Journal of Surgical Oncology 2008;97:136–140

Isolated Paraaortic Lymph-Node Recurrence After

the Curative Resection of Colorectal Carcinoma

BYUNG SOH MIN, MD, NAM KYU KIM, MD, PhD,* SEUNG KOOK SOHN, MD, CHANG HWAN CHO, MD,KANG YOUNG LEE, MD, AND SEUNG HYUK BAIK, MD

Department of Surgery, Yonsei University College of Medicine, Seoul, Korea

Background and objectives: Isolated paraaortic lymph-node recurrence (IPLR) after curative surgery for colorectal carcinoma is rare and no

previous report has specifically addressed this type of recurrence. We investigated the clinical features of IPLR and analyzed prognostic factors.

Methods: Of 2,916 patients who underwent curative surgery for colorectal carcinoma, IPLR was identified in 38 patients (1.3%). The clinical

features and prognostic factors of these patients were analyzed.

Results: IPLR was first detected by increased serum carcinoembryonic antigen (CEA) levels (63.2%) or by routine follow-up computed

tomography (CT) (36.8%). Curative resection of IPLR was performed in six patients (15.8%). A total of 19 patients (50.0%) received

chemoradiation therapy and 13 patients (34.2%) received chemotherapy only. The median survival from IPLR was 13 months (range: 5–

60 months). The median survival time from IPLR for the resected patients was 34 months, whereas it was 12 months for those who did not undergo

resection (P¼ 0.034). The factors associated with the prognosis were histological grade (P¼ 0.003), location (P¼ 0.032), and resection of IPLR

(P¼ 0.034).

Conclusions: IPLR after curative surgery for colorectal carcinoma is rare. Although it is generally associated with poor prognosis, better survival

might be achieved through curative resection in selected cases.

J. Surg. Oncol. 2008;97:136–140. � 2007 Wiley-Liss, Inc.

KEY WORDS: paraaortic lymph-node recurrence; locoregional recurrence; colorectal cancer

INTRODUCTION

Colorectal cancer is the third most frequent cancer worldwide [1].

Nearly 0.6 million new cases are diagnosed each year. During the past

few decades, the incidence of colorectal cancer in Asian populations

has increased by two to four times, and South Korea is no exception

[2]. Colorectal cancer is the fourth leading cause of cancer-related

mortality in South Korea.

Although primary colorectal adenocarcinoma is known to have a

relatively better prognosis than other malignancies in the gastro-

intestinal tract, approximately 20–40% of patients undergoing curative

resection will eventually suffer some form of recurrence [3]. Therefore,

understanding the patterns and prognoses of various forms of recurrence

is essential to improve the oncologic outcomes of patients with colorectal

cancer.

Isolated paraaortic lymph-node recurrence (IPLR) is a very rare

type of recurrence. In previous reports it was often classed as a retro-

peritoneal recurrence, which is a subtype of distal local recurrences [4–

6]. However, retroperitoneal recurrences sometimes represent growth of a

tumor deposit or residual from surgery [6], and there are no published

data specifically addressing the patterns and prognosis of IPLR.

Furthermore, it is questionable whether IPLR should be categorized as

a distal regional recurrence because, according to the American Joint

Committee on Cancer (AJCC) staging system [7], paraaortic lymph-

node metastasis is categorized as M1.

Thus, in the present study we investigated clinical characteristics of

patients with IPLR after curative surgery for colorectal cancer and

analyzed factors that affect the prognosis of this subgroup of patients.

PATIENTS AND METHODS

We defined IPLR as isolated recurrences of colorectal cancer in an

area adjacent to the abdominal aorta without evidence of recurrence at

any other sites after curative surgery for colorectal carcinoma. Biopsy

confirmation was not always required provided the results of conven-

tional radiologic studies such as abdominal-computed tomography (CT),

magnetic resonance imaging (MRI), and positron emission tomography

(PET) were highly indicative of recurrence according to radiologists

(Fig. 1). The location of IPLR was classified as A (above renal vessels)

or B (below renal vessels) according to the classification of the

Japanese Society of Clinical Oncology [8].

