Repeat hepatic surgery for colorectal cancer metastasis to the liver.

OBJECTIVE: The authors addressed whether a repeat hepatic operation is warranted in patients with recurrent isolated hepatic metastases. Are the results as good after second operation as after first hepatic operation? SUMMARY BACKGROUND DATA: Five-year survival after initial hepatic operation for colorectal metastases is approximately 33%. Because available alternative methods of treatment provide inferior results, hepatic resection for isolated colorectal metastasis currently is well accepted as the best treatment option. However, the main cause of death after liver resection for colorectal metastasis is tumor recurrence. METHODS: Records of 95 patients undergoing initial hepatic operation and 10 patients undergoing repeat operation for isolated hepatic metastases were reviewed for operative morbidity and mortality, survival, disease-free survival, and pattern of failure. The literature on repeat hepatic resection for colorectal metastases was reviewed. RESULTS: The mean interval between the initial colon operation and first hepatic resection was 14 months. The mean interval between the first and second hepatic operation was 17 months. Operative mortality was 0%. At a mean follow-up of 33 +/- 27 months, survival in these ten patients was 100% at 1 year and 88% +/- 12% at 2 years. Disease-free survival at 1 and 3 years was 60% +/- 16% and 45% +/- 17%, respectively. After second hepatic operation, recurrence has been identified in 60% of patients at a mean of 24 +/- 30 months (median 9 months). Two of these ten patients had a third hepatic resection. Survival and disease-free survival for the 10 patients compared favorably with the 95 patients who underwent initial hepatic resection. CONCLUSIONS: Repeat hepatic operation for recurrent colorectal metastasis to the liver yields comparable results to first hepatic operations in terms of operative mortality and morbidity, survival, disease-free survival, and pattern of recurrence. This work helps to establish that repeat hepatic operation is the most successful form of treatment for isolated recurrent colorectal metastases.     

Patterns of failure after surgical cure of large liver tumors: A change in the proximate cause of death and a need for effective systemic adjuvant therapy

During a period of 7 years, we have aggressively treated liver tumors whether primary or metastatic. Our experience after 43 curative major liver resections has shown an excellent overall survival: 34 of 43 patients still alive a median of 12 months after liver resection (patient ages ranged from 21 to 85 years, median 57 years). Nineteen patients underwent right hepatic lobectomy, 9 trisegmentectomy, 5 left hepatic lobectomy, 5 extended left hepatic lobectomy, 4 right lobectomy plus left lobe wedge resection, and 1 patient underwent a major hilar wedge resection. Two patients died from sepsis and hepatic failure on or before the 60th postoperative day. One patient with no evidence of recurrent colorectal cancer was lost to follow-up after 2.5 years. One patient died without cancer 12 months after left hepatic lobectomy for colon cancer metastases. Cumulative survival for the entire series and for patients after resection of colorectal cancer metastases was the same: 1 year survival 90 percent; 2 year survival 75 percent, and 3 year survival 65 percent. Seventeen of 30 patients remain disease-free after resection of liver metastases. Of the 13 who had recurrence, 8 are still alive. Ten recurrences were outside of the residual liver (predominantly multiple pulmonary metastases). One recurrence was in the right hemidiaphragm, and only three were in the residual or regenerated liver. Serial carcinoembryonic antigen analysis was the best indicator of recurrence in these 13 patients, 12 of whom were asymptomatic. These data confirm that major liver resection can be performed with minimum postoperative mortality (4.7 percent in this series). More importantly, the majority of patients were cured of their liver metastases. The next goal should be the initiation of adjuvant systemic therapy trials after liver resection in such patients.     

