European Journal of Cancer (2013) 49, 1104– 1108

Available at www.sciencedirect.com

journal homepage: www.ejcancer.info

Predicting lymph node metastases in early rectal cancer

Deborah Saraste ⇑, Ulf Gunnarsson, Martin Janson

Division of Surgery, CLINTEC, Karolinska Institutet, Sweden

Department of Surgical Gastroenterology, Karolinska University Hospital, Huddinge, Sweden

Available online 31 October 2012

KEYWORDS Abstract Aim: In this population-based study, the aim was to investigate risk factors for
Rectal cancer lymph node metastases and to construct a risk stratification index with relevance for pre-oper-
Rectal neoplasia ative planning in T1 and T2 rectal cancers.
Lymph node metastases Methods: Data were retrieved from The Swedish Rectal Cancer Register, a mandatory,
Risk national, prospectively collected data base. All T1 and T2 rectal cancers treated with abdom-
inal resection surgery without neo-adjuvant or adjuvant radio-chemotherapy from 2007 to
2010 were analysed. T-stage, sm-level, histologic differentiation, mucinous tumour type, blood
vessel- and perineural infiltration, tumour location (in cm from the anal verge), age and gender
were evaluated as potential predictors of lymph node metastases, using uni- and multivariate
logistic regression.
Results: T2-stage (odds ratio [OR] = 2.0), poor differentiation (OR = 6.5) and vascular infil-
tration (OR = 4.3) were identified as significant risk-factors for lymph node metastases in the
multivariate analysis. The risk stratification index shows the risk for lymph node metastases
gradually increasing from 6% to 65% and 11% to 78% in T1 and T2 cancers respectively, when
adding the risk factors one by one.
Conclusion: There is a considerable span in the risk for lymph node metastases between low
risk T1 and high risk T2 rectal cancer. Using the risk stratification-model, with the concept
of local excision as a macro-biopsy with standby for subsequent immediate radical resection
surgery in high-risk cases, could benefit patients by providing the advantages of local excision
yet ensuring adequate oncologic outcome.
Ó 2012 Elsevier Ltd. All rights reserved.

⇑ Corresponding author: Address: Department of Surgical Gastroenterology, Karolinska University Hospital, Huddinge, 141 86 Stockholm,
Sweden. Tel.: +46 0 8 58580000; fax: +46 0 8 58582340.
E-mail address: deborah.saraste@karolinska.se (D. Saraste).

