|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
Journal of Clinical Oncology, Vol 26, No 22 (August 1), 2008: pp. 3681-3686 © 2008 American Society of Clinical Oncology. DOI: 10.1200/JCO.2007.15.0904 Usefulness of the Serum Carcinoembryonic Antigen Kinetic for Chemotherapy Monitoring in Patients With Unresectable Metastasis of Colorectal Cancer
From the Digestive Oncology Unit, Hepato-Gastroenterology Department, and Biostatistics Unit, Inserm U 657, Rouen University Hospital; and Laboratoire de Biologie Clinique et de Radioanalyse, Centre Henri-Becquerel, Rouen, France Corresponding author: Pierre Michel, MD, PhD, Digestive Oncology Unit, Hepato-Gastroenterology Department, Rouen University Hospital–Charles Nicolle–France, 1 rue de Germont, 76031 Rouen Cedex, France; e-mail: pierre.michel{at}chu-rouen.fr
Purpose The aim of the study was to evaluate the relationship between serum carcinoembryonic antigen (CEA) kinetic and response to chemotherapy in patients with unresectable metastasis of colorectal cancer. Patients and Methods The kinetic was calculated using the slope of an exponential-regressive curve connecting the semi-logarithmic values of CEA. Receiver operating characteristic (ROC) curves were drawn to select the CEA slope thresholds to define patients with progressive or responsive disease with the highest sensitivity, specificity, and diagnosis accuracy odds ratio (DOR). The correlation between the CEA slopes and progression-free survival (PFS) was evaluated by the Cox model and Kaplan-Meier methods. Results A total of 122 patients were included. Progression defined by CEA slope greater than +0.05 resulted in sensitivity of 85.7%, specificity of 85.1%, and DOR of 34. The area under the ROC (AUROC) curve was 0.885 (95% CI, 0.815 to 0.936; P = .0001). Response defined by CEA slope less than –0.2 resulted in sensitivity of 74.7%, specificity of 82.5%, and DOR of 16. The AUROC curve was 0.847 (95% CI, 0.770 to 0.906; P = .0001). The difference between AUROC curves calculated with six or four CEA values was not significant. PFS was correlated with CEA slopes (hazard ratio, 4.6; 95% CI, 2.48 to 8.57). The median PFS was 10 months for patients with CEA slope values less than –0.2 months versus 6 months for patients with CEA slope values greater than –0.2 (P < .0001). Conclusion These results suggest that the CEA kinetic is an accurate, simple, and noninvasive method to identify the disease progression in patients with unresectable metastasis of colorectal cancer.
Colorectal cancer is one of the most common causes of cancer death in Western countries.1 The median overall survival of patients with metastatic colorectal cancer has increased from 12 months to approximately 20 months in the past decade as a result of improvement in systemic therapies.1 To date, only 10% of patients are eligible for basic surgery to treat metastatic disease, and 10% to 20% are eligible after chemotherapy.2 The guidelines for treatment monitoring is based on imaging evaluation every 2 or 3 months using standardized criteria.3 The main challenge is to identify the disease progression at an early stage using a simple method to shift treatment in patients with unresectable metastasis colorectal cancer treated with chemotherapy. More than 80% of patients with a metastatic colorectal cancer have a serum carcinoembryonic antigen (CEA) value greater than the normal rate (5 ng/mL).4,5 According to American Society of Clinical Oncology recommendations, CEA is considered the marker of choice for monitoring metastatic colorectal cancer. During the past decade, several studies have suggested that the CEA serum level variation could be used in chemotherapy monitoring.6-13 However, recent guidelines from the American Society of Clinical Oncology have stressed that these data are not yet sufficient to recommend the routine use of CEA serum level variation for monitoring response to chemotherapy.4 This cautionary advice is based on two points: published studies previously included only a small population, and the method was not relevant because only the variation between two serum levels of CEA was used.11-13 It has been shown that the evaluation of the tumor marker kinetic is biologically more relevant than the serum level variation calculated between two points.14,15 The kinetic of a tumor marker is obtained with the slope of the regressive exponential curve based on semi-logarithmic values of at least three marker levels obtained during treatment. Furthermore, the half-life or doubling time of the tumor marker level is calculated based on this regressive exponential curve. In contrast with other malignancies such as ovarian cancer, the kinetic of tumor markers in patients with metastasis colorectal cancer, to our knowledge, has never previously been evaluated.4,15 The aim of the present prospective study was to evaluate the CEA kinetic to detect disease progression in patients with metastatic colorectal cancer treated with chemotherapy.
Patients The study included all patients treated with first-line chemotherapy for colorectal adenocarcinoma with unresectable metastasis between January 1, 1998, and December 31, 2005, in our unit. The therapeutic strategy was defined in agreement with the operating surgeon. All patients had histologically proven colorectal adenocarcinoma without prior treatment for metastasis, measurable liver and/or lung metastasis (>10 mm), a WHO performance status of 0 to 2, and baseline CEA serum level measurement greater than 5 ng/mL. Patients with a previous history of carcinoma during the past 5 years were excluded. Each eligible patient received as first-line treatment a fluoropyrimidine-based chemotherapy regimen with or without oxaliplatin or irinotecan (14-day treatment cycle). A computed tomography (CT) scan was performed at baseline and every 3 months for each patient. CT scan evaluation was performed every 3 months until disease progression. Patients were considered in response, disease progression, or stable disease according to the Response Evaluation Criteria in Solid Tumors Group criteria at the CT scan evaluation performed at 3 months.16 Major response in this study was defined as a reduction in tumor size to allow for resection of metastasis considered initially unresectable.
Methods The kinetic of CEA level was calculated using the slope of an exponential-regressive curve connecting the semi-logarithmic values of CEA (Ln CEA). As shown in Appendix Fig A1 (online only), Excel software (Microsoft office Excel 2003, Redmond, WA) was used to calculate the equation of this curve: y = a.exp (px; p = slope of the curve; x = number of days between two chemotherapies; a was the corresponding CEA value). On the basis of this curve, tumor marker half-time (t1/2 = Ln2/–p) or doubling time (td = Ln2/p) were calculated. In the first stage, CEA slopes were calculated using the CEA serum level obtained at cycles 1 through 6. In the second stage, only the four initial CEA serum levels obtained in cycles 1 through 4 were considered for the slope calculation. Moreover, the kinetic of CEA was compared with a simple variation of the marker between 2 months (samples 1 and 5) using the cutoff previously defined.13
Statistical Analysis The DOR was compared between the CEA slopes (1 through 6 and 1 through 4) and the variation between samples 1 and 5 (2 months). The correlation between the CEA slopes and progression-free survival (PFS) was evaluated by two methods. A Cox model was used to test the correlation between PFS and the CEA slope values. PFS time was calculated from the day of first chemotherapy cycle until radiologic progression or death. The Kaplan-Meier method was used to estimate survival curves according to the threshold used for patients in response. The log-rank test was used to compare the curves. MEDCALC (Mariakerke, Belgium) and STATVIEW (Cary, NC) statistical software was used.
From 1998 to 2005, 122 patients who met the inclusion criteria were observed (sex ratio = 1.7 men/one woman; median age, 62.7 years). On the CT evaluation after 3 months of chemotherapy, patients in disease progression, stable disease, and response were, respectively, 28 patients (22%), 51 patients (41%), and 43 patients (35%). A major response was obtained in seven patients, six patients had undergone surgery for their metastasis (4%), and one patient had percutaneous radiofrequency (0.8%). Patient characteristics are listed in Table 1.
The range of CEA slopes of the regressive exponential curves was –1.3551 to +0.6439, with a mean of –0.1449 and a median of –0.1127. The median slopes for patients with progressive, responsive, and stable disease were significantly different (P < .001) at 0.2239, –0.3384, and –0.0616, respectively. For the seven patients with major response who underwent subsequent resection or destruction of their metastasis showed a median slope of –0.2463, which was not significantly different from the median slope as compared with the other patients achieving response to treatment (P = .95). In the first stage, six values of CEA were used to calculate the slope of the regressive curve. For patients with progressive disease, a CEA slope threshold greater than +0.05 corresponding to a doubling time of 13.8 days was chosen, with sensitivity of 85.7% (95% CI, 67.3% to 95.9%), specificity of 85.1% (95% CI, 76.3% to 91.6%), PPV of 63.6% (95% CI, 60.2 to 66.2%), NPV of 95.3% (95% CI, 94.0% to 96.6%), and a DOR of 34 (Table 2). The area under the ROC curve was 0.885 (95% CI, 0.815 to 0.936; P = .0001; Fig 1A). For patients achieving response, a CEA slope threshold less than –0.2 corresponding to a half-life of 3.4 days was chosen, with sensitivity of 74.5% (95% CI, 60.4% to 85.7%), specificity of 84.7% (95% CI, 74.3% to 92.1%), PPV of 77.6% (95% CI, 74.6% to 80.6%), NPV of 82.4% (95% CI, 80.4% to 84.4%), and a DOR of 16 (Table 2). The area under the curve was 0.847 (95% CI, 0.770 to 0.906; P = .0001; Fig 1B).
In the second stage, CEA slopes were calculated with the four first CEA values and compared with the results with six CEA values. For patients with progressive disease, a CEA slope threshold greater than +0.05 corresponding to a doubling time of 13.8 days was chosen, with sensitivity of 88.9% (95% CI, 70.8% to 97.5%), specificity of 77.6% (95% CI, 67.9% to 85.6%), PPV of 53.3% (95% CI, 50.4% to 56.2%), NPV of 96.1% (95% CI, 94.7% to 95.5%), and DOR of 28.4 (Table 3). The area under the ROC curve was 0.846 (95% CI, 0.770 to 0.805; P = .0001; Fig 2A). For patients achieving response, a CEA slope threshold less than –0.2 corresponding to a half-life of 3.4 days was chosen, with sensitivity of 60.7% (95% CI, 46.1% to 74.1%), specificity of 80% (95% CI, 68.7% to 87.7%), PPV of 68.9% (95% CI, 66.5% to 71.3%), NPV of 73.7% (95% CI, 71.3% to 76.1%), and DOR of 6.2 (Table 3). The area under the curve was 0.847 (95% CI, 0.770 to 0.906; P = .0001; Fig 2B).
Analysis of CEA was also performed with a simple variation of the marker between two months (samples 1 and 5). For patients with progressive disease, the area under the ROC curve was 0.863 (95% CI, 0.788 to 0.920; P = .0001). With the threshold of simple variation of marker previously reported (> 200%), sensitivity was 4.0% (95% CI, 0.7% to 20.4%), specificity was 96.7% (95% CI, 90.9% to 99.3%), PPV was 25.0% (95% CI, 22.4% to 27.6%), NPV was 78.8% (95% CI, 77.5% to 82.2%), and DOR was 1.2.13 However, according to the results of ROC curve, the threshold of +15% was more clinically relevant, with sensitivity of 88.0% (95% CI, 68.8% to 97.3%), specificity of 82.6% (95% CI, 73.3% to 89.7%), and a DOR of 33.7. For patients achieving response, the area under the ROC curve was 0.799 (95% CI, 0.714 to 0.867; P = .0001). The most relevant threshold was 30%, with sensitivity of 76.0% (95% CI, 61.8% to 86.9%) and specificity of 71.6% (95% CI, 59.3% to 82.0%), PPV of 66.7% (95% CI, 64.3% to 69.1%), NPV of 80.0% (95% CI, 77.8% to 82.2%), and a DOR of 7.9. No significant difference was found between ROC curves calculated with six and four CEA values for patients with disease progression or response, respectively (P = .54 and P = .1). The difference between ROC curves calculated with four CEA values and the simple variation of two samples (samples 1 and 5) were considered statistically significant for patients with response (P = .038) and not significant for patients with disease progression (P = .370). A significant correlation between PFS and CEA slopes (one through six) was shown with Cox model (P < .0001; hazard ratio, 4.6; 95% CI, 2.48 to 8.57). The median PFS was 10 months for patients with CEA slope values less than –0.2 months versus 6 months for patients with CEA slope values more than –0.2 (P < .0001; Fig 3).
This study demonstrates that the kinetic of CEA allows an accurate evaluation of progression, response, and PFS during chemotherapy in patients with metastatic colorectal cancer. The CEA slopes under the threshold of less than –0.2 and a half-life less than 3.4 days were significantly correlated with response to treatment, and the CEA slopes over the threshold of greater than +0.05 and a doubling time greater than 13.8 days were significantly correlated with disease progression. The study was specifically designed to evaluate the CEA kinetic in chemotherapy response and PFS, which is an appropriate surrogate for overall survival.17 To our knowledge, the present study is the first report series to evaluate the kinetic of CEA in the follow-up of patients with colorectal cancer treated with palliative chemotherapy. Our results suggest that the kinetic of CEA may be useful in monitoring chemotherapy in patients with unresectable metastatic colorectal cancer. However, the correlation between CEA kinetic and PFS should be assessed in patients treated with new drugs as angiogenesis inhibitors. To be effective, CEA slopes should be calculated from the onset of chemotherapy based on a minimum of three serial samples (CEA value > 5 ng/mL), enabling the calculation of a regression line. This method is simple and easily accessible using computer software.15 This method to calculate the kinetic of a tumor marker with the slope of the regressive exponential curve has been previously used in other solid tumors. The tumor marker CA-125 kinetic is now widely accepted to be an accurate predictive and prognostic factor in CA-125–positive ovarian cancers.14,15 The prostate-specific antigen kinetic has also been used as a prognostic factor after prostatectomy and in the follow-up of treatment.15 The kinetic of CA 15-3 is commonly used for the monitoring of breast cancer.15 In comparison with other tumor markers, the CEA kinetic approach can be useful to optimize patient management during chemotherapy. Our results also suggest that the simple variation of the marker during the first 2 months and kinetic of CEA produce similar results to predict disease progression; however, CEA kinetic is considered more accurate to predict response to chemotherapy. These findings support that the evaluation of the tumor CEA kinetic may be clinically more relevant than the serum level variation in patients with chemotherapy for unresectable metastasis of colorectal cancer. Although CEA kinetic seems to be biologically relevant for tumor monitoring during chemotherapy, some limits should also be mentioned. In fact, the weakness of tumor markers is that individual factors such as the pathophysiologic state of the patient or treatment may affect the measurement and the interpretation of marker levels. A false decrease in tumor markers may be attributable to procedures leading to hemodilution (parenteral nutrition and blood transfusion). CEA levels may be transiently affected by renal and hepatic diseases because these organs are involved in the metabolism of markers. Tumor metastasis may also exhibit patterns of marker secretion that are different from that of primary tumors. Moreover, aggressive chemotherapy may provoke massive destruction of cancer cells, leading to a transient increase in serum markers that should not be interpreted as tumor escaping eradication via chemotherapy resistance. However, these factors still remain effective by calculating the slope with a minimum of three serial semi-logarithmic tumor marker values as reported in other series.14,15 Furthermore, only 80% to 90% of patients with metastasis of colorectal cancer have a CEA level greater than 5 ng/mL; therefore, we cannot use the CEA kinetic for all patients with metastatic colorectal cancer. However, the present study suggests that the CEA kinetic is an accurate, simple, and robust method for managing chemotherapy in patients with unresectable metastasis of colorectal cancer. Currently, other methods are under evaluation, primarily because the standard evaluation method by CT scan remains debated.18,19 To define progression at an early stage during treatment, new methods are being developed, such as metabolic imaging (positron emission tomography) or detection of cancer cells in serum.18,20,21 In comparison with these new technologies, the main advantages of CEA kinetic are its low cost, ease of use, and accessibility using a standard personal computer. In conclusion, the CEA kinetic method may be considered to optimize treatment monitoring of patients with metastatic colorectal cancer. A future multicenter study with a large patient population is warranted to confirm our results.
The author(s) indicated no potential conflicts of interest.
Conception and design: Bernard Paillot, Jean Pierre Basuyau, Pierre Michel Administrative support: Pierre Michel Provision of study materials or patients: Frédéric Di Fiore, Isabelle Roque, Emilie Astruc, Bernard Paillot, Pierre Michel Collection and assembly of data: Isabelle Iwanicki-Caron, Isabelle Roque, Emilie Astruc, Monica Stetiu, Aude Duclos, David Tougeron, Sandrine Saillard, Sébastien Thureau, Pierre Michel Data analysis and interpretation: Isabelle Iwanicki-Caron, Frédéric Di Fiore, Emilie Astruc, Jacques Benichou, Pierre Michel Manuscript writing: Isabelle Iwanicki-Caron, Frédéric Di Fiore, Pierre Michel Final approval of manuscript: Isabelle Iwanicki-Caron, Frédéric Di Fiore, Jacques Benichou, Bernard Paillot, Pierre Michel
We thank Richard Medeiros, Rouen University Hospital Medical Editor, for his valuable help in editing the manuscript.
I.I.-C. and F.D.F contributed equally to this work. Authors’ disclosures of potential conflicts of interest and author contributions are found at the end of this article.
1. Meyerhardt JA, Mayer RJ: Systemic therapy for colorectal cancer. N Engl J Med 352:476-487, 2005 2. Jemal A, Murray T, Ward, et al: Cancer statistics. CA Cancer J Clin 55:10-30, 2005 3. Jaffe CC: Measures of response: RECIST, WHO, and new alternatives. J Clin Oncol 24:3245-3251, 2006 4. Locker GY, Hamilton S, Harris J, et al: ASCO 2006 update of recommendations for the use of tumor markers in gastrointestinal cancer. J Clin Oncol 24:5313-5327, 2006 5. Holt AD, Kim JT, Murrel Z, et al: The role of carcinoembryonic antigen as a predictor of the need for preoperative computed tomography in colon cancer patients. Am Surg 72:897-901, 2006[Medline] 6. Mayer RJ, Garnick MB, Steele GD Jr, et al: Carcinoembryonic antigen (CEA) as a monitor of chemotherapy in disseminated colorectal cancer. Cancer 42:1428-1433, 1978 (suppl 3)[CrossRef][Medline] 7. Shani A, O'Connell MJ, Moertel CG, et al: Serial plasma carcinoembryonic antigen measurements in the management of metastatic colorectal carcinoma. Ann Intern Med 88:627-630, 1978 8. 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 28:1625-1629, 1987 9. Ward U, Primrose JN, Finan PJ, et al: The use of tumour markers CEA, CA-195 and CA-242 in evaluating the response to chemotherapy in patients with advanced colorectal cancer. Br J Cancer 67:1132-1135, 1993[Medline] 10. Tsavaris N, Vonorta K, Tsoutsos H, et al: Carcinoembryonic antigen (CEA), alpha-fetoprotein, CA 19.9 and CA 125 in advanced colorectal cancer (ACC). Int J Biol Markers 8:88-93, 1993[Medline] 11. 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 12:221-226, 2001 12. Wang WS, Lin JK, Lin TC, et al: Tumor marker CEA in monitoring of response to tegafur-uracil and folinic acid in patients with metastatic colorectal cancer. Hepatogastroenterology 49:388-392, 2002[Medline] 13. Trillet-Lenoir V, Chapuis F, Touzet S, et al: Any clinical benefit from the use of oncofoetal markers in the management of chemotherapy for patients with metastatic colorectal carcinomas? Clin Oncol (R Coll Radiol) 16:196-203, 2004[Medline] 14. Riedinger JM, Wafflart J, Ricolleau G, et al: CA125 half-life and CA125 nadir during induction chemotherapy are independent predictors of epithelial ovarian cancer outcome: Results of a French multicentric study. Ann Oncol 17:1234-1238, 2006 15. Bidart JM, Thuillier F, Augereau C, et al: Kinetics of serum tumor marker concentrations and usefulness in clinical monitoring. Clin Chem 45:1695-1707, 1999 16. Therasse P, Susan G, Arbuck SG, et al: New guidelines to evaluate the response to treatment in solid tumors. J Natl Cancer Inst 92:205-216, 2000 17. Buyse M, Burzykowski T, Carroll K, et al: Progression-free survival is a surrogate for survival in advanced colorectal cancer. J Clin Oncol 25:5218-5224, 2007 18. Dimitrakopoulou-Strauss A, Strauss LG, Dimitri J: PET-FDG as predictor of therapy response in patients with colorectal carcinoma. Q J Nucl Med 47:8-13, 2003[Medline] 19. Tang PA, Bentzen SM, Chen EX, et al: Surrogate end points for median overall survival in metastatic colorectal cancer: Literature-based analysis from 39 randomized controlled trials of first-line chemotherapy. J Clin Oncol 25:4562-4568, 2007 20. Wang J-Y, Hsieh J-S, Chang M-Y, et al: Molecular detection of APC, K-ras, and p53 mutations in the serum of colorectal cancer patients as circulating biomarkers. World J Surg 28:721-726, 2004[Medline] 21. Meropol NJ, Cohen SJ, Iannotti N, et al: Circulating tumor cells (CTC) predict progression free (PFS) and overall survival (OS) in patients with metastatic colorectal cancer. J Clin Oncol 25:18S, 2007 (suppl, abstr 4010) Submitted October 26, 2007; accepted April 4, 2008.
Related Correspondence
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2008 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
