• Open Access

Circulating tumor cells as a surrogate marker for determining response to chemotherapy in Japanese patients with metastatic colorectal cancer

Authors

  • Satoshi Matsusaka,

    1. Department of Medical Oncology, Cancer Institute Hospital of Japanese Foundation for Cancer Research, Tokyo
    2. Cancer Chemotherapy Center, Clinical Chemotherapy, Japanese Foundation for Cancer Research, Tokyo, Japan
    Search for more papers by this author
  • Mitsukuni Suenaga,

    1. Department of Medical Oncology, Cancer Institute Hospital of Japanese Foundation for Cancer Research, Tokyo
    Search for more papers by this author
  • Yuji Mishima,

    1. Cancer Chemotherapy Center, Clinical Chemotherapy, Japanese Foundation for Cancer Research, Tokyo, Japan
    Search for more papers by this author
  • Ryoko Kuniyoshi,

    1. Cancer Chemotherapy Center, Clinical Chemotherapy, Japanese Foundation for Cancer Research, Tokyo, Japan
    Search for more papers by this author
  • Koichi Takagi,

    1. Department of Medical Oncology, Cancer Institute Hospital of Japanese Foundation for Cancer Research, Tokyo
    Search for more papers by this author
  • Yasuhito Terui,

    1. Department of Medical Oncology, Cancer Institute Hospital of Japanese Foundation for Cancer Research, Tokyo
    2. Cancer Chemotherapy Center, Clinical Chemotherapy, Japanese Foundation for Cancer Research, Tokyo, Japan
    Search for more papers by this author
  • Nobuyuki Mizunuma,

    1. Department of Medical Oncology, Cancer Institute Hospital of Japanese Foundation for Cancer Research, Tokyo
    Search for more papers by this author
  • Kiyohiko Hatake

    Corresponding author
    1. Department of Medical Oncology, Cancer Institute Hospital of Japanese Foundation for Cancer Research, Tokyo
    2. Cancer Chemotherapy Center, Clinical Chemotherapy, Japanese Foundation for Cancer Research, Tokyo, Japan
    Search for more papers by this author

To whom correspondence should be addressed. E-mail: khatake@jfcr.or.jp

Abstract

The purpose of this study was to investigate the potential of circulating tumor cells (CTC) as a surrogate marker of the clinical outcome in metastatic colorectal cancer (mCRC) patients in order to identify Japanese patients responsive to oxaliplatin-based chemotherapy. Between January 2007 and April 2008, 64 patients with mCRC were enrolled in this prospective study. The treatment regimen was oxaliplatin-based chemotherapy. Collection of CTC from whole blood was performed at baseline and at 2 and 8–12 weeks after initiation of chemotherapy. Isolation and enumeration of CTC was performed using immunomagnetics. Patients with ≥3 CTC at baseline and at 2 and 8–12 weeks had a shorter median progression-free survival (8.5, 7.3 and 1.9 months, respectively) than those with <3 CTC (9.7, 10.4 and 9.1 months, respectively) (log-rank test: P = 0.047, P < 0.001 and P < 0.001, respectively). Patients with ≥3 CTC at 2 and 8–12 weeks had a shorter median overall survival (10.2 and 4.1 months, respectively) than those with <3 CTC (29.1 and 29.1 months, respectively) (P < 0.001 and P = 0.001, respectively). A spurious early rise in carcinoembryonic antigen level was observed in 11 patients showing a partial response. In contrast, no rise in early CTC level was observed among responders. Our data support the clinical utility of CTC enumeration in improving our ability to accurately assess treatment benefit and in expediting the identification of effective treatment regimens for individual Japanese patients. (Cancer Sci 2011; 102: 1188–1192)

Circulating tumor cells (CTC) have been documented in the peripheral blood from patients with various cancers(1–3). Attempts to isolate CTC have led to the development of two leading procedures: density–gradient centrifugation(4–6) and flow cytometry(7). The number of CTC, as quantified by the CellSearch (Veridex LLC, Raritan, NJ, USA) methodology, has been shown to have prognostic significance in patients with breast cancer, prostate cancer and colorectal cancer, so recent efforts have concentrated on detecting CTC in the peripheral blood of cancer patients.

Cohen et al.(8) reported that the number of CTC before and during treatment was an independent predictor of progression-free survival (PFS) and overall survival (OS) in patients with metastatic colorectal cancer (mCRC). Detection of three or more CTC versus fewer than three CTC before and after initiation of a new systemic treatment regimen was associated with shorter median PFS and OS. These observations led us to conduct a validation study with the hypothesis that the number of CTC in Japanese patients relative to a threshold of three would correlate strongly with disease progression, allowing decisions on treatment efficacy to be made earlier than would normally be possible with imaging alone.

