[Surgical infections: the role of laparoscopic surgery]. / Chirurgische Infektionen - Stellenwert der laparoskopischen Chirurgie.

Surgical diagnostic and therapy in intraabdominal infections should be fast and definite. Infectious diseases of the abdomen are appendicitis, cholecystitis, peptic ulcer perforations, colonic perforations (mostly diverticulitis), gynaecologic infections and other rare indications. The role of laparoscopy especially in the treatment of intraabdominal infections is presented for different diagnoses. Besides its therapeutic options, laparoscopy plays a crucial role as diagnostic tool in intraabdominal infections. The decision for a laparoscopic or open access to the abdomen should be made individually under consideration of the diagnoses and the laparoscopic training of the operating team.

[Influence of adjuvant radio-chemotherapy for rectal cancer on quality of life]. / Einfluss der adjuvanten Radiochemotherapie beim Rektumkarzinom auf die Lebensqualität.

INTRODUCTION: Modern therapy for rectal cancer is associated with functional disorders. Dysfunction as a consequence of surgery has to be distinguished from disorders caused by postoperative therapy. Therefore we have compared the long-term functional results of patients who received postoperative radio-chemotherapy or no therapy in conjunction with low anterior resection of the rectum. PATIENTS AND METHOD: From 1997 to 2002, a total of 32 patients (16 males and 16 females) after low anterior rectal resection and postoperative radio-chemotherapy or surgical therapy alone was compared using standardized and validated instruments (Short-Form-36-Health-Survey, EORTC QLQ-C30, QLQ-CR 38 and ASCRS fecal incontinence questionnaire) in a matched-pair analysis (age, gender and time of surgery). Mean age was 61.8 (62.1) years and mean follow-up was 4 (3.8) years. RESULTS: Two out of the 40 examined parameters differed significantly. There were no significant differences in Short-Form-36-Health-Survey and EORTC QLQ-C30 scales between both groups. The QLQ-CR38 scale sexual enjoyment differed significantly, whereas future perspectives, sexual functioning, micturition problems, symptoms in the area of the gastrointestinal tract, weight loss, defecation problems, male and female sexual problems did not differ significantly. The scales Lifestyle, Coping/Behavior and Depression/Self-Perception of the ASCRS fecal incontinence questionnaire also did not differ significantly. The difference for embarrassment was significant. DISCUSSION: No differences in quality of life after postoperative radio-chemotherapy or no postoperative therapy in conjunction with low anterior rectal resection can be found. There are, however, tendencies that postoperative radio-chemotherapy has more adverse effects on continence and sexual function than low anterior rectal resection alone.

[Functional results after transvaginal, transperineal and transrectal correction of a symptomatic rectocele]. / Funktionelle Ergebnisse nach transvaginaler, transperinealer und transrektaler Korrektur einer symptomatischen Rektozele.

INTRODUCTION: The aim of this study was to compare functional outcome after transvaginal, transperineal and transrectal repair of a symptomatic rectocele and to develop the ideal surgical approach. PATIENTS AND METHOD: 28 patients (27 female, 1 male) who had undergone rectocele repair from 1996 to 2003 were analysed. Mean age was 59 years (range 30-79 years), follow-up was 24 months (range 3 to 70 months) and mean appearance of symptoms was 4 years prior to the operation (6 months-32 years). Transvaginal repair was performed in 13 cases, transperineal repair in 8 cases and transrectal repair in 7 cases. RESULTS: 24 of 28 patients (85.7 %) are satisfied with the operation-result (transvaginal 12 of 13 patients [92.3 %], transperineal 7 of 8 patients [87.5 %] and transrectal 5 of 7 patients [71.4 %]). 25 patients (89.3 %) are free of complaints or describe an evident improvement of symptoms (transvaginal 12 of 13 patients [92.3 %], transperineal 7 of 8 patients [87.5 %] und transrectal 6 of 7 patients [85.7 %]). There is one postoperative dyspareunia. DISCUSSION: Best treatment of a rectocele starts with patients selection. Considering pelvic floor as functional unity, concomitant urologic-gynaecologic lesions and proximal intraabdominal disturbances the appropriate surgical procedure is selected. CONCLUSION: Surgical approach to correct a symptomatic rectocele depends on the concomitant lesion.

Stoma formation for fecal diversion: a plea for the laparoscopic approach.

BACKGROUND: The aim of this study was to assess the results of laparoscopic stoma creation for fecal diversion, specifically focussing on feasibility, safety, and efficacy, as well as indications and techniques. METHODS: Within a 10-year-period, all patients requiring laparoscopic stoma creation were evaluated prospectively. Patients' profiles and indications, procedures and results of operation, conversion, morbidity, mortality and short-term complications (stoma-related, laparoscopy-associated) were analyzed. RESULTS: A total of 80 patients (39 males, 41 females) with a mean age of 55.5 years (range, 17-91) underwent laparoscopic stoma creation. Most