Phase II study of weekly oxaliplatin and high-dose infusional 5-fluorouracil plus leucovorin in pretreated patients with metastatic colorectal cancer.

BACKGROUND: Chemotherapy with oxaliplatin, fluorouracil (5-FU) and leucovorin (LV) has proven efficacy in patients with advanced colorectal carcinoma (CRC), although the optimal dosage and administration schedule are still unclear. This phase II trial investigated the tolerability and activity of weekly oxaliplatin, high-dose infusional 5-FU and LV in pretreated patients with metastatic CRC. MATERIALS AND METHODS: Patients received weekly courses of i.v. oxaliplatin 50 mg/m2 (1-h infusion), LV 100 mg/m2 (1-h infusion) and 5-FU 2100 mg/m2 (24-h infusion) until disease progression or unacceptable toxicity. NCI-CTC criteria were used for assessment of side-effects (at each cycle) and WHO criteria for assessment of tumour response (every 8 cycles). For descriptive purposes, time to progression, overall survival and duration of objective response were also calculated. RESULTS: Forty-four patients were enrolled and received a total of 606 cycles (median 13/patient, range 4-33), and 70% of courses (421/606) were delivered at 100% of the planned dose. The most frequent side-effects were gastrointestinal and neurological and incidence rates were: diarrhoea 66% (grade III: 29%), nausea/vomiting 54%, neurotoxicity 34% (grade III: 2%), fatigue 27%, mucositis 22%, leucopenia 14%. No grade IV toxicity was observed. Objective response rates were: partial response 23% (10 patients), stable disease 59% (26) and progressive disease 11% (5). Median time to progression was 7 months, overall survival 13 months and the duration of partial response and stable disease were 9 and 6 months, respectively. CONCLUSION: The study demonstrated that this regimen has a favourable tolerability profile and is an active combination in the pretreated metastatic CRC patient, deserving further evaluation in phase III trials.

Sentinel lymph node mapping in early-stage breast cancer: technical issues and results with vital blue dye mapping and radioguided surgery.

BACKGROUND AND OBJECTIVES: Axillary lymph node status is the most important prognostic factor in patients with operable breast cancer. Recent studies have demonstrated the possibility of identifying the sentinel lymph node (sN) as a reliable predictor of axillary lymph node status in both cutaneous melanoma and breast cancer. Sentinel lymph node identification proved feasible by either peritumoral dye injection (Patent Blue-V) or radiodetection, with identification rates of 65-97% and 92-98%, respectively. However, some important issues need further definition, namely (a) optimization of the technique for intraoperative detection of the sN, (b) predictive value of the sN with regard to axillary lymph node status, and (c) reliability of intraoperative histology of the sN. We reviewed our experience in sN detection in patients with stage I-II breast cancer to assess the feasibility and accuracy of lymphatic mapping, by vital blue dye or radioguided surgery, and sN histology as a predictor of axillary lymph node status. METHODS: Two groups of patients (55 and 48) were recruited between May 1996 and May 1997 and between October 1997 and February 1998; the patients of the first series underwent vital blue dye lymphatic mapping only, whereas those of the second series had a combined approach with both vital blue dye mapping and radioguided detection of the sN. RESULTS: In the first set of patients, the sN was identified in 36/55 patients (65.4%); sN histology predicted axillary lymph node status with a 77% sensitivity (10/13), a 100% specificity (23/23), an 88.5% negative predictive value (23/26), and an overall 91.5% accuracy (33/36). The sN was the quasi-elective site of lymph node metastases because in clinically N0 patients nodal involvement was 20-fold more likely at histology in sN than in non-sN (30% and 1.5%, respectively). In the second set of patients, 49 lymphadenectomies were performed because 1 patient had bilateral breast cancer; the sN was identified in 45/49 lymphadenectomies (92%). The sN was intraoperatively negative at frozen-section examination in 33 cases, and final histology confirmed the absence of metastases in 31/33 cases (94%), whereas in 2 cases (6%) micrometastases only were detected. Final histology of the sN predicted axillary lymph node status with an 87.5% sensitivity (14/16), a 100% specificity (29/29), a 93.5% negative predictive value (29/31), and an overall 95.5% accuracy (43/45). CONCLUSIONS: Sentinel lymphadenectomy can be better accomplished when both mapping techniques (vital blue dye and radioguided surgery) are used. In this group of patients, agreement of intraoperative histology of the sN with the final diagnosis was 94%, and sN histology accurately predicted axillary lymph node status in 43/45 lymphadenectomy specimens (95.5%) in which an sN was identified.

Selective lymph node dissection in patients with intermediate thickness melanoma: our experience.

The role of elective lymph node dissection (ELND) for the treatment of cutaneous melanoma has been debated for many years. Actually, the value of ELND is seriously questioned because an increasing of overall survival rates has not been demonstrated. The lymphatic mapping technique, initially performed by an intradermic injection of vital blue dye, subsequently improved by the use of radioguided surgery (RGS), proved effective for the detection of clinical occult lymph node metastasis. We performed a sentinel node biopsy on 71 patients with stage pT2/T3N0M0 melanoma. Vital blue dye mapping alone was performed on 39 patients; the remaining 32 patients had a combined lymphatic mapping with both blue dye and RGS. The sentinel node (SN) was complexively identified in 69/71 (97.2%) patients. Sixteen patients (23.2%) were found to have metastatic melanoma cells in their SN(s); all these patients underwent lymph node dissection of the affected basin. Our experience confirmed that the intraoperative detection of sentinel nodes using both blue dye and radio-guided surgery is an effective and reliable technique for selecting patients to be submitted to lymph node dissection.

Sentinel lymph node mapping and biopsy for breast cancer: a review of the literature relative to 4791 procedures.

The status of axillary nodes is the most important prognostic factor in breast cancer to select patient subgroups for adjuvant chemotherapy; the current standard of care for surgical management of invasive breast cancer is complete removal of the tumor by either mastectomy or lumpectomy followed by axillary lymph node dissection (ALND). The recent introduction of intraoperative lymphatic mapping and sentinel lymph node biopsy (SLND) represents a major new opportunity for appropriate and less invasive surgical management of many tumors. There is an almost uniformly enthusiasm concerning the potential of this technique in breast carcinoma management, shown by published data. A peculiar attention to the so-called "sentinel node debate" in breast cancer surgery is a constant in the last years issues of the major medical journals. Even patients have become more aware about medical enthusiasm and their request of concise information on the topic and the possibilities of this approach is an increasing reality in medical practice. The aim of this paper is to review recent literature to offer an overview about the main controversial methodological aspects and a wide analysis of reported results. The most significative international literature papers from Medline were retrieved from 1993 to September 1999, and 4782 procedures were analysed. This extensive review of the literature has confirmed accuracy, feasibility and reliability of the SN detecting technique in axillary mapping. Provided a good proficiency in SN localisation and pathological evaluation, human resources and efforts should be mainly focused on its clinical validation as an alternative to ALND instead of on further phase I-lI clinical studies.

Radioimmunoguided surgery with different iodine-125 radiolabeled monoclonal antibodies in recurrent colorectal cancer.

Sixty-four patients with recurrent or metastatic colorectal cancer underwent radioimmunoguided surgery (RIGS). Thirty patients (Group A) were preoperatively injected with radiolabeled monoclonal antibody (MAb) B72.3, a whole IgG1 that reacts with tumor-associated glycoprotein (TAG-72) antigen. Thirty-four patients (Group B) were given monoclonal antibody FO23C5, an F(ab')2 which reacts with the carcinoembryonic antigen (CEA). The use of F(ab')2 antibodies ensured a lower time interval from the preoperative injection of the radiolabeled MAb to surgery. This interval was 22.7 days for Group A patients and 10.9 days for Group B patients. The correct RIGS identification of tumor sites occurred in 80.4% of Group A patients and in 92.6% of Group B patients. Additional information capable of modifying surgical strategy was obtained in 23.3% of Group A patients and in 8.8% of Group B patients. This difference was due to the different patterns of biodistribution and pharmacokinetics of the two MAbs. Although FO23C5 yields an improved diagnostic resolution for macroscopic tumor sites, we believe that B72.3 or other whole IgG1 should be the first choice for RIGS in recurrent or metastatic colorectal cancer patients.

Use of radioimmunoguided surgery after induction chemotherapy in locally advanced breast cancer.

Twenty-one patients with histologically proven locally advanced breast cancer (LABC) were treated with a combined modality approach based on primary chemotherapy and radical modified mastectomy followed by adjuvant chemotherapy. Surgery was performed by using radioimmunoguided surgery (RIGS) technique with the preoperative injection of Iodine-125 labeled monoclonal antibodies (MoAbs) B72.3 anti-TAG (11 patients, Group A) and FO23C5 anti-carcinoembryonic antigen (CEA; 10 patients, Group B). The role of RIGS was defined at surgery by using an intraoperative hand-held gamma-detecting probe (GDP) to locate the primary tumor, possible clinically occult multicentric foci and ipsilateral lymph node metastases. In Group A, RIGS correctly defined the primary tumor in seven out of 11 patients (63.3%) and was able to find multicentric tumors in two out of four patients (50%). Positive lymph nodes were identified by RIGS in three out of eight patients (37.5%). In Group B, patients RIGS correctly located the primary in 4/10 cases (40%); in two RIGS-positive cases, the tumor was clinically not evident after primary chemotherapy (yT0). RIGS correctly identified multicentric foci of tumor in one out of two cases (50%). Correct lymph nodal RIGS assessment was observed in three out of nine patients (33.3%). No RIGS false-positive findings occurred in the 21 patients included in the study. RIGS appears to be a reliable technique for the intraoperative diagnosis and staging of breast cancer with a potential role especially when conservative surgery is planned after primary chemotherapy in LABC.

Role of tumor-associated antigen expression in radioimmunoguided surgery for colorectal and breast cancer.

One hundred thirty-six patients with colorectal and breast cancer were enrolled in a retrospective study using radioimmunoguided surgery (RIGS) with Iodine-125 (I125) radiolabeled B72.3 (Group A, 73 patients) and F023C5 (Group B, 63 patients) monoclonal antibodies (MAbs). The correlation between intraoperative tumor-to-normal tissue (T/NT) gamma-detecting probe (GDP) counts ratio and the expression of tumor-associated glycoprotein (TAG)-72 (GroupA patients) and carcinoembryonic antigen (CEA; Group B patients) tumor-associated antigens (TAA) expression of 209 resected or biopsy tumor specimens was assessed. Ex vivo radioimmunolocalization index (R.I.) was carried out on the same specimens as a control of intraoperative GDP ratio values. RIGS positive definition of tumor occurred in 80/113 (70.8%) tumor sites of Group A patients and in 84/96 (87.5%) tumor sites of Group B patients. Mean percent B72.3 TAA expression of 113 tumor sites of Group A patients was 62.74 +/- 28.79% vs. 73.00 +/- 26.28% of 96 tumor sites of Group B patients (P < 0.05). The higher incidence of positive RIGS results was observed in tumor sites with the higher expression of the relative TAA. A statistically significant correlation between RIGS ratios and B72.3 and CEA expression was observed in the 113 tumor sites of Group A (P < 0.05) and in the 96 tumor sites of Group B (P < 0.01), respectively. The role of a preoperative evaluation of TAA expression in patients undergoing RIGS is discussed. Its assessment, whenever possible, may help to select those patients who will benefit more from this immunodiagnostic technique.

Radioimmunoguided surgery benefits in carcinoembryonic antigen-directed second-look surgery in the asymptomatic patient after curative resection of colorectal cancer.

Radioimmunoguided surgery (RIGS) with radiolabeled monoclonal antibodies (MoAbs) has been reported as useful in second-look colorectal cancer procedures to improve surgical decision-making by helping avoid needless extensive surgery and expanding curative resection to sites of recurrence that have been missed previously. Sixteen asymptomatic patients with an history of colorectal cancer surgery underwent second-look surgery using the RIGS system, solely on the basis of rising serum levels of carcinoembryonic antigen (CEA). All patients were injected preoperatively with the anti-tumor-associated glycoprotein (TAG) 125I-labeled MoAb B72.3. Both traditional and RIGS exploration were used to determine the extension of a possible recurrence and its resectability for cure. Recurrent disease was observed in 14 of the 16 patients as the result of this combined exploration. Exploration alone showed the presence of recurrent disease in 9 of 16 patients (56.2%). Thus, RIGS found overlooked tumor in five patients (31.2%). All the additional RIGS-detected tumor sites were locoregional recurrences resectable for cure; conversely, no diagnostic improvements were shown in patients with liver metastases. Resection for cure was obtained by this approach in 9 of 16 patients (56.2%). Two patients without disease at the exploratory laparotomy recurred within 2 months at sites away from the abdomen. RIGS improved the results of colorectal cancer CEA-guided second-look procedures in asymptomatic patients by recruiting one-third of patients to curative resections.

Activity of high dose 24 hour 5-fluorouracil infusion plus L-leucovorin in advanced colorectal cancer.

BACKGROUND: Modulation of 5-fluorouracil (5-FU) by leucovorin (L-LV) in patients (pts) with advanced colorectal cancer has been demonstrated to produce a highly significant benefit over single-agent 5-FU in terms of tumor response rate, but this advantage does not translate into an evident improvement of overall survival. To improve the clinical efficacy of the 5-FU plus L-LV regimen a phase II study of weekly 24-hour high-dose 5-FU infusion with L-LV was undertaken. PATIENTS AND METHODS: Seventy advanced colorectal patients were enrolled and treated by a weekly outpatient combination regimen according to the following schedule: L-LV 100 mg/sqm by 4 hours i.v. infusion followed by 5-FU 2600 mg/sqm over a 24 hours infusion combined with a fixed dose of oral L-LV (50 mg) every 4 hours for 5 times. Forty-four pts did not receive any previous CT and 26 pts were pretreated with fluoropyrimidines. RESULTS: The overall objective response rate (OR) was 35.3%; 7 CR and 11 PR (42.8% OR) were observed in the group of untreated pts, and 6 PR (23% OR) were reported among previously treated pts. Major side effects were represented by diarrhoea (grade III: 26%, grade IV: 1%), hand-foot syndrome (grade III: 4%, grade IV: 1%) and mucositis (grade III: 4%); however, this did not significantly influence the therapeutic programme. Median 5-FU dose intensity was 100% and 80% at 4 weeks, 87% and 75% at 8 weeks in untreated and pretreated pts, respectively. CONCLUSIONS: L-Leucovorin modulation of weekly short-term continuous infusion of high-dose 5-fluorouracil appeared a well-tolerated outpatient regimen; it demonstrated a high activity in advanced colorectal cancer, both in untreated pts and in pts resistant to 5-FU-based chemotherapy.

Pseudomixoma peritonei: a case report.

Pseudomixoma peritonei is a rare neoplasm characterized by mucinous ascites and the mucinous involvement of peritoneal surfaces, omentum and bowel loops. Usually pseudomixoma peritonei is associated with benign or malignant mucinous tumor of the appendix or ovary. The diagnosis of pseudomixoma peritonei is difficult because laboratory and radiology results are frequently nondiagnostic. We report a case that was initially mistaken for carcinomatosis of unknown origin and that underwent cytoreductive procedure and omentectomy as the treatment of choice.

[Long-term central venous access in oncology. Review of a caseload of 750 cases]. / Accessi venosi centrali a lungo termine in oncologia. Revisione di una casistica relativa a 750 casi.

Totally implantable central venous accesses systems are presently widely used in cancer patients. Perioperative and long-term morbidity of 750 consecutive implants performed from 1985 to 1994 were retrospectively reviewed. Our Series consisted in 616 (82.1%). Ports and 134 (17.9%) external tunnelled catheters with subcutaneous Dacron cuff (Hickmann or similar catheters) implanted by percutaneous access route in local anesthesia. Subclavian vein was the elective access route in 700 (93.3%) patients while in the remaining 50 (6.7%) the access was performed in the femoral vein, due to peculiar clinical conditions. One-day surgery was only required for 19 (2.6%) patients while outpatient surgery was the routine in 731 (97.4%) patients. Perioperative morbidity was at all similar both for Port and external catheters Series (p > 0.07). Late morbidity requiring the removal of the device occurred in 27/134 patients (18.7%) and in 40/616 (6.5%) for external catheters and Port Series, respectively (p < 0.002). The rate of infection, 2.3% for ports (p < 0.002). Patient's compliance was higher for Ports compared to external catheters. Apart from peculiar clinical conditions such as hematologic cancers, bone marrow transplantation or short life expectancy, Ports seem to guarantee for a lower morbidity coupled with better long-term results.

Radiation-associated angiosarcoma: diagnostic and therapeutic implications--two case reports and a review of the literature.

BACKGROUND: Angiosarcoma (AS) accounts for 1 to 2% of all soft tissue sarcoma. Both primary and secondary AS may occur, the latter being reported in the upper extremity with lymphedema after extended radical mastectomy for breast cancer (postmastectomy AS) or following radiotherapy of the breast, the thoracic wall, or other sites (radiation-associated AS). The authors report two cases of cutaneous radiation-associated AS and review literature regarding treatment planning and follow-up data to define the most appropriate therapy for cutaneous and noncutaneous radiation-associated AS. METHODS: The clinical records of two patients with radiation-associated AS were analyzed and previously reported cases were reviewed. RESULTS: Case 1: a female age 67 years developed cutaneous AS in the residual breast 27 months after breast-conserving therapy and conventional external beam radiotherapy (EBR). She underwent chemotherapy followed by simple mastectomy and chemotherapy with the same regimen but developed early recurrence that was treated with hyperthermia and EBR, wide excision, and second-line chemotherapy. She died 30 months after primary diagnosis of AS with multiple metastases. Case 2: a male age 59 years developed cutaneous AS in the left groin, 10 years after conservative surgery and EBR for a penile carcinoma. Early recurrence following wide excision was treated with chemotherapy, re-excision, and immunochemotherapy but the patient died 24 months after the primary diagnosis of cutaneous AS with local progression and distant metastases. CONCLUSIONS: The prognosis of radiation-associated AS is dismal, due mostly to its poor differentiation and frequent diagnostic delay. Simple mastectomy is advised for patients with cutaneous AS after breast-conserving surgery with wide tumor-free margins. If primary surgery fails, survival is seriously compromised because adjuvant or palliative treatments are not effective.

Radioimmunoguided surgery after primary treatment of locally advanced breast cancer.

PURPOSE: To assess the role of radioimmunoguided surgery (RIGS) using a handheld intraoperative gamma-detecting probe (GDP) to identify neoplastic disease after primary chemotherapy in locally advanced breast cancer (LABC) patients injected with iodine 125-labeled monoclonal antibodies (MAbs). PATIENTS AND METHODS: Twenty-one patients with histologically documented LABC were treated with a combined modality approach. After three courses of primary chemotherapy and before modified radical mastectomy, the 125I-radiolabeled MAbs B72.3 (anti-TAG72) and FO23C5 (anti-carcinoembryonic antigen [CEA]) were administered to 11 patients (group A) and 10 patients (group B), respectively. At surgery, a GDP was used to locate the primary tumor and to assess possible tumor multicentricity and the presence of ipsilateral axillary metastases. Routine pathologic examination was performed in neoplastic and normal tissue specimens of all 21 patients. In addition, immunohistochemical assay for TAG72 and CEA expression was performed. RESULTS: In group A patients, RIGS identified primary tumor in seven of 11 patients (63.3%) and unpalpable multicentric tumor lesions were located in two of four (50%). Positive axillary lymph nodes were histologically documented in eight of 11 patients (72.7%) and RIGS identified three of eight (37.5%). In group B, RIGS located the primary tumor lesion in four of 10 patients (40%); in two cases, the tumor was not clinically evident. Multicentricity was observed in one of two patients and lymph node involvement in three of nine (33.3%). No false-positive results were observed in either group A or B. CONCLUSION: RIGS appears to be a safe and reliable technique. However, the MAbs used in this study are not sufficiently specific. RIGS represents a technique for which the full potential for intraoperative assessment of breast cancer lesions can be reached when more specific antibodies become readily available.

Long-term femoral vein central venous access in cancer patients.

Subclavian percutaneous access with reservoir placement has been shown to be difficult or contraindicated in some patients. Of 465 cancer patients who required a port placement between January 1992 to January 1995, 41 (8.8%) had alternative percutaneous femoral access with a totally implantable port reservoir located in the abdomen because of the inaccessibility to subclavian or jugular veins and/or the presence of massive cutaneous metastases or severe radiodermitis in the upper part of the torso. Overall implant days was 9880, with an average of 241 days (range: 65-445). Ports were alternatively used for chemotherapy and nutritional purposes in 11 of 41 patients. Late morbidity causing the removal of the implanted ports was observed in two of 41 (4.9%) and 25 of 424 (5.9%) patients in the femoral and subclavian series, respectively (P = 0.86). The femoral percutaneous access for totally implantable port devices appears to be a safe alternative for cancer patients when subclavian and/or jugular vein catheterization and reservoir in the upper part of the torso is contraindicated.

Role of B72.3 iodine 125-labeled monoclonal antibody in colorectal cancer detection by radioimmunoguided surgery.

BACKGROUND: Radioimmunoguided surgery (RIGS) by means of radiolabeled monoclonal antibodies and a probe has been reported to be useful in recognizing subclinical tumor deposits during operation. Aim of this study was to understand the limits of this technique and to assess the potential diagnostic use of RIGS in colorectal cancer surgery. METHODS: Monoclonal antibody B72.3 reacting with TAG 72 antigen, labeled with iodine 125, was injected in 32 patients with primary tumors and in 22 patients with recurrent colorectal cancer. One hundred thirty-three suspected tumor sites were evaluated during operation by means of probe and resected with immunohistochemistry as control. RESULTS: Primary tumor sites were localized by RIGS in 60% of cases, and recurrent sites were localized in 82% of cases. There was a significant correlation both for primary (p < 0.001) and recurrent (p < 0.001) tumor sites between intraoperative RIGS findings and TAG 72 tumor antigen expression. Results obtained with the probe were instrumental in modifying the surgical approach in six (27%) of 22 patients with recurrences, allowing the removal of tumor masses that would otherwise have been overlooked. CONCLUSIONS: The results of RIGS seems to be encouraging in terms of clinical use. The potential high diagnostic resolution appears to improve surgical ablation of colorectal cancer, especially in patients with recurrent cancer or suspected recurrent tumors who have negative results for intraabdominal disease by all other roentgenographic criteria with rising carcinoembryonic antigen or TAG 72 antigen levels.

Liver and tumor uptake and plasma pharmacokinetic of arterial cisplatin administered with and without starch microspheres in patients with liver metastases.

Arterial chemoembolization of liver tumors should improve regional treatment by reducing native blood flow of the whole organ and redistributing residual flow toward hypovascular masses. Plasma cisplatin pharmacokinetics and its tissue uptake and relative tumor and liver vascularity were studied during surgical placement of arterial catheters in four patients and in four patients with colorectal metastases given intraoperative arterial cisplatin (DDP, 25 mg/m2), with an without coadministration of 600 mg degradable starch microspheres (DSM). Mean (+/- standard deviation) filterable plasma platinum levels peaked later (2 minutes) and were significantly lower after DDP with DSM (1.23 +/- 0.69 micrograms/ml) than after DDP alone (2.13 +/- 0.43 micrograms/ml, P less than 0.05), with the area under the curve (AUC0-30 min) values of 15.8 +/- 5.5 and 25.1 +/- 3.8 micrograms x min/ml (P less than 0.05), respectively. No differences in urine excretion, total body clearance, or plasma protein binding of platinum were observed. Tissue biopsies were started 15 minutes after DDP administration and completed in all cases within 5 minutes. Tumor platinum concentrations were significantly higher after DDP with DSM (3.03 +/- 1.60 micrograms/g) than after DDP alone (0.67 +/- 0.49 micrograms/ml, P less than 0.05). Liver concentrations and tumor-liver ratios of platinum also were higher, although not significantly, after DDP with DSM. Preoperative vascularization, studied with arterial perfusion scan, influenced individual tissue drug uptake in cases given DDP alone, with the lowest tumor levels in cold masses. Very high and almost superimposable liver and tumor concentrations were measured in those receiving DDP and DSM. The latter phenomenon was irrespective of native vascularization, indicating that DSM administration induced both an increased whole-liver extraction of the drug and a redistribution of blood flow and flow-dependent tissue uptake of platinum.

A phase I study of recombinant interleukin-2 intraperitoneal infusion in patients with neoplastic ascites: toxic effects and immunologic results.

A phase I study to evaluate the use of i.p. infusion of recombinant interleukin-2 (rIL-2) was planned. The following dose levels were calculated: 0.1, 0.3, 1.0, 3.0 and 10 mg/m2/day for 14 days, but only the second levels were reached. In this trial the acute toxic effects at this dosage included cardiac ischemia, transient liver impairment and septic peritonitis. The maximum tolerated dose (MTD) was 0.3 mg/m2/day for 14 days. In addition, two patients developed peritoneal fibrosis. No objective responses were observed. Therefore, in order to explore the biological activity of low (nontoxic) doses, three patients (one untreated and two previously treated with rIL-2) were infused with 0.01 and 0.03 mg/m2/day for 7 days. Potentiation of cytolytic activities in peritoneal lymphocytes and activation of a lymphokine cascade in the ascitic fluid were observed at doses ranging from 0.03 mg/m2/day to 0.3 mg/m2/day. These findings in association with the toxic effects observed at the MTD suggest the use of the minimum effective dose for future locoregional immunotherapeutic protocols.