Conservative treatment of rectal adenocarcinoma.

Endocavitary radiotherapy and transrectal excision are highly effective treatments for properly selected patients with favorable early-stage rectal adenocarcinoma. The likelihood of local control and survival after treatment with either modality is similar, and differences among various series probably reflect selection. The parameter most predictive of local control and survival in the authors' series was tumor configuration. As has been previously observed, "selection is the silent partner of success." Suitable candidates for endocavitary radiotherapy or wide local excision are patients whose tumors are 3 cm or less in diameter, well-to-moderately differentiated, exophytic, mobile, limited to the submucosa on transrectal ultrasound, and within 10 cm of the anal verge. The advantages of endocavitary irradiation are (1) it is an outpatient procedure, (2) it does not require anesthesia, and (3) it is less expensive than transrectal excision. The advantages of transrectal excision are (1) it may be performed during one brief hospitalization (as opposed to four outpatient visits), and (2) a small subset of patients will have pathologic findings predicting an increased risk of regional lymph node involvement, revealing the need to treat the nodes with external-beam radiotherapy. A disadvantage of wide local excision is that some patients who would be suitable for a local procedure alone must be subjected to a course of external-beam radiotherapy when they are found to have equivocal or positive margins. Patients who are treated with transrectal excision and external-beam radiotherapy have less favorable lesions and are not comparable with patients who are treated with endocavitary radiotherapy or wide local excision alone. They are best compared with patients who have undergone major surgery consisting of abdominoperineal resection or low anterior resection. Because the risk of positive nodes is significantly increased with adverse pathologic findings such as poor differentiation, invasion of the muscularis propria, and endothelial-lined space invasion, a subset of these patients treated with wide local excision would have positive nodes. This subset of patients is not comparable with patients with stage pT1N0 and pT2N0 tumors treated with major surgery. The latter group of patients undergo complete surgical staging, whereas the pathologic staging for patients who undergo wide local excision and radiotherapy is limited to the extent of the primary tumor. With this caveat in mind, wide local excision and radiotherapy seem to result in locoregional control and survival rates similar to the rates obtained with major surgery for patients with pT1 and pT2 cancers (Table 5). Patients who should receive postoperative irradiation have tumors that exhibit one or more of the following characteristics: size greater than 3 cm in diameter, poorly differentiated, invasion of the muscularis propria, endothelial-lined space invasion, fragmented resection, equivocal or positive margins, or perineural invasion. Patients with gross residual disease are not suitable candidates for radiotherapy and require further surgery. The authors' policy is to treat these patients with chemoradiation followed by resection. Patients thought to have transmural invasion before treatment are probably best treated with preoperative chemoradiation combined with major surgery, although a subset of patients can be downstaged and rendered suitable for a wide local excision.

Local excision and postoperative radiation therapy for rectal adenocarcinoma.

Sixty-seven patients with early-stage adenocarcinoma of the rectum who had lesions thought to be unsuitable for either local excision alone or endocavitary irradiation were treated with local excision followed by postoperative radiation therapy. The purpose of this study was to evaluate the effectiveness of local excision followed by radiation therapy for treatment of rectal adenocarcinoma. The patients were treated between 1974 and 1999; follow-up time was 6 to 273 months (median, 65 months). All living patients had follow-up for at least 2 years. The indications for postoperative irradiation included equivocal or positive margins, invasion of the muscularis propria, endothelial-lined space invasion, poorly differentiated histology, and perineural invasion. Cox proportional hazards regression analysis was performed using six explanatory variables including tumor size, configuration (exophytic vs. ulcerative), histologic differentiation, pathologic T stage, endothelial-lined space invasion, and margin status. The time interval between treatment and development of recurrent disease was in the range of 11 to 48 months. The 5-year results were as follows: local-regional control, 86%; ultimate local-regional control, 93%; distant metastasis-free survival, 93%; absolute survival, 80%; and cause-specific survival, 90%. When the Cox proportional hazards regression analysis was performed for these endpoints, margin status influenced absolute survival (P = 0.0074), cause-specific survival (P = 0.0405), and ultimate local-regional control (P = 0.0439). Tumor configuration marginally influenced cause-specific survival (P = 0.0577). None of the variables had an influence on the endpoints' local-regional control, ultimate local-regional control with sphincter preservation, or distant metastasis. Five patients (7%) had severe complications; no complication was fatal. Local excision and postoperative radiation therapy results in a high probability of local-regional control and survival for selected patients with relatively early-stage rectal adenocarcinoma. Patients with ulcerative tumors may have a lower likelihood of cause-specific survival.

