Prognostic impact of hematogenous tumor cell dissemination in patients with

stage II colorectal cancer

Moritz Koch1, Peter Kienle1, Dorejd Kastrati1, Dalibor Antolovic1, Jan Schmidt1, Christian Herfarth1,Magnus von Knebel Doeberitz2 and J€urgen Weitz1*

1Department of Surgery, University of Heidelberg, Heidelberg, Germany2Department of Molecular Pathology, University of Heidelberg, Heidelberg, Germany

Adjuvant chemotherapy is not routinely recommended in patientswith colorectal cancer stage UICC II. Some of these patients, how-ever, develop recurrent disease. Therefore, valid prognostic crite-ria are needed to identify high-risk patients who might benefitfrom adjuvant therapy. Disseminated tumor cells, detected inblood and bone marrow, may prove to be a valid marker, how-ever, the prognostic relevance of these cells remains debated. Inour study, we examined the prognostic significance of dissemi-nated tumor cells in blood and bone marrow of patients with stageII colorectal cancer. Ninety patients with potentially curative (R0)resection of colorectal cancer stage II were prospectively enrolledinto the study. Bone marrow and blood samples were examinedfor disseminated tumor cells by CK 20 RT-PCR. Patient, tumorand treatment factors were analyzed as prognostic factors. Multi-variate analysis confirmed tumor cell detection in blood (hazardratio 2.1, p 5 0.03) and T-category (hazard ratio 2.2, p 5 0.02) tobe independent prognostic factors for relapse-free survival. Tu-mor cell detection in postoperative blood samples (hazard ratio7.7, p < 0.001) and number of removed lymph nodes (hazard ratio6.4, p < 0.001) were independent prognostic factors for disease-specific survival. Detection of circulating tumor cells in blood sam-ples of patients with stage II colorectal cancer identifies patientswith poor outcome. This finding should be confirmed by furtherstudies and could then be used as a basis for conducting a random-ized trial evaluating the effect of adjuvant chemotherapy in stageII patients.' 2006 Wiley-Liss, Inc.

Key words: colorectal cancer; stage II; RT-PCR; prognosis

Colorectal cancer is one of the most common malignanciesworldwide with �490,000 of affected patients dying every year.1

Patients with colorectal cancer UICC stage II are generally consid-ered to be at low risk for developing recurrent or metastatic dis-ease.2 Therefore, patients with colon cancer in this stage, in con-trast to patients with positive lymph nodes (UICC stage III), arenot recommended to undergo routine adjuvant chemotherapy.2–6

For patients with rectal cancer, there is now a trend towards neo-adjuvant radiation therapy. In rectal cancer patients with negativelymph node status, who have undergone neoadjuvant radiation, thevalue of postoperative chemotherapy again is not yet adequatelydefined.3 Despite the low tumor stage some of the stage II patientssubsequently develop recurrent disease. It remains under debate,whether all stage II patients or at least stage II patients with addi-tional risk factors should receive adjuvant chemotherapy.7–9 It istherefore of high importance to define prognostic criteria for thispatient group to help identify high-risk patients for tumor relapse,who might benefit from adjuvant therapeutic regimes.10,11

Many molecular markers have been evaluated as putative prog-nostic markers in patients with colorectal cancer; however, none ofthese are currently in clinical use regarding the decision whether apatient with colorectal cancer stage II should receive adjuvantchemotherapy.1 Tumor cell dissemination and formation of meta-stases is known to be a sequence of complex events, therefore, sin-gle molecular markers might not be able to serve as adequateprognostic markers. In addition, other factors, such as the immunecompetence of the patients, will influence prognosis.12 We, havetherefore used a different approach hypothesizing that detectionof circulating tumor cells might identify patients at high risk forrelapse. The potential advantage of this approach is that tumor and

patient related factors that may be of prognostic relevance wouldboth influence the presence of circulating tumor cells, therefore,detection of these cells is a logical approach to identify patients athigher oncological risk. PCR based protocols have been proven tobe sensitive and specific assays for detecting disseminated colo-rectal cancer cells. The prognostic significance of these cells, how-ever, remains under critical debate in the literature.13 Recently, wedemonstrated the sensitivity and specificity of a CK 20 reversetranscriptase-polymerase chain reaction (RT-PCR) assay in detect-ing colorectal cancer cells in blood and bone marrow.14–19

The aim of our prospective study was to evaluate the prognosticsignificance of disseminated tumor cells detected by CK 20 RT-PCRin blood and bone marrow in patients with stage II colorectal cancer.