Using a prospective clinical database at Severance Hospital, Yonsei

University Healthcare System, the clinicopathological data of 2,916

patients who underwent curative surgery for colorectal cancer between

1992 and 2004 were retrieved. From this population, 201 patients

(6.9%) were noted to have local recurrences and 455 patients (15.6%)

had systemic recurrences. Thirty-eight patients (1.3%) were then

identified as having IPLR fitting this definition.

The clinical and histopathological variables were investigated and

analyzed for prognostic significance. Patient, tumor, and treatment

factors were correlated using the Fisher exact test and Mann-Whitney

test. Statistical analyses were performed using the log-rank test and

Kaplan–Meier estimates with cancer-specific survival as the primary

endpoint. A two-sided P-value of less than 0.05 was considered

significant. All statistical tests were performed using SPSS software

(version 12.0, SPSS, Chicago, IL).

Grant sponsor: Korea Health 21 R&D Project, Ministry of Health andWelfare, Republic of Korea; Grant numbers: 0412-CR01-0704-0001, 0405-BC01-0604-00020.

*Correspondence to: Nam Kyu Kim, MD, PhD, Department of Surgery,Yonsei University College of Medicine, Seodamun-gu Shincheon-dong134, Seoul, Korea. Fax: 82-2-313-8289.E-mail: namkyuk@yumc.yonsei.ac.kr

Received 22 May 2007; Accepted 14 September 2007

DOI 10.1002/jso.20926

Published online 26 October 2007 in Wiley InterScience(www.interscience.wiley.com).

� 2007 Wiley-Liss, Inc.

RESULTS

Demographics and Primary Tumors

The mean age of enrolled patients was 54.6 years and most patients

were male (60.5%) (Table I). The colon (55.3%) was a more common

primary tumor location than the rectum (44.7%). The mean serum

carcinoembryonic antigen (CEA) level measured 7 days after the

curative resection of the primary tumor was 9.4 ng/ml (reference range:

�5 ng/ml). Moderately differentiated adenocarcinoma (78.9%) was

the most common histological-type, followed by poorly differentiated

adenocarcinoma (7.9%) and mucinous adenocarcinoma (7.9%). Most

patients (86.8%) had lymph-node metastasis and 73.7% of the patients

had more than three metastatic lymph nodes (Table I). All the patients

received either chemoradiation or chemotherapy postoperatively after

resection of the primary tumor except one patient who had a primary

tumor stage of pT2N0M0. Neither before nor after IPLR patients

received an operation for separate metastases unrelated to the lymph

node recurrence. The mean and median follow-up periods were

30.9 months and 30 months, respectively.

The Clinical Patterns of IPLR

The mean and median disease-free intervals were 19.9 months and

14 months, respectively. No patients showed any symptoms. An

increased serum CEA level at routine check-up was the first

presentation in 63.2% of the patients, whereas recurrence was found

incidentally in routine check-up abdominal CT scans for the other

36.8% of patients. The mean serum CEA level at the time of recurrence

was 40.0 ng/ml. One half of the recurrences occurred in the A region

and the other half in the B region. In six patients (15.8%), R0 resection

was performed (Table II). All patients underwent postoperative

chemotherapy. Radiation therapy plus concurrent or sequential

chemotherapy was performed in 19 patients (50.0%), and systemic

chemotherapy only was performed in 13 patients (34.2%). The mean

and median survival times after the recurrence were 17.7 months

and 13 months, respectively. All the enrolled patients eventually

developed secondary metastases. All of the patients who had received

chemotherapy alone or with radiation developed another multiple

systemic metastases in other solid organs. Multiple liver metastases

were developed in 30 patients (78.9%), lung metastases in 28 patients

(73.7%), brain metastases in 3 patients (7.9%), and bone metastases in

9 patients (23.7%) during the course of treatment. For those who

underwent surgical resection, the mean and median disease-free

survival after the resection of IPLR were 28 and 21 months,

respectively. Four out of six patients developed multiple secondary

metastases. The most frequent site of secondary metastases were liver

(six patients) followed by lung (three patients), bone (two patients).

The Analyses of Prognostic Factors

Factors analyzed included age, gender, histological grade, patho-

logic T and N stages of primary tumor, the location of primary tumor,

performance of surgical resection, serum CEA level 7 days after the

operation and at the time of recurrence, disease-free interval, and the

location of IPLR. Histological grade (P¼ 0.003), the location of IPLR

(P¼ 0.032), and surgical resection (P¼ 0.034) were found to be

significantly associated with survival after the recurrence according to

univariate analyses (Table III). The median survival time after the

Journal of Surgical Oncology DOI 10.1002/jso

Fig. 1. Abdominal-computed tomography (CT) scan (A) and positron emission tomography (PET) (B) images of isolated paraaortic lymph-node recurrence (white and black arrows) in a 56-year-old male patient who had undergone left hemicolectomy due to adenocarcinoma of thedescending colon 2 years earlier. Abdominal CT revealed a possible slightly enhanced, 1-cm nodule conglomerate around the left paraaortic area(A, white arrow), which showed uptake of 18F-fluoro-deoxyglucose (18F-FDG) in the PET images (B, black arrow), suggesting paraaortic lymph-node recurrence of the cancer.

TABLE I. Patient Characteristics

Factors N (%)

Age, mean (range) 54.6 years (15–87)

Male-to-female ratio 23:15

Location

Colon 21 (55.3%)

Rectum 17 (44.7%)

Histology

Well differentiated 2 (5.3%)

Moderately differentiated 30 (78.9%)

Poorly differentiated 3(7.9%)

Mucinous 3 (7.9%)

Mean serum CEA level at POD #7 9.4 ng/ml (0.1–121.7)

Primary tumor stage

pT2N0 1 (2.6%)

pT2N1 1 (2.6%)

pT2N2 2 (5.3%)

pT3N0 2 (5.3%)

pT3N1 4 (10.5%)

pT3N2 24 (63.1%)

pT4N0 2 (5.3%)

pT4N1 0

pT4N2 2 (5.3%)

TABLE II. Summary of Treatment Methods

Treatment method N

Surgery plus chemotherapy 6

Paraaortic lymph-node dissection with en-bloc

left nephrectomy and resection of small bowel and

remnant colon

2

Paraaortic lymph-node dissection en-bloc resection of

small bowel and/or remnant colon

4

Radiation plus chemotherapy

Concurrent chemoradiation 12

Chemotherapy with sequential radiation therapy 7

Systemic chemotherapy only 13

Isolated Paraaortic Lymph-Node Recurrence 137

recurrence was 34 months for the patients who underwent surgical

resection and 14 months for those who did not (Fig. 2). Among the

three variables, histological grade, IPLR location, and surgical

resection, only histological grade was found to be an independent

prognostic factor for survival after recurrence according to multivariate

analyses (hazard ratio: 2.844; 95% confidence interval: 1.037–7.797;

P¼ 0.042).

Treatment Details

The clinical characteristics of the patients who underwent surgical

resection of IPLR are summarized in Table IV. Resectability was

determined on the basis of radiologic findings by CT scan. Radiologic

findings were reviewed by both radiologists and surgeons at the

colorectal tumor board of our institution. Extensive involvement of

vascular structures such as superior mesenteric artery and celiac axis,

adjacent organs such as pancreas and bile duct, and poor performance

status were considered as contraindications for surgical resection. Most

of the IPLRs occurred in the region A (at the level of or above the level

of renal vessels) were unresectable because most cases involved either

celiac axis or the root of superior mesenteric artery. However, in two

cases, metastatic lymph nodes were confined to around left renal artery

and curative lymph node dissections were successfully performed with

combined left nephrectomies. En-bloc resection of small bowel and/or

the remnant colon was performed when needed. Curative resection can

be performed in four patients with IPLR in region B because their

IPLRs were confined around or below inferior mesenteric artery. In two

cases, the root inferior mesenteric artery had not been ligated at

the time of initial surgery, thus ligation of inferior mesenteric artery

and resection of remnant colon were necessary. In the rest of the cases

with IPLR in the region B unresectability was due to encasement or

invasion of aorta and/or vena cava, and due to poor performance status.

Curative resection was confirmed in all six patients by pathologically

negative resection margins. There was no operative mortality.

Operative morbidity was 33.3%. Two out of six patients had an early

intestinal obstruction complication, which was successfully managed

by conservative measures. In all six patients who underwent surgical

resection, postoperative chemotherapy with 5-fluorouracil–leucovorin

(5-FU/LV) and oxaliplatin was performed.

Radiation therapy was performed in 19 patients (50.0%). A total of

50.4 Gy was delivered in 14 (73.7%) of 19 patients. Five patients

(26.3%) failed to complete the radiation therapy and received between

3,240 and 4,500 cGy. Twelve (63.2%) of 19 patients received

concurrent chemoradiation therapy with the 5-FU/LV regimen. In 7

of 19 patients (36.8%), systemic chemotherapy with sequential radiation

therapy was performed. Systemic chemotherapy was performed using a

combination of oxaliplatin (n¼ 4; 21.1%) or irinotecan (n¼ 3; 15.8%)

and 5-FU/LV. Thirteen patients (34.2%) received only systemic

chemotherapy for the treatment of IPLR. Eight of 13 patients (61.5%)

received oxaliplatin-containing chemotherapy, whereas five patients

(38.5%) received irinotecan-containing chemotherapy.

DISCUSSION

Although IPLR is a rare type of recurrence for all cancers, IPLR is

well characterized in cervical cancer [9,10]. IPLR in cervical

carcinoma is known to occur in about 2% of patients and can be

successfully treated with concurrent chemoradiation therapy. In

colorectal cancer, the concept of IPLR is not established and has

previously been categorized as a retroperitoneal recurrence, which is a

type of locoregional recurrence [4–6]. Retroperitoneal recurrences

include not only IPLR, but also growth of a tumor deposit or residual

from surgery [6], so we can only infer the pattern and prognosis of

IPLR from previous reports.

We found the incidence of IPLR to be 1.3%, but because our

hospital is a tertiary referral hospital in Korea the current study

population cannot represent all patients who underwent curative

resection of colorectal cancer. Therefore, the actual incidence of IPLR

after the curative resection of colorectal cancer might be lower than

recorded in our study.

Poor prognosis of IPLR can be deduced from previous reports

[6,11]. Shibata et al. [6] reported the median survival time after

recurrence to be 15 months. However, the median survival time for

patients with resectable recurrent disease is 40–44 months, which is

similar to the reported survival time after resection of all types of

locoregional recurrences [3–6,11].

Previous studies [6,12] reported an association between the disease-

free interval and survival after recurrence. Bowne et al. [11] could not

find any correlation between length of the disease-free interval and

post-salvage survival, which they attributed to patient selection or a

small sample size. The results from the present study were in agree-

ment with those of Bowne et al. However, we found that the mean

disease-free interval of the patients who underwent surgical resection

of IPLR was significantly longer than that of those who did not undergo

resection.

Journal of Surgical Oncology DOI 10.1002/jso

TABLE III. Univariate Analyses of Prognostic Factors

Factors

Mean/median survival

(months) P-value

Surgical resectiona

Yes 42.5/34 0.034

No 18.7/14

CEA at POD#7b

Within normal rangec 21.3/15 0.555

Above normal 25.5/16

CEA at recurrence

Within normal rangec 21.1/14 0.803

Above normal 23.4/15

Primary tumor location

Colon 19.9/14 0.582

Rectum 24.2/15

Disease-free interval

�24 months 20.7/14 0.233

>24 months 24.0/28

Age

�55 years 22.2/15 0.796

23.2/14

Histology

Well/mod. diff.d 26.2/23 0.003

Poorly diff./mucinous 7.8/4

Gender

Male 24.4/15 0.543

Female 21.3/14

Primary T stage

pT2-3 23.3/15 0.707

pT4 19.3/11

Primary N stage

pN(�) 20.2/9 0.461

pN(þ) 23.9/15

IPLR locatione

A 15.9/14 0.032

B 29.7/23

a‘‘Yes’’ means surgery plus chemotherapy and ‘‘No’’ means chemotherapy with

or without radiation therapy.bSerum carcinoembryonic antigen (CEA) level at the 7th postoperative day

(POD).cThe reference range was within 5 ng/ml.dModerately differentiated.eThe location was categorized according to the classification of lymph nodes by

the Japanese Society of Clinical Oncology. A was defined as above the renal

vessels and B as below the renal vessels; IPLR, isolated paraaortic lymph-node

recurrence.

138 Min et al.

We observed that, in resected cases, the median survival after IPLR

was 34 months, whereas it was 14 months for those who did not

undergo resection. When considering patients for resection of IPLR,

careful selection is important. Previous studies [6] have reported that

potential factors associated with resectability and subsequent favorable

outcome include longer disease-free interval and smaller tumor size.

Complete resection with negative margins was crucial for durable

tumor control. The current study has the limitations such as small

patients number, heterogeneous treatment protocols (inconsistent

chemotherapy regimens and radiation dose), and retrospective nature,

but on the basis of our results we may suggest some predilections for

surgical resection of IPLR: (1) The IPLRs occurred in region A are

frequently unresectable. However when IPLR is confined to either

sides of aorta involving neither celiac axis nor superior mesenteric

artery, curative resection may be possible with combined resections.

(2) Curative resection can be performed for IPLRs in region B

especially when they are confined around or below inferior mesenteric

artery. The root inferior mesenteric artery was not ligated at the time of

initial surgery, ligation of inferior mesenteric artery, and resection of

remnant left colon may be necessary. We also identified the patient

characteristics that indicate improved prognosis after surgical resection

of IPLR such as longer disease-free interval, lower serum CEA levels,

and more favorable tumor histology, and our observations might reflect

that less aggressive biology might have an important role. This was

supported by multivariate analyses that showed tumor histological

grade, an indicator of tumor biology, to be a single independent

prognostic factor associated with survival after IPLR.

We observed improved survival when IPLR was located below the

renal vessels (location B). This may be due to the fact that more

surgical resections were performed for cases of IPLR located below

the renal vessels. IPLR in the A region (above renal vessels) was more

likely to involve adjacent organs such as the pancreas, the root of the

superior mesenteric artery (SMA), the duodenum, the stomach and

the renal hilum, which may have made complete removal by surgery or

irradiation with adequate margins difficult and might have influenced

the oncologic outcomes of the subgroups.

Journal of Surgical Oncology DOI 10.1002/jso

Fig. 2. The survival rates after recurrence. The median survival time after the recurrence was 34 months for the patients who underwent surgicalresection and 14 months for those who did not (P¼ 0.034).

TABLE IV. Comparison of Clinical and Histopathologic Factors Between

the Patients Who Underwent Surgical Resection of IPLR and Those Who

Did Not

Factors

Resected

(n¼ 6)

Not-resected

(n¼ 32) P-value

Mean age 58.2 years 53.9 years 0.427

Gender

Male 3 (50.0%) 20 (62.5%) 0.663

Female 3 (50.0%) 12 (37.5%)

Primary tumor location

Colon 3 (50.0%) 14 (43.8%) 0.778

Rectum 3 (50.0%) 18 (56.3%)

Histology

Well/mod. diff.a 6 (100%) 26 (81.3%) 0.562

Poorly diff./mucinous 0 6 (18.8%)

Mean serum CEA

At POD#7b 2.66 ng/ml 5.97 ng/ml 0.047

At recurrence 22.04 ng/ml 49.83 ng/ml 0.050

Mean disease-free interval 22 months 18 months 0.049

Primary T stage

pT2-T3 6 (100%) 28 (77.5%) 0.428

pT4 0 4 (12.5%)

Primary N stage

pN(�) 0 5 (15.6%) 0.570

pN(þ) 6 (100%) 27 (84.4%)

Location of IPLR recurrencec

A 2 (33.3%) 17 (53.1%) 0.660

B 4 (66.7%) 15 (46.9%)

aModerately differentiated.bSerum carcinoembryonic antigen (CEA) level at the 7th postoperative day

(POD). The reference range was within 5 ng/ml.cThe location was categorized according to the classification of lymph nodes by

the Japanese Society of Clinical Oncology. A was defined as above the renal

vessels, and B as below the renal vessels; IPLR, isolated paraaortic lymph-node

recurrence.

Isolated Paraaortic Lymph-Node Recurrence 139

The median survival after IPLR in this study was 13 months, which

is poorer than results from recent studies on systemic chemotherapy for

metastatic colorectal cancer [13]. This might be partly owing to the

poorer prognosis associated with IPLR, but might also be because this

study included data collected over many years. Most of the patients

who were treated in the early- and mid-1990s did not receive benefits

from the recent developments in chemoradiation treatment.

The new biologic agents have improved the survival of patients with

metastatic colon cancer [14,15], and new radiation techniques such as

intensity-modulated radiotherapy (IMRT), tomotherapy, and robotic

linear accelerators (CyberKnife, Accuray Inc., Sunnyvale, CA) have

made it possible to deliver high-dose radiation to focused areas without

damaging adjacent normal organs [16]. These new treatment modalities

could all improve the survival of the patients with IPLR. In particular, in

some cases, state-of-the-art radiation therapy might play a role as

important as that of surgery in the future.

CONCLUSIONS

IPLR after curative surgery for colorectal carcinoma is a very rare

type of recurrence and is associated with poor prognosis. However, in

cautiously selected cases, better survival might be expected by

potentially curative resection of IPLR.

ACKNOWLEDGMENTS

This work was selected as an outstanding presentation at 2007

Spring Meeting of Korean Society of Coloproctology.

REFERENCES

1. Coleman MP, Esteve J, Arslan PDA, et al.: Trend in cancerincidence and mortality. Lyon: IARC 1993.

2. Sung JJ, Lau JY, Goh KL, et al.: Increasing incidence of colorectalcancer in Asia: Implications for screening. Lancet Oncol 2005;6:871–876.

3. Galandiuk S, Wieand HS, Moertel CG, et al.: Patterns ofrecurrence after curative resection of carcinoma of the colonand rectum. Surg Gynecol Obstet 1992;174:27–32.

4. Delpero JR, Pol B, Le Treut P, et al.: Surgical resection of locallyrecurrent colorectal adenocarcinoma. Br J Surg 1998;85:372–376.

5. Salo JC, Paty PB, Guillem J, et al.: Surgical salvage of recurrentrectal carcinoma after curative resection: A 10-year experience.Ann Surg Oncol 1999;6:171–177.

6. Shibata D, Paty PB, Guillem JG, et al.: Surgical management ofisolated retroperitoneal recurrences of colorectal carcinoma. DisColon Rectum 2002;45:795–801.

7. Greene FL: The American Joint Committee on Cancer: Updatingthe strategies in cancer staging. Bull Am Coll Surg 2002;87:13–15.

8. Committee on Classification of Regional Lymph Nodes of JapanSociety of Clinical Oncology. Classification of regional lymphnodes in Japan. Int J Clin Oncol 2003;8:248–275.

9. Chou HH, Wang CC, Lai CH, et al.: Isolated paraaortic lymphnode recurrence after definitive irradiation for cervical carcinoma.Int J Radiat Oncol Biol Phys 2001;51:442–448.

10. Singh AK, Grigsby PW, Rader JS, et al.: Cervix carcinoma,concurrent chemoradiotherapy, and salvage of isolated paraaorticlymph node recurrence. Int J Radiat Oncol Biol Phys 2005;61:450–455.

11. Bowne WB, Lee B, Wong WD, et al.: Operative salvagefor locoregional recurrent colon cancer after curative resection:An analysis of 100 cases. Dis Colon Rectum 2005;48:897–909.

12. Gwin JL, Hoffman JP, Eisenberg BL: Surgical management ofnonhepatic intra-abdominal recurrence of carcinoma of the colon.Dis Colon Rectum 1993;36:540–544.

13. Kelly H, Goldberg RM: Systemic therapy for metastatic color-ectal cancer: Current options, current evidence. J Clin Oncol2005;23:4553–4560.

14. Cunningham D, Humblet Y, Siena S, et al.: Cetuximabmonotherapy and cetuximab plus irinotecan in irinotecan-refractory metastatic colorectal cancer. N Engl J Med 2004;351:337–345.

15. Hurwitz H, Fehrenbacher L, Novotny W, et al.: Bevacizumab plusirinotecan, fluorouracil, and leucovorin for metastatic colorectalcancer. N Engl J Med 2004;350:2335–2342.

16. Fenwick JD, TomeWA, Soisson ET, et al.: Tomotherapy and otherinnovative IMRT delivery systems. Semin Radiat Oncol 2006;16:199–208.

Journal of Surgical Oncology DOI 10.1002/jso

140 Min et al.