Background
Colorectal cancer is the second most commonly diagnosed malignancy in men and women in the Netherlands with increasing incidence due to growth and ageing of the general population [
1]. The presence of lymph node metastases remains the most reliable prognostic predictor and the gold indicator for adjuvant treatment in colon cancer (CC) [
2]. Interestingly, in a large percentage of patients without lymph node metastases in the surgical specimen, who are therefore not subjected to adjuvant chemotherapy, represent with disease recurrence. In about 10% of the patients with stage I (Dukes A) and 15-30% with stage II (Dukes B) disease recurrent locoregional or distant metastases develop within 5 years [
2‐
4]. One possible factor could be the presence of occult lymph node metastases at the time of presentation and surgical resection. Evidence has emerged showing a significant amount of nodal metastases being smaller than 2 mm or less (<0.2 mm isolated tumor cells [ITC]; 0.2 - 2 mm micrometastasis [MM]) and, therefore, likely to be missed during conventional gross pathological specimen examination. }[
3‐
5] Focused examination methods, such as more extensive nodal examination by serial sectioning or step sectioning, or molecular detection of metastatic nodal cells by immunohistochemistry (IHC) or reverse transcriptase-polymerase chain reaction (RT-PCR), increase the likelihood of finding these tumour deposits. However, these focused examination methods are expensive and time consuming, and therefore not applicable to all lymph nodes derived from the surgical specimen. By using sentinel lymph node mapping (SLNM) the nodes at highest risk of harbouring tumour deposits can be potentially detected and more thoroughly examined. The
ex vivo SLNM procedure is technically easy, as the procedure is executed extracorporally by injection peritumoral blue dye subserosally or submucosally after which gentle massage of the injection site is performed [
5‐
11]. Blue coloured lymph nodes are excised or marked by sutures. The
ex vivo SLNM procedure is characterized by a high accuracy of 90-100%, and negative predictive value of 80-100%[
6‐
11]. A rate of 19-57% upstaging is observed [
6‐
11].
Current knowledge about the prognostic relevance of nodal micrometastases and isolated tumor cells has adequately been reviewed in 2004 separately by Iddings and Nicastri [
3,
4]. They concluded that a suggestion of prognostic relevance could be made for micrometastatic disease related to worsened disease-free survival (DFS) and overall survival (OS). The meta-analysis performed by Iddings showed a decreased 3-year DFS and OS of respectively 78% and 78% in pN0
micro+ patients compared to 90% and 97% in pN0
micro- patients [
4]. New evidence from two large international prospective observational studies shows a clear negative prognostic effect of micrometastatic nodal disease [
5]. The 4-year DFS detoriated from 94% to 78% with the presence of nodal micrometastasis [
5]. A solid conclusion, however, could not be made because of the lack of well-designed, well powered, controlled clinical trials. Data from other solid organ malignancies like breast cancer links micrometastatic disease to a worsened prognosis [
12]. Several prospective trials are currently recruiting (see table
1). However, an randomized, controlled clinical trial of significant magnitude is needed to answer this clinically relevant question.
Table 1
Currently recruiting studies
Jonsson Comprehensive Cancer Center (A. Bilchik) | 2009 - 2011 | N = 300 |
John Wayne Cancer Institute (S. Baker) | 2000 - 2011 | N = 200 |
University Hospital, Basel, Switzerland (C. Viehl) | 2004 - 2007 | N = 225 |
Nowadays, adjuvant chemotherapy is only offered to high risk stage I-II colon cancer patients in the Netherlands [
2]. Stage I-II CC patients without risk factors are thought not to benefit from adjuvant treatment. Individual randomized trials did not show a significant survival benefit in stage II colon cancer patients [
2]. The results of meta-analyses and systematic reviews show at the most a slight disease-free survival benefit of adjuvant chemotherapy in stage II disease [
2]. However, to stress out the importance of further investigation, stage I-II CC patients do suffer from disease recurrence and their overall 5-year survival is just around 70-80%. Even stage II CC patients without risk factors have been shown in several studies to benefit from adjuvant treatment [
13,
14]. It is because of these results that in Eastern countries and the United States stage II CC patients with or without micrometastatic disease do receive adjuvant treatment. Thus, there is an international need for better delineation of high-risk stage I-II CC patients who probably should be offered adjuvant treatment.
Discussion
Recurrent locoregional and/or systemic disease in stage I/II colon cancer is a significant clinical and social health care problem. The fact of high disease recurrence in stage I-II colon cancer patients is the basis of the study idea and the justification for the study. Around 7000 new colon cancer patients are diagnosed yearly in the Netherlands at this moment and the incidence will rise in the coming decades [
2]. Of these new colon cancer patients respectively 16 and 38% have stage I or stage II disease [
15]. Therefore, yearly approximately 3500 new stage I-II patients are at risk for disease recurrence. A upstaging rate of 30% is expected to be observed by using SLNM, serial sectioning and IHC for micrometastatic disease detection [
6‐
11]. Approximately 1050 are at high risk for disease recurrence. Nowadays, adjuvant chemotherapy is only offered to high risk stage II colon cancer patients in the Netherlands [
2]. Stage II CC patients without risk factors are thought not to benefit from adjuvant treatment. Individual randomized trials did not show a significant overall survival benefit in stage II colon cancer patients [
2]. The results of meta-analyses and systematic reviews show at the most a slight disease-free survival benefit of adjuvant chemotherapy in stage II disease of 3-6% absolute risk reduction in disease-free survival (3.8 - 6% ARR in 5-yr DFS by Figueredo et al, 3% ARR in 5-yr DFS in the IMPACT-B2 study, 4% ARR in 5-yr DFS in the Gill-study [
2,
16‐
18].
