Systematic review of fibroadenoma as a risk factor for breast cancer.

Fibroadenoma (FA) is a common breast lesion occurring in 25% of asymptomatic women. Several studies considered FA as a risk factor for breast cancer, prompting a systematic review of the literature. We selected cohort and case-control studies. Seven studies satisfied the selection criteria. Dupont et al. (1994) and McDivitt et al. (1992) were considered to provide the strongest evidence, showing the relative risk for excised FA without hyperplasia in the range 1.48-1.7, with hyperplasia 3.47-3.7, and with hyperplasia and atypia 6.9-7.29, persisting for more than 20 years. Five studies by Levi et al. (1994), Ciatto et al. (1997), Moskowitz et al. (1980), Carter et al. (1988), and Levshin et al. (1998), were considered to provide weaker evidence, although they showed similar results. None of the results could be used to quantify the risks of excised, non-excised, and asymptomatic FA. More care should be taken in managing patients with complex FAs and FAs with hyperplasia with or without atypia. Core biopsy may be a better option in diagnosis of FAs. Regular screening may be advisable in patients with FAs and a family history of breast cancer in a first-degree relative. More rigorous research is needed in this area.

Emergency surgery: atavistic refuge of the general surgeon?

A prospective audit of emergency soft-tissue surgery for an eight-week period revealed that general surgical emergency operations were more than twice as common as those undertaken in other soft-tissue specialties. The audit reveals that emergency general surgery needs an increase in resources, an increase in available staff and an increase in the role of the consultant general surgeon on call. An alternative solution would be to admit soft-tissue emergencies by specialty and develop specialist emergency services.

Perioperative blood transfusion increases the risk of recurrence in colorectal cancer.

A retrospective study was conducted on 519 patients undergoing curative resection for colorectal carcinoma between 1969 and 1980. Recurrence was diagnosed in 214 patients (41.2 percent), 179 of whom (34.5 percent) had received blood transfusions and 35 of whom (6.7 percent) had not (P less than .001). Exclusion of the right-sided colonic tumors still showed that recurrence was more common in transfused than nontransfused patients (135 [47.2 percent] vs. 25 [22.5 percent]; P less than .001). Recurrence in patients transfused only during surgery (N = 201) was higher than in nontransfused patients (P less than .001) and, similarly, all patients transfused during surgery (N = 297) had an increased risk (P less than .001). Among patients with rectal cancer, transfusion increased the risk of recurrence in those treated by abdominoperineal resection (P less than .02), but this was not the case in those treated by sphincter-saving resection (P = .2). Hierarchical log linear analysis of all dependent factors (Dukes' stage, histologic grade, age, sex, site, elective, or emergency procedure) showed that Dukes' stage and blood transfusion had the most significant effects on the development of recurrence (chi 2 = 54.04, df = 6, P less than .0001 and chi 2 = 13.93, df = 3, P less than .003). The risk of recurrence following curative surgery for colorectal cancer is markedly increased by blood transfusion on the day of operation.

An evaluation of the preoperative diagnosis and management of cystosarcoma phyllodes.

Cystosarcoma phyllodes is a rare tumour of the breast which is notoriously difficult to diagnose accurately preoperatively. In this review we report the clinical, imaging and histopathological features of 20 patients who have presented in our Centre over a 10-year period.

Wide excision of primary breast cancer: the incidence of residual carcinoma at the site of excision.

Fifty-one consecutive women with early breast cancer underwent wide excision and axillary clearance. After wide excision five biopsies were taken from the walls of the cavity. On histological examination tumour was present in the cavity biopsies in 13 cases (25%). The presence of ductal carcinoma in situ in the primary tumour was not associated with an increased number of positive cavity biopsies. In nine cases incomplete excision was due to separate foci of invasive or in situ carcinoma, in two cases tumour was contiguous with the primary carcinoma and in two, separate foci and contiguous disease both occurred. The high incidence of residual local tumour after wide excision of breast cancer demands the need for postoperative irradiation to the breast and frequent review.

