71
Curative Treatment of Peritoneal Carcinomatosis Arising from Colorectal Cancer by Complete Resection and Intraperitoneal Chemotherapy D. Elias, M.D. F. Blot, M.D. A. El Otmany, M.D. S. Antoun, M.D. P. Lasser, M.D. V. Boige, M.D. P. Rougier, M.D. M. Ducreux, M.D. Department of Oncological Surgery, Gustave Roussy Institute, Villejuif, France.
Supported in part by grants from the French Association for Research Against Cancer (Association Franc¸aise de Recherche contre le Cancer [ARC]). Address for reprints: D. Elias, M.D., De´partement de Chirurgie Ge´ne´rale Carcinologique, Rue Camille Desmoulins, 94805 Villejuif cedex, France; Fax: (11) 33 1 42 11 52 56; E-mail:
[email protected] Received November 6, 2000; revision received February 14, 2001; accepted March 20, 2001. © 2001 American Cancer Society
BACKGROUND. Peritoneal carcinomatosis (PC) is fatal, despite standard systemic chemotherapy. A new approach that combines maximal surgery with maximal regional chemotherapy has potential to cure selected patients who have colorectal PC. The authors have reported the oncologic results of this combined treatment. METHODS. The authors performed a retrospective study of 64 patients who had PC arising from colorectal adenocarcinomas, 19 (29.6%) of whom also had other metastases. These patients were treated by complete resection of all detectable tumors and by a 5-day course of early intraperitoneal chemotherapy (EPIC) with mitomycin C, then by 5-fluorouracil (n ⫽ 37), or by intraoperative intraperitoneal chemohyperthermia (IPCH) with mitomycin C, alone or combined with cisplatin (n ⫽ 27), in 2 separate trials. In the trial of IPCH, aimed at selecting the most reliable procedure in terms of spatial diffusion and thermal homogeneity, the 27 patients were treated with 7 different procedures. The extent of PC was assessed precisely by using a peritoneal index. The median follow-up period for the entire patient population was 51.7 months. RESULTS. The postoperative mortality and morbidity rates were 9.3% and 54.6%, respectively. Most severe complications occurred in patients who required extensive cytoreductive surgery. Global and disease-free survival rates were respectively 60.1% and 54.7% at 2 years and were 27.4% and 18.4% at 5 years. Results were significantly better (P ⫽ 0.04) when patients were metastasis-free (apart from PC) and when the peritoneal index was lower than 16 (P ⫽ 0.005). IPCH seemed to be more effective than EPIC for treatment of PC. CONCLUSION. This treatment plan, which combined maximal surgery with maximal regional chemotherapy, cured approximately 25% of patients. This strategy was mainly applicable to patients with limited intraperitoneal cancer volume and no extraperitoneal involvement. IPCH proved to be more effective than EPIC but more difficult to use correctly. Future results should improve through routine use of the optimal hyperthermia procedure, with improvements in the composition of instillate, better patient selection, and the reduction in the rate of complications that occurs with physician experience. Cancer 2001;92:71– 6. © 2001 American Cancer Society. KEYWORDS: peritoneal carcinomatosis, colorectal cancer, intraperitoneal chemotherapy, hyperthermia, cytoreductive surgery, peritonectomy.
P
eritoneal carcinomatosis (PC) arising from gastrointestinal cancer is invariably fatal. In a recent prospective study, the mean and median overall survival times were 6.9 and 5.2 months in patients with colorectal primary tumors.1 Information on the natural spread of gastrointestinal cancer has been published recently,2 and a potentially curative approach to PC has been developed.3 Intraperitoneal chemotherapy has the advantages of bathing the entire cavity and
72
CANCER July 1, 2001 / Volume 92 / Number 1
FIGURE 1. Peritoneal index defining precisely the extent of the peritoneal carcinomatosis.17 Each of the 13 regions are classified as follows: 0 if no tumor; 1 if ⱕ 5 mm tumor; 2 if 0.5–5.0 tumor; 3 if ⬎ 5 cm tumor or confluence. [Reprinted with permission from].