0959-8049/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.ejca.2012.10.005
 D. Saraste et al. / European Journal of Cancer 49 (2013) 1104–1108
1. Introduction
Colorectal cancer is the most common form of cancer
in Europe, with an incidence (2008) of 436,000 cases.1
The presence of lymph node metastases in rectal can-
cer affects long-term survival and the risk for local recur-
rence. Previous studies have shown that the T-stage and
sm-level of the tumour are important predictors of
lymph node metastases.2–4 Infiltration of nerves and ves-
sels, lymphatic invasion, histologic differentiation,
tumour size and tumour location in the rectum are other
predisposing factors,2,3,5,6 though results are not unani-
mous. There is some evidence that gender and age could
influence the risk for lymph node metastases.4,7
Accurate preoperative assessment of the nodal status
and T-stage of the tumour are crucial when planning
surgery, but despite improvements in radiology, the pre-
operative staging of nodal status is still insufficient.8,9
In local excision procedures, where the mesorectum is
left, no lymph nodes are harvested and the postoperative
nodal status remains uncertain. Furthermore there is the
risk of leaving possible positive nodes. These factors lie
behind the debate that still surrounds the oncologic
results of local surgery in T1 and T2 cancers.
Local surgery, as compared to abdominal resection
surgery, shows excellent results in terms of postoperative
morbidity, mortality and function.10,11 Dissatisfying
oncologic results after local excision regarding local
and overall recurrence have been the major drawback,
though many studies show survival rates comparable
to abdominal resection surgery.10–12
Adopting the idea of local excision as a ‘macro-
biopsy’ in T1 and T2 cancers with standby to perform
immediate resection surgery, should histopathology
reveal a high risk for lymph node metastases, would pro-
vide the advantages of local surgery in terms of low mor-
bidity and good functional results, while still ensuring
adequate oncologic outcome.
The aim of this study was to investigate potential risk
factors for lymph node metastases, and to construct a
risk stratification index relevant for pre-operative plan-
ning and clinical decision-making in T1 and T2 rectal
cancers. By stratification of risk, not only low-risk T1,
but probably also low-risk T2 tumours could benefit
from the advantages of local excision.
The gathering of data from a large, population-based
and validated register provides a picture that reflects the
risk profile for lymph node metastases in routine hospi-
tal care.
2. Patients and methods
The Swedish Rectal Cancer Register (SRCR) is a
mandatory, population based register covering 97% of
rectal cancers operated on in Sweden. The register
comprises information concerning the preoperative
1105
investigation, peroperative surgical details, postopera-
tive histopathology and complications during a 5-year
follow-up period.
A high accuracy of data has been shown when vali-
dating the register.13,14
Rectal cancer was defined as a tumour located within
15 cm from the anal verge, measured at rigid rectoscopy.
From the SRCR, 2007–2010, data on all patients
treated with abdominal resection surgery for rectal can-
cer without neo-adjuvant or adjuvant radio-chemother-
apy were analysed.
The T-stage distribution for the entire material
(n = 1664) is shown in Table 1.
Inclusion criteria for this study were histopathologi-
cally confirmed R0-resections in T1 and T2 cancers.
Tx (n = 14), Nx (n = 26) and T0 (n = 12) were excluded,
leaving 677 cases for further analysis.
Uni- and multivariate analyses were performed on
the following possible risk factors for lymph node
metastases: T-stage, sm-grade, histologic differentiation,
mucinous tumour type, blood vessel- and perineural
infiltration, tumour location (in cm from the anal verge),
age and gender. Risk factors identified as statistically
significant in the univariate analysis were included in
the multivariate analysis.
Tumour size and lymph vessel infiltration were not
recorded in the register.
2.1. Statistical analysis
The primary effect variable analysed was risk for
lymph node metastases.
Uni- and multivariate logistic regression analyses
were used to calculate the predictive power of each risk
factor, expressed as odds ratio (OR).
The Chi-square test was used for detection of differ-
ences between groups.
Age was calculated as a continuous predictor.
Parameters were judged from the 95% confidence
intervals and when considered significant included in
the multivariate analysis.
A risk stratification index was calculated by summa-
tion of variables with a statistically significant increase
in risk according to the multivariate analysis.
The risk associated with each variable was calculated
using the intercept and estimate derived from the multi-
variate analysis using the following formula:
Table 1
Distribution of T-stage and lymph node metastases.
T-stage Distribution of Proportion with lymph
T-stage (%) node metastases (%)
T1, n = 205 13 12
T2, n = 472 29 22
T3, n = 830 51 46
T4, n = 105 7 65
1106 D. Saraste et al. / European Journal of Cancer 49 (2013) 1104–1108

Risk ¼ expg =ð1 þ expg Þ 3.2. Multivariate analysis

g = intercept + estimate1  variable 1 + estimate2 
variable 2. . .
(The y intercept is the value of y when x equals zero.)
Statistical analyses were performed using the Statisti-
ca Release 10 (Statsoft, Tulsa, United States of America
[USA]).
3. Results
The distribution of T-stage and the proportion of
positive lymph nodes in relation to T-stage are shown
in Table 1.
3.1. Univariate analysis
T2-stage, vascular infiltration, intermediate and poor
differentiation were independent, statistically significant
risk factors (Table 2).
Missing data in each group of variables were ana-
lysed as possible predictors of lymph node metastases.
Risk factors identified as statistically significant in the
univariate analysis were included in the multivariate
analysis.
T2, poor differentiation and vascular infiltration
were identified as statistically significant risk factors
(Table 2).
3.3. Risk stratification index
The index shows the increase in risk for lymph
node metastases in T1 and T2-tumours respectively,
by summation of poor differentiation and vascular
infiltration. The variables chosen for inclusion in the
risk index are the variables that showed significance
in the multivariate analysis. Minimum risk is T1 or
T2 tumour without poor differentiation and vascular
infiltration. This risk is thus lower than the total risk
for lymph node metastases in T1 and T2 tumours,
respectively (Table 3).