Arterioenteric fistulas.

Any patient with an abdominal aortic aneurysm, aortic graft or previously resected infected graft who presents with gastrointestinal bleeding or sepsis must be assumed to have an arterioenteric fistula. Warning bleeds are common. Oesophagogastroduodenoscopy may be diagnostic and should exclude other causes of upper gastrointestinal haemorrhage. Urgent investigation and surgery are essential for a successful outcome.

Thrombosis of the superior mesenteric vein.

Four patients under 45 years old presented with small bowel infarction secondary to superior mesenteric venous thrombosis. Diagnosis was not made pre-operatively in three patients and delay to operation of over 4 days occurred in two patients. In each case infarcted bowel was resected and a primary anastomosis fashioned. A further resection of infarcted small bowel was necessary in three patients. Each patient was discharged on long-term warfarin therapy. Two patients required permanent total parenteral nutrition, one of whom died of liver failure at one year. The remaining patients remain well.

Large bowel cancer in the young.

Eighty-five patients aged 40 or less who presented with colorectal carcinoma over a 32-year period were reviewed. The incidence was 2.5 per cent of all patients with large bowel cancer (n = 3426). Predisposing causes included familial polyposis (eight patients), panprotocolitis (ulcerative, one, Crohn's, one) and irradiation (one), four patients were pregnant, one-third of the patients presented as emergencies, and 43 per cent of these had intestinal obstruction. Five-year survival rates were 41 per cent overall and 59 per cent after "curative" resection. Survival was equivalent both for elective and emergency admission and for mucinous carcinomas (n = 16) and those non-mucinous carcinomas of moderate histological differentiation. Five-year survival rates were poorer when the history was less than three months in duration (20 per cent vs 45 per cent: P = 0.02) and for rectal and rectosigmoid tumours than colonic tumours (31 per cent vs 50 per cent: P = 0.05). Radical resection is indicated when feasible: four of five patients with involvement of adjacent viscera and four of six patients with resectable recurrence survived beyond 10 years. The outcome is similar to that at all ages, any unfavourable pathological features being balanced by improved survival following emergency operations.

Secondary arterio-enteric fistulae: a surgical challenge.

Ten cases of secondary arterio-enteric fistulae are described. There were nine graft enteric fistulae and one fistula involving the aortic suture line following elective resection of an infected graft. Only four of the patients initially received prophylactic antibiotics (single dose) at the original aortic reconstruction, and the vascular suture line had only been protected in two. Eight patients presented with bleeding and two with groin abscesses. One patient died before operation. Graft resection was undertaken in all patients and organisms were grown from six of eight grafts cultured. No patient died during operation but one died after 3 days. Axillofemoral bypass grafts were constructed in seven patients (four immediately after resection of prosthetic grafts and three within 4 days of operation). Only three of the eight patients who survived operation are still alive; two died of a ruptured aorta and one from a recurrent fistula. Two patients died of other causes. Four of five axillofemoral grafts in surviving patients subsequently occluded.

Proliferative and metastatic potential of exfoliated colorectal cancer cells.

Exfoliated colorectal carcinoma cells were obtained by lavage of 27 freshly resected tumor-bearing segments of human bowel with the use of either Hartmann's solution or medium 199. The tumor cells were isolated from the lavage fluid on Nycodenz (Nyegaard, Oslo, Norway) columns. Their proliferative and metastatic potentials were investigated by their ability to incorporate tritiated thymidine ([3H]dThd) and by their ability to form experimental pulmonary tumors following iv injection into immune-deprived stain A mice. Tumor cells from 7 of 12 patients incorporated [3H]dThd as detected by autoradiography of the cells. Pulmonary nodules of colorectal carcinoma were seen in a single mouse after iv injection of tumor cells from 6 of 17 patients and were histologically distinct from spontaneous lung cancers seen in 3 other animals. No tumors were seen in 12 immune-deprived mice receiving tissue culture medium 199 only. Thus exfoliated colorectal carcinoma cells can undergo further division and might give rise to implantation metastases in humans.

