Int J Colorectal Dis (2014) 29:847–852
DOI 10.1007/s00384-014-1885-z
ORIGINAL ARTICLE
Impact of D3 lymph node dissection on survival for patients
with T3 and T4 colon cancer
Kenjiro Kotake & Tomoka Mizuguchi & Konosuke Moritani &
Osamu Wada & Heita Ozawa & Izumi Oki & Kenichi Sugihara
Accepted: 23 April 2014 / Published online: 6 May 2014
# Springer-Verlag Berlin Heidelberg 2014
Abstract
Purpose The clinical significance of D3 lymph node dissec-
tion for patients with colon cancer remains controversial. This
study aims to clarify the impact of D3 lymph node dissection
on survival in patients with colon cancer.
Methods This is a retrospective cohort study from a prospec-
tively registered multi-institutional database of colorectal can-
cer in Japan. Propensity score matching method was applied
to balance potential confounders of the treatment. A cohort of
10,098 patients who underwent radical colectomy for pT3 and
pT4 colon cancer between 1985 and 1994 were identified. A
total of 3,425 propensity score matched pairs were extracted
from the entire cohort. The primary outcome measure was
overall survival (OS).
Results In the entire cohort, there was a statistically significant
difference in overall survival (OS) between the patients who
had D3 and D2 lymph node dissection (p=0.00003). The
estimated hazard ratio (HR) for OS of patients who had D3
versus D2 lymph node dissection was 0.827 (95 % confidence
interval, 0.757 to 0.904). In the matched cohort, there was also
a significant difference in OS between the two groups (p=
0.0001), and the estimated HR for OS was 0.814 (95 %
confidence interval, 0.734 to 0.904).
K. Kotake (*) : T. Mizuguchi : K. Moritani : O. Wada : H. Ozawa
Department of Colorectal Surgery, Tochigi Cancer Center,
4-9-13 Yohnan, Utsunomiya, Tochigi 320-0834, Japan
e-mail: kkotake@tcc.pref.tochigi.lg.jp
I. Oki
Epidemiology Unit, Tochigi Cancer Center Research Institute,
Utsunomiya, Japan
K. Sugihara
Department of Surgical Oncology, Graduate School,
Tokyo Medical and Dental University,
1-5-45 Yushima, Bunkyo-ku, Tokyo 113-8519, Japan
Conclusions We found D3 lymph node dissection for pT3 and
pT4 colon cancer to be associated with a significant survival
advantage in a large-scale database, even after adjusting po-
tential confounders of lymph node dissection. This finding
may provide a rationale for D3 lymph node dissection in
radical surgery for pT3 and pT4 colon cancer.
Keywords Colon cancer . D3 lymph node dissection . Overall
survival . Propensity score matching
Introduction
Colorectal cancer (CRC) is the fourth leading cause of cancer
death worldwide [1]. There are approximately 110,000 new
cases diagnosed with CRC annually in Japan, of which approx-
imately two thirds occur in the colon [2]. Although adjuvant
chemotherapy has the potential to decrease disease recurrence
for patients with stages II and III colon cancer [3–5], radical
surgery is the only treatment modality for cure, and over 90 %
of new cases undergoes surgical resection in Japan [6].
The number of lymph node metastasis (NLNM) is the most
robust determinant of survival for patients with localized CRC
[7]. In addition to NLNM, it has been recognized that the
number of lymph nodes retrieved (NLNR) is closely related to
prognosis of CRC, and increased NLNR is significantly asso-
ciated with decreased risk of disease recurrence and death
[8–12]. This may be partly due to stage migration as well as
decreased risk of recurrence by the resection of metastatic
lymph nodes. The NLNR is naturally related to the extent of
lymph node resection, providing that the methods of patho-
logical examination are consistent [13]. Although en bloc
resection of the primary tumor with its draining lymph nodes
is a widely accepted principle of treatment for CRC world-
wide, optimal extent of lymph node dissection, especially that
along the supplying vessels to the involved bowel segment,
848
remains controversial. Under those circumstances, this study
was conducted to clarify whether D3 lymph node dissection
(LND) improves survival of patients with surgically curable
T3 and T4 colon cancer using large-scale multi-institutional
database.