permitting very high local drug concentrations to be reached.3,4 However, it is ineffective when used alone, for the following reasons: 1) Cytotoxic agents only penetrate tumor nodules superficially, to a depth of less than 2 mm.5,6 2) Tumor cells colonize injured surfaces rapidly and massively,7 a phenomenon that cannot be prevented by irrigating the peritoneum.8 3) Postoperative adhesions occur very rapidly, considerably reducing the efficacy of delayed postsurgical intraperitoneal chemotherapy by creating a sanctuary where implanted tumor cells capitalize on growth factors generated during the healing process.10,11 In addition, they are inaccessible to systemic chemotherapy, because neovascularization has not yet occurred. Thus, intraperitoneal chemotherapy will be effective on PC after resection of tumor nodules exceeding 1 mm only if the procedure is performed very early after surgery, before adhesions emerge.3 Chemotherapy can take the form of early postoperative intraperitoneal chemotherapy (EPIC) or intraoperative chemohyperthermia (IPCH).4,5,12,13 However, even if the PC is cured, patients may develop extraperitoneal metastases. The aim of this study was to assess the oncologic results of this new approach based on a combination of maximal surgery with maximal regional chemotherapy in patients with colorectal PC.
PATIENTS AND METHODS From January 1993 to December 1999, sixty-four patients with colorectal PC were treated curatively in our center. There were 27 men and 37 women, with a mean age of 47.1 ⫾ 12.9 years (range, 16 –75 yrs). The PC arose from adenocarcinomas of the colon in 46 patients, the rectum in 9, and from mucinous adenocarcinomas of the appendix in 9 patients (Grade 3 pseudomyxomas with solid areas of cancer, necrosis, atypia, and high mitotic activity). Patients were en-
rolled in two prospective trials focusing on the treatment of PC arising from gastrointestinal carcinomas, malignant mesotheliomas, and sarcomas. The first trial was a Phase II study of EPIC (without hyperthermia), based on the initial schedule used by Sugarbaker9 and consisting of mitomycin C on postoperative Day 1, and 5-fluorouracil during the next 4 days.9 We have published the preliminary results of this study.14 The second trial was a Phase I–II study of IPCH aimed at establishing the most reliable technique to obtain complete spatial diffusion of the infusate and thermal homogeneity within the abdominal cavity; seven different procedures were tested before finding an acceptable one.15 The two trials were pooled for this analysis. Patients with liver or lymph nodes metastases in addition to PC were not treated with this protocol unless all distant sites of involvement were considered amenable to resection. Patients with computed tomography findings of lung metastases or invaded lymph nodes along the aorta or vena cava were not treated. Treatment combined maximal surgery and immediate maximal regional chemotherapy.
Surgery Complete resection (or electrocauterisation)16 of all macroscopically visible lesions of PC was required before intraperitoneal chemotherapy. If complete resection (residual tumor ⬍ 1 mm in diameter) was not possible or would have resulted in extensive bowel resection necessitating permanent total parenteral nutrition, the procedure was contraindicated and the laparotomy was closed. Of course, peritonectomy was performed only on sites with tumor deposits. The normal peritoneum was left in place. The extent of PC was scored using the peritoneal index described by Sugarbaker.17 Briefly, the abdominal cavity was divided into 13 regions (Fig.1), each assigned a score
Curative Treatment of Colorectal Peritoneal Carcinomatosis//Elias et al.
73
TABLE 1 Different IPCH Procedures Performed Successively during the Trial, Which Aimed to Select the Best Technical Procedurea No
No. of cases
Type of circuit
Drugs
Cause of stop
1 2b
5 7
Complete closure of the wound, 1 circuit ⫹ 4 drains ⫹ Y connections Peritoneal cavity expander, same as above
Mito: 10 mg/L Mito: 10 mg/L
3
4
Mito: 5 mg/L
4 5 6
5 4⫹4 4
Only skin closure, double circuit, 4 drains without Y connections, with maximum tolerable volume Same as above, but standard volume of 3.5 L/m2 Same as above Same as above ⫹ blue staining of the instillate
Thermal inhomogeneity Parietal wound recurrences No volumetric standardization
7b
6
c
Open wound, traction of skin upwards, double circuit, 2 L/m2 ⫹ blue staining of the instillate
Mito: 5 mg/L Mito: 10 mg/L then 8 mg/L Mito: 8 mg/L
No complete blue staining
Mito: 20 mg/m2 Cisplat: 200 mg/m2
a
Only 27 patients of this trial had colorectal cancers; 9 patients had another type of tumor. Procedures performed with an open cavity. c This procedure was selected because it allowed complete bating of the abdominal cavity and good thermal homogeneity. b
from 0 –3 according to the size and extent of tumor implants (total score 1–39). It is important to underline that very large and diffuse PCs were resected during the first 3 years of these trials (14 patients had a peritoneal index exceeding 25, with a maximum of 34).
