• E-mail
• Print
• Comment
• Font Size
• Digg
• del.icio.us
• Discuss article

Biological Variation of Tumor Markers and Its Application in the Detection of Disease Progression in Patients With Non-Small Cell Lung Cancer

Posted on: Sunday, 9 January 2005, 03:00 CST

Patients with advanced non-small cell lung cancer (NSCLC) may benefit from chemotherapy (7, 2). This type of treatment is highly toxic and expensive, and patients should be monitored to detect those in whom the disease progresses despite treatment to avoid ineffective treatment and a loss of quality of life. Tumor markers are useful for detecting progression in cancer patients undergoing chemotherapy (3, 4), but there are no uniform criteria establishing how large a difference between two consecutive measurements should be to indicate progression of the disease.

The first step in dynamic assessment of a specific biological constituent is to define the difference between two consecutive results that indicates a change in the patient's health status. The most widely accepted approach for this purpose is the so-called reference change value (RCV), a concept described by Harris and Brown (5). Using serial analytic results from the same individual, it is possible to calculate the RCV that defines the critical difference between two consecutive results for a constituent at a specific probability. The RCV encompasses both biological and analytical variation and has been determined for several constituents in healthy patients and in some diseases (6-14).

The aims of this study were (a) to estimate the biological variation (BV) of the tumor markers CYFRA 21-1, CA 125, and carcinoembryonic antigen (CEA) in a control group and in patients with NSCLC; (b) to establish analytical quality goals for CYFRA 21- 1 measurements; and (c) to calculate the RCVs for these tumor markers and assess their application for detecting disease progression in NSCLC patients under treatment.

The BV components of CYFRA 21-1, CA 125, and CEA were determined in two groups. The control group included 30 patients [15 men; mean (SD) age, 61.1 (7.2) years] with colon adenocarcinoma (Astler- Coller classification stages A or B1) treated surgically. Three samples were obtained from each patient every 6 months for 3 years after the procedure. There was no evidence of recurrence (by clinical examination and imaging methods) during the study period and up to at least 1 year after the last determination. The results from this group also were used to establish the analytical quality goals.

The lung cancer group comprised 10 NSCLC patients [5 men; mean (SD) age, 59 (8.1) years; 4 with stage IIIA and 1 with stage IIIB disease] presenting with complete remission treated with chemotherapy and 5 patients [4 men; mean age, 57 (5.2) years; 3 with stage IA and 2 with stage IIA disease] who underwent total tumor resection. Tumor markers were measured every 3 months for all patients of this group, with a mean of 4.8 determinations (range, 2- 11) per patient. Samples in the treated patients were obtained before administration of chemotherapy. Results of tests performed in the time period from 6 months before the detection of progression to the time of progression were excluded from the BV calculations. Possible significant differences in BV values in this group compared with the control group were assessed with the F-test. The variation obtained in this group included the between-day analytical variation, and baseline factors (e.g., smoking habits) that can affect tumor marker concentrations in patients with NSCLC; the values from this group were therefore used to calculate the RCV.

To investigate the usefulness of the tumor marker RCVs for the detection of disease progression, a study was performed in 44 patients (42 men), ranging in age from 38 to 84 years (mean, 61.8 years). The selection criteria were the following: diagnosis of NSCLC stages IIIA to IV; performance status (Eastern Cooperative Oncology Group) <3, and platin-based chemotherapy. The clinical characteristics of the patients are described in the Data Supplement that accompanies the online version of this Technical Brief at http:/ /www.clinchem.org/content/ vol51/issue1/. The mean monitoring period was 13.5 months (range, 2-72 months), and the mean number of tumor marker measurements per patient was 6.9 (range, 3-23). Markers were measured every 4 weeks during chemotherapy, every 3 months on completion of chemotherapy over 1 year, and every 6 months thereafter until recurrence. In patients undergoing second-line treatment, monitoring was repeated every 4 weeks. Disease progression, was defined as lesion growth >25% on two-dimensional imaging or >40% on volumetric imaging, or the appearance of new lesions.

Tumor markers were measured by an electrochemiluminescent method on an ELECSYS 2010 analyzer (Roche Diagnostics). Control group samples were analyzed in a single run. For the control group, within- run imprecision (CV; n = 10) was 1.8% for CYFRA 21-1 (mean concentration, 2.1 g/L), 2.1% for CEA (2.9 g/L), and 1.7% for CA 125 (21 kilounits/L), obtained with replicates of patient samples. Samples from the NSCLC groups were analyzed on the day of collection. The between-run CVs (n = 620), i.e., the mean imprecision for all lots of Precicontrol tumor marker 1 (Roche Diagnostics) used during the study, were 4.3% for CYFRA 21-1 (2.8 g/ L), 3.9% for CEA (4.8 g/L), and 4.8% for CA 125 (31 kilounits/L).

Variations in tumor marker values around the means in the control and lung cancer groups did not follow a gaussian distribution (Kolmogorov-Smirnov test, P <0.2); thus, logarithmic transformation was applied (6).

There were no significant differences in tumor marker BV values between the control and lung cancer groups: F-test results for CYFRA 21-1, P = 0.17; CA 125, P = 0.28; and CEA, P = 0.19. The BV values for the two study groups are shown in Fig. 1. The II was estimated as 0.55, 0.57, and 0.28 for CYFRA 21-1, CA 125, and CEA, respectively, in the lung cancer patients. The performance of the tumor marker RCVs for the detection of disease progression at the different probabilities is shown in Table 1. There was a decrease in sensitivity and an increase in specificity at increasing probabilities.

Application of the tumor marker RCVs (P <0.05) to the detection of disease progression in a separate cohort of NSCLC patients showed that CYFRA 21-1 provided the highest diagnostic performance, followed by CEA and CA 125. In 18 of the 44 patients (40.9%), disease progression was detected at the first clinical evaluation (3 months after beginning treatment). RCVs for CYFRA 21-1, CA 125, and CEA detected progression at the first clinical evaluation in 7 of 18 (38.8%), 6 of 18 (33.3%), and 10 of 18 (55.5%) patients, respectively, with detection by all three RCVs in 12 (66.6%) patients. The mean lead time of the patients showing progression by RCVs was 2 months (range, 1-6 months), with 11 of 44 (25.0%), 7 of 44 (15.9%), and 15 of 44 (34.1%) detected by CYFRA 21-1, CA 125, and CEA, respectively, and 18 of 44 (40.9%) detected by all three markers. Disease progression was simultaneously detected by imaging methods and tumor marker RCVs in 12 of 44 (27.2%) patients by CYFRA 21-1, 10 of 44 (22.7%) by CA 125, 7 of 44 (15.9%) by CEA, and 15 of 44 (34.1%) by a combination of the three markers. No critical increases in tumor markers were found at the time of imaging progression in 11 of 44 (25.0%) patients. In six cases, the increases were all lower than the RCV, and in five patients values were within the reference intervals at initial diagnosis and showed no increases during the study. On completion of the study, only two patients did not experience disease progression.

Table 1. Diagnostic performance of tumor markers at different probabilities for RCV.

The CV^sub w^ values for CA 125 and CEA in this study were similar to those reported previously (15-17). The components of BV for CYFRA 21-1 in healthy individuals are unknown (15). We established the CV^sub w^ of CYFRA 21-1 in the control group at 22.5%. The desirable analytical performance for CYFRA 21-1, defined as CV^sub a^ = 0.5CV^sub w^ (18-21), was 11.2%.

Previous studies have applied the criterion used in imaging findings (40% increase in tumor volume) to define the critical difference in consecutive analyses. With this criterion, the corresponding RCV for CYFRA 21-1 is closest to the 90% probability and that for CEA is closest to the 99% probability value. If the RCV at the 90% probability value is used for CEA, the critical difference to assess would be 22.6%, and the sensitivity would increase to 52%. Application of the RCV at the 95% probability value (the most widely used value) allows detection of disease progression with a lead time similar to that reported by other authors (3, 4) and a sensitivity slightly lower than for CYFRA 21-1 in other reports (3, 4, 22, 23) and higher for CEA (22).

In conclusion, the criteria we have established for the magnitude of the difference in serial tumor marker values based on BV at 95% probability allowed the detection of 75% of cases of disease progression. Furthermore, when none of the markers increased above the RCV, the probability of progression was low. Monitoring of lung cancer patients undergoing chemotherapy with the use of these three tumor markers is helpful for the detection and exclusion of disease progression and allows a point of reference for detection of progression to be established specifically for each tumor marker.

References

1. \Souquet PJ, Chauvin F, Boisel JP, Bernard JP. Meta-analysis of randomised trials of systemic chemotherapy versus supportive treatment in non-resectable non-small cell lung cancer. Lung Cancer 1995;12(Suppl 1):S147-54.

2. Cesano A, Lane SR, Ross GA, Fields SZ. Stabilization of disease as an indicator of clinical benefit associated with chemotherapy in non-small cell lung cancer patients. Int J Oncol 2000;17:587-90.

3. Ebert W, Muley T. CYFRA 21-1 in the follow-up of inoperable non-small cell lung cancer patients treated with chemotherapy. Anticancer Res 1999;19:2669-72.

4. van der Gaast A, Kok TC, Kho GS, Blijenberg BG, Splinter TAW. Disease monitoring by tumour markers cyfra 21-1 and TPA in patients with non small cell lung cancer. Eur J Cancer 1995;31A:1790-3.

5. Harris EK, Brown SS. Temporal changes in the concentration of serum constituents in healthy men. Distributions of within-person variances and their relevance to the interpretation of differences between successive measurements. Ann Clin Biochem 1979;16:169-76.