Materials and methods
We used the same database of the Japanese Society for Cancer
of the Colon and Rectum (JSCCR) as our previous report [12].
The member hospitals of the JSCCR, which are located all
over Japan, have been voluntarily registering the clinical and
pathological information of patients with CRC who were
treated in each hospital. The database currently contains in-
formation for about 160,000 CRC patients treated between
1974 and 2004. In this study, a total of 10,098 patients who
had pT3 or pT4 colorectal adenocarcinoma and underwent
R0-resection with D2 or D3 LND between 1985 and 1994
were extracted from the database. To eliminate the confound-
ing effects of new chemotherapeutic agents and laparoscopic
surgery, we used the relatively old data of the database. During
this study period from 1985 to 1994, T3 and T4 CRCs were
treated with open surgery, and neither chemotherapeutic reg-
imens containing 5-fluorouracil plus leucovorin nor
oxaliplatin were available in Japan.
The classification of JSCCR clearly states the scope of
LND [14]. D2 LND is defined as removal of peri-/para-colic
and intermediate nodes, and D3 LND as removal of main
lymph nodes at the root of the supplying artery in addition
to D2 dissection. Patients with unspecified age, synchronous
multiple cancers, no pathologic report of lymph nodes, un-
known followed up status, and receiving perioperative radio-
therapy were excluded [12].
To identify potential predictors to undergoing D3 LND,
demographics, clinical, and pathological characteristics of the
entire cohort were identified among available variables and
were assessed using univariate analysis. Summary statistics of
the univariate analysis were constructed using frequency and
proportional for categorical variables and mean with standard
deviations for continuous variables.
Predictive variables for D3 LND, including treatment year,
sex, age at diagnosis, tumor location, gross appearance, tumor
histology, depth of tumor invasion, lymph node metastasis,
and adjuvant chemotherapy were used as covariates in a
multivariable logistic regression in which LND was the de-
pendent variable. The estimated probabilities were used as
propensity score. Using the propensity score, the entire cohort
was matched by a 1:1 nearest neighbor matching method with
a caliper of 0.01. Then, we examined the balance of the
covariates between the matched pairs.
The primary outcome of interest of this study was the
overall survival (OS), which was calculated in months relative
Int J Colorectal Dis (2014) 29:847–852
to date-of-surgery information available from the database.
Survival analysis was performed by the Kaplan and Meier
method for the entire cohort and the propensity score matched
cohort. The log-rank test was used for comparison of survival
curves. In addition to the propensity score matching analysis,
multivariate analyses were performed using Cox proportional
hazards and propensity-adjusted Cox proportional hazards
analyses to determine the effect of D3 LND on survival. In
the latter analysis, propensity score was added to the covari-
ates as a continuous variable.
In this study, a p value of less than 0.05 was considered
statistically significant. Statistical analyses were performed
using SPSS Statistic version 20 (IBM Corporation, Somers,
NY,), SPSS plug-in of PSMATCHING.3, and R version
2.12.1 (R Foundation for Statistical Computing; http://www.
r-project.org).
Results
In a total of 10,098 eligible patients who underwent curative
colectomy for pT3 or pT4 colon cancer, 6,580 patients (65.2 %)
had D3 LND. Figure 1 shows the proportion of D3 LND
performed by each institute. The rate ranged from 3/76 (4 %)
to 79/79 (100 %). Table 1 shows the selected covariates of the
entire cohort including 6,580 patients with D3 LND and 3,518
with D2 LND. In this cohort, all but one covariate, significantly
differed between the two groups. Both NLNM and NLNR were
significantly larger for D3 LND than that of D2 LND. In the
entire cohort, overall survival of the patients who underwent D3
LND was significantly higher than that of D2 LND (p=
0.00003). The estimated hazard ratio (HR) for OS of patients
who had D3 versus D2 LND was 0.827 (95 % confidence
interval, 0.757 to 0.904). Among 326 patients with metastases
to the main nodes, 122 (37 %) survived 5 years or more.
