The clinical records of 61 patients who underwent extended surgery, including resection of parietes or other viscera, from 1965 to 1977 for cancer of the rectum and sigmoid were reviewed. Abdominoperineal resection was performed in 41 patients, anterior resection in 18 and Hartmann's resection in 2. The postoperative mortality rate was 8.2 per cent, the non-lethal morbidity rate 30.3 per cent, but 4 patients presented multiple complications. The 5-year survival rate was evaluated separately for patients with and without microscopic evidence of neoplastic involvement of the simultaneously excised structures; in the first group it was 32 per cent, in the second 75 per cent. Local or distant recurrence occurred in 66 per cent of patients with microscopic infiltration and in 24 per cent of patients without microscopic infiltration. These results compare favourably with those reported after ordinary resections of Dukes' C cancers of the rectum and sigmoid, and seem to justify the use of extended surgery when cancer of the rectum and sigmoid has invaded contiguous structures.

Previous studies indicated that peripheral blood lymphocytes from patients (Pt-PBL) with lymph node metastatic melanomas proliferated in vitro and developed into tumor-restricted cytotoxic lymphocytes in response to alloantigens or interleukin 2 (IL-2). However, Pt-PBL were not stimulated by irradiated autologous metastatic melanoma (Auto-Me) cells. In the present study we report that the lack of stimulatory activity of Auto-Me cells may be due to a suppressive effect exerted by Auto-Me cells on the responder lymphocytes. In fact, we found that in 62% of cases examined, the addition of 5-10% Auto-Me cells to Pt-PBL cultures strongly inhibited both proliferation and the generation of tumor cytotoxic lymphocytes induced by alloantigens or IL-2. The inhibition was dose-dependent and tumor-restricted, and was not due either to toxicity, medium depletion or IL-2 absorption by Auto-Me cells. Normal fibroblasts, K562 cells and autologous E-lymphocytes were not suppressive. Auto-Me cells were able to inhibit Pt-PBL responses only when added during the first 24 h of culture and not later. Phenotypic analysis of Auto-Me cells using monoclonal antibodies directed against HLA-A,B,C, HLA-DR and melanoma-associated antigens revealed that the expression of high levels of DR antigens on Auto-Me cells was associated with an elevated suppressive activity. Conversely, Auto-Me cells with low or undetectable levels of DR antigens were not inhibitory. Furthermore, the increased expression of DR antigens on Auto-Me cells obtained by in vitro treatment with human interferon gamma (IFN-gamma) also resulted in an increased suppressive activity. We conclude that HLA-DR+ metastatic melanoma cells can interfere with the generation of an anti-tumor immune response, thus potentially favoring the escape of the tumor from the host's control mechanism.

566 stage-II melanoma patients treated at the National Cancer Institute of Milan, Italy, were analyzed to evaluate the prognosis. Among the criteria considered, four were significantly associated with survival when considered as single factors: growth pattern, levels of invasion, the number of involved lymph nodes, and the extent of metastatic growth. As regards growth pattern, the observed 5-year survival rates were 41.9% for superficial spreading melanoma and 20.5% for nodular melanoma (P = 10(-3)). As regards levels, the 5-year survival rates were as follows: level II, 20.9%; level III, 33.1%; level IV, 43.2%; level V, 10.2% (P = 10(-3)). Patients with a partial node metastasis had 64.5% 10-year survival, while those with extension beyond the capsule had 32.6% 10-year survival (P = 10(-9). Patients with one metastatic node had 43.4% 10-year survival, and patients with three or more positive nodes had 26.0% 10-year survival (P = 10(-9)). Multifactorial analysis shows that growth pattern and extent of node metastases significantly affect survival (P = 10(-2) and P = 10(-4), respectively) while the number of involved nodes turns to borderline P-value (0.051) and the levels are no longer significant (P = 0.4).

A phase I study was carried out to test the feasibility and toxicity of infusing large numbers of autologous, alloactivated helper lymphocytes into patients with metastatic melanoma. Patient peripheral blood lymphocytes (Pt-PBL) obtained by lymphopheresis and expressing the helper phenotype BT5/9 were separated and stimulated for 48 or 72 h with a pool of PBL from four to six healthy donors. Patients were then infused with such activated lymphocytes over a 2-3 h period. A total of 4 phereses and infusions (2/week for 2 weeks) were carried out for each cycle in each patient. Of the five patients treated, two received a second round of infusions. Infusion of autologous PBL stimulated in vitro for 48 h caused chills, fever, headache, and increased blood pressure. All symptoms disappeared in 2-3 h and were easily controlled by appropriate therapy. When lymphocytes were given after 72 h of allostimulation, no or very mild toxicity was observed. Serum chemistry, coagulation, autoimmunity, and urine analysis showed no gross abnormalities during therapy or follow-up of the patients. Immunological parameters (OKT4/OKT8 ratio, NK activity and cytotoxic T cell activity to autologous melanoma) were evaluated before starting the therapy, during its course and during the 3 to 6 months follow-up. The OKT4/OKT8 ratio increased significantly but transiently soon after the first course of infusions in one of the two patients tested. NK activity increased after 75-100 days in the three patients tested and in one of them it was high even after 180 days. No correlation between NK activity and prognosis was apparent. Cytotoxicity to autologous tumor was assessed in two patients, only of one of whom exhibited an increased activity from 75 to 180 days, which was associated with a prognosis better than that of the negative patient. Five patients were treated: two had progressive disease, two had stable disease for 5 and 6 months, respectively. In the first of these patients, a new cycle of lymphocyte infusions was carried out which caused a measurable reduction of lung tumor nodules whose growth, however, resumed 4 months later. This patient died 14 months after the onset of therapy. The fifth patient had a partial regression of pulmonary and intracranial metastases after therapy, but eventually died 3 months later. These results indicate that infusion of a high numbers of autologous, allostimulated helper PBL is a feasible and safe procedure, which could therefore be used in future studies of adoptive immunotherapy of cancer.

Early surgical complications following colostomy closure in 65 cancer patients operated on at the Istituto Nazionale Tumori of Milan were evaluated retrospectively. The overall complication rate was 24.6 per cent, including infections (13.8 per cent), fistulas (6.1 per cent), wound dehiscence (3.0 per cent), and distal stenosis (1.5 per cent). Type and rate of complications were analyzed to find a correlation with type, site, and location of colostomy, technique of closure, presence or absence of drains, or time interval between construction and closure of colostomy. No statistically significant association between the aforementioned factors and occurrence and rate of complications was found. The authors think, therefore, that surgical attention, including meticulous manipulation of the stoma, avoidance of contamination of the wound, tension of sutures, dead spaces, and collection of blood in the wound, and use of antibiotics and antiseptics are the most important principles to minimize postoperative complications.