Material and methods

Patients

Included in this study were 90 consecutive patients undergoingelective curative (R0) resection of colorectal adenocarcinomastage II at the Department of Surgery, University of Heidelberg,Germany, between September 1995 and February 2000. Patientswith other malignant disease in their medical history were ex-cluded. Informed consent was obtained from all patients; the studyprotocol was approved by the ethics committee of the Universityof Heidelberg. The PCR-results of a subgroup of these patients (n525) were already reported in earlier works from our group.14,15

Tumors located less than 16 cm above the anal verge were clas-sified as rectal cancers. Tumor height of rectal cancers was deter-mined using rigid endoscopy performed by the surgeon preopera-tively. All patients were staged preoperatively with colonoscopy,abdominal ultrasound and/or CT scan and chest X-ray.

No patient received treatment prior to surgery. According to rec-ommendations for stage II patients, valid at the time patients wereincluded in the study, adjuvant therapy was not recommended forpatients with colon cancer and adjuvant chemoradiation was recom-mended for patients with rectal cancer. Adjuvant therapy was for themost part performed at the discretion of the local oncologist.

The length of follow-up was calculated from the date of opera-tion at our institution. Follow-up was done according to the guide-lines of the German Cancer Society, which consists of periodicclinical evaluation, serum CEA levels, imaging of the abdomen,chest X-ray and colonoscopy.

Occurrence of distant metastases and local recurrences, follow-up time and reason for death was obtained for each patient toassess relapse-free and disease-specific survival.

Tumor stage and grade were classified according to the 5th edi-tion of the TNM classification of the UICC (International UnionAgainst Cancer).20

*Correspondence to: Head Section of Surgical Oncology, Departmentof Surgery, University of Heidelberg, Im Neuenheimer Feld 110, D-69120Heidelberg, Germany. Fax:149-6221-565506.E-mail: juergen_weitz@med.uni-heidelberg.deReceived 27 July 2005; Accepted after revision 18 November 2005DOI 10.1002/ijc.21784Published online 19 January 2006 inWiley InterScience (www.interscience.

wiley.com).

Int. J. Cancer: 118, 3072–3077 (2006)' 2006 Wiley-Liss, Inc.

Publication of the International Union Against Cancer

Blood and bone marrow samples

Through a central venous catheter in the superior vena cava,which is routinely placed during colorectal cancer operations inour hospital, 3 blood samples (10 ml) were obtained from eachpatient: the first after induction of anesthesia, the second afterresection of the tumor and the third 24 hr after the operation.

Bone marrow samples (10 ml) were obtained after induction ofgeneral anesthesia by aspiration from both iliac crests. The bloodand bone marrow samples were diluted with 10 ml phosphate buf-fered saline (PBS). After density centrifugation through Ficoll-Paque (Pharmacia) (30 min, 400 g), mononuclear cells were har-vested from the interphase and washed twice in PBS. The cell pelletwas then shock frozen in liquid nitrogen and stored at270�C.Seventy-one patients consented to bone marrow puncture. Of

the blood samples, 89 preoperative samples, 88 intraoperativesamples and 82 postoperative samples could be evaluated by CK20 RT-PCR. The missing data in regard to the blood samples areeither due to unsuccessful RNA extraction or reverse transcriptionor due to organizational or patient related reasons, resulting innon-acquisition of the blood samples.