A total of 3,425 propensity score matched pairs were
extracted from the entire cohort. Table 2 shows the selected
covariates in the propensity score adjusted cohort. NLNM did
not differ between the two groups, but NLNR for D3 LND
Proportion of D3 LND
1
0.8
0.6
0.4
0.2
0
Fig. 1 Proportion of D3 lymph node dissection in each institution
Int J Colorectal Dis (2014) 29:847–852 849

Table 1 Characteristics of the 10,098 patients undergoing curative sur- Table 2 Characteristics of 6,850 patients undergoing curative surgery for
gery for T3 and T4 colon cancer, according to the scope of lymph node T3 and T4 colon cancer, according to the scope of lymph node dissection
dissection in the propensity score matched cohort

Characteristics Lymph node dissection p value Characteristics Lymph node dissection p value

D2 D3 D2 D3

Sex Sex
Male 1,917 3,556 0.666 Male 1,864 1,890 0.272
Female 1,601 3,024 Female 1,561 1,535
Age Age
Mean 65.01 62.21 <0.001 Mean 64.6 64.6 0.928
SD 0.197 0.138 SD 11.5 10.5
Gross appearance Gross appearance
Type 1 295 454 0.004 Type-1 288 245 0.026
Type 2 2,686 5,222 Type-2 2,611 2,710
Type 3 448 753 Type-3 438 381
Others 89 151 Others 88 89
Histology Histology
Well differentiated 1,503 2,614 0.014 Well differentiated 1,443 1,482 0.344
Moderately differentiated 1,715 3,382 Moderately differentiated 1,684 1,675
Others 300 584 Others 298 268
Tumor site Tumor site
Rigt side 1,923 3,358 0.0005 Right side 1,843 1,847 0.471
Left side 1,595 3,222 Left side 1,582 1,578
pT-category pT-category
pT3 2,234 3,996 0.003 pT3 2,154 2,187 0.499
pT4a 1,108 2,162 pT4a 1,097 1,054
pT4b 176 422 pT4b 174 184
pN-category pN-category
pN0 2,121 3,674 0.01 pN0 2,036 2,135 0.014
pN+ 1,397 2,906 pN+ 1,389 1,290
pN1 941 1,654 pN1 935 817
pN2 453 926 pN2 451 373
pN3 3 326 pN3 3 100
Adjuvant chemotherapy Adjuvant chemotherapy
No 1,203 2,044 0.001 No 1,145 1,179 0.200
Yes 2,315 4,536 Yes 2,280 2,246

was statistically significantly higher than that of D2 LND Discussion

(Table 3).
In the matched cohort, overall survival of the patients who
underwent D3 LND was significantly higher than that of D2
LND (p=0.0001), with the estimated HR of 0.814 (95 %
confidence interval, 0.734 to 0.904) (Fig. 2). Subset analysis
revealed that there were significant differences in OS both of
patients with stages II and III patients between D3 and D2
LND (log-rank p=0.016, p=0.013, respectively).
Multivariate survival analysis of the entire cohort using
Cox proportional hazard modeling, controlled for the propen-
sity score as the only covariate, revealed the estimated HR for
OS of D3 versus D2 LND as 0.810 (95 % confidence interval,
0.741 to 0.886).
This study showed that D3 LND is significantly superior to
D2 LND in terms of overall survival for patients with pT3 and
pT4 colon cancer with a relative reduction in the risk of death
by 18 %. To our knowledge, this is the first report to estimate
the impact of the extent of LND up to the main node on
survival using large-scale retrospective cohort study.
The extent of LND for CRC had been mainly rationalized
by studies on lymph flow of the bowel segment, and incidence
and distribution of positive lymph nodes [15–17]. And gener-
ally, lymph node metastasis has been considered as an indica-
tor, but not governor of survival as articulated by Cady [18].
