostic value in this large series of patients, whereas seven others, i.e. lactate dehydrogenase (LDH), protein, albumin, carcinoembryonic antigen (CEA), bilirubin, and aspartate or alanine aminotransferases, did not reach statistical significance in multivariate analysis.

Even if we would quite agree with Koehne et al. that their model might serve as a point of reference for any other model using different parameters, we would have a number of arguments against the exclusion of several of the cheap laboratory investigations mentioned above from any reference model.

First, one can note that the four laboratory variables that were of independent prognostic significance in Koehne et al.’s multivariate analysis were the four with the lowest percentages of missing data, i.e. 30% for WBC or haemoglobin, 32% for platelets, and 38% for ALP, compared with 72% for LDH, 67% for albumin, 53% for CEA, 46% for bilirubin, and 48% and 44% for the aminotransferases. This might have flawed their analysis.

Secondly, a number of other studies have used multivariate statistical analysis to compare the prognostic values of ALP or WBC with those of one or several of the other afore-mentioned laboratory variables. The many studies published so far, taking into account only overall survival (the objective outcome), include: Fountzilas et al.’s study in which CEA or albumin were of independent significance, whereas WBC, ALP or LDH were not [2]; Saltz et al.’s study in which LDH, bilirubin, WBC or haemoglobin were of independent significance, whereas CEA was not [3]; and Steinberg et al.’s study in which albumin or aspartate aminotransferase were of independent significance, whereas ALP was not [4]. These examples are a short summary of the quite contradictory results that have been published so far on the subject (a list of other references is available from us on request).

In order to explain such controversial results, various hypotheses have been proposed, among these the likely heterogeneity in the laboratory techniques used in these numerous studies deserves a mention. The problem is that, as far as we know, none of the studies published so far clearly indicated its laboratory methods. In this situation, it would not be possible for independent teams to try to precisely reproduce the laboratory techniques used in these studies—a basic principle of good science. Some of the problems that may be caused by this situation are quite well established; in particular, for some laboratory variables, randomly distributed variations of >50% from one study to another, can not be excluded [5].

In summary, Koehne et al. suggested that in order to stratify metastatic colorectal cancer patients in clinical trials, one would have to measure ALP, WBC, haemoglobin and platelets. Taking into account the other studies published so far (i.e. evidence-based laboratory medicine) would suggest that at least LDH, transaminases, bilirubin, albumin and CEA should be added to this list of laboratory variables that should be mandatory to measure. Also, the laboratory methodologies should be indicated in any ensuing publication in order to eventually make it possible to better establish not only the most useful laboratory tests in this context, but also the best way to measure them.

J. WatineMember of the Committee on Evidence-Based Laboratory Medicine of the International Federation of Clinical Chemistry and Laboratory Medicine, Laboratoire de Biologie Polyvalente, Hôpital Général, Rodez, France (E-mail: watine61@hotmail.com)

References

1. Kohne CH, Cunningham D, Di CF 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: 308–317.[Abstract/Free Full Text]

2. Fountzilas G, Gossios K, Zisiadis A et al. Prognostic variables in patients with advanced colorectal cancer treated with fluorouracil and leucovirin-based chemotherapy. Med Pediatr Oncol 1996; 26: 305–317.[CrossRef][ISI][Medline]

3. Saltz LB, Cox JV, Blanke C et al. Irinotecan plus fluorouracil and leucovorin for metastatic colorectal cancer. Irinotecan Study Group. N Engl J Med 2000; 343: 905–914.[Abstract/Free Full Text]

4. Steinberg J, Erlichman C, Gadalla T et al. Prognostic factors in patients with metastatic colorectal cancer receiving 5-fluorouracil and folinic acid. Eur J Cancer 1992; 28A: 1817–1820.

5. Friedberg B, Watine J, Miédougé M. Unresected colorectal liver metastases: prognostic value of laboratory variables. Gastroenterol Clin Biol 2001; 25: 962–966 [article in French; full English text on-line] http://www.e2med.com/gcb[ISI][Medline]





This Article
Full Text (PDF)
E-letters: Submit a response
Alert me when this article is cited
Alert me when E-letters are posted
Alert me if a correction is posted
Services
Email this article to a friend
Similar articles in this journal
Similar articles in ISI Web of Science
Similar articles in PubMed
Alert me to new issues of the journal
Add to My Personal Archive
Download to citation manager
Search for citing articles in:
ISI Web of Science (2)
Disclaimer
Request Permissions
Google Scholar
Articles by Watine, J.
PubMed
PubMed Citation
Articles by Watine, J.