Conservative treatment of rectal adenocarcinoma with endocavitary irradiation or wide local excision and postoperative irradiation.

PURPOSE: To evaluate the role of endocavitary irradiation and wide local excision followed by irradiation in the treatment of early-stage rectal adenocarcinoma. MATERIALS AND METHODS: Sixty-five patients with early-stage adenocarcinoma of the rectum were treated with endocavitary irradiation (n = 20) or wide local excision followed by external-beam irradiation (n = 45) between 1974 and 1994 at the University of Florida. All patients were monitored for a minimum of 2 years or until death. RESULTS: The rates of local-regional control at 5 years were 80% after endocavitary irradiation and 86% after wide local excision and radiotherapy. The ultimate 5-year local-regional control rates were 85% and 92%, respectively. Multivariate analysis of local-regional control with sphincter preservation showed that tumor configuration (exophytic v ulcerative) significantly influenced this end point; local-regional control was decreased in patients with ulcerated cancers. Five-year cause-specific survival rates were 84% after endocavitary irradiation and 88% after wide local excision and radiotherapy. Multivariate analysis revealed that tumor configuration significantly influenced cause-specific survival; patients with ulcerated tumors had a worse prognosis. CONCLUSION: Endocavitary irradiation is a highly effective treatment for properly selected patients with early-stage rectal adenocarcinoma. Patients with less favorable lesions that appear to be limited to the muscularis propria have a high chance of cure with sphincter preservation after wide local excision and external-beam irradiation.

Marginal ulcer following gastric bypass for morbid obesity.

Four hundred twelve patients underwent gastric bypass for treatment of morbid obesity between 1981 and 1985 at the University of Florida Affiliated Hospitals. Thirty-four patients (8.2%) developed marginal ulcers, considerably higher than the 0-3 per cent ulcer occurrence commonly reported in the literature. Factors predisposing to ulcer formation include: (1) a large gastric pouch; (2) a vertically oriented pouch; and (3) staple-line dehiscence. Twenty-two of 34 patients (65%) with symptomatic marginal ulcers were noted to have staple-line disruption. Twenty-one of these patients (95%) eventually required operative therapy for their ulcers compared with four of 12 patients (33%) with an intact gastric staple line. Surgical therapy consisted of takedown of the Roux-en-Y limb with resection of the ulcer and gastrogastrostomy. Staple-line dehiscence is a significant etiologic factor in the development of marginal ulcer following gastric bypass and when present constitutes an indication for reoperation.

Pouch outlet obstruction following vertical ring gastroplasty for morbid obesity.

We have experienced a 14% (38 of 264 patients) incidence of pouch outlet obstruction following vertical ring gastroplasty. Initial management consisted of dilatation in 34 of 38 patients (94%). Ten of 34 patients (29%) were spared reoperation by 1 to 3 dilatations. Non-passage of an endoscope through the stoma immediately following dilatation predicted the need for surgery; 4 of 11 patients (36%) with passage underwent reoperation compared with 17 of 20 patients (85%) without passage (p less than 0.02). Surgical findings included "tipped bands" in 9 of 28 patients (32%); fibrous reaction to the band in 10 of 28 patients (36%); adhesions with angulation of the pouch in 2 of 28 patients (7%); and no identifiable cause of obstruction in 7 of 28 patients (25%). Surgical therapy consisted of removal of the band (2 patients), removal of the band and replacement with a similar length or larger band (20 patients), "tacking" the band in the horizontal position (4 patients), or conversion to a Roux-Y bypass (2 patients). The first three options were associated with an unacceptably high rate of weight regain and/or continued symptoms, whereas the last-named procedure met with good success.

Pathogenicity of the enterococcus in surgical infections.

The enterococcus has been relegated to a position of unimportance in the pathogenesis of surgical infections. However the increasing prevalence and virulence of these bacteria prompt reconsideration of this view, particularly because the surgical patient has become increasingly vulnerable to infectious morbidity due to debility, immunosuppression, and therapy with increasingly potent antibiotics. The enterococcus is a versatile opportunistic nosocomial pathogen, causing such diverse infections as wound, intra-abdominal, and urinary tract infections; catheter-associated infection; suppurative thrombophlebitis; endocarditis; and pneumonia. Although surgical drainage remains the cornerstone of therapy for enterococcal infections involving a discrete focus, in the circumstances typified by the compromised surgical patient, specific antibacterial therapy directed against the enterococcus is warranted. Recent evidence indicates that parenteral antibiotic therapy for enterococcal bacteremia is mandatory and that appropriate therapy clearly reduces the number of deaths.