Meanwhile, excellent long-term outcomes of complete
850 Int J Colorectal Dis (2014) 29:847–852

Table 3 Number of lymph node retrieved and metastasis of patients 36.4 %. In this study, incidence of the main lymph node
undergoing curative surgery for T3 and T4 colon cancer, in the entire
metastasis was 4.9 % (326/6,580) in patients who had D3
cohort and the propensity-matched cohort
LND and among those, 37 % (122/326) survived more than
Entire cohort Propensity-matched cohort 5 years. The prognosis of patients with metastasis to main
(10,098 patients) (6,850 patients) nodes was comparable to that after hepatectomy for liver
Lymph node dissection Lymph node dissection metastasis in CRC. Although Coller and colleagues [23] ex-
amined surgically resected specimens and speculated that “If
D2 D3 p value D2 D3 p value cancer has metastasized to the main group of lymph node at
the origin of the various vessels, then generalized metastases,
Number of lymph node retrieved
undoubtedly, have already occurred, and any operative proce-
Mean 14.9 22.3 <0.001 14.9 21.8 <0.001
dure would only be palliative in character”, considering our
SD 0.2 0.2 11.2 14.3
results, there is a chance that some patients with main lymph
Number of lymph node metastasis
node metastasis may be cured by D3 LND.
Mean 1.01 1.31 <0.001 1.04 1.05 0.819
SD 0.034 0.032 2.02 2.09 In this study, two thirds of the entire cohort underwent D3
dissection. According to the nationwide survey by the JSCCR,
the proportion of D3 LND performed for stages II and III
colon cancer in 2001 were 55 and 59 %, respectively, which is
mesocolic excision (CME) with central vascular ligation similar to the proportion of D3 LND performed in our study.
(CVL) technique that was applied to colon cancer as a similar We believe that adjuvant chemotherapy will have little
concept of total mesorectal excision for rectal cancer, have effect on the outcome of this study, because active chemother-
been reported from German surgeons [19, 20]. The essentials apeutic regimen, namely 5-fluorouracil and leucovorin with or
of the CME with CVL are the mobilization of the bowel by without oxaliplatin were not available at this study period.
dissecting along the anatomical planes to obtain negative Other possible mechanism of the survival benefit brought on
circumferential resection margin, and high ligation of the by D3 LND than removal of positive main nodes would be
supplying vessels for adequate LND, which would be almost improved clearance of micro-metastasis which is reported to
identical to D3 LND [21]. Recently, Kanemitsu and col- be associated with increased risk of disease recurrence and
leagues reported excellent long-term outcome of D3 LND in poor survival in CRC patients [24–26]. In D3 LND, the
right hemicolectomy from single institute non-randomized mesocolon including drainage lymphatic vessels and lymph
retrospective study in Japan [22]. In the study, the 5-year nodes up to the origin of the feeding vessels are mobilized by
overall survival of patients with stages II and III were 94.5 anatomical dissection along with embryonic planes, so scope
and 85.0 %, respectively. Among the patients with pT3 and of D3 LND could work as an anatomical landmark to indicate
pT4 colon cancer, 3.8 % (11/284) had metastasis to the main the possible extent of such invisible micro-metastases at sur-
lymph node, and the 5-year disease-free survival of those was gery. In this study, NLNR in D3 LND was significantly larger
than that of D2 LND after adjusting possible confounders
using propensity score, and survival benefit of D3 LND could
Survival probability be seen in patients without lymph node metastasis as well as
1.0 patients with positive nodes. From this result, maximal re-
D3 LND
trieval of micro-metastases to the regional lymph nodes by D3
.8 LND may be associated with improved survival.
Although recent progress in adjuvant chemotherapy for
D2 LND colon cancer has improved survival, 5-year risk of cancer-
.6
related death for stage III colon cancer following high-
.4 intensity regimen such as FOLFOX remains as high as ap-
proximately 30 % [4]. Adjuvant chemotherapy based on suf-
ficient local control with adequate LND would be essential for
.2
further improvement of OS.