Repeat hepatic cryotherapy for metastatic colorectal cancer

This study evaluated the risks and benefits of repeat hepatic cryotherapy for recurrent, unresectable hepatic metastases from colorectal carcinoma. Review of a prospective database identified 195 patients who underwent hepatic cryotherapy for metastatic colorectal carcinoma during a 7-year period. Of the 14 patients who underwent successful repeat cryotherapy for recurrences confined to the liver, 86% had Duke’s stage D colorectal carcinoma at initial diagnosis. The median age of the 14 patients was 58 years (range 41 to 77 years). The median number of hepatic metastases was three at the first cryotherapy and two at the second cryotherapy. At a median follow-up of 71 months, the mean survival times from original diagnosis, first cryotherapy, and second cryotherapy were 53,42, and 19 months, respectively. At the most recent follow-up, eight patients (57%) have died of their disease, four (29%) are alive with disease, and two (14%) have no evidence of disease. The mean interval between the first and second cryotherapies was 23 months. The complication rates after the first and second cryotherapies were 7% and 14%, respectively. One patient developed a wound dehiscence after the first cryotherapy. Following the second cryotherapy, one patient had a small bowel obstruction and another had a pleural effusion. There was no perioperative mortality. Repeat cryotherapy for recurrent, unresectable hepadc metastases from colorectal cancer is safe and improves survival. However, a prospective trial is needed to validate the efficacy of systemic therapy and to better define the indications for repeat hepatic cryotherapy. Supported in part by grant T32 CA 09689 from the National Cancer Institute and by funding from the Rogovin-Davidow Foundation, Los Angeles, Calif. Presented at the Forty-First Annual Meeting of The Society for Surgery of the Alimentary Tract, San Diego, Calif., May 2l–24, 2000.     

Repeat hepatic resections for metastatic colorectal cancer.

OBJECTIVE: The authors weighed the risks and benefits of repeat liver resections for colorectal metastatic disease. METHOD: In the 6-year period between January 1985 and June 1991, 499 patients underwent liver resections for colorectal metastases at the Memorial Sloan-Kettering Cancer Center. Of these, 25 patients had repeat surgical resections for isolated recurrent disease to the liver. The clinical data for these patients were reviewed. RESULTS: The median interval between the two resections was 11 months. There were no perioperative deaths, and the complication rate was 28%. Median follow-up after the second liver resection is 19 months, with median survival of 17 months for nonsurvivors. Although the median survival after the second resection is 30 months, 20 of the 25 patients have had recurrences with a median disease-free interval of only 9 months. No characteristic of primary or metastatic disease predicted outcome, including time between presentation of the primary and development of liver metastases, disease-free interval after the first liver resection, and bilobar liver involvement. CONCLUSIONS: Although repeat liver resections can be performed safely and improves survival, the likelihood of cure from such resection therapy is low. This likelihood of further recurrences encourage studies of adjuvant or alternative treatments of this population.     

Repeat hepatic resection as effective treatment for recurrent colorectal liver metastases

Background: Approximately 20–40% of patients who undergo liver resection for colorectal metastases develop recurrent disease confined to the liver. The goals of this study were to determine whether the survival benefit of repeat hepatic resection justified the potential morbidity and mortality. Methods: A retrospective review was performed on all patients who underwent liver resection for colorectal cancer metastases between 1983 and 1995 (N=202). Repeat liver resections were performed on 23 patients for recurrent metastases. Results: There were no operative deaths in the 23 patients, and the postoperative morbidity rate was 22%. The 5-year actuarial survival rate after repeat resection was 32%, with a median length of survival of 39.9 months. There were three patients who survived for >5 years after repeat resection. Sixteen patients (70%) developed recurrent disease at a median interval of 11 months after the second resection; 10 of these 16 patients (62%) had new hepatic metastases. No clinical or pathological factors were significant in predicting long-term survival. Conclusions: Repeat liver resection for recurrent colorectal metastases (a) can be performed safely with acceptable mortality and morbidity rates and (b) may result in long-term survival in some patients.Presented at the 49th Annual Cancer Symposium of The Society of Surgical Oncology, Atlanta, Georgia, March 21–24, 1996.     

Repeat hepatectomy for colorectal liver metastases

This study includes 16 patients (9 men, 7 women; mean age 64 years) who underwent a total of 19 repeat hepatectomies for metastasis after colon (n=7) or rectal (n=9) carcinoma. All patients were reoperated for recurrent liver metastasis after the first resection (mean, 21 months; range, 7-40 months), and three had a third hepatectomy 13, 24, or 65 months after the second. Perioperative mortality was 0% and morbidity was 37%. The 3- and 5-year survival rates after the second resection were 56.8% and 28.4%, respectively, with a median survival of 42.3 months. Seven patients died (mean survival, 25.7 months; range, 9-58 months) before the end of the study. Six patients were alive with one or more recurrences, and three (24, 51, and 173 months of follow-up) were alive without known recurrence. Survival rates for repeat resections of colorectal liver metastases in selected patients were comparable with those obtained after resection of a first liver metastasis.     

Hepatic resection for breast cancer metastases

INTRODUCTION: Hepatic resection is an established modality of treatment for colorectal cancer metastases. Resection of breast cancer liver metastases remains controversial, but has been shown to be an effective treatment in selected cases. This study reports the outcome of 8 patients with liver metastases from breast cancer. PATIENTS & METHODS: 8 patients with liver metastases from previously treated breast cancer were referred for hepatic resection between September 1996 and December 2002. Six were eligible for liver resection. The mean age was 45.8 years. The resections performed included 1 segmentectomy and 5 hemihepatectomies of which one was an extended hemihepatectomy. One patient had a repeat hepatectomy 44 months after the first resection. RESULTS: There were no postoperative deaths or major morbidity. The resectability rate was 75%. Follow-up periods range from 6 to 70 months with a median survival of 31 months following resection. There have been 2 deaths, one died of recurrence in the residual liver at 6 months and one died disease-free from a stroke. Of the remaining 4 patients, 1 has had a further liver resection at 44 months following which she is alive and ''disease-free'' at 70 months. The one patient with peritoneal recurrence is alive 49 months after her liver resection with 2 patients remaining disease-free. CONCLUSION: Hepatic resection for breast cancer liver metastases is a safe procedure with low morbidity and mortality.     

Repeated resection for malignant liver tumours.

Fifteen repeat hepatic resections were performed on 12 patients with either recurrent or residual malignant tumours of the liver. Of these, one patient underwent three repeat resections and another underwent two. Five had primary liver liver tumours and seven had liver metastases. Planned, ''staged'', repeat resections were performed on three patients because of multiple deposits of tumour, cirrhosis or extensive disease at initial presentation. There was no operative mortality. The period of follow-up from the time of repeat sections ranged between 4 months and 36 months during which two patients died from recurrent disease. The mean survival after the repeat resection was 16.8 months (range 4-36 months). Although technically demanding, repeat hepatectomy is feasible and provide similar benefits.     

Surgical therapy for recurrent liver metastases from colorectal cancer

Forty percent of patients whose disease recurs after hepatic resection for liver metastases from colorectal cancer initially will have liver-only metastases. We have retrospectively reviewed our experience with repeated surgical treatment for liver-only recurrence after previous hepatic resection for colorectal metastases. Repeated hepatic procedures were performed with no mortality in 10 patients. Intraoperative ultrasound allowed identification of three unsuspected metastases and determination of unresectability of two metastases during 11 procedures. Three patients were free of disease at 31, 41, and 48 months from the first hepatic procedure and at 15, 31, and 43 months from the second procedure. Five patients have remained free of hepatic disease. Patients whose initial metastases were less than 6 cm in diameter and had single liver recurrences after hepatic resection appeared to be the best candidates for further surgical therapy. These data and a review of the literature suggest that surgical treatment of recurrent liver metastases from colorectal cancer can be performed safely, and it is associated with long-term disease-free survival in up to 38% of highly selected patients.     

Hepatic resections for bilobar liver metastases from colorectal cancer

The present study was performed to assess survival benefits in patients who underwent a hepatic resection for isolated bilobar liver metastases from colorectal cancer. Thirty-eight patients underwent a curative hepatic resection for isolated colorectal liver metastasis. Among them, 11 patients had bilobar liver metastases and 19 had a solitary metastasis. The remaining 8 patients had unilobar multiple lesions. We investigated survival in two groups those with bilobar and those with solitary metastatic tumors. Survival and disease-free survival were 36% and 18% at 5 years, respectively, in the patients with bilobar liver metastases, while these survivals were 43% and 34% in the patients with solitary liver metastasis. In the 38 patients, repeated hepatic resections were performed in 15 patients with recurrent liver disease. The 5-year survival and disease-free survival rates for these patients were 38% and 27%, respectively, after the second hepatic resections. Of the 11 patients with bilobar liver metastases, 5 underwent a repeated hepatic resection, and they all survived for over 42 months. Based on our observations, a hepatic resection was thus found to be effective even in selected patients with either bilobar nodules or recurrence in the remnant liver. Received: February 7, 2000 / Accepted: April 26, 2000     

Surgical treatment for colorectal liver metastases involving the paracaval portion of the caudate lobe

BACKGROUND: Hepatic neoplasms in the paracaval portion of the caudate lobe (S1r) are usually difficult to treat surgically because such neoplasms often invade the hepatic veins and/or inferior vena cava (IVC). We reevaluated resected cases of colorectal liver metastases involving S1r to confirm the significance of aggressive surgical treatments. METHODS: Between July 1977 and December 2002, 95 consecutive patients with colorectal liver metastases underwent hepatic resection. Seven patients with liver metastases involving the S1r underwent resection. RESULTS: The surgical procedures for liver metastases comprised 3 isolated caudate lobectomies, 2 right hepatectomies, and 2 right hepatic trisectionectomies with caudate lobectomy. Combined resections included partial resection of the hepatic vein in 2 patients, wedge resection of the IVC in 3, and segmental resection of the IVC in 1. Six of the 7 patients with S1r metastasis had recurrent disease in liver and/or lung. A second hepatectomy was carried out in 4 patients and a partial lung resection in 2 patients. Four of the 7 patients survived more than 5 years, but 2 of them died of recurrent disease at 61 and 95 months after initial hepatectomy. The remaining 2 patients are alive 72 and 118 months without any sign of recurrence. The median survival time of the 7 patients was 60 months. CONCLUSION: Liver metastases involving the S1r could be resected radically with en bloc resection of the major hepatic veins and/or the inferior vena cava. An aggressive surgical approach with combined resection of the adjacent major vessels may offer a better chance of long-term survival in selected patients with caudate lobe metastasis from colorectal cancer.     

Surgery After Downstaging of Unresectable Hepatic Tumors With Intra-Arterial Chemotherapy

Background: This retrospective study was performed to assess the outcome among patients who underwent hepatic resection or tumor ablation after hepatic artery infusion (HAI) therapy downstaged previously unresectable hepatocellular carcinoma (HCC) or liver metastases from colorectal cancer (CRC).Methods: Between 1983 and 1998, 25 patients with HCC and 383 patients with hepatic CRC metastases were treated with HAI therapy for unresectable liver disease. We retrospectively reviewed the records of 26 (6%) of these patients who underwent subsequent surgical exploration for tumor resection or ablation.Results: At a median of 9 months (range 7–12 months) after HAI treatment, four patients (16%) with HCC underwent exploratory surgery; two underwent resection with negative margins, and the other two were given radiofrequency ablation (RFA) because of underlying cirrhosis. At a median postoperative follow-up of 16 months (range 6–48 months), all four patients were alive with no evidence of disease. At a median of 14.5 months (range 8–24 months) after HAI therapy, 22 patients with hepatic CRC metastases underwent exploratory surgery; 10 underwent resection, 6 underwent resection and RFA or cryotherapy, and 2 underwent RFA only. At a median follow-up of 17 months, 15 (83%) of the 18 patients with CRC who had received surgical treatment had developed recurrent disease; the other 3 died of other causes (1 of postoperative complications) within 7 months of the surgery. One patient in whom disease recurred underwent a second resection and was disease-free at 1 year follow-up.Conclusions: Hepatic resection or ablation after tumor downstaging with HAI therapy is a viable option for patients with unresectable HCC. However, given the high rate of recurrence of metastases from CRC, hepatic resection or ablation after downstaging with HAI should be used with caution.Presented at the 53rd Annual Meeting of the Society of Surgical Oncology, New Orleans, Louisiana, March 16–19, 2000.     

Prognostic factors and evaluation of surgical management of hepatic metastases from colorectal origin: A 10-year single-institute experience

The aim of this study was to determine prognostic factors and outcome after liver resection for colorectal metastases in 102 patients over a period of 10 years. A stepwise procedure using proportional hazard regression analysis was used to identify prognostic factors. Estimated survival at 2 years was 71%, and at 5 years, 29% (Kaplan-Meier). Of 19 patients with isolated liver recurrence, 6 had a second metastasectomy; 4 of the 6 are still alive.Wefound that the number of hepatic lesions on computed tomography (P = 0.012), the interval between resection of the primary colon tumor and the hepatic metastasectomy (P = 0.012), and synchronicity of the primary and the hepatic metastasis (P = 0.048) showed evidence of independent prognostic value regarding survival. Resection of hepatic colorectal metastases may result in long-term survival. Patients with recurrence after a first liver resection may benefit from a repeat metastasectomy. Our data suggest there is no strong predictor of survival. Survival seems to decrease with increasing number of metastases found on computed tomography.     

Hepatic resections: an eight year experience at a community hospital

Between April 1979 and March 1987 24 patients underwent 26 hepatic resections. Colorectal liver metastases constituted the largest group (n = 18), followed by hepatocellular carcinoma (n = 2), Echinococcal liver cyst (n = 1), cholangiocarcinoma (n = 1), and leiomyosarcoma (n = 1). The mean age was 41.8 +/- 14.6 years (range: 23-69 years). Fifteen women and nine men comprised the group. The operative morbidity was 21 per cent, the 30-day operative mortality was 8 per cent (two deaths). Both operative deaths occurred in patients with colorectal liver metastases. The 18 patients with colorectal liver metastases included ten women and eight men. The mean age was 59.1 +/- 6.5 years (range: 46-69 years). There were seven synchronous and 11 metachronous liver metastases. Carcinoembryonic antigen (CEA) was found elevated in 14 of the original primary colonic carcinomas, and in all but one patient with metachronous liver metastases. The mean time from colorectal carcinoma resection to occurrence of metachronous metastases was 17.1 +/- 5.8 months. To date, 10 patients have had recurrences of liver metastases after hepatic resection for colorectal liver metastases. The mean time of recurrence was 12.6 +/- 11.9 months. The size of the metastases was 3.8 +/- 3.2 cm (range: 0.2-17 cm). The mean number of lesions present was 1.5 +/- 1.0. The 1 year and 2 year actuarial survival rates were 87.5 and 43.8 per cent respectively. The longest survivor is alive 54 months after his hepatic resection for colorectal liver metastases and remains to this date disease free.(ABSTRACT TRUNCATED AT 250 WORDS)     

Second hepatic resection for recurrent hepatocellular carcinoma

During the last 5 years, radical hepatic resection was performed in 91 patients with hepatocellular carcinoma (HCC). Thirty-one of them had tumour recurrence in the remaining liver during the follow-up period. Second hepatic resection was carried out on nine of them 4-38 months after the first hepatectomy. The ages of these patients ranged from 39 to 65 years with an average of 53.7. There were six men and three women. Eight patients had underlying cirrhosis of the liver and one chronic active hepatitis. Six patients are alive, four being free of HCC and two with disease, for 15-45 months after the first operation. Two patients died of systemic cancer dissemination. The remaining patient had tumour recurrence in the liver again and died of hepatic failure after the third laparotomy. The survival rate of these nine patients was significantly better than that of twenty-two patients who were treated by other palliative methods. The present result shows that a second hepatic resection is a possible and meaningful method of treatment for the patients with recurrent HCCs in the liver remnant.     

Complete hepatic resection of metastases from leiohyosarcoma prolongs survival

Although liver resection has been shown to prolong survival in selected patients with metastases from colorectal cancer, the benefit for other metastatic tumors is unproved. To determine whether hepatic resection has a role in the management of metastatic leiomyosarcoma, medical records from 11 consecutive patients who underwent resection of isolated metastases from leiomyosarcoma between 1984 and 1995 were reviewed. All liver resections were for leiomyosarcomas originating in the viscera (n = 6) or retroperitoneum (n = 5). The average disease-free interval was 16 months. Five of 11 primary tumors were classified as low grade, whereas six were high grade. Hepatic resections included lobectomy or extended lobectomy (n = 4), segmentectomy and/or wedge resection (n = 5), and complex resection (n = 2). There were no operative deaths. Median survival of all patients after liver resection was 39 months. Patients who underwent complete resection of hepatic metastases (n = 6) had a significantly longer survival than those who had incomplete resections (n = 5) (P = 0.03, log-rank test). Furthermore, five of six patients who underwent complete resection are alive after hepatectomy with a median follow-up of 53 months. Therefore, in selected patients with isolated liver metastases from visceral and retroperitoneal leiomyosarcomas, complete resection of hepatic metastases results in prolonged survival. Presented in part at the Fiftieth Annual Cancer Symposium of the Society of Surgical Oncology, Chicago, Ill., March 20–23, 1997.     

CRYOTHERAPY OF THE RESECTION EDGE AFTER LIVER RESECTION FOR COLORECTAL CANCER METASTASES

Background : An involved or inadequate (< 1 cm) resection margin is associated with a high rate of local tumour recurrence and reduced survival rates after liver resection for colorectal metastases. This paper assesses whether or not hepatic cryotherapy of the resection edge is suitable to improve local disease control. Methods : From April 1990 to May 1997, we performed cryotherapy of the resection edge in 44 patients after liver resection for colorectal liver metastases with an involved or inadequate resection margin. The reasons for performing edge cryotherapy instead of extension of resection were: proximity of hepatic veins or portal sheath (n= 12); avoidance of extended left or right hemihepatectomy (n= 15); inadequate liver tissue reserve after resection (n= 16); and patient unfit to undergo further major resection (n= 1). Histological examination showed the resection margin to be involved in 24 patients and close (< 1 cm) in 20 patients. Results : Two patients died after surgery. Morbidity consisted of intra-abdominal collections (n= 6), postoperative bleeding (n= 1), wound infection (n= 1) and transient liver failure (n= 1). At a median follow-up of 19 months, 16 patients are alive and disease-free, 26 patients developed recurrence and 15 of them died. Nineteen patients developed recurrence which involved the liver but only five of these were at the resection edge. Median overall and liver disease-free survival was 33 and 23 months, respectively. Conclusions : Cryotherapy of the resection edge after resection of colorectal liver metastases with involved or inadequate resection margins considerably improves local disease control and may allow a greater proportion of patients with liver metastases to undergo potentially curative treatment.     

Second liver resections are safe and effective treatment for recurrent hepatic metastases from colorectal cancer: a bi-institutional analysis

OBJECTIVE: To determine the value of repeat liver resection for recurrent colorectal metastases to the liver. SUMMARY BACKGROUND DATA: Liver resection represents the best and a potentially curative treatment for metastatic colorectal cancer to the liver. After resection, however, most patients develop recurrent disease, often isolated to the liver. METHODS: This study reports the combined experience of repeat liver resection for recurrent liver metastases at an American and a European surgical oncology center. Patients were identified from prospective databases and records were retrospectively reviewed. A total of 126 patients (American n = 96, 1986-2001; European n = 30, 1985-1999) underwent repeat liver resection. Patient characteristics were similar in the two institutions. Median follow-up from first liver resection was 88 and 105 months, respectively. RESULTS: Operations performed included 90 minor resections and 36 resections of a lobe or more. The 1-, 3-, and 5-year survival rates were 86%, 51%, and 34%. There were 19 actual 5-year survivors to date. By multivariate regression analysis (proportional hazard model), more than one lesion and tumor size larger than 5 cm were independent prognostic indicators of reduced survival. The interval between the first and second liver resection was not predictive of outcome. CONCLUSIONS: Repeat liver resection for colorectal liver metastases is safe. Patients with a low tumor load are the best candidates for a repeat resection. In well-selected patients, further resection of the liver can provide prolonged survival after recurrence of colorectal liver metastases.     

Repeat liver resection for recurrent colorectal liver metastases

BACKGROUND: This study aimed to delineate the role of surgery for recurrent colorectal cancer in the liver and to identify prognosticators for better patient selection and outcome. METHODS: Data from 90 repeat hepatectomies (second = 75; third = 12; fourth = 3) for recurrent colorectal cancer were collected. RESULTS: After the second hepatectomy, the 3-and 5-year survival rates were 48% and 31%, respectively. Twenty-seven percent (20 of 75) of patients are alive without recurrence after a median follow-up of 27 months, and 9 survived more than 5 years. Four or more tumors, positive regional lymph node metastases, concomitant extrahepatic disease, and residual tumor were independent poor prognostic factors after the second hepatectomy. CONCLUSIONS: Repeat hepatectomy should be applied for recurrent colorectal cancer, when curative removal of the tumor is possible, although the benefit from treatment was limited in a patient with regional lymph node metastases, 4 or more metastases, or extrahepatic disease.     

Aggressive repeat liver resection for hepatic metastases of colorectal carcinoma

Although hepatectomy for liver metastases from colorectal carcinoma is an effective treatment, recurrence in the liver is still the most common site after hepatectomy. Thirty patients underwent hepatectomy for hepatic metastases and 17 of them had recurrence in the remnant liver during the following 12-year period. Six of the 17 patients underwent a removal of isolated hepatic recurrences. Two of the six patients underwent a third hepatectomy, and three patients underwent partial lung resection on a total of five occasions. There were no operative deaths while complications after a third hepatectomy contributed to a high morbidity rate of 40 per cent. The mean length of survival of the six patients was 28.5 months from the second hepatectomy. The prognosis of the six patients who underwent a repeat hepatectomy was significantly better than that of patients with unresectable recurrence after an initial hepatectomy (p<0.01). The overall 5-year survival of 29 patients excluding one inhospital death was 44.7 per cent. Our results reveal that aggressive removal of isolated and resectable recurrent disease has the potential to improve the prognosis of selected patients with metastatic cancer.