Peculiarities of mucinous colorectal carcinoma.

The clinical and pathological features of 54 mucinous carcinomas of the large intestine were compared with those of 576 non-mucinous carcinomas. Tumours were only categorized as mucinous if they contained at least 60 per cent of mucin by volume. Those with a moderate mucin content (60-80 per cent) were indistinguishable in behaviour from 'non-mucinous' tumours. By contrast, those with a high mucin content (greater than 80 per cent) showed several differences from non-mucinous cancers: they had a more proximal distribution through the large intestine, they comprised a greater fraction of cancers in the under 50 age group (24 versus 7 per cent: P less than 0.01), they were more likely to be Dukes' stage 'D' (58 versus 31 per cent: P less than 0.01) and local fixity was commoner (70 versus 37 per cent: P less than 0.001). Consequently the overall resection rate was reduced from 90 to 73 per cent (P less than 0.01), the curative resection rate from 69 to 42 per cent (P less than 0.01) and the 5-year survival rate from 37 to 18 per cent (P less than 0.05). Colorectal carcinomas of high mucin content require wide excision, tend to recur locally and carry a poor prognosis.

Expression of histocompatibility antigens and characterization of mononuclear cell infiltrates in normal and neoplastic colorectal tissues of humans.

Serial frozen sections were prepared from 22 colorectal carcinomas. Additional samples were obtained from the adjacent normal bowel in 10 patients, from 6 concomitant adenomas in 5 patients, and from another 4 isolated adenomas. Mononuclear cell infiltrates were stained by the indirect immunoperoxidase technique with the use of a panel of 6 mouse monoclonal antibodies to human leukocyte antigens. The degree of infiltration was graded from 4 (heavy) to 0 (nil). The colorectal carcinomas and adjacent normal bowel showed an equal degree of leukocyte infiltration (HLe-1), graded 3-4 in 8 cases and 2-3 in the other 2 cases. In 7 carcinomas cytotoxic-suppressor T-lymphocytes (UCHT-4) graded 2-3 predominated over helper T-cells (OKT-4) graded 0-1. By contrast, in the adjacent normal bowel cytotoxic and helper cells were present in equal numbers. Among the adenomas leukocyte infiltration was grade 4 in 9 and grade 3 in 1. In 9 of the 10 adenomas cytotoxic cells graded 2 predominated over helper cells graded 0-1. The number of helper cells was equivalent among 6 concomitant adenomas and carcinomas from 5 patients. Adenomatous epithelial cells expressed class II major histocompatibility complex antigens (OKIa-1). However, carcinomatous or normal epithelium showed only faint staining with OKIa-1. The similarity in cell infiltration is consistent with an adenoma-carcinoma sequence. The predominance of cytotoxic cells in carcinomas that expressed class I major histocompatibility complex supports the association between lymphocyte infiltration and a favorable prognosis.

Survival of 727 patients with single carcinomas of the large bowel.

The outcome of 727 patients presenting with solitary colorectal carcinoma over a seven-year period is reviewed. Of the patients, 52 per cent were females 45 per cent were over 70 years and 31 per cent had an emergency admission. Of the tumors, 43 per cent occurred in the rectum and 40 per cent were stage D (not treated curatively). Predisposing causes included inflammatory bowel disease (n = 12) and abdominal irradiation (n = 6); associated adenomatous polyps were present in 22 per cent of resection specimens. Hospital mortality rates (20 per cent overall) were adversely affected by emergency admission (36 per cent), age greater than 70 years (29 per cent) and advanced, stage D disease (31 per cent). Corrected overall five-year survival rate was 32 per cent and, after curative resection, 59 per cent. Of patients in whom curative resection included contiguous organs, 47 per cent survived five years. Survival was reduced in patients over 70 years (26 per cent), in emergency admissions (24 per cent), in poorly differentiated tumors (18 per cent), and if tumor fixity was present (14 per cent). Factors contributing to a favorable outlook included a long history (greater than one year) and a tumor situated in the left colon. Recurrence developed in 47 per cent of patients surviving curative resection and was seldom diagnosed at a curable stage.

Viability of exfoliated colorectal carcinoma cells.

The viability of tumour cells shed into the intestinal lumen was determined in 49 patients with carcinoma of the large bowel. Preoperative colorectal lavage was performed in 19 patients and irrigation of the cut ends of the operative specimen in 30 patients. The resulting cell suspensions were centrifuged on a Nycodenz linear density gradient column so that tumour cells, being larger, were concentrated in a band at the top. In 14 of 19 colorectal lavage cases viable tumour cells were recovered, as assessed by their characteristic morphology and ability to exclude trypan blue. A median of 0.78 X 10(6) viable tumour cells was recovered. The median percentage cell viability in the suspension was 92, i.e. 8 per cent of the tumour cells were dead (stained with trypan blue). In eight specimens viability was confirmed by the ability of tumour cells to hydrolyse fluorescein diacetate. In 17 of 30 proximal resection margin irrigations a median of 0.55 X 10(5) viable tumour cells was recovered, with a median percentage viability of 92.5. In 15 specimens the neoplastic cells showed fluorescence. In 21 of 25 distal resection margin irrigations a median of 1.92 X 10(5) viable tumour cells was recovered with a median percentage cell viability of 79.3, and fluorescence was observed in all specimens. The number of viable tumour cells did not correlate with the stage, differentiation, diameter or fixity of the tumour. However, the number of tumour cells recovered from the distal resection margin was inversely related to the distance of the tumour from that margin (Rank Difference Coefficient R = -0.6). Thus viable exfoliated tumour cells were demonstrated in 52 of 74 specimens (70 per cent). Their presence in large numbers at the site of intestinal anastomoses supports a potential role in the aetiology of suture-line recurrence.

Survival in acute obstructing colorectal carcinoma.

Acute intestinal obstruction was the presenting feature in 124 (19 per cent) of 646 patients with colorectal carcinoma seen over a six-year period. Forty-two per cent of tumors were incurable at presentation. Obstruction was complicated by perforation in 22 patients (18 per cent). Only 15 per cent of tumors occurred in the rectum. Although the postoperative mortality rate was higher in patients with coincidental perforation than in those without (52 vs. 26 per cent: P = 0.03), five-year survival rates were the same: 18 per cent overall, rising to 29 to 34 per cent after "curative" resection. Five-year survival rates were best for right colon tumors and worst for rectal tumors (36 vs. 5 per cent: P = 0.01). The overall hospital mortality rates for colostomy and delayed resection, resection with colostomy, and resection with anastomosis were equivalent (18 to 22 per cent), but following "curative" resection the hospital mortality rate was higher for resection with colostomy than with other treatments (29 vs. 15 per cent), since two patients died following early closure of colostomy. Five-year survival was better following resection with anastomosis (48 per cent) than staged procedures (18 per cent: P = 0.01), since two patients died following late closure of colostomy.

The efficacy of agents employed to prevent anastomotic recurrence in colorectal carcinoma.

Forty-eight of 72 surgeons canvassed in the South West of England (67%) routinely use an intraluminal cytotoxic agent to prevent suture-line recurrence following partial resection of the large bowel for cancer. The most popular agents are chlorhexidine-cetrimide preparations (n = 14), mercuric perchloride (12), povidone-iodine (7) and water (12); noxythiolin, sodium hypochlorite and silver nitrate are used occasionally. The mean duration of treatment is 2 minutes. When assayed for cytotoxity against tumour cells freshly prepared from human colorectal carcinomas (n = 10), both chlorhexidine-cetrimide and povidone-iodine were rapidly lethal at a wide range of concentrations (5-100%). Mercuric perchloride (0.2%) was similarly effective, but up to 20% of tumour cells remained viable after exposure to noxythiolin and nearly 30% with water alone. Chlorhexidine-cetrimide and povidone-iodine are the agents of choice to kill malignant cells exfoliated into the colorectal lumen.