The JSCCR strongly recommended D3 LND for T3 and T4
0
colon cancer for the first time in its guidelines issued in 2005
0 10 20 30 40 50 60 70 [27]. Hopefully, D3 LND will be widely disseminated from
Time after surgery (months) now onward. Further analyses using a large-scale prospective
Fig. 2 Overall survival for patients with pT3 and pT4 colon cancer
registration data of the JSCCR will be conducted in the future.
according to D3 and D2 lymph node dissection in the propensity score Finally, several limitations in this study inherent to retro-
matched cohort spective study and non-randomized design should be
Int J Colorectal Dis (2014) 29:847–852
considered. These limitations include patient and treatment
selection bias. The reason why they were offered D2 or D3
LND was not recorded in our database. Patients selected to
undergo D3 LND could have had more favorable clinical
characteristics, including fewer comorbidities, better perfor-
mance status, and less need of emergency for surgical opera-
tion. In addition, non-randomized study may be confounded
by other important contributing factors due to insufficient
information on variables such as histological prognostic fac-
tors (i.e., venous/lymphatic invasion, perineural invasion, and
so forth), CEA and molecular markers, quality of surgery, and
pathological examination that might have differed between
each institute and individual surgeon. This study is also lim-
ited in its generalizability because the study population was
patients who presented to the leading institutes of colorectal
cancer surgery in Japan. Further, since patients without
follow-up information were excluded from the analyses, sur-
vival probabilities in the present study could be over- or
underestimated. In spite of these limitations, our finding was
significant and warrant further investigations.
Conclusion
D3 LND for pT3 and pT4 colon cancer was associated with a
significant survival advantage. This finding provides a rationale
for D3 LND in radical surgery for T3 and T4 colon cancer.
Ethical statement This study was exempted from review by the
Tochigi Cancer Center’s institutional review board, because it used pre-
existing data with no personal identifiers.
Conflict of interest None declared.
References
1. Ferlay J, Shin HR, Bray F, Forman D, Mathers C, Parkin DM (2010)
Estimates of worldwide burden of cancer in 2008: GLOBOCAN
2008. Int J Cancer 127:2893–2917
2. Matsuda T, Marugame T, Kamo K, Katanoda K, Ajiki W, Sobue T,
Japan Cancer Surveillance Research Group (2012) Cancer incidence
and incidence rates in Japan in 2006: based on data from 15
population-based cancer registries in the Monitoring of Cancer
Incidence in Japan (MCIJ) Project. Jpn J Clin Oncol 42:139–147
3. Quasar Collaborative Group, Gray R, Barnwell J, McConkey C, Hills
RK, Williams NS, Kerr DJ (2007) Adjuvant chemotherapy versus
observation in patients with colorectal cancer: a randomised study.
Lancet 370(9604):2020–2029
4. André T, Boni C, Navarro M, Tabernero J, Hickish T, Topham C,
Bonetti A, Clingan P, Bridgewater J, Rivera F, de Gramont A (2009)
Improved overall survival with oxaliplatin, fluorouracil, and
leucovorin as adjuvant treatment in stage II or III colon cancer in
the MOSAIC trial. J Clin Oncol 27(19):3109–3116
5. Yothers G, O’Connell MJ, Allegra CJ, Kuebler JP, Colangelo LH,
Petrelli NJ, Wolmark N (2011) Oxaliplatin as adjuvant therapy for
851
colon cancer: updated results of NSABP C-07 trial, including surviv-
al and subset analyses. J Clin Oncol 29(28):3768–3774
6. National Cancer Center, Center for Cancer Control and Information
Services: the national database of hospital-based cancer registry in
Japan. (in Japanese) http://ganjoho.jp/professional/statistics/hosp_c_
registry.html. Accessed 20 January 2014
7. Gospodarowicz MK, O’Sullivan B, Sobin LH (eds) (2006)
Prognostic factors in cancer, 3rd edn. John Wiley & Sons, Inc,
Hoboken
8. Baxter NN, Virnig DJ, Rothenberger DA, Morris AM, Jessurun J,
Virnig BA (2005) Lymph node evaluation in colorectal cancer pa-
tients: a population-based study. J Natl Cancer Inst 97(3):219–225
9. Le Voyer TE, Sigurdson ER, Hanlon AL, Mayer RJ, Macdonald JS,
Catalano PJ, Haller DG (2003) Colon cancer survival is associated
with increasing number of lymph nodes analyzed: a secondary survey
of intergroup trial INT-0089. J Clin Oncol 21(15):2912–2919
10. Joseph NE, Sigurdson ER, Hanlon AL, Wang H, Mayer RJ,
MacDonald JS, Catalano PJ, Haller DG (2003) Accuracy of deter-
mining nodal negativity in colorectal cancer on the basis of the
number of nodes retrieved on resection. Ann Surg Oncol 10(3):
213–218
11. Swanson RS, Compton CC, Stewart AK, Bland KI (2003) The
prognosis of T3N0 colon cancer is dependent on the number of
lymph nodes examined. Ann Surg Oncol 10(1):65–71
12. Kotake K, Honjo S, Sugihara K, Hashiguchi Y, Kato T, Kodaira S,
Muto T, Koyama Y (2012) Number of lymph nodes retrieved is an
important determinant of survival of patients with stage II and stage
III colorectal cancer. Jpn J Clin Oncol 42(1):29–35
13. Miller EA, Woosley J, Martin CF, Sandler RS (2004) Hospital-to-
hospital variation in lymph node detection after colorectal resection.