RNA extraction, reverse transcriptase-polymerase chain reaction

RNA extraction from peripheral mononuclear blood cells,from bone marrow samples and frozen tissue sections of tumorswas done as previously described.14 CK 20-RT-PCR was per-formed as previously described.14 PCR-products were analyzedby electrophoresis on 2% agarose gels. RNA quality and per-formance of reverse transcription of all analyzed samples wasconfirmed by RT-PCR amplification of glyceraldehyde phosphatedehydrogenase (GAPDH) transcripts as previously described.14

Sensitivity and specificity of CK 20 RT-PCR

The sensitivity of the CK 20 RT-PCR assay was determined inprevious cell spiking experiments, allowing the detection of 10 co-lon cancer cells (HT 29) in 10 ml blood.14

The specificity of the CK 20-RT-PCR assay was determined inprevious studies, in which 174 blood samples from 98 healthy indi-viduals and bone marrow samples from 30 patients without malig-nant disease consistently tested negative for CK 20 expression.14,15,21

TABLE I – DISTRIBUTION OF CLINICAL, PATHOLOGICAL AND TREATMENTFACTORS IN 90 PATIENTS WITH STAGE II COLORECTAL

CANCER UNDERGOING R0-RESECTION

Variable No. ofpatients

%

Sex Male 59 66Female 31 34

Age �65 43 48>65 47 52

Tumor position Colon 52 58Rectum 38 42

T-category T3 80 89T4 10 11

Grading G2 65 72G3 25 28

Tumor size �5cm 53 59<5cm 37 41

Lymphaticvessel invasion

Present 7 8Absent 83 92

Mucin producingcarcinoma

Yes 17 19No 73 81

Number of removedlymph nodes

�12 lymph nodes 80 89<12 lymph nodes 10 11

Adjuvant therapy None 62 69Radiotherapy 2 2Chemotherapy 5 6Chemoradiation 21 23

TABLE II – UNIVARIATE ANALYSIS OF PROGNOSTIC FACTORS IN 90 PATIENTS WITH STAGE IICOLORECTAL CANCER UNDERGOING R0-RESECTION

RFS1 DSS2

5-yearsurvival

p-value3 5-yearsurvival

p-value3

Sex Male 89 ns (0.12) 97 ns (0.09)Female 79 86

Age �65 84 ns (0.4) 93 ns (0.9)>65 88 94

Tumor position Colon 86 ns (0.8) 95 ns (0.5)Rectum 86 91

T-category T3 89 0.006 97 ns (0.12)T4 60 74

Grading G2 89 ns (0.13) 95 ns (0.8)G3 77 90

Tumor size �5cm 88 ns (0.4) 97 ns (0.7)<5cm 83 89

Lymphaticvessel invasion

Present 83 ns (0.9) 75 ns (0.4)Absent 86 95

Mucin producingcarcinoma

Yes 86 ns (0.9) 93 ns (0.9)No 87 93

Number of removedlymph nodes

�12 lymph nodes 88 0.04 97 <0.0001<12 lymph nodes 69 60

Adjuvant Therapy None 86 ns (0.9) 96 ns (0.7)Radiotherapy 100 100Chemotherapy 80 80Chemoradiation 85 88

Tumor cell detectionany blood sample

Yes 79 0.03 87 ns (0.06)No 94 100

Tumor cell detectionpreoperative blood

Yes 72 0.05 87 ns (0.5)No 90 95

Tumor cell detection Yes 72 0.05 92 ns (0.6)intraoperative blood No 90 94

Tumor cell detection Yes 70 0.003 77 0.0006postoperative blood No 93 100

Tumor cell detection Yes 95 ns (0.3) 94 ns (0.8)bone marrow No 83 91

1RFS, relapse-free survival.–2DSS, disease-specific survival.–3Log-rank test. ns, not significant.

3073PROGNOSIS OF STAGE II COLORECTAL CANCER

Statistical analyses

Statistical computations were done using the software packageJMP (JMP, Cary, NC). Survival was estimated according to theKaplan-Meier method and compared using the log-rank test.22 Amultivariate proportional hazards model was built using the varia-bles that were significant prognostic parameters in the univariateanalysis (p � 0.05 at univariate analysis).23 A result was consid-ered statistically significant when the p-value was less than orequal to 5% (p � 0.05).

Results

Patient characteristics

The patient characteristics of the 90 patients included in thisstudy are displayed in Table I. The median age of the patients was66 (range: 29–85 years). The median number of lymph nodesremoved was 19 (range: 6–65). Of 52 patients with colon cancer,5 received adjuvant 5-FU-based chemotherapy; of 38 patients withrectal cancer, 21 received adjuvant chemoradiation and 2 adjuvantradiation therapy.

Tumor cell detection in blood and bone marrow samples

Tumor cells were detected in preoperative blood samples of 21of 89 (24%) patients. Twenty-two of 88 (25%) patients revealeddisseminated tumor cells in blood samples taken intraoperatively.Twenty-eight of 82 (34%) patients showed tumor cells in post-operative blood samples (taken 24 hr after the operation). Of the11 patients with a tumor recurrence, 9 patients (82%) had at least1 positive blood sample; 5 of 11 patients (45%) revealed a positivepreoperative blood sample, 5 of 11 patients (45%) a positive intra-operative blood sample and 8 of 11 patients (73%) a positive post-operative blood sample.

Twenty of 71 (28%) patients tested positive for disseminated tu-mor cells in their bone marrow samples. Patients with positiveblood samples had positive bone marrow samples in 35% of casescompared to 20% in patients with negative blood samples (p5 ns).

Follow-up and univariate characterization of prognostic factors

The median follow-up for all patients was 58 months with arange from 3 to 81 months. The actuarial 5-year relapse-free sur-vival was 85% (95% CI, 77–93%), and the actuarial 5-year dis-ease-specific survival was 93% (95% CI, 87–99%). At the time ofthe last follow-up, 70 patients (78%) had no evidence of disease (5of these patients had recurred, however, were rendered free ofdisease by surgical resection), 6 patients (7%) had died of diseaseand 12 patients (13%) had died of unrelated causes. Two patients(2%) were lost to follow-up; these patients were censored at the timeof last follow up (35 and 50 months postoperatively). Table II de-picts the univariate analysis of the different factors for relapse-freeand disease-specific survival. Regarding relapse-free survivalT-category, number of removed lymph nodes and tumor cell de-tection in blood samples showed prognostic relevance. The actua-rial 5-year relapse-free survival was 93% compared to 70% inpatients with negative versus positive postoperative blood samples(p 5 0.003) (Fig. 1). Number of removed lymph nodes and tumorcell detection in postoperative blood samples were also prognosticcriteria for disease-specific survival. In patients with negativepostoperative blood samples, the actuarial 5-year disease-specificsurvival was 100% compared to 77% in patients with a positivepostoperative blood sample (p 5 0.0006) (Fig. 1). Positive postop-erative blood samples were associated with a worse disease-spe-cific survival irregardless of whether the patients had received ad-juvant therapy or not. This was also the case when looking at co-lon and rectal cancer patients separately (Figs. 2 and 3). Positivepostoperative blood samples were also associated with a worse relapse-free survival in patients with rectal cancer and those receiving adju-vant therapy.

Multivariate characterization of prognostic factors

Separate multivariate analyses were performed for tumor celldetection in any blood sample (pre-, intra- and postoperative sam-ple) and for tumor cell detection in postoperative blood samples.

The model for relapse-free survival is shown in Table III. Thismodel identified tumor cell detection in blood samples as an inde-pendent risk factor. The T-category reached statistical significanceonly in the model analyzing tumor cell detection in any bloodsample. The number of removed lymph nodes failed to reach sta-tistical significance, but showed a strong trend.

Table IV depicts the results of the multivariate model for dis-ease-specific survival. Number of removed lymph nodes, and tu-mor cell detection in postoperative blood samples were indepen-dently associated with disease-specific survival.

Discussion

In contrast to colorectal cancer patients with positive lymphnodes (stage III), where the benefit of adjuvant chemotherapy hasbeen proven, the use of adjuvant chemotherapy for patients withstage II remains controversial. This is due to the favorable progno-sis of stage II patients; optimized surgery without adjuvant chemo-therapy can result in disease-specific survival rates of over 90% in

FIGURE 1 – Effect of tumor cell detection in postoperative bloodsamples on relapse-free and disease-specific survival.

3074 KOCH ET AL.

this patient group.24 Therefore, many patients who will never de-velop recurrent disease will have to undergo adjuvant chemother-apy to treat one patient who might benefit from adjuvant therapy.This dilemma is reflected by the available data from randomizedtrials studying the effect of adjuvant chemotherapy in colorectalcancer stage II patients; no consistent significant benefit of adju-vant chemotherapy has yet been demonstrated.3–8,10 Some authors,however, suggest that chemotherapy can lead to a similar improve-ment of survival for stage II patients as demonstrated for stage IIIpatients.25 For patients with rectal cancer, the situation is evenmore complicated as there are several different radiation regimesavailable. For patients receiving neoadjuvant radiation, the role ofadjuvant chemotherapy is currently unclear, again underlining theimportance of defining valid prognostic criteria for these patients.Prognostic markers may help to identify a subgroup of stage IIpatients at high risk for disease relapse who may also benefit fromadjuvant therapy. Several clinical, pathological and molecularprognostic factors have been discussed in the literature; however,

no single factor has yet been identified which is generally acceptedas an indicator for initiating adjuvant chemotherapy.1,4,8,11,25,26

Using single molecular markers to describe the prognosis ofpatients with colorectal cancer might not be successful, as tumorcell dissemination and formation of metastases are the result ofmultiple steps involving tumor and patient related factors. In addi-tion, cancers may be very heterogeneous; therefore, assessment ofthe primary tumor may not accurately predict the presence or be-havior of disseminated cancer cells, which are the primary target ofadjuvant chemotherapy. The occurrence of circulating tumor cellsis the result of 2 basic competing factors: first, the ability of the can-cer cells to disseminate, and secondly, the effectiveness of tumorcell elimination mechanisms, especially of the immune system.Therefore, detection of circulating cancer cells seems to be a logicalapproach to assess prognosis of the patients to define surrogatemarkers for indicating adjuvant therapy.

Lindemann et al. were able to detect disseminated colorectalcancer cells in bone marrow aspirates of patients with colorectal

FIGURE 2 – Effect of tumor cell detection in postoperative bloodsamples on disease-specific survival regarding to adjuvant therapy. FIGURE 3 – Effect of tumor cell detection in postoperative blood

samples on disease-specific survival in colon and rectal cancer patients.

3075PROGNOSIS OF STAGE II COLORECTAL CANCER

cancer using immunocytological methods.27 The authors also dem-onstrated the prognostic relevance of these cells; however, otherauthors could not confirm this finding making the detection ofthese cells not a generally accepted prognostic marker.13,26,28 Theprognostic impact of tumor cells circulating in the blood stream iseven more controversial. Opponents of the prognostic value of cir-culating tumor cells argue that implantation of circulating tumorcells is a very inefficient process with less than 0.01% of circulat-ing tumor cells actually establishing metastases.29,30 Similar to tu-mor cell detection in lymph nodes and bone marrow, detection oftumor cells in the blood stream is still not an accepted prognosticmarker.13,26,28 A major problem of most of the published studies isthat only small, inhomogeneous patient groups with short follow-up periods were evaluated. Recently published studies report con-flicting results regarding the prognostic value of circulating tumorcells.31–34 Moreover, the methods used for tumor cell detectionalso need to be taken into account, as sensitivity and specificity areof major importance and may differ significantly.

Our study prospectively examined a homogenous group of patientswith stage II colorectal cancer undergoing R0-resection with a me-dian follow up of almost 5 years. We used a CK 20 RT-PCR assayfor tumor cell detection in blood and bone marrow samples. In ourprevious studies, we have demonstrated the sensitivity and specificityof this assay. In these studies, 174 blood samples of 98 individualsand bone marrow samples of 30 patients without malignant diseaseconsistently tested negative for CK 20 expression.14–16 Blood sam-ples taken intraoperatively from patients having surgery for benigncolorectal diseases also revealed a negative finding in these studies.

Some authors argue that colon and rectal cancer should not bestudied together as they might show a different biologic behavior.We, therefore, analyzed site of tumor (colon versus rectum) as apotential prognostic marker. No prognostic relevance could bedemonstrated in our patient cohort on univariate analysis; there-

fore, site of tumor did not influence the outcome in this study andwas not included in the multivariate model.