EPIC and IPCH The chemotherapeutic regimens and the volume of instillate were based on body surface area, resulting in similar drug concentrations in all the patients. EPIC began as soon as the wound was closed. This therapy was based on mitomycin C (10 g/m2) on Day 1 and on 5-fluorouracyl (500 mg/m2) in 1 L/m2 of Ringer lactate solution from Day 2 to Day 6. Each infusion lasted 23 hours, and the instillate was removed with drains over the next hour. IPCH was performed intraoperatively and involved seven successive techniques. At least four patients were treated with each technique before changing to the next technique if a techique proved unsatisfactory. A continuous closed-circuit was used, with six thermal probes placed inside the peritoneal cavity. Blue-stained instillate was used in the two last groups of procedures to assess the distribution of the liquid.15 We successively used complete and permanent closure of the abdomen, an open technique with a peritoneal cavity expander, a closure of the skin alone (followed by reopening of the cavity, visceral suturing, and wound closure), and an open procedure, without an expander, the skin being pulled upwards with a specific retractor (Table 1).15 The first 21 patients were treated with mitomycin C alone (5, 8, or 10 mg/L, in 3.5 L/m2), and the other 6 patients received mitomycin C (20 mg/m2) plus cisplatin (200 mg/m2) in 2.5 L/m2 of perfusate. IPCH lasted 1 hour, at a peritoneal temperature of between 41 °C and 44 °C. Even if
the technique of IPCH changed during the study, with only 6 patients over 27 undergoing the best procedure, we decided to group the results of all patients who underwent IPCH. This forms the IPCH group.
Data Analysis Data were recorded prospectively. No patient was lost to follow-up. The cutoff date for this analysis was June 15, 2000. The median follow-up was 51.7 months (range, 8.1– 89.3 mos). A Kaplan–Meier survival curve was fitted to the data and a log-rank test was used to identify differences between curves. The difference was considered significant if P ⬍ 0.05.
RESULTS Surgical Procedures The mean number of tumoral regions was 7.9 ⫾ 4.4; the mean peritoneal index was 15.3 ⫾ 8.9; the mean number of resected organs was 4.7 ⫾ 2.4; the mean number of bowel anastomoses was 2.6 ⫾ 2.27; (median, 2; range, 0 –7); and the mean number of lateral bowel sutures was 2.1 ⫾1.5. There were 14 (21.8%) total colectomies with resection of the upper part of the rectum, 29 (45.3%) anastomoses on the lower part of the rectum after rectal resection with the Douglas pouch, 9 urinary tract sutures, and 15 (23.4%) iatrogenic pleural breaches. The mean operating time was 444 ⫾ 159 minutes (median, 440 min; range, 130 –780 min), and mean blood loss was 1260 ⫾1169 mL (median, 800 mL; range, 0 – 4500 mL). In 19 (29.6%) patients, other metastases (liver ⫽ 11, ovary ⫽ 9, spleen ⫽ 3) were resected at the same time, and positive lymph nodes were found on pathologic examination of resection specimens in 27 (42.1%) cases. Details of the procedures in the two groups (EPIC and IPCH) are
74
CANCER July 1, 2001 / Volume 92 / Number 1
TABLE 2 Intraoperative Characteristics of the 64 Patients Treated for Colorectal Peritoneal Carcinomatosis
Peritoneal index (mean ⫾ SD) Invaded regions (mean ⫾ SD) Resected organs (mean ⫾ SD) Bowel anastomosis (mean ⫾ SD) Low rectal anastomosis (%) Total colectomy (%) Diaphragmatic break (%) Visceral metastasis (%) Lymph node metastases (%) Duration of surgerya (mean ⫾ SD) Blood loss (mL) (mean ⫾ SD) a
EPIC (n ⴝ 37)
IPCH (n ⴝ 27)
P
13.5 ⫾ 8.3 6.8 ⫾ 4.2 4.3 ⫾ 2.6 2.9 ⫾ 2.4 17 (45.9) 6 (18.6) 7 (18.9) 12 (32.4) 20 (54) 417 ⫾ 189 1660 ⫾ 1152
18 ⫾ 5.6 9.3 ⫾ 4.1 5.4 ⫾ 1.8 2.4 ⫾ 1.9 12 (44.4) 8 (29.6) 8 (29.6) 7 (25.9) 7 (25.9) 541 ⫾ 130 1303 ⫾ 1234
NS NS NS NS NS NS NS NS 0.01 NS NS
SD: standard deviation; NS: nonsignificant.
