| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
NEW APPROACHES TO THE TREATMENT OF HEPATIC MALIGNANCIES |
From the Centre de Chirurgie Viscérale et de Transplantation, Hôpital de Hautepierre, Strasbourg, France.
Correspondence: Address correspondence and reprint requests to: Prof. Daniel Jaeck, Centre de Chirurgie Viscérale et de Transplantation, Hôpital de Hautepierre Strasbourg, Avenue Molière, 67098 Strasbourg Cedex, France; Fax: +33 3 88 12 72 86; E-mail: Daniel.Jaeck{at}chru-strasbourg.fr
ABSTRACT
Background: Surgical resection of colorectal liver metastases (CLM) is the only hope for cure, with a 5-year survival rate ranging from 20% to 54%. However, the resectability rate of CLM is reported to be <20%. This limitation is mainly due to insufficient remnant liver and to extrahepatic disease. Among extrahepatic locations, lymph node metastases are often considered indications of a very poor prognosis and a contra-indication to resection.
Methods and Results: Our studies showed that the prevalence of hepatic pedicle lymph node metastases ranges from 10% to 20%. When located near the hilum and along the hepatic pedicle (area 1) they should not be considered an absolute contra-indication to resection of CLM, and an extended lymphadenectomy should be performed. However, when they reach the celiac trunk (area 2), there is no survival benefit after resection of CLM. For other cases of liver malignancies, lymph node dissection seems justified only in cases of fibrolamellar hepatocellular carcinoma and in case of hilar cholangiocarcinoma. However, few data are available, and they are controversial.
Conclusions: There is a need for more evaluation of lymph node involvement, at least in patients with high risk of such an extension, i.e., patients with more than three metastases, located in segment 4 or 5. There is also a need for prospective trials in order to evaluate the survival benefit of liver resection in such circumstances and the impact of extensive lymphadenectomy.
Key Words: Colorectal liver metastasis Hepatic pedicle lymph node involvement Hepatic resection Lymphadenectomy Prognostic factors
Liver metastases from colorectal cancer (colorectal liver metastases, or CLM) have become the most frequent indication for liver resections. Indeed, nearly half of all patients with colorectal cancer develop liver metastases in the course of the disease. Surgical resection of these liver metastases is considered the "gold standard" for patients in whom all disease can be excised. Many large series of liver resection for metastatic colorectal cancer have been published in which 5-year survival rates ranged from 20% to 54%.17 Conversely, there were very few survivors at 3 years among unresected patients in historical series.8
LIMITS IN INDICATIONS FOR RESECTION
The indications for resection are limited, and the resectability rate is reported to be <20%.4 This limitation is mainly due to two obstacles that have to be overcome. Indeed, major involvement of the liver with multiple and/or bilobar metastases and presence of extrahepatic disease are considered to be the two main contraindications to resection. Concerning the major involvement of the liver, new strategies have been developed in order to extend the frontiers of surgical treatment. These mainly involve preoperative effective chemotherapy,9,10 which may lead to downstaging in some patients; portal vein embolization, which may induce hypertrophy of the remnant liver and increase the safety of hepatectomy1113; and the combination of resection and tumor ablation with radiofrequency or cryotherapy for nonresectable liver metastases.14,15
Recently, it has even been suggested that a margin of 2 mm appears safe enough for curative resection of colorectal liver metastases.16
Unfortunately, recurrences are still observed in about two-thirds of patients after resection of liver metastases. Various attempts are being made to reduce this risk. One method would be to improve the selection of patients who undergo resection. It has been shown that patients with more than three metastases, with deposits >5 cm in diameter, occurring within a short interval (612 months) after resection of a stage III primary colorectal tumor are at higher risk of developing recurrence after liver resection than are patients with small solitary metastasis developing several years after resection of a stage III primary tumor.1,5,7 However, even in cases of higher risk of recurrence, surgical resection offers better long-term survival results than other treatments. With progress in surgical technique and with improved surgical skill, morbidity and mortality after liver resection have been significantly reduced during the past decade; operative mortality rates have decreased to well below 5%.17 As a consequence, the trend is to be more aggressive and to increase the indications for surgical resection of colorectal metastases. However, it does not mean that the indications should be determined only by technical feasibility.
EXTRAHEPATIC DISEASE
Major involvement of the liver can be overcome in selected cases with all these improvements, but the second obstacle, which involves control of extrahepatic disease, must be further investigated before any conclusion about resection can be made. Extrahepatic disease can be located either inside the abdomen or outside it (lungs, bone).