6. Queralt JM, Boyd JC, Harris EK. On the calculation of reference change values, with examples from long-term study. Clin Chem 1993;39:1398-403.

7. Harris EK, Yasaka T. On the calculation of a "reference change" for comparing two consecutive measurements. Clin Chem 1983;29:25-30.

8. Philipou G, Philips PJ. Intraindividual variation of glycohemoglobin: implications for interpretation and analytical goals. Clin Chem 1993;39:2305-8.

9. Hzel WGE, Beer R, Deschner W, Griesmacher A, Mller MM. Individual reference ranges of CA15-3, MCA and CEA in recurrence of breast cancer. Scand J Clin Lab Invest Suppl 1995;221:93-101.

10. Panteghini M Pagani F, Bonora R. Pre-analytical and biological variability of prostatic acid phosphatase and prostate specific antigen in serum from patients with prostatic pathology. Eur J Clin Chem Clin Biochem 1992;30:135-9.

11. Mora J, Queralt JM. Within-subject variation of carcinoembryonic antigen in colorectal cancer-application of reference change values and individual reference ranges to patient follow-up. Eur J Clin Chem Clin Biochem 1997;35:453-7.

12. Trap J, Aliart M, Brunei M, Dern E, Abadal E, Queralt JM. Reference change value for HbA1c in patients with type 2 diabetes mellitus. Clin Chem Lab Med 2000;38:1283-7.

13. Biosca C, Rics C, Jimenez CV, Lauzurica R, Galimany R. Model for establishing biological variation in non-healthy situations: renal posttransplantation data [Technical Brief]. Clin Chem 1997;43:2206-8.

14. Trap J, Botargues JM, Porta F, Rics C, Badal JM, Salinas R, et al. Reference change value for α-fetoprotein in early detection of hepatocellular carcinoma in patients with hepatic diseases [Technical Brief]. Clin Chem 2003;49:1209-11.

15. Rics C, Alvarez V, F Cava, Lario JV, Hernndez A, Jimnez CV, et al. Desirable specifications for total error, and bias, derived from biologic variation, http://www.westgard.com/biodatabasel.htm (accessed December 2003).

16. Tuxen M, Sltormos G, Petersen PH, Schioler V, Dombernowsky P. Assessment of biological variation and analytical imprecision of CA 125, CEA and TPA in relation to monitoring of ovarian cancer. Gynecol Oncol 1999;74:12-22.

17. Tuxen MK, Sltorrnos G, Petersen PH, Dombernowsky P. Interpretation of sequential measurements of cancer antigen 125 (CA 125) Carcinoembryonic antigen (CEA) and tissue polypeptide antigen (TPA) based on analytical imprecision and biological variation in the monitoring of ovarian cancer. Clin Chem Lab Med 2001;39:531-8.

18. Rics C, Baadenhuijsen H, Libeer JC, Libeer CJ, Petersen PH, Stockl D, et al. Currently used criteria for evaluating performance in EQA in European countries and a proposal for harmonization. Eur J Clin Chem Clin Biochem 1996;34:159-65.

19. Fraser CG. General strategies to set quality specifications for reliability performance characteristics. Scand J Clin Lab Invest 1999;59:487-90.

20. Kenny D, Fraser CG, Petersen PH, Kallner A. Consensus agreement. Scand J Clin Lab Invest 1999;59:585.

21. Fraser CG, Harris EK. Generation and application of data on biological variation in clinical chemistry [Review]. Crit Rev Clin Lab Sci 1989;27:409-37.

22. Ebert W, Hoppe M, Muley T, Drings P. Monitoring of therapy in inoperable lung cancer patients by measurement of CYFRA 21-1, TPA- TP CEA and NSE. Anticancer Res 1997;17:2875-8.

23. Nisman B, Lafair J, Heching N, Lyass O, Baras M, Peretz T, et al. Evaluation of tissue polypeptide specific antigen, CYFRA 21-1, and carcinoembryonic antigen in non-small cell lung cancer. Cancer 1998;82:1850-9.

DOI: 10.1373/clinchem.2004.040659

Jaume Trap,1* Joaquim Prez de Olaguer,2 Josep Bux,2 and Laura Lpez1 (1 Clinical Laboratory and 2 Service of Oncology, Hospital Sant Joan de Du, Althaia Xarxa Assistencial de Manresa, Barcelona, Spain; * address correspondence to this author at: Clinical Laboratory, Hospital Sant Joan de Du, Althaia Xarxa Assistencial de Manresa, Dr Joan Soler, s/n, 08243 Manresa (Barcelona), Spain; fax 34-93-8743859, e-mail jtrape@yahoo.es)

Copyright American Association for Clinical Chemistry Jan 2005

Source: Clinical Chemistry

More News in this Category