When evaluating a new prognostic marker, it is essential to de-monstrate independence from other prognostic parameters by multi-variate analysis. Multivariate analysis revealed 3 independent prog-nostic markers in our patient cohort: number of removed lymphnodes, T-category and tumor cell detection in the blood stream. Thefirst 2 parameters are well-defined markers, with proven prognosticrelevance in the literature.26,35 The present study is the first studydemonstrating tumor cell detection in blood samples to be an inde-pendent prognostic marker in stage II colorectal cancer patients forboth relapse-free and disease-specific survival. Of special interest isthat tumor cell detection in blood obtained 24 hr postoperatively wasthe most relevant time point of sampling in regard to prognosis. The5-year disease-specific survival for patients with a negative postoper-ative blood sample was 100% compared to only 77% in patients withdetectable tumor cells at that sampling time. It is well known fromthe literature that the majority of circulating tumor cells are clearedfrom the blood stream within 24 hr.36,37 Possible reasons for the per-sistence of disseminated tumor cells in the circulation may be animmune escape mechanism or a higher malignant potential of thisspecific clone. One could therefore contemplate that circulating tu-mor cells which survive for a longer period of time in the bloodstream have higher prognostic relevance. In contrast to the prognos-tic significance of tumor cells detected in the blood stream, we werenot able to find the same effect for tumor cells detected in bone mar-row samples. We can only speculate on the reason for this discrepan-cy. It is well known that the vast majority of tumor cells in the bonemarrow are in a dormant state and do not proliferate.38 One couldhypothesize that tumor cells that are able to survive in the bloodstream as described above might have a greater prognostic impactcompared to dormant tumor cells in the bone marrow. Further stud-ies have to clarify this issue.

In conclusion, the results of this study support a prognostic sig-nificance of tumor cells detected in blood samples of patients withstage II colorectal cancer. These findings, once confirmed by fur-ther studies, could be used as a basis for conducting a randomizedtrial evaluating the effect of adjuvant chemotherapy in stage IIpatients with proven circulating tumor cells.

Acknowledgements

The authors thank Professor M.W. B€uchler, M.D. for his sup-port of our work.

References

1. Weitz J, Koch M, Debus J, H€ohler T, Galle PR, B€uchler MW. Colo-rectal cancer. Lancet 2005;365:153–65.

2. McDonald JS. Adjuvant therapy of colorectal cancer. CA CancerJ Clin 1999;49:202–19.

3. Chawla AK, Kachnic LA, Clark JW, Willett CG. Combinedmodality therapy for rectal and colon cancer. Semin Oncol 2003;30:101–12.

4. Haydon A. Adjuvant chemotherapy in colon cancer: what is the evi-dence? Intern Med J 2003;33:119–24.

5. Benson AB, Schrag D, Somerfield MR, Cohen AM, Figueredo A,Flynn PJ, Krzyzanowska MK, Maroun J, McAllister P, van Cutsem E,Brouwers M, Charette M, et al. American Society of Clinical Oncol-ogy recommendations on adjuvant chemotherapy for stage II coloncancer. J Clin Oncol 2004;22:3408–19.

6. Figueredo A, Charette M, Maroun J, Brouwers M, Zuraw L. Adjuvanttherapy for stage II colon cancer: a systematic review from the cancercare Ontario program in evidence-based care’s gastrointestinal cancerdisease site group. J Clin Oncol 2004;22:3395–407.

TABLE III – MULTIVARIATE ANALYSIS OF FACTORS ASSOCIATED WITH RELAPSE-FREE SURVIVALINCLUDING TUMOR CELL DETECTION IN ANY BLOOD SAMPLE (MODEL I) OR IN

POSTOPERATIVE BLOOD SAMPLES (MODEL II)

Model I Model II

Hazardratio (95% CI)

p-value Hazardratio (95% CI)

p-value

T-category T3 Reference 0.02 Reference ns (0.15)T4 2.2 (1.1, 4.1) 1.8 (0.8, 3.2)

Number of removedlymph nodes

�12 Reference ns (0.16) Reference ns (0.08)<12 1.7 (0.8, 3.2) 2.0 (0.9, 4.0)

Tumor cell detection No Reference 0.03any blood sample Yes 2.1 (1.1, 5.4)

Tumor cell detection No Reference 0.008Postop. blood sample Yes 2.4 (1.3, 5.3)

ns, not significant.

