SUMMARYWe previously reported that the number of TNF-a-producing cells was increased in the liver of patients with type C chronic liver disease. To understand further the pathophysiology of this change, we examined serum levels of two soluble TNF receptors, TNF-aRI (p55) and -aRII (p75), and IL-10, all of which act as TNF-a buffer, and IL-15, a novel cytokine sharing many immunological activities with IL-2, using ELISA methods. We studied control individuals and patients with type C chronic liver disease, including asymptomatic hepatitis C virus (HCV) carriers with persistently normal serum ALT values, and those with chronic hepatitis (CH), liver cirrhosis (LC) and hepatocellular carcinoma (HCC). Both types of sTNF-aR closely correlated with disease progression. Patients with LC and HCC had significantly elevated levels for sTNF-aRII compared with the other patient groups and controls. Serum IL-10 levels were significantly greater in all chronic liver disease groups than in controls. With respect to IL-15, the values were high in CH, LC and HCC compared with those of controls. Notably, HCC patients showed highest values for both IL-10 and IL-15, with significant differences from the other patient groups. Serial determinations revealed that interferon (IFN) treatment for CH patients resulted in the suppression of circulating IL-10 and IL-15 levels along with decrease in serum aminotransferase values. Both cytokines remained at decreased levels after cessation of therapy in patients who went into clinical and virological remission. On the other hand, treatment did not affect serum levels of sTNF-aRs. These findings indicate that serum levels of these molecules correlated with disease progress in chronic HCV infection, and that IL-10 and IL-15 may reflect the degree of inflammation in the liver. It is also suggested that both cytokines may be related to the development of HCC.

These findings indicated that a proximal shift in the subsite distribution of colorectal cancer has occurred in Japan. This rightward shift of colorectal cancer is due to the decreasing proportion of rectal cancer. Furthermore, the increasing proportion of older patients, especially females, may be another major determinant of the changing colon cancer subsite distribution.

Clinical, physiologic, and anatomic assessments were carried out in 22 female patients with symptomatic rectocele (Group A), 15 patients with asymptomatic rectocele (Group B), and 14 subjects having no rectocele (Group C). Resting and pressure, rectal pressure, rectal compliance, anorectal inhibitory reflex, and rectal sensation did not differ among the groups. Proctography revealed that the lengths of the rectocele during attempted defecation in groups A (1.6 [1.0-3.5] cm) (median and range) and B (1.6 [1.0-3.0] cm) were significantly greater than that in Group C (0.4 [0.1-0.9] cm) (P less than 0.001 in both groups). Median pelvic floor descent at rest in Groups A (4.3 [1.6-7.5] cm) (median and range) and B (4.3 [1.3-6.9] cm) were significantly greater than that in Group C (2.5 [1.2-5.0] cm) (P less than 0.001 and P less than 0.02, respectively). These results indicate that rectocele is not associated with any physiologic change apart from a significant increase of pelvic floor descent.

Subtotal colectomy has been performed in 40 patients with severe constipation. Only one patient was male. Five patients (13 per cent) had a history of a serious psychiatric disorder. Twenty-six patients (65 per cent) could not expel a 100-ml air-filled balloon and 19 patients (48 per cent) had electromyographic evidence of abnormal puborectalis contraction during attempted defaecation before operation. Median preoperative passage of 50 radio-opaque markers over 5 days was 16 per cent. Sixteen patients had had a previous anorectal myectomy to exclude Hirschsprung's disease. Initial resections were subtotal colectomy and ileorectal anastomosis (n = 34), caecorectal (n = 5) or ileosigmoid (n = 1) anastomosis. Secondary operations included restorative proctocolectomy and ileal pouch-anal anastomosis (n = 6) and six patients eventually had an ileostomy. Median bowel frequency per week significantly increased after operation (0.3 (range 0-1) preoperatively to 21 (range 2-70) postoperatively, P less than 0.005), the percentage of patients with abdominal pain fell after operation from 93 to 39 per cent but symptoms of abdominal distension remained the same (86 per cent preoperatively and 82 per cent postoperatively).

Tumor progression could be recognized and occult metastases were detected during the interval between operations. Delayed resection of synchronous LM may be useful to reduce the risk of rapid recurrence in the remnant liver. Patients with poorly differentiated or mucinous adenocarcinoma and those with larger tumors who undergo delayed liver resection should receive neoadjuvant chemotherapy during the interval between operations.

One hundred sixty-five cases of abdominal rectopexy using polypropylene (Marlex) mesh for rectal prolapse were reviewed. Six patients were men and 159 were women. Thirty patients have not been evaluated after surgery, 22 having died of interrecurrent disease and 8 have had their surgery during the last two months. Incontinence was observed in 95 patients (58 per cent) before surgery, whereas it persisted in only 21 of 135 patients (16 percent) after surgery. Forty patients (24 percent) claimed constipation before surgery, whereas 60 of 135 patients (44 percent) had constipation after rectopexy. Recurrence of full-thickness rectal prolapse was found in only 2 patients-(1.5 percent). Mucosal prolapse occurred in 9 patients (7 percent) after surgery. These results indicate that abdominal posterior rectopexy using Marlex mesh is an effective operation for rectal prolapse, but persistent incontinence occurs in one third of patients and almost half become constipated after the procedure.

The occurrence of IF and IF-related death in CD patients is not rare over the long term. There is a pressing need to develop strategies for the prevention and management of IF.

Early recurrence is the leading cause of death within 5 years after curative resection of CRLM. Postoperative morbidity increases the risk of early recurrence in these patients. A reduction in perioperative morbidity may, therefore, improve the outcome of curative resection, as well as reducing medical costs.