Table 2
Uni-and multivariate analyses of risk factors for lymph node metastases.
Univariate analysis Multivariate analysis
Number OR 95% CI OR 95% CI
T-stage/sm-level
Sm1 54 1 Ref
Sm2 24 0.54 (0.06–5.12)
Sm3 50 2.38 (0.67–8.46)
Sm missing 77 2.08 (0.63–6.93)
T2 472 3.45 (1.22–9.77) 1.97 (1.19–3.25)
Tumour differentiation
High 114 1 Ref 1 Ref
Intermediate 498 1.98 (1.04–3.75) 1.72 (0.93–3.18)
Low 39 7.29 (3.06–17.4) 6.47 (2.71–15.4)
Differentiation missing 26 0.71 (0.15–3.38)
Vascular infiltration
Yes 61 4.81 (2.75–8.40) 4.34 (2.46–7.65)
No 492 1 Ref 1 Ref
Missing 124 0.96 (0.56–1.66)
Perineural infiltration
Yes 10 1.85 (0.46–7.31)
No 458 1 Ref
Missing 209 0.93 (0.61–1.42)
Mucinous type
Yes 52 1.87 (0.99–3.55)
No 539 1 Ref
Missing 86 0.97 (0.54–1.77)
Tumour location (cm from anal verge)
0–5 cm 118 1.03 (0.59–1.84)
6–10 cm 259 1.23 (0.80–1.88)
11–15 cm 300 1 Ref
Gender
Male 389 0.94 (0.63–1.39)
Female 288 1 Ref
Agea 0.99 (0.98–1.01)
OR, odds ratio; CI, confidence interval.
a
Age was set as a continuous predictor. OR was not calculated for missing data n = 1.
 D. Saraste et al. / European Journal of Cancer 49 (2013) 1104–1108
Table 3
Risk stratification index showing the increase in risk for lymph node
metastases (%) in T1 and T2 tumours respectively when adding risk
factors one at a time.
a = poor differentiation; b = vascular invasion.
The variables chosen for inclusion into the risk index are the variables
with statistical significance in the multivariate analysis.
4. Discussion
In this population-based study, T2-stage, poor differ-
entiation and vascular infiltration were identified as sig-
nificant predictors of lymph node metastases. Sm-level
was not a significant predictor.
The risk stratification-index shows a large span in the
risk for lymph node metastases in T1 (6–65%) and T2
(11–78%) tumours respectively, when significant predic-
tors in the multivariate analysis are summated.
The findings in the present study, where depth of
invasion, poor differentiation and vessel infiltration are
risk factors for lymph node metastases, are in accor-
dance with results from previous studies.3,4,6,7,15 In T1
tumours, increasing depth of submucosal penetration,
sm-1–3, has previously been identified as a risk factor
for lymph node metastases (2), but this could not be
confirmed in the present study.
Interestingly, Rasheed et al.5 could not demonstrate
depth of tumour invasion as a risk factor for lymph
node metastases. In the study by Rasheed et al., the
T1 group was relatively small (T1 = 18%, T2 = 81%)
and in the study population there was a large number
of patients with FAP (familial adenomatous polyposis)
which could have affected the outcome.
In addition to the depth of penetration, other poten-
tial risk factors for lymph node metastases such as poor
differentiation, vessel infiltration, age, gender etc. have
been addressed in a large number of studies. However,
not many studies have tried to summate the different
risk factors into a coherent tool for assessment of the
aggregated risk, which is the intention of our risk strat-
ification index. It is difficult to compare previous studies
1107
on risk stratification with this one because of the heter-
ogeneous selection of potential risk factors. Ding et al.16
studied T2 tumours finding an interval of risk for lymph
node metastases between 3.4% and 67% depending on
tumour location in the rectum, age, depth of invasion
within the muscularis propria (T2) and histopatholo-
gical characteristics. In a study by Blumberg et al.6 on
T1 and T2 tumours, only blood vessel invasion reached
statistical significance as a single predictive factor. Nev-
ertheless, a model for assessment of risk for lymph node
metastases was constructed. The risk span was 7–33%
and 14–30% in T1 and T2 tumours respectively, depend-
ing on the presence of adverse features, i.e. poor differ-
entiation, lymphatic vascular and blood vessel
invasion. Kobayashi et al.7 also studied T1 and T2
tumours, with an intricate model of risk assessment
showing a span of risk from 1% to 30% and 16% to
37% in T1 and T2 tumours, respectively.
Even if the studies are heterogeneous and use differ-
ent models for risk calculation, the point of interest is
that there is less chance of low-risk T2 tumours having
lymph node metastases than high-risk T1 tumours. Con-
sequently there are likely to be low-risk T2 tumours that
can be treated curatively with local excision alone had
we the means of assessing the risk for lymph node
metastases more accurately.
The concept of using local excision as a first line
treatment or a ‘macro-biopsy’ in early rectal cancer with
readiness to perform immediate surgical re-intervention
in case of unfavourable histopathology is not without
complications oncologically. Many studies have shown
poor survival following salvage surgery for local recur-
rence after local excision.17–19 On the other hand
Hahnloser et al.20 showed no difference in distant metas-
tases or 5-year survival after immediate re-intervention
compared to primary radical surgery.
Clearer knowledge of the risk factors for lymph node
metastases and future improvements in the diagnostic
accuracy of radiology regarding lymph node metastases
will further improve the chances of local excision alone
being sufficient for cure in early rectal cancer.
In the present study, many of the demonstrated risk
factors are known from previous studies, but few studies
have explored the additive effect of risk factors in order
to calculate the total risk using population based data
showing outcome in routine care. Results from the pres-
ent study and the consequent risk stratification index
aims at facilitating pre-operative planning and clinical
decision-making when using local excision procedures
as first line treatment, and to guide the decision whether
to proceed with ‘immediate’ radical surgery after pri-
mary local excision in a high-risk scenario.
Acceptable level of risk must be evaluated for each
patient taking into account not only the risk index but
also concomitant disease and the patient’s own assess-
ment of risk; i.e. the risk of recurrence after local excision
1108 D. Saraste et al. / European Journal of Cancer 49 (2013) 1104–1108
versus the risk of mortality, morbidity and functional
loss associated with major abdominal surgery.
The current Swedish National Guidelines for rectal
cancer21 do not support the use of neo-adjuvant or adju-
vant treatment in T1–T2NO tumours amenable for local
surgery and this subject is not discussed further.
Conflict of interest statement
None declared.
Acknowledgements
Robert Johansson, statistician at the National
Oncology Centre. The study received financial support
from Stockholm County Council, Karolinska Institutet
(ALF, Grant number 20100123) and the Bengt Ihre
foundation (Grant numbers SLS-94401, SLS-248341
and SLS-171781).
References
1. Ferlay J, Parkin DM, Steliarova-Foucher E. Estimates of cancer
incidence and mortality in Europe in 2008. Eur J Cancer
2010;46(4):765–81.
2. Kikuchi R, Takano M, Takagi K, et al. Management of early
invasive colorectal cancer. Risk of recurrence and clinical guide-
lines. Dis Colon Rectum 1995;38(12):1286–95.
3. Saclarides TJ, Bhattacharyya AK, Britton-Kuzel C, Szeluga D,
Economou SG. Predicting lymph node metastases in rectal cancer.
Dis Colon Rectum 1994;37(1):52–7.
4. Sitzler PJ, Seow-Choen F, Ho YH, Leong AP. Lymph node
involvement and tumor depth in rectal cancers: an analysis of 805
patients. Dis Colon Rectum 1997;40(12):1472–6.
5. Rasheed S, Bowley DM, Aziz O, et al. Can depth of tumour
invasion predict lymph node positivity in patients undergoing
resection for early rectal cancer? A comparative study between T1
and T2 cancers. Colorectal Dis 2008;10(3):231–7.
6. Blumberg D, Paty PB, Guillem JG, et al. All patients with small
intramural rectal cancers are at risk for lymph node metastasis. Dis
Colon Rectum 1999;42(7):881–5.
7. Kobayashi H, Mochizuki H, Kato T, et al. Is total mesorectal
excision always necessary for T1–T2 lower rectal cancer? Ann Surg
Oncol 2010;17(4):973–80.
8. Bipat S, Glas AS, Slors FJM, et al. Rectal cancer: local staging
and assessment of lymph node involvement with endoluminal US,
CT, and MR imaging – a meta-analysis. Radiology 2004;232(3):
773–83.
9. Al-Sukhni E, Milot L, Fruitman M, et al. Diagnostic accuracy of
MRI for assessment of T category, lymph node metastases, and
circumferential resection margin involvement in patients with
rectal cancer: a systematic review and meta-analysis. Ann Surg
Oncol 2012;19(7):2212–23.
10. De Graaf EJ, Doornebosch PG, Tollenaar RA, et al. Transanal
endoscopic microsurgery versus total mesorectal excision of T1
rectal adenocarcinomas with curative intention. Eur J Surg Oncol
2009;35(12):1280–5.
11. Sgourakis G, Lanitis S, Gockel I, et al. Transanal endoscopic
microsurgery for T1 and T2 rectal cancers: a meta-analysis and
meta-regression analysis of outcomes. Am Surg 2011;77(6):761–72.
12. You YN, Baxter NN, Stewart A, Nelson H. Is the increasing rate
of local excision for stage I rectal cancer in the United States
justified?: a nationwide cohort study from the National Cancer
Database. Ann Surg 2007;245(5):726–33.
13. Jorgren F, Johansson R, Damber L, Lindmark G. Risk factors of
rectal cancer local recurrence: population-based survey and vali-
dation of the Swedish Rectal Cancer Registry. Colorectal Dis 2009.
14. Gunnarsson U, Seligsohn E, Jestin P, Pahlman L. Registration
and validity of surgical complications in colorectal cancer surgery.
Br J Surg 2003;90(4):454–9.
15. Ueno H, Mochizuki H, Hashiguchi Y, et al. Risk factors for an
adverse outcome in early invasive colorectal carcinoma. Gastro-
enterology 2004;127(2):385–94.
16. Ding P-R, An X, Cao Y, et al. Depth of tumor invasion
independently predicts lymph node metastasis in T2 rectal cancer.
J Gastrointest Surg 2011;15(1):130–6.
17. Doornebosch PG, Ferenschild FT, de Wilt JH, Dawson I,
Tetteroo GW, de Graaf EJ. Treatment of recurrence after
transanal endoscopic microsurgery (TEM) for T1 rectal cancer.
Dis Colon Rectum 2010;53(9):1234–9.
18. Mellgren A, Sirivongs P, Rothenberger DA, Madoff RD, Garcia-
Aguilar J. Is local excision adequate therapy for early rectal
cancer? Dis Colon Rectum 2000;43(8):1064–71.
19. Paty PB, Nash GM, Baron P, et al. Long-term results of local
excision for rectal cancer. Ann Surg 2002;236(4):522–9 [discussion
529–30].
20. Hahnloser D, Wolff BG, Larson DW, Ping J, Nivatvongs S.
Immediate radical resection after local excision of rectal cancer: an
oncologic compromise? Dis Colon Rectum 2005;48(3):429–37.
21. National Center of Oncology. National guidelines for rectal cancer
in Sweden. <http://www.karolinska.se/upload/Onkologiskt%20
centrum/NationellaVardprogram/KolorektalNatVP2008.pdf>; 2008