PREVALENCE OF LYMPH NODE METASTASES
We shall focus on the abdomen and particularly on lymph node metastases. Indeed, hepatic metastases of colorectal carcinoma can lead to infiltration of regional lymph nodes in the hepatoduodenal ligament via the lymphatic drainage route of the liver. Colorectal tumors that spread to the lymph nodes of the hepatic pedicle or celiac region are generally considered to be metastases from liver metastases.18,19 The prevalence of regional hepatic lymph node involvement in the presence of colorectal liver metastases has been reported from several series and recently has been summarized.20 This prevalence ranges from 3% to 33%, with a mean value of 9.6%, in the 10 main series reported in the litterature.20
LYMPH NODES METASTASES AND PROGNOSIS
The presence of extrahepatic disease at the time of liver resection indicates a dismal prognosis.21,22 Most surgeons consider hepatic pedicle lymph node (HP-LN) involvement as extrahepatic disease23 and would not perform liver resection in such circumstances. Indeed, the 5-year survival rate is poor for these patients. A systematic review recently analyzed 15 series that provided survival data on 145 node-positive patients.24 Five patients were reported to have survived 5 years after liver resection: 1 was disease-free, 2 had recurrent disease, and 2 had undescribed disease status. However, 5 studies involving 83 patients specified a formal lymph node dissection as part of the surgical procedure, and 4 of the 5 node-positive 5-year survivors were in these studies. The authors of the review concluded that there are few 5-year survivors after liver resection, with or without lymph node dissection, for colorectal hepatic metastases involving the hepatic lymph nodes.
However, of the large number of series of resection of colorectal liver metastases published, only a few analyzed the prevalence of HP-LN involvement and the outcome of surgical resection of CLM in cases of positive lymph nodes. Of many prognostic factors that have been tested, HP-LN involvement is known as one of the most significant factors that affect patient survival.25 As surgeons have become more aggressive in the field of resection of colorectal liver metastases, it has become essential to address this issue and to decide whether to resect CLM in cases of HP-LN involvement. Most surgeons would not proceed with liver resection in cases of HP-LN involvement.24 However, when we analyzed the results of a retrospective multicenter French study2 of 1818 cases of liver resection for CLM, we noted that 12 of 100 patients with positive hepatic nodes were alive 5 years after liver resection. In order to clarify the debate we decided (1) to perform a multicenter prospective study in order to determine prevalence of microscopic HP-LN involvement in patients undergoing curative hepatectomy26 and (2) to investigate, subsequently, whether HP-LN involvement is a more significant prognostic factor and whether HP-LN dissection could be efficient in patients with positive HP-LNs.20
Before discussing briefly the results of these two studies, we summarize below the current knowledge about lymphatic drainage of the liver.
THE LYMPHATICS OF THE LIVER AND LYMPHATIC SPREAD
Several lymphatic pathways from the liver have been described18,27: (1) through the hepatic hilum and the retropancreatic portion, along the hepatic artery and the celiac axis; (2) along the falciform ligament, across the diaphragm, into the mediastinum (precardiac and juxtaesophageal lymph nodes); (3) through the esophageal hiatus and the caval foramen, also into the mediastinum; (4) along the lesser omentum and the upper gastric portion; and (5) along the phrenic artery into the lymph nodes surrounding the celiac axis.