TABLE IV – MULTIVARIATE ANALYSIS OF FACTORS ASSOCIATEDWITH DISEASE-SPECIFIC SURVIVAL

Hazard ratio(95% CI)

p-value

Number of removedlymph nodes

�12 Reference 0.0004<12 7.7 (2.5, 38)

Tumor cell detection No Reference 0.0003Postop. blood sample Yes 6.4 (2.2, 34)

3076 KOCH ET AL.

7. Wein A, Hahn EG, Merkel S, Hohenberger W. Adjuvant chemother-apy for stage II (Dukes’B) colon cancer: too early for routine use. EurJ Surg Oncol 2000;26:730–2.

8. Mamounas EP. Adjuvant chemotherapy for stage II colon cancer: thetime has come. Eur J Surg Oncol 2000;26:725–9.

9. Weitz J, Koch M, Debus J, H€ohler T, Galle PR, B€uchler MW. Colo-rectal cancer. Lancet, in press.

10. Cascinu S, Georgoulias V, Kerr D, Maughan T, Labianca R, Ychou M.Colorectal cancer in the adjuvant setting: perspectives on treatment andthe role of prognostic factors. Ann Oncol 2003;14:25–9.

11. Graziano F, Cascinu S. Prognostic molecular markers for planningadjuvant chemotherapy trials in Dukes’B colorectal cancer patients:how much evidence is enough? Ann Oncol 2003;14:1026–38.

12. Menon AG, van Rhihn CM, Morreau H, Putter H, Tollenaar RAEM,van de Velde CJH, Fleuren GJ, Kuppen PJK. Immune system andprognosis in colorectal cancer: a detailed immunohistochemical anal-ysis. Lab Invest 2004;84:493–501.

13. Kienle P, Koch M. Minimal residual disease in gastrointestinal can-cer. Semin Surg Oncol 2001;20:282–93.

14. Weitz J, Kienle P, Lacroix J, Willeke F, Benner A, Lehnert T,Herfarth Ch, von Knebel Doeberitz M. Dissemination of tumor cellsin patients undergoing surgery for colorectal cancer. Clin Cancer Res1998;4:343–8.

15. Weitz J, Kienle P, Magener A, Koch M, Schr€odel A, Willeke F,Autschbach F, Lacroix J, Lehnert T, Herfarth Ch, von Knebel Doeber-itz M. Detection of disseminated colorectal cancer cells in lymphnodes, blood and bone marrow. Clin Cancer Res 1999;5:1830–6.

16. Weitz J, Koch M, Kienle P, Schr€odel A, Willeke F, Benner A, Lehnert T,Herfarth Ch, von Knebel Doeberitz M. Detection of hematogenic tumorcell dissemination in patients undergoing resection of liver metastasesof colorectal cancer. Ann Surg 2000;232:66–72.

17. Kienle P, Koch M, Autschbach F, Benner A, Treiber M, Wannen-macher M, von Knebel Doeberitz M, B€uchler MW, Herfarth Ch,Weitz J. Decreased detection rate of disseminated tumor cells ofrectal cancer patients after preoperative chemoradiation—a first steptowards a molecular surrogate marker for neoadjuvant treatment incolorectal cancer. Ann Surg 2003;238:324–4.

18. Soeth E, Vogel I, Roder C, Juhl H, Marxsen J, Kruger U, Henne-Bruns D, Kremer B, Kalthoff H. Comparative analysis of bone mar-row and venous blood isolates from gastrointestinal cancer patientsfor the detection of disseminated tumor cells using reverse transcrip-tion PCR. Cancer Res 1997;57:3106–10.

19. Vogel I, Kalthoff H. Disseminated tumor cells. Their detection andsignificance for prognosis of gastrointestinal and pancreatic carcino-mas. Virchows Arch 2001;439:109–17.

20. Sobin LH, Wittekind C. UICC: TNM-classification of malignanttumors. London: Wiley, 1997.

21. Koch M, Weitz J, Kienle P, Benner A, Willeke F, Lehnert T, HerfarthCh, von Knebel Doeberitz M. Comparative analysis of tumor cell dis-semination in mesenteric, central and peripheral venous blood inpatients with colorectal cancer. Arch Surg 2001;136:85–9.

22. Kaplan EL, Meier P. Nonparametric estimation from incompleteobservations. J Am Stat Assoc 1958;53:457–81.

23. Cox DR. Regression models and life-tables. J R Stat Soc B 1972;34:187–220.

24. Merkel S, Wein A, G€unther K, Papadopoulos T, Hohenberger W,Hermanek P. High-risk groups of patients with stage II colon carcino-ma. Cancer 2001;92:1435–43.

25. Mamounas E, Wieand S, Wolmark N, Bear HD, Atkins JN, Song K.Comparative efficacy of adjuvant chemotherapy in patients withDukes’B versus Dukes’C colon cancer: results from four NationalSurgical Adjuvant Breast and Bowel Project adjuvant studies (C-01,C-02, C-03, and C-04). J Clin Oncol 1999;17:1349–55.

26. Compton CC, Fielding LP, Burgart LJ, Conley B, Cooper HS, Hamil-ton SR, Hammond EH, Henson DE, Hutter P, Nagle RB, Nielsen ML,Sargent DJ, et al. Prognostic factors in colorectal cancer. Arch PatholLab Med 2000;124:979–94.

27. Lindemann F, Schlimock G, Dirschedl P, Witte J, Riethm€uller G.Prognostic significance of micrometastatic tumor cells in bone mar-row of colorectal cancer patients. Lancet 1992;340:685–9.

28. Vlems FA, Wobbes T, Punt CJ, Van Muijen GN. Detection and clini-cal relevance of tumor cells in blood and bone marrow of patientswith colorectal cancer. Anticancer Res 2003;23:523–30.

29. Weiss L. Metastatic inefficiency. Adv Cancer Res 1990;54:159–211.30. Fidler IJ. The biology of human cancer metastasis. Acta Oncol 1991;

30:669–75.31. Guller U, Zajac P, Schnider A, B€osch B, Vorburger S, Zuber M,

Spagnoli GC, Oertli D, Maurer R, Metzger U, Harder F, Heberer M,et al. Disseminated single tumor cells as detected by real-time quanti-tative polymerase chain reaction represent a prognostic factor inpatients undergoing surgery for colorectal cancer. Ann Surg 2002;236:768–76.

32. Bessa X, Pinol V, Castelli-Bel S, Piazuelo E, Lacy AM, Elizalde JI,Pique JM, Castells A. Prognostic value of postoperative detection ofblood circulating tumor cells in patients with colorectal cancer oper-ated for cure. Ann Surg 2003;237:368–75.

33. Bessa X, Elizalde JI, Boix L, Pinol V, Lacy AM, Salo J, Pique JM,Castells A. Lack of prognostic influence of circulating tumor cells inperipheral blood of patients with colorectal cancer. Gastroenterology2001;120:1084–92.

34. Bosch B, Guller U, Schnider A, Maurer R, Harder F, Metzger U,Marti WR. Perioperative detection of disseminated tumor cells is anindependent prognostic factor in patients with colorectal cancer. Br JSurg 2003;90:882–8.

35. LeVoyer TE, Sigurdson ER, Hanlon AL, Mayer RJ, Macdonald JS,Catalano PJ, Haller DG. Colon cancer survival is associated withincreasing number of lymph nodes analyzed: a secondary survey ofintergroup trial INT-0089. J Clin Oncol 2003;21:2912–9.

36. Fidler IJ. Metastasis. Quantitative analysis of distribution and fate oftumor emboli labelled with 125 I-5-iodo-20deoxyuridine. J Natl Can-cer Inst 1970;45:773–82.

37. Patel H, LeMarer N, Wharton RQ, Khan ZAJ, Araia R, Glover C,Henry MM, Allen-Mersh TG. Clearance of circulating tumor cellsafter excision of primary colorectal cancer. Ann Surg 2002;235:226–31.

38. Pantel K, Schlimok G, Braun S, Kutter D, Schaller G, Funke I, IzbickiJ, Riethm€uller G. Differential expression of proliferation-associatedmolecules in individual micrometastatic carcinoma cells. J Natl Can-cer Inst 1993;85:1419–24.

3077PROGNOSIS OF STAGE II COLORECTAL CANCER