FIGURE 2. Survival rates of patients treated with resection and EPIC or IPCH (n ⫽ 64).
TABLE 3 Postoperative Mortality and Morbidity of the 64 Patients Treated for Colorectal Peritoneal Carcinomatosis
Death (%) Abdominal complication (%) Fistula Abscess Reoperation Extraabdominal complication (%) Pulmonary Aplastic Miscellaneous Food intolerance (in days) mean (range) Hospital stay (in days) mean (range)
EPIC (n ⴝ 37)
IPCH (n ⴝ 27)
P
3 (8.1) 17 (45.9) 7 5 10 20 (54.1) 9 0 10
3 (11.1) 12 (44.4) 5 3 8 16 (59.2) 8 2 7
NS NS
10 (6–35)
12 (7–62)
NS
25 (12–99)
30 (14–87)
NS
NS
NS: nonsignificant.
reported in Table 2, together with the extent of resection.
Postoperative Course Six (9.3%) patients died during the postoperative period, 3 after EPIC and 3 after IPCH. These 6 patients had major intraoperative risk factors (diffuse PC with peritoneal index ⬎ 28 in 4 patients and associated right lobectomy of the liver in 1 of these 4), and 2 patients also had preoperative risk factors (obesity 134 kg, clotting factor deficit). Minor or major complications occurred in 42 (65.6%) patients (Table 3). There was no significant difference between EPIC and IPCH in regard to the rate of complications, which correlated with the peritoneal index (P ⫽ 0.01). Intraabdominal complications occurred in 29 (45.3%) patients, with 12 gastrointestinal fistulae, 8 deep abscesses, and reoperation in 18 (28.1%) patients. Oral
FIGURE 3.
Overall survival rates according to the presence or not of associated metastases (P ⫽ 0.04).
food intolerance lasted a median of 10 days after EPIC and 12 days after IPCH. Extraabdominal complications, mainly of a pulmonary nature, occurred in 36 (56.2%) patients. The median hospital stay was longer for IPCH (30 days) than for EPIC (25 days) patients.
Survival Overall and disease-free survival rates among the entire population (Fig. 2) were respectively 60.1% and 54.7% at 2 years, 47.1% and 39.4% at 3 years, 36% and 30.1% at 4 years, and 27.4% and 18.4% at 5 years (median survival time, 35.9 ⫾ 8.5 mos). The presence of associated metastases had a significantly (P ⫽ 0.04) negative influence on survival (Fig. 3), with 3-year overall and disease-free survival rates of 54.9% and 25.1%, respectively. The presence of involved lymph nodes in resected tissues did not affect survival (P ⫽ 0.14). The peritoneal index, based on an arbitrary
Curative Treatment of Colorectal Peritoneal Carcinomatosis//Elias et al.
75
DISCUSSION
FIGURE 4.
Overall survival rates according to the peritoneal index (P
⫽ 0.019).
FIGURE 5.
Overall survival rates according to the type of intraperitoneal chemotherapy (P ⫽ 0.13).
cutoff of 15 (yielding similar numbers of patients in the 2 groups), had a major impact on survival (P ⫽ 0.01) (Fig.4), with 3-year survival rates of 60.3% and 32.5%, respectively, among patients with values below and above this cutoff. Among the 14 patients with a peritoneal index exceeding 25, only 2 were alive and only 1 was disease-free after 49 months of follow-up. Survival rates were lower in the EPIC group than in the IPCH group (Fig. 5), but not significantly (P ⫽ 0.13).
Recurrences After a median follow-up of 51.6 months, cancer recurrence was detected in 42 (65.6%) patients. No peritoneal recurrence was detected in half of these 42 patients. Thus, only 21 (31.3%) patients had peritoneal recurrences.
These results confirm the efficacy of a curative approach to peritoneal carcinomatosis based on maximal surgery plus immediate maximal regional chemotherapy, according to the therapeutic principle described by Sugarbaker et al.3 Several recent trials of intraperitoneal chemotherapy without preliminary maximal cytoreductive surgery have shown a 5-year survival rate of 0%.18,19 The study by Eilber et al.,20 involving resection of recurrent abdominal sarcomas and delayed intraperitoneal mitoxantrone instillation, showed that postoperative adhesions prevented complete bathing of the peritoneal cavity and that subsequent recurrent peritoneal disease was located in drug “no go” areas rather than in blue-stained areas of loculated mitoxantrone. Likewise, in our study of IPCH, blue staining of the instillate showed that abdominal closure prevented complete bathing of the cavity15 as areas behind the liver and stomach and at the root of the mesenterium frequently were not stained. This is probably one reason for the superior efficacy of IPCH over EPIC. The only survival rates so far published are those of Sugarbaker et al.18 concerning colorectal patients with colorectal PC treated by resection and EPIC. In this latter series of 64 patients, the 5-year survival rate was close to 20% in the overall population, compared with 0% in the 28 patients with incomplete cytoreduction and 37% in the 36 patients with complete cytoreduction. Our 5-year survival rate was 27.4% among the 64 patients with complete cytoreductive surgery, despite the higher frequency of other metastases, which had a significantly deleterious impact on survival (Fig. 3). In our previous study of 22 patients with moderate PC and liver metastases, only the 3 patients with a single liver metastasis were free of disease at 24, 27, and 46 months after this new approach.21 Like Sugarbaker’s research group, we found no prognostic impact from invaded lymph nodes. In contrast, the extent of PC before resection had a major prognostic influence (Fig.4). Early in this study, we treated cases of massive diffuse PC, but we subsequently excluded such patients because of their high complication rate and poor survival. The peritoneal index is very useful for evaluating the precise extent of PC and for comparing the results of different series. Currently, we are unable to define a cutoff value above which this combined approach is not indicated for colorectal cancer, but it appears to be between 20 and 25. In our study, the oncologic results of IPCH were apparently better than those of EPIC (Fig.5), although the difference was not significant because of an inadequate number of patients and because only 6 patients in the EPIC group were treated with the optimal
76
CANCER July 1, 2001 / Volume 92 / Number 1
hyperthermia procedure. Although this was not a randomized comparison of the two procedures, it is noteworthy that the 2 groups of patients were comparable (Table 1). The superiority of IPCH seems to be due to the complete bathing of all at-risk surfaces of the abdominal cavity in the intraoperative open-cavity technique,15 and the drug potentiation by hyperthermia.4,5,12,13 The main advantage of EPIC is its simplicity. In our opinion, EPIC is preferable to no treatment and to poorly performed IPCH. The postoperative mortality and morbidity rates were high in this study, and mainly they were attributed to patients who underwent extensive cytoreductive surgery for massive diffuse PC. However, we recently performed a pharmacokinetic study of IPCH with oxaliplatin in 22 patients and observed no postoperative deaths and only two gastrointestinal fistulae. The complication rate fell with experience. We think that a five-day course of EPIC, with gastrointestinal sutures floating in the chemotherapy solution, is more dangerous than IPCH alone. Future results with IPCH should improve through routine use of the optimum hyperthermia procedure identified here, through improvements in the composition of the instillate, and through better patient selection. This new combined curative approach to PC is indicated only for selected patients with moderate or minimal PC and with no distant metastases. We currently are able to cure more than 25% of such patients with PC arising from colorectal cancer, which is incurable by standard systemic chemotherapy. However, in the last 2 years, rare patients who have undergone resection of the primary and 1–5 small deposits of PC followed by the most recent systemic chemotherapy (including oxaliplatin or irinotecan) have not had a recurrence in the peritoneum (negative second-look laparotomy performed 8 –10 mos latter). These findings provide justification for a randomized trial that compares treatment by complete resection of PC followed by IPCH plus systemic chemotherapy with treatment by complete resection plus systemic chemotherapy alone.
4.
5. 6.
7.
8.
9.
10.
11.
12. 13.
14.
15.
16. 17.
18.
19.
REFERENCES 1.
2.
3.
Sadeghi B, Arvieux C, Glehen O, Beaujard AC, Rivoire M, Baulieux J, et al. Peritoneal carcinomatosis from non-gynecologic malignancies: resuts of the EVOCAPE 1 multicentric prospective study. Cancer 2000;88:358 – 63. Sugarbaker PH. Observations concerning cancer spread within the peritoneal cavity and concepts supporting an ordered pathophysiology. Cancer Treat Res 1996;82:79 –100. Sugarbaker PH, Cuniffe W, Belliveau JF, de Bruin E, Graves T. Rationale for perioperative intraperitoneal chemotherapy as a surgical adjuvant for gastrointestinal malignancy. Reg Cancer Treat 1988;1:66 –79.
20.
21.
Elias D, Detroz B, Debaene B, Damia E, Leclercq B, Rougier P, et al. Treatment of peritoneal carcinomatosis by intraperitoneal chemo-hyperthermia: reliable and unreliable concepts. Hepatogastroenterology 1994;41:207–13. Los G, McVie JG. Experimental and clinical status of intraperitoneal chemotherapy. Eur J Cancer 1990;26:755– 62. Didkhoff T, van der Heider J, Dubbelman R, ten Bokkel Huinink WW. Tissue concentration of platinum after intraperitoneal cisplatin administration in patients. Proc AACR 1985;26:162. Zoetmulder FA. Cancer cell seeding during abdominal surgery: experimental studies. In: Sugarbaker PH, editor. Peritoneal carcinomatosis: principles of management. Boston: Kluwer, 1996:155– 62. Sweitzer KL, Nathanson SD, Nelson LT, Zachary C. Irrigation does not dislodge or destroy tumor cells adherent to the tumor bed. J Surg Oncol 1993;53:184 –90. Sugarbaker PH. Intraperitoneal chemotherapy for treatment and prevention of peritoneal carcinomatosis and sarcomatosis. Dis Colon Rectum 1994;37:(2 Suppl):S115–22. Jacquet P, Elias D, Sugarbaker PH. [Tumor implantation in cicatrization sites following surgery for digestive cancers]. J Chir (Paris) 1996;133:175– 82. Murakami A, Koga S, Maeta M. Thermochemosensitivity: augmentation by hyperthermia of cytotoxicity of anticancer drugs against human colorectal cancer, measured by the human tumor clonogenic assay. Oncology 1988;45:236 – 41. Hahn GM. Potential for therapy of drugs and hyperthermia. Cancer Res 1979;39(6 Pt 2):2264 – 8. Shimizu T, Maeta M, Koga S. Influence of local hyperthermia on the healing of small intestinal anastomosis in the rat. Br J Surg 1991;78:57–9. Elias D, Dube´ P, Blot F, Bonvallot S, Eggenspieler P, Sabourin JC, et al. Peritoneal carcinomatosis treatment with curative intent: the Institut Gustave-Roussy experience. Eur J Surg Oncol 1997;23:317–21. Elias D, Antoun S, Goharin A, El Otmany A, Puizillout JM, Lasser P. Research on the best chemohyperthermia technique for treatment of peritoneal carcinomatosis after complete resection. Int J Surg Invest 2000;1:431–9. Sugarbaker PH. Dissection by electrocautery with a ball tip. J Surg Oncol 1994;56:246 – 8. Sugarbaker PH. Intraperitoneal chemotherapy and cytoreductive surgery for the prevention and treatment of peritoneal carcinomatosis and sarcomatosis. Semin Surg Oncol 1998;14:254 – 61. Sugarbaker PH, Schellinx ME, Chang D, Koslowe P, von Meyerfeldt M. Peritoneal carcinomatosis from adenocarcinoma of the colon. World J Surg 1996; 20:585–91; discussion 592. Beaujard AC, Glehen O, Caillot JL, Francois Y, Bienvenu J, Panteix G, et al. Intraperitoneal chemohyperthermia with mitomycin C for digestive tract cancer patients with peritoneal carcinomatosis. Cancer 2000;88:2512–9. Eilber FC, Rosen G, Forscher C, Nelson SD, Dorey FJ, Eilber FR. Surgical resection and intraperitoneal chemotherapy for recurrent abdominal sarcomas. Ann Surg Oncol 1999;6:645– 50. Elias D, Goharin A, El Otmany A, Bonvallot S, Meshaka P, Attalah D, et al. Treatment of liver metastases associated with moderate peritoneal carcinomatosis by hepatectomy and cytoreductive surgery followed by immediate intraperitoneal chemotherapy: results in 22 cases. Int J Surg Invest 2001 (in publication).