The American Society for Clinical Oncology recommends carcinoembryonic antigen (CEA) as the marker of choice for monitoring the response of metastatic disease to systemic therapy. However, Sorbye and Dahl(9) reported a transient increase in CEA level despite an objective response among patients receiving oxaliplatin-based chemotherapy for colorectal cancer.

In the present study, using the CellSearch system, we investigated the potential of CTC level in comparison with CEA level as a surrogate marker of clinical outcome in order to identify Japanese patients responsive to chemotherapy.

Materials and Methods

Patients.  All patients were enrolled using institutional review board-approved protocols at the Cancer Institute Hospital of the Japanese Foundation for Cancer Research. Informed consent was obtained from all patients. The study population consisted of patients aged 18 years or older with histologically proven mCRC. Other inclusion criteria were an Eastern Cooperative Oncology Group (ECOG) performance status (PS) of 0 or 1 and adequate organ function. The chemotherapy regimen was FOLFOX4 with or without bevacizumab.

Sample preparation for isolation of CTC from blood.  For isolation of CTC from mCRC patients, 10-mL samples of blood were drawn into a Cell Save Preservative Tube (Veridex LLC). Blood was drawn before initiation of treatment (baseline) and at 2 and 8–12 weeks after administration of FOLFOX4 with or without bevacizumab. The CellSearch system (Veridex LLC) consists of the CellPrep system, the CellSearch Epithelial Cell kit (for measurement of CTC) and the CellSpotter Analyzer. The CellPrep system is a semi-automated sample preparation system, and the CellSearch Epithelial Cell kit consists of ferrofluids coated with epithelial cell-specific EpCAM antibodies to immunomagnetically enrich epithelial cells, a mixture of two phycoerythrin-conjugated antibodies that bind to cytokeratin 8, 18 and 19, an antibody to CD45 conjugated to allophycocyanin, nuclear dye 4′,6-diamidino-2-phenylindole (DAPI) to fluorescently label the cell, and buffers to wash, permeabilize and resuspend the cells. Sample processing and evaluation were performed as described by Allan et al.(10) Briefly, 7.5 mL blood was mixed with 6 mL buffer, centrifuged at 800 g for 10 min and then placed on the CellPrep system. After aspiration of the plasma and buffer layer, ferrofluids were added. After incubation and subsequent magnetic separation, unbound cells and remaining plasma were aspirated. The staining reagents were then added in conjunction with a permeabilization buffer to fluorescently label the immunomagnetically labeled cells. After incubation in the system, magnetic separation was repeated and excess staining reagents aspirated. As the final step in the procedure, the cells were resuspended in the MagNest Cell Presentation Device (Veridex LLC). This device consists of a chamber and two magnets that orient the immunomagnetically labeled cells for analysis using the CellSpotter Analyzer.

Sample analysis.  The MagNest is placed on the CellSpotter Analyzer, a four-color, semi-automated fluorescence microscope. Image frames covering the entire surface of the cartridge are captured. Captured images containing objects that meet predetermined criteria are automatically presented in a web-enabled browser; an operator makes the final selection of cells. The criteria for an object to be defined as a CTC include round-to-oval morphology, a visible nucleus (DAPI-positive), positive staining for cytokeratin and negative staining for CD45. Results of cell enumeration are always expressed as the number of cells per 7.5 mL blood (Fig. 1).

Figure 1.

 Image galleries after CellSearch processing. Circulating tumor cells were cytokeratin (CK) and DAPI positive, but CD45 negative.

Statistical analysis.  Progression-free survival was defined as the time elapsed from blood collection to progression. Each time blood was collected, Kaplan–Meier survival plots were generated based on CTC levels and curves were compared using log-rank testing. A P-value of <0.05 was considered significant. The Cox proportional-hazards regression model was used to determine univariate and multivariate hazard ratios for selected potential predictors of PFS and OS. The distribution of patients above and below the CTC threshold and their clinical response were compared using the Fisher exact test.

Results

Patient characteristics.  A total of 64 patients were enrolled. Patient characteristics at baseline, which are summarized in Table 1, were as follows: median age, 59 years (range, 18–72 years); PS 0/1, 61/3; primary site rectum/colon, 36/28; and bevacizumab +/−, 31/33. Among the 64 patients, the objective response rate was 56%.