Cancer 101(5):1065–1071
14. Japanese Society for Cancer of the Colon and Rectum (2009)
Japanese classification of colorectal carcinoma. Second English ed.
Kanehara-& Co., Ltd., Tokyo
15. Jamieson JK, Dobson JF (1909) The lymphatics of the colon. Proc R
Soc Med 2:149–174
16. Grinnell RS (1950) Lymphatic metastases of carcinoma of the colon
and rectum. Ann Surg 131(4):494–506
17. Grinnell RS (1966) Lymphatic block with atypical and retrograde
lymphatic metastasis and spread in carcinoma of the colon and
rectum. Ann Surg 163(2):272–280
18. Cady B (1997) Basic principles in surgical oncology. Arch Surg
132(4):338–346
19. Hohenberger W, Reingruber B, Merkel S (2003) Surgery for colon
cancer. Scand J Surg 92(1):45–52
20. Hohenberger W, Weber K, Matzel K, Papadopoulos T, Merkel S
(2009) Standardized surgery for colonic cancer: complete mesocolic
excision and central ligation–technical notes and outcome. Colorectal
Dis 11(4):354–364
21. West NP, Kobayashi H, Takahashi K, Perrakis A, Weber K,
Hohenberger W, Sugihara K, Quirke P (2012) Understanding opti-
mal colonic cancer surgery: comparison of Japanese D3 resection
and European complete mesocolic excision with central vascular
ligation. J Clin Oncol 30(15):1763–1769
22. Kanemitsu Y, Komori K, Kimura K, Kato T (2013) D3 lymph node
dissection in right hemicolectomy with a no-touch isolation technique
in patients with colon cancer. Dis Colon Rectum 56(7):815–824
23. Coller FA, Kay EB, Macintyre RS (1941) Regional lymphatic me-
tastases of carcinoma of the colon. Ann Surg 114(1):56–67
24. van der Zaag ES, Bouma WH, Tanis PJ, Ubbink DT, Bemelman WA,
Buskens CJ (2012) Systematic review of sentinel lymph node map-
ping procedure in colorectal cancer. Ann Surg Oncol 19(11):3449–
3459
25. Sirop S, Kanaan M, Korant A, Wiese D, Eilender D, Nagpal S, Arora
M, Singh T, Saha S (2011) Detection and prognostic impact of
micrometastasis in colorectal cancer. J Surg Oncol 103(6):534–537
852
26. Rahbari NN, Bork U, Motschall E, Thorlund K, Büchler MW, Koch
M, Weitz J (2012) Molecular detection of tumor cells in regional
lymph nodes is associated with disease recurrence and poor survival
in node-negative colorectal cancer: a systematic review and meta-
analysis. J Clin Oncol 30(1):60–70
Int J Colorectal Dis (2014) 29:847–852
27. Watanabe T, Itabashi M, Shimada Y et al (2012) Japanese
Society for Cancer of the Colon and Rectum. Japanese
Society for Cancer of the Colon and Rectum (JSCCR) guide-
lines 2010 for the treatment of colorectal cancer. Int J Clin
Oncol 17(1):1–29