Accelerated lung amino acid release in hyperdynamic septic surgical patients.

Amino acid flux across the lungs was studied in humans to gain further insight into the altered nitrogen metabolism that characterizes catabolic disease states. Lung flux of glutamine, glutamate, and alanine was determined in three groups of surgical patients with indwelling pulmonary artery catheters: (1) preoperative controls (n = 14), (2) postoperative elective general surgical patients (n = 10, and (3) hyperdynamic septic surgical patients (n = 17). In controls the lung was an organ of amino acid balance. These exchange rates did not change in general surgical patients. In the septic group, glutamine release by the lung increased markedly from a control value of 0.80 +/- 0.99 mumol/kg per minute to 6.80 +/- 1.32 mumol/kg per minute. This accelerated release rate was secondary to both an increase in total pulmonary blood flow and an increase in the pulmonary artery-systemic arterial concentration difference. The lung also became an organ of significant alanine release in septic patients. The lung plays an active metabolic role in the processing of amino acids and may be a key regulator in interorgan nitrogen flux after major injury and infection.

The case for prophylactic cholecystectomy concomitant with gastric restriction for morbid obesity.

The pre and postoperative incidence of cholelithiasis were investigated in patients undergoing bariatric surgery at the University of Florida. The first part of the study was retrospective and revealed a pre and 24-month postoperative incidence of cholelithiasis of 30 and 40 percent respectively. Age and postoperative interval were not predictive of cholelithiasis. Patients with cholelithiasis had a significantly greater weight loss (130 +/- 61.0 lbs) than those without stones (109 +/- 59.9 lbs) P = 0.04. Men had a significantly greater weight loss than women (160 +/- 15 lbs SEM versus 99 +/- 7 lbs SEM) as well as a higher incidence of cholelithiasis (53 and 24%, respectively). In the second, prospective part of the study, cholecystectomy was performed in 73 consecutive patients concomitant with their bariatric procedure. Ninety six per cent of removed gallbladders had gross or histologic abnormalities including cholelithiasis in 27 per cent and cholesterolosis/cholecystitis in 69 per cent. The incidence of cholelithiasis was higher than that found in the retrospective series by preoperative ultrasound. The bariatric surgical patient is clearly at risk for the development of postoperative cholelithiasis and cholecystitis. The risk appears to be related to the amount of weight loss. In addition, some gallstones may remain undetected at the time of surgery. We therefore recommend prophylactic cholecystectomy at the time of bariatric surgery.

Clinically resectable adenocarcinoma of the rectum treated with preoperative irradiation and surgery.

Seventy-four patients with clinically resectable adenocarcinoma of the rectum were treated with preoperative irradiation and surgery at the University of Florida between August 1975 and February 1982. All patients have been followed for at least five years. Between 1975 and 1978, 29 patients received 3500 cGy; thereafter the dose was increased to 4000 to 5000 cGy for the remaining 45 patients. All patients were treated at 180 cGy per fraction. Following preoperative irradiation, 65 of 74 patients (88 percent) underwent complete resection of their lesions. Compared with a series of historical controls treated with surgery alone, the local recurrence rate at five years was 5 of 65 (7.7 percent) vs. 39 of 135 (29 percent) (P = .001), and the five-year absolute survival was 43 of 65 (66 percent) vs. 51 of 135 (38 percent) (P less than .001). The local recurrence rate was 13 percent for patients receiving 3500 cGy and 5 percent for doses of 4000 to 5000 cGy. There was no apparent increased incidence in postoperative complications in the preoperatively irradiated patients.

Effect of indomethacin on intestinal water transport in salmonella-infected rhesus monkeys.

Indomethacin, a nonsteroidal anti-inflammatory agent, will abolish salmonella-induced rabbit ileal secretion when given prior to the establishment of infection. In the present study, we assessed whether indomethacin can inhibit salmonella-induced intestinal secretion when administered after infection and net intestinal secretion are well established. A physiological model of salmonellosis, salmonella-infected rhesus monkeys, was used. This model also permitted an examination of the effects of indomethacin in both the small and large intestines. The effect of indomethacin in control monkeys was also studied. Indomethacin caused a striking enhancement of net intestinal water transport in the jejunum,, ileum, and colon of salmonella-infected monkeys. These effects occurred promptly and were of sufficient magnitude in the ileum and colon to cause a reversal in the direction of net transport from net secretion to net absorption. Indomethacin also enhanced net water transport in the jejunum ileum, and colon of normal animals. These data show that indomethacin markedly enhances net intestinal water transport in both the small and large intestines of salmonella-infected monkeys, even when administered after salmonella infection and intestinal secretion are well established. Similar enhancement also occurs in the normal intestine. The mechanism(s) by which indomethacin produces these effects is not known.