Table 1.   Patient characteristics
 Oxaliplatin-based regimen
  1. CR, complete response; PD, progressive disease; PR, partial response; PS, performance status; SD, stable disease.

Median age (range) (years)59 (18–72)
Sex (male/female)31/33
PS: 0/161/3
Primary site: rectum/colon36/28
No. lines: 1st/2nd49/15
Bevacizumab: +/−33/31
Site of metastasis
 Liver34
 Lung32
 Bone4
 Lymph node25
 Local9
 Peritoneum20
 Metastases to more than two organs46
Best objective response (CR/PR/SD/PD)2/34/20/8

CTC level and imaging to assess response to therapy.  Fifty-six of 64 patients were classified as having no progressive disease (PD) (non-PD, including stable disease, partial or complete response), with 47 of these patients having <3 CTC and nine patients having ≥3 CTC before initiation of therapy. Eight patients were classified as having PD, with five of these having <3 CTC and three having ≥3 CTC before initiation of therapy. The difference between the clinical response and CTC level was not significant. In contrast, 55 of 63 patients were classified as having non-PD, with 51 of these patients having <3 CTC and four patients having ≥3 CTC at 2 weeks. Eight of 63 patients were classified as having PD, with five of these having <3 CTC and three having ≥3 CTC at 2 weeks. The difference between the clinical response and CTC level was highly significant (= 0.038, Fisher’s exact test). Fifty-three of 60 patients were classified as having non-PD, with 52 of these patients having <3 CTC and one patient having ≥3 CTC at 8–12 weeks. Seven of 60 patients were classified as having PD, with four of these having <3 CTC and three having ≥3 CTC at 8–12 weeks. The difference between best overall response and CTC level was highly significant (= 0.004, Fisher’s exact test) (Table 2).

Table 2.   CTC and correlation with response assessment by imaging
 Non-PDPDFisher’s exact P
No. patientsCTC <3CTC ≥3No. patientsCTC <3CTC ≥3
  1. CTC, circulating tumor cells; PD, progressive disease.

Baseline564798530.164
2 weeks555148530.038
8–12 weeks535217430.004

Spurious early rise in CEA and CTC levels.  A spurious early rise in CEA level was observed in 11 patients showing a partial response. In contrast, no rise in CTC levels at 2 weeks was observed in any patient showing either a partial or complete response (Table 3).

Table 3.   Spurious early rise in CEA level or CTC level
 No. patients with a transient rise
CEA levelCTC level
  1. CEA, carcinoembryonic antigen; CR, complete response; CTC, circulating tumor cells; PR, partial response.

CR00
PR110

Analysis of PFS according to CTC level. Figure 2 shows the Kaplan–Meier plots for prediction of PFS using the CTC counts at baseline (Fig. 2a) and at 2 weeks (Fig. 2b) and 8–12 weeks (Fig. 2c). Patients with ≥3 CTC at baseline had a shorter median PFS (8.5 months; 95% CI, 7.4–9.6 months) than those with <3 CTC at baseline (9.7 months; 95% CI, 7.3–12.0 months) (P = 0.047) (Fig. 2a). Patients with ≥3 CTC at 2 weeks had a shorter median PFS (7.3 months; 95% CI, 0–21.0 months) than those with <3 CTC at 2 weeks (10.4 months; 95% CI, 7.5–13.3 months) (P < 0.001) (Fig. 2b). Patients with ≥3 CTC at 8–12 weeks had a shorter median PFS (1.9 months; 95% CI, 0.5–3.3 months) than those with <3 CTC at 8–12 weeks (9.1 months; 95% CI, 7.6–10.7 months) (P < 0.001) (Fig. 2c).

Figure 2.

 Kaplan–Meier plots of progression-free survival in metastatic colorectal cancer patients with fewer than three circulating tumor cells (CTC) or ≥3 CTC at baseline (a), 2 weeks (b) and 4 weeks (c).

Analysis of OS according to CTC level. Figure 3 shows the Kaplan–Meier plots for prediction of OS using baseline CTC counts (Fig. 3a) at 2 weeks (Fig. 3b) and at 8–12 weeks (Fig. 3c). A shorter median OS was observed in patients who had ≥3 CTC at all time points. Patients with ≥3 CTC at 2 weeks had a significantly shorter median OS (10.2 months; 95% CI, 0–25.7 months) than those with <3 CTC at 2 weeks (29.1 months; 95% CI, 21.5–36.8 months) (P < 0.001) (Fig. 3b). Patients with ≥3 CTC at 8–12 weeks had a significantly shorter median OS (4.1 months; 95% CI, 0–11.7 months) than those with <3 CTC at 8–12 weeks (29.1 months; 95% CI, 20.3–38.0 months) (P = 0.001) (Fig. 3c).

Figure 3.

 Kaplan–Meier plots of overall survival in metastatic colorectal cancer patients with fewer than three circulating tumor cells (CTC) or ≥3 CTC at baseline (a), 2 weeks (b) and 4 weeks (c).

Univariate and multivariate analysis of predictors of PFS and OS.  Univariate and multivariate Cox proportional-hazards regression was performed to assess the association between factors of interest and PFS or OS. In the univariate analyses, PS, lung metastasis, bevacizumab and CTC levels at baseline and at 2 and 8–12 weeks predicted PFS, and PS, bevacizumab and CTC levels at 2 and 4 weeks predicted OS (Table 4). In order to evaluate the independent predictive effect of chemotherapy, multivariate Cox regression analysis was carried out (Table 5). Levels of CTC at 2 and 4 weeks were the strongest predictors.

Table 4.   Independent predictive factors by univariate Cox regression analysis for PFS and OS
ParameterNo. patientsHR95% CIP-valueχ2
  1. CI, confidence interval; CTC, circulating tumor cells; HR, hazard ratio; OS, overall survival; PFS, progression-free survival; PS, performance status.

PFS
 PS644.4181.338–14.5890.0150.008
 Bevacizumab: +/−640.2760.155–0.493<0.001<0.001
 Lung metastasis641.9881.138–3.4730.0160.016
 CTC at baseline641.0851.026–1.1470.0040.003
 CTC at 2 weeks631.1791.089–1.276<0.001<0.001
 CTC at 8–12 weeks611.2111.099–1.334<0.001<0.001
OS
 PS6420.4165.172–80.592<0.001<0.001
 Bevacizumab: +/−640.4490.222–0.9050.0250.022
 CTC at 2 weeks631.1921.090–1.303<0.001<0.001
 CTC at 8–12 weeks611.3391.119–1.601<0.001<0.001
Table 5.   Independent predictive factors by multivariate Cox regression analysis for PFS and OS
ParameterNo. patientsHR95% CIP-valueModel χ2
  1. CI, confidence interval; CTC, circulating tumor cells; HR, hazard ratio; OS, overall survival; PFS, progression-free survival; PS, performance status.

PFS
 PS6474.4210.063–550.35<0.001<0.001
 Liver metastasis1.8971.057–3.4060.032
 Bone metastasis0.1360.024–0.7590.023
 Bevacizumab: +/−0.1690.088–0.324<0.001
 CTC at baseline 1.0580.977–1.1450.164 
 PS630.080.014–0.4620.005<0.001
 LN metastasis0.5420.297–0.986 
 Bevacizumab: +/−0.1620.081–0.322<0.001
 CTC at 2 weeks 1.1441.047–1.2510.003 
 Lung metastasis611.8361.013–3.3290.045<0.001
 Bevacizumab: +/−0.2690.147–0.492<0.001
 CTC at 8–12 weeks 1.2111.092–1.344<0.001 
OS
 PS6346.1949.401–226.971<0.001<0.001
 Peritoneum2.7871.331–5.8390.007
 Bevacizumab: +/−0.4680.221–0.9900.047
 CTC at 2 weeks 1.2361.100–1.387<0.001 
 PS6122.1422.415–203.0350.006<0.001
 Bevacizumab: +/−0.3460.154–0.7770.010
 CTC at 8–12 weeks1.4411.143–1.8170.002

Discussion

To our knowledge, this is the first study to validate the clinical use of CTC for monitoring the response of mCRC to systemic therapy in Japanese patients. A cut-off of three CTC was chosen based on the results of an earlier study by Cohen et al.(8) We determined the relationship among patients with no CTC, those with one or two CTC and those with three or more CTC on the clinical outcome. Patients with one or two CTC had a similar PFS and OS to those with more than three CTC at baseline. In contrast, patients with one or two CTC at 2 and 8–12 weeks had a similar PFS and OS to those with no CTC (Table 6). A cut-off of three CTC not before but during treatment was an independent predictor of PFS and OS in patients with mCRC in the present study. Therefore, we analyzed the relationship between the change in CTC levels from baseline to 2 or 8–12 weeks and the clinical outcome in oxaliplatin-based chemotherapy. Kaplan–Meier plots were generated for patients with <3 CTC at both time points (group 1), patients with three or more CTC at baseline and fewer than three CTC at 2 or 8–12 weeks (group 2), patients with fewer than three CTC at baseline and three or more CTC at 2 or 8–12 weeks (group 3), and patients with three or more CTC at both time points (group 4). Median PFS in group 2 was not significantly different from that in group 1. However, the median PFS in group 2 was significantly longer than that in groups 3 or 4 (Fig. 4). Median OS in group 2 was not significantly different from that in group 1. However, the median OS in group 2 was significantly longer than that in groups 3 or 4 (Fig. 5). The results of the present study clearly show that persistent achievement of fewer than three CTC at 2 weeks after initiating chemotherapy is a strong indicator that the current therapy is effective, whereas three or more CTC is a strong indicator that any benefits are likely to be short-term only. These data suggest that CTC counts are valuable in the identification of chemotherapy-resistant patients, irrespective of ethnicity, who could thus benefit from early treatment change and/or different investigational approaches.

Table 6.   Relationship between CTC levels and outcome
 0 CTC1 or 2 CTC≥3 CTCP-value
Median PFS (95% CI)
Baseline11.7 (9.5–13.9)7.3 (5.4–9.2)8.5 (7.4–9.6)0.002
2 weeks11.3 (7.9–14.2)9.7 (2.2–17.1)7.3 (0–21.0)<0.001
8–12 weeks9.7 (6.3–13.0)8.7 (2.6–14.7)1.9 (0.5–3.3)<0.001
Median OS (95% CI)
  1. CI, confidence interval; CTC, circulating tumor cells; OS, overall survival; PFS, progression-free survival.

Baseline31.1 (27.2–34.9)18.7 (5.8–31.6)15.1 (12.3–17.8)0.058
2 weeks29.1 (22.1–36.2)22 (12.9–31.1)10.2 (0–25.7)0.001
8–12 weeks31.1 (20.2–41.9)23.3 (20.4–26.2)4.1 (0–11.7)<0.001
Figure 4.

 Kaplan–Meier plots of progression-free survival in metastatic colorectal cancer patients with circulating tumor cell (CTC) change from baseline to 2 weeks (a) and 8–12 weeks (b).

Figure 5.

 Kaplan–Meier plots of overall survival in overall survival metastatic colorectal cancer patients with circulating tumor cells (CTC) change from baseline to 2 weeks (a) and 8–12 weeks (b).

The 2006 update of ASCO recommended CEA as the marker of choice for monitoring the response of metastatic disease to systemic therapy. However, caution should be exercised in interpreting a rise in CEA level during the first 4–6 weeks of a new therapy, as a spurious rise might occur early on in treatment, especially with oxaliplatin(11). We observed a transient increase in CEA level in 11 patients, despite an objective response among those receiving oxaliplatin-based chemotherapy. The observation here of a transient increase in CEA level, even among patients responsive to oxaliplatin-based chemotherapy, agrees with the results of Locker et al.(11) In contrast, to our knowledge, no other studies to date have reported such a surge phenomenon in the CTC levels in patients receiving oxaliplatin-based chemotherapy for mCRC. These results suggest that CTC are a more effective marker than CEA for monitoring the response of metastatic disease to systemic therapy.

The strongest data have been provided by analyses from several prospective studies(12–14) that used the US Food and Drug Administration-approved CellSearch system. A previous study showed that CTC detection also provided significant prognostic information for patients with advanced gastric cancer.(15) However, the CellSearch system is yet to be approved for use in Japan. We anticipate that CTC counts for monitoring patients with colorectal, gastric, breast and prostate cancer will eventually be approved in Japan.

In conclusion, our data support the clinical utility of CTC enumeration in improving our ability to accurately assess the treatment benefit and in expediting the identification of effective treatment regimens for individual Japanese patients. In further studies, patients should be randomly assigned to continue current therapy or start a new treatment regimen if they have three or more CTC at 2 weeks before typical imaging intervals.

Acknowledgments

This work was supported by AstraZeneca Research Grant 2007, the Kobayashi Institute for Innovative Cancer Chemotherapy and a Grant-in-Aid for Scientific Research (Japan Society for the Promotion of Science) (grant numbers 19790963, 21591741 and 17016077). The excellent technical assistance of Sayuri Minowa, Harumi Shibata and Mariko Kimura (Cancer Chemotherapy Center, Clinical Chemotherapy, Japanese Foundation for Cancer Research, Tokyo, Japan) is greatly appreciated.

Disclosure Statement

The authors have no conflict of interest.

Ancillary