According to SEER data 5-year overall survival data of stage IIa (T3N0M0), IIb (T4N0M0) and IIIa (T1-2N1M0) and b (T3-4N1M0) are respectively 84.7%, 72.2%, 83.4% and 64.1% [
19]. The MOSAIC-trial observed a 4-year DFS of untreated versus treated colon cancer stage II and III of respectively 85-87.6% and 66.5-74.1% [
20]. The XELOXA-trial found a 3-year DFS of 66.5-70.9% (untreated versus treated) in stage III colon cancer [
21]. An 9% absolute risk reduction on 5-year OS (85.4% in the surgery-only group and 94.2% adjuvant chemotherapy group) was observed in T3N0M0 patients following adjuvant chemotherapy in a large retrospective study [
13]. A similar result was observed in 468 stage II patients (i.e. 8% 4-year OS risk reduction of 70% to 78%) [
22]. In an Australian population-based cohort-study patients with good prognosis (T3, no lymphovascular invasion) showed improved 5-year survival rate with chemotherapy from 85 to 95% (P = 0.064, log-rank test) in women but showed no change in men (84 vs 82%). The poor prognosis (T4 and/or presence of lymphovascular invasion) patients, survival rates for both women (65 vs 79%) and men (72 vs 78%) improved with the use of chemotherapy, however, did not reach significance in either group (P = 0.22 and P = 0.26, respectively) [
14]. It is therefore assumed that the use of adjuvant chemotherapy in high-risk pN0 colon cancer patients could result in clinical significant health benefits for individual patients.
The detection of sentinel nodes using a SLNM procedure has been extensively investigated [
5‐
9,
11,
23‐
29]. For correct interpretation of these data a gross distinction has to be made between i)
in vivo and
ex vivo procedures; ii) immunohistochemical (IHC) or reverse-transcriptase polymerase chain reaction (RT-PCR) occult nodal metastatic disease detection.
The
in vivo SLNM procedure is characterized by a more difficult technical procedure, most notably during a laparoscopic colonic resection, as a subserosal dye injection is performed preoperatively [
23‐
29]. The accuracy and negative predictive value of
in vivo SLNM procedures are high and range from 90 to 95% and 93 to 97% respectively [
23‐
29]. Upstaging was observed in 28-35% of H&E pN0 patients [
23‐
29]. The
ex vivo SLNM procedure is technically much easier, as the procedure is executed extracorporally by injection peritumoral blue dye subserosally or submucosally after which gentle massage of the injection site is performed [
6‐
11]. The
ex vivo SLNM procedure is characterized also by a high accuracy of 90-100%, and negative predictive value of 80-100%[
6‐
11]. A rate of 19-57% upstaging is observed [
6‐
11].
Although most studies combine data of colon and rectal cancer, a clear distinction in sentinel lymph node mapping between both cancers has been observed in some well-designed studies [
6,
11,
26]. The accuracy of negative predictive value of SLNM in rectal cancer is estimated to be as low as 76% and 65% respectively, and rectal cancers are therefore excluded from inclusion [
6].
The method of examination, i.e. IHC or RT-PCR, is also an important determinant of outcome [
3,
4]. Cytokeratin is a cellular marker for epithelial origin, adhering to constituents of the cytoskeleton of both normal and malignant cells. However, epithelial cells are not normal constituents of lymphoid tissue, and morphological examination of all cytokeratin-positive cells for secondary malignant characteristics is mandatory. Most studies used an IHC methodology, with little similarity in immunohistochemical markers unfortunately [
3,
4,
6‐
9,
24,
30]. A high variety in the use of molecular markers for RT-PCR inhibits the selection of the one with highest accuracy and negative predictive value [
3,
4,
23]. The best correlation between micrometastatic disease and worsened DFS and OS has been observed in RT-PCR detected occult disease in comparison to IHC [
3,
4]. The strength of the evidence of this observed correlation is weak though. Furthermore, RNA is subtracted from halve of a sentinel node in RT-PCR procedures. This unables pathologists the examination of secondary malignant characteristics with H&E colouring. Moreover, the authors of the most recent prospective clinical trial using RT-PCR stated that quantified RT-PCR assays may be of value when cytokeratin immunohistochemistry fails to detect sentinel lymph node metastasis [
23]. Therefore, no definitive preference for IHC or RT-PCR methods can be given. For practical arguments, including the possibility for all Dutch hospitals to participate in the study, a controlled clinical trial using an
ex vivo SLNM procedure with focussed nodal examination by cytokeratin IHC techniques is proposed.
Adjuvant chemotherapy is provided according to the CAPOX (XELOX) protocol (capecitabine (Xeloda
®) and oxaliplatin(Eloxatin
®)). This CAPOX (XELOX) protocol is assumed to be comparable in therapeutical effect and toxicity to the currently used FOLFOX-4 protocol, and more patient friendly because of the oral supplementation of chemotherapy (capecitabine) [
2].
As primary endpoint is 3-year DFS chosen as surrogate endpoint for overall survival. Colon cancer recurrence is most notably observed in the first years after initial treatment [
2,
31]. Furthermore, 3-year DFS is recognized as an appropriate surrogate end point for overall 5-year survival in colon cancer trials [
32‐
34].
Daniel J. Lips1, Boukje Koebrugge1, Gerrit Jan Liefers2, Johannes C. van der Linden3, Vincent T.H.B.M. Smit4, Hans F.M. Pruijt5, Hein Putter6, Cornelis J.H. van de Velde2, Koop Bosscha1.
1 Department of Surgery, Jeroen Bosch Hospital, 's-Hertogenbosch, the Netherlands
2 Department of Surgery, Leiden University Medical Center, Leiden, the Netherlands
3 Department of Pathology, Jeroen Bosch Hospital, 's-Hertogenbosch, the Netherlands
4 Department of Pathology, Leiden University Medical Center, Leiden, the Netherlands
5 Department of Medical Oncology, Jeroen Bosch Hospital, 's-Hertogenbosch, the Netherlands
6 Department of Medical Statistics, Leiden University Medical Center, Leiden, the Netherlands
Acknowledgements
The EnRoute⊕ Study is funded by the Dutch Cancer Society (grant-number 2010-4788) and by Roche Pharmaceuticals through an Educational Grant (grant-number ML25369). Protocol developed during the 12th joint ECCO-AACR-EORTC-ESMO Workshop 'Methods in Clinical Cancer Research', Waldhaus Flims, Switzerland, 19 - 25 June 2010 by D.J. Lips as a fellow of the European Society of Surgical Oncology.
Ethics
The study protocol is approved by the Dutch central medical-ethical body METOPP (Tilburg, the Netherlands) referred to as protocol number NL31438.028.10.
EnRoute+ study committee members
Steering Committee and Key Staff at coordinating centres
K. Bosscha (principal investigator), D.J. Lips (study coordinator), B. Koebrugge (Ph.D.-student), H. Pruijt, J.C. van der Linden, Jeroen Bosch Hospital; C.J.H. van de Velde (co-principal investigator), G.J. Liefers, V.T.H.B.M. Smit, H. Putter, LUMC.
Clinical centres and investigators
Jeroen Bosch Hospital, K. Bosscha
Leiden University Medical Centre, C.J.H. van de Velde
Viecuri Medical Centre, J.L.M. Konsten
Medical Spectrum Twente, J. M. Klaase
Academic Hospital Maastricht, M.L. Smidt
Sint Lucas Andreas Hospital, B.C. Vrouenraets
Diakonessen Hospital Utrecht, P.H.P. Davids
Amphia Hospital, R.M.P.H. Crolla
Elkerliek Hospital, J.A. Wegdam
Ikazia Hospital, W.J. Vles
Amstelveen Hospital, S.C. Veltkamp
Groene Hart Hospital, L.N.L. Tseng
Deventer Hospital, M.S.L. Liem
Hospital Group Twente, E.A. Kouwenhoven
Onze Lieve Vrouwe Gasthuis, M.F. Gerhards
Tergooi Hospitals location Hilversum 50, A. van Geloven
Rode Kruis Hospital, H.A. Cense
Mesos Medical Centre, A. Smits
Ropcke-Zweers Hospital, J.H. Tomee
Isala Clinics, G. A. Patijn
University MediCal Centre Groningen, B. van Leeuwen
Stichting Christelijke Zorgvoorzieningen, Talma Sionsberg, H. Bos
Diaconessen hospital Leiden, R. Vree
HagaHospital, W.H. Steup
Bronovo Hospital, F. van Duijnhoven
Gemini Hospital, N. Bode
St. Jansgasthuis Weert, N.A.J.B. Peters
Westfries Gasthuis, J.W. de Waard
Medical Centre Alkmaar, A.P.J. Houdijk
Orbis Medical Centre, J. Stoot
Medical Centre Leeuwarden, S.A. Koopal
Oosterschelde Hospitals, H. de Morree
University Medical Centre Utrecht, W.M.U. van Grevenstein
Heerenveen Hospital De Tjongerschans, F. Wit
Antonius Hospital Sneek, D. Hess
Spaarne Hospital, B. de Valk
Martini Hospital, W. Kelder
Canisius Wilhelmina Hospital, C. Rosman
Competing interests
The EnRoute⊕ Study is supported through an Educational Grant by Roche Pharmaceuticals.
The authors declare that they have no other competing interests.
Authors' contributions
DJL drafted the manuscript. KB, CJHV, GJL edited the manuscript. All authors participated in the design of the study. HP performed the statistical analysis. All authors read and approved the final manuscript.