Role of plasma filtration in the intestinal fluid secretion mediated by infection with Salmonella typhimurium.

The mechanisms whereby invasive enteropathogens, e.g., Salmonella typhimurium, induce intestinal secretion are largely unknown. Since these organisms penetrate the intestinal epithelium, disrupt the brush border, and evoke an acute inflammatory reaction, increased plasma filtration through a damaged, more permeable epithelium might contribute to the secretory process. To examine this possibility, the plasma-to-lumen clearance of two different sized molecules, [51Cr]albumin and [14C]mannitol, was measured in the in vivo rabbit ileal loop and in vivo rhesus monkey models of salmonellosis. In the rabbit ileal loop model, the clearance of neither molecule was increased when compared to cholera toxin-exposed loops. In the rhesus monkey, clearance of [14C]mannitol into the jejunum, ileum, and colon of Salmonella-infected animals did not differ from the observed in control animals. These data indicate that invasion of the intestinal mucosa by S. typhimurium has not substantially altered the permeability characteristics of the intestinal mucosa and that plasma filtration through a damaged, more permeable mucosa does not contribute to the Salmonella-induced intestinal secretory process.

Small intestinal mucosal injury in the experimental blind loop syndrome. Light- and electron-microscopic and histochemical studies.

Microscopic (light and electron) and histochemical abnormalities have been demonstrated in the jejunum of rats with the blind loop syndrome. Three groups of animals were studied: normal control animals, and animals with either self-filling (SF) or self-emptying (SE) blind loops. Vitamin B12 malabsorption and bacterial overgrowth occurred only in those animals with SF blind loops. Three jejunal segments were studied: the blind loop segment and the jejunal segments proximal and distal to the blind loop. In the animals with the blind loop syndrome, those with SF blind loops, the most striking findings occurred in the blind loop itself, with similar but less marked changes in the jejunum distal but not proximal to the blind loop segment. Hypertrophy of both crypts and villi was evident with focal abnormalities of villus architecture. Approximately 10 to 20% of the columnar cells in the upper half of the villi were swollen and vesiculated. By electron microscopy microvilli demonstrated a variety of degeneration changes and the glycocalyx and terminal web were disrupted. Mitochondria and endoplasmic reticulum (ER), both smooth and rough, were swollen. Concentric whorls of parallel membranes and long, curvilinear rough ER were present in the cytoplasm. Histochemically, there was loss of enzymatic activity in the epithelial brush border, mitochondria and ER. Inasmuch as bacterial invasion of the jejunal mucosa was not seen, the etiology of these changes is not known but may involve bacterial "toxins" or products of bacterial metabolism. These morphological observations demonstrate that both brush border and intracellular injury occur in the jejunal epithelial cell of rats with the experimental blind loop syndrome.

Pathophysiology of Shigella diarrhea in the rhesus monkey: intestinal transport, morphological, and bacteriological studies.

In contrast to the "toxigenic diarrheas" caused by Vibrio cholerae and Escherichia coli, the site and mechanism of fluid loss in shigellosis are unknown. The occurrence of watery diarrhea in shigellosis suggests involvement of the small bowel. Therefore, jejunal, ileal, and colonic water and electrolyte transport was studied in Shigella flexneri 2a-infected monkeys. Infected animals fell into three groups: dysentery alone, diarrhea alone, or diarrhea and dysentery. In controls, net water, sodium, and chloride absorption was seen in the jejunum, ileum, and colon. All infected animals demonstrated diminished colonic absorption or net colonic secretion. In monkeys with dysentery alone, this was the only transport defect observed. In contrast, animals with diarrhea either alone or in combination with dysentery, exhibited net jejunum secretion. Ileal transport was normal in all animals. A severe colitis with intramucosal shigellae was seen in all symptomatic animals. In the jejunum or ileum, however, morphological changes were minimal and bacterial invasion was not seen. Therefore, unlike the "toxigenic" diarrheas," shigellosis is both a small and large intestinal disease. Mucosal invasion of the colon is essential to the development of a morphological and transport defect. Dysentery results from a colonic transport defect, while diarrhea is secondary to jejunal secretion superimposed on the defect in colonic absorption.