According to the studies on lymphatics of the liver, it appears that in the first step, lymph nodes of the hepatoduodenal ligament and retropancreatic area are involved (area 1 in our study)20 (Fig. 1), whereas in the second step, nodes around the common hepatic artery and the celiac axis (area 2 in our study) are involved. The studies on lymphatic spread of gallbladder cancer,28 in particular, showed this way of extension. However, skip metastases and overlapping of the spreading route have been described. In our own study, six of nine patients presenting with area 2 lymph node involvement did not have area 1 lymph node involvement. Consequently, lymph node sampling does not appear to be an accurate method for evaluating lymph node involvement. Only a complete lymphadenectomy of both area 1 and area 2 enables accurate assessment of the lymph node status. However, this procedure cannot be recommended routinely, and currently we use it only in cases involving more than three metastases, cases involving metastases in segment 4 or 5, and cases in which a solitary peritoneal deposit can be completely resected or a poorly differentiated carcinoma is noted. Indeed, all these risk factors were significantly more frequently associated with positive HP-LN in our study.20
|
OTHER LIVER MALIGNANCIES
Lymphatic spread of primary liver malignancies along the described routes has been previously demonstrated at autopsy.29 It has also been assessed in cases of intrahepatic3032 and hilar33 cholangiocarcinoma. The HP-LNs are frequently involved, and long-term survival after extended lymph node dissection has been demonstrated (14.7% 5-year survival for 39 patients with regional node metastases,33 not so different from the 30.5% survival for 52 patients without node involvement). To the contrary, another study suggested that lymph node dissection did not prolong survival.31 In 25% to 33% of patients with advanced hepatocellular carcinoma (HCC), HP-LN metastases have been observed at autopsy.34 In one series, for instance, whereas some patients survived for more than 3 years following resection of HCC with solitary lymph node involvement, no patient survived more than 25 months following resection of HCC with more than two lymph node metastases.35
Fibrolamellar HCC is known to be frequently associated with HP-LN metastases. To increase the chances of cure, an extended lymph node dissection at the time of resection of the liver tumor has been recommended as a means of preventing later lymph node metastases.36 Furthermore, such lymph node recurrence can be successfully treated by lymphadenectomy.36
SIGNIFICANCE OF LYMPH NODE INVOLVEMENT IN CASES OF COLORECTAL LIVER METASTASES
In order to try to build an appropriate strategy for treatment of CLM, with special attention to HP-LN, we analyzed the results of our two prospective studies in which we participated26 or conducted.20
Prevalence of Microscopic HP-LN Involvement in Patients Undergoing Curative Hepatectomy for CLM
Extensive lymph node dissection of the hepatic pedicle was undertaken in 100 consecutive patients undergoing curative hepatectomy for CLM26 in whom HP-LN involvement was not macroscopically detectable. Seven institutions participated in this study. Microscopic lymph node involvement, at different lymph node sites, was found in 14 patients and was related to the number of metastases, extent of liver involvement, and carcinoembryonic antigen level (all P < .05). It is noteworthy that 100 consecutive curative hepatectomies with extensive lymphadenectomy were performed and resulted in no deaths in seven different hospitals. Extensive lymphadenectomy should be performed carefully to avoid any morbidity.
Significance of HP-LN Involvement in Patients Undergoing Curative Hepatectomy for CLM
Among 174 patients undergoing hepatectomy for CLM20 in our institution during the study period (19931998), we included 160 patients who underwent extensive lymphadenectomy of the hepatoduodenal ligament and retropancreatic portion as well as around the common hepatic artery and the celiac trunk. The 160 patients were included according to the following criteria: first hepatectomy; age <75 years; absence of severe cardiovascular, pulmonary, or other associated disease; absence of other malignant disease; absence of distant metastases secondary to colorectal cancer other than liver metastases (except a solitary peritoneal tumor that could be completely resected during hepatectomy); and complete resection of liver metastases confirmed during surgery.
The survival of patients with HP-LN involvement limited to the hepatoduodenal ligament and retropancreatic portion (area 1) was compared with that of patients with HP-LN involvement spreading over the common hepatic artery and the celiac axis (area 2). HP-LN involvement was detected in 17 patients: for 8, in area 1, and for 9, in area 2. HP-LN involvement was significantly more frequent in patients with more than three metastases, with metastases located in segment 4 or 5, with a solitary resectable peritoneal deposit, or with poorly differentiated adenocarcinoma. The survival rate was significantly lower among patients with HP-LN involvement (62% 3-year survival in the negative group of 143 patients, versus 18.9% in the positive group of 17 patients; P < .0001). Moreover, univariate analysis showed that HP-LN involvement was a more significant prognostic factor than others such as synchronous metastases, bilobar disease, presence of four or more metastatic nodules, and presence of a resectable solitary peritoneal nodule. No retroperitoneal, mesenteric, or para-aortic lymph node metastases were found in the 17 patients with positive HP-LNs. No patient with area 2 involvement survived longer than 1 year after hepatectomy, whereas two of eight patients with HP-LN involvement limited to area 1 were still alive 3 years after hepatectomy. The survival rate was significantly higher among patients in whom HP-LN involvement was limited to area 1 than among patients with area 2 metastases (3-year survival: 38% vs. 0%; P < .001).
Coxs proportional hazard test showed that in multivariate analysis, HP-LN involvement, existence of a resectable solitary peritoneal nodule, presence of four or more metastatic nodules, and synchronous metastases significantly predicted poorer survival. The relative risk of HP-LN involvement was the greatest (15.1) of all the significant prognostic factors, even greater than the presence of a resectable solitary peritoneal nodule (3.5). Our study leads us to conclude that involvement of HP-LN is one of the most significant prognostic indicators in patients with CLM. Sampling and intraoperative histological control of HP-LN during surgery for CLM appear justified, at least for patients who are at high risk of HP-LN involvement, i.e., those with more than three metastases, with metastases located in segment 4 or 5, with a solitary resectable peritoneal deposit, or with a poorly differentiated adenocarcinoma. It is likely that in patients with metastases located in segment 7 or 8 and with HP-LN involvement, HP-LN involvement also develops in the mediastinum.
Recently, other prognostic factors have been studied, and our group also noted, for instance, a significant correlation between proliferation index (Ki67)37 or microsatellite instability38 and progression of colorectal cancer.
CONCLUSION
The treatment of lymph node metastases secondary to liver tumors is still controversial. There are two opposite strategies: either consider lymph node metastases as a sign of generalized disease and a contraindication to surgery (and therefore offer only palliative chemotherapy to these patients) or try to achieve a radical resection (as with any other extrahepatic deposit) and perform a lymphadenectomy from the hilum of the liver to the celiac area. This latter strategy is widely accepted for other tumors such as esophageal and gastric tumors, colorectal cancer, and hilar cholangiocarcinoma. It is also accepted for primary tumors of the liver or of the biliary tract, but it remains highly controversial for liver metastases because the lymph node involvement appears as metastases of metastases. Whether extended lymph node dissection provides a survival benefit is a question requiring further studies, ideally prospective randomized trials. The fact that long-term survival has been observed despite involvement of area 1 HP-LN should encourage to undertake these trials.
In conclusion, it appears mandatory to pay attention to the status of the HP-LN, first to establish a more precise prognosis and, second, in selected cases for prospective trials, to evaluate the impact of extensive lymphadenectomy on survival. Complete eradication of the disease can offer to the patients the best chance for cure or at least a prolonged survival.
FOOTNOTES
Hepatic pedicle lymph nodes involvement from colorectal liver metastases or from other liver malignancies appears to be one of the most significant factors of poor prognosis. However, when limited to the proximal area, it should not contraindicate liver resection associated with lymphadenectomy.
Received for publication September 6, 2002. Accepted for publication September 8, 2002.
REFERENCES
This article has been cited by other articles:
![]() |
J. J. Bennett, C. R. Schmidt, D. S. Klimstra, S. R. Grobmyer, N. M. Ishill, M. D'Angelica, R. P. DeMatteo, Y. Fong, L. H. Blumgart, and W. R. Jarnagin Perihepatic Lymph Node Micrometastases Impact Outcome after Partial Hepatectomy for Colorectal Metastases Ann. Surg. Oncol., April 1, 2008; 15(4): 1130 - 1136. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. Mulier, Y. Ni, J. Jamart, L. Michel, G. Marchal, and T. Ruers Radiofrequency Ablation Versus Resection for Resectable Colorectal Liver Metastases: Time for a Randomized Trial? Ann. Surg. Oncol., January 1, 2008; 15(1): 144 - 157. [Abstract] [Full Text] [PDF] |
||||
![]() |
T. M. Pawlik, R. D. Schulick, and M. A. Choti Expanding Criteria for Resectability of Colorectal Liver Metastases Oncologist, January 1, 2008; 13(1): 51 - 64. [Abstract] [Full Text] [PDF] |
||||
![]() |
G. Poston, R. Adam, and J.-N. Vauthey Downstaging or Downsizing: Time for a New Staging System in Advanced Colorectal Cancer? J. Clin. Oncol., June 20, 2006; 24(18): 2702 - 2706. [Full Text] [PDF] |
||||
![]() |
W. J. Lesterhuis, I. J. M. de Vries, D. H. Schuurhuis, A. C. I. Boullart, J. F. M. Jacobs, A. J. de Boer, N. M. Scharenborg, H. M. H. Brouwer, M. W. M. M. van de Rakt, C. G. Figdor, et al. Vaccination of colorectal cancer patients with CEA-loaded dendritic cells: antigen-specific T cell responses in DTH skin tests Ann. Onc., June 1, 2006; 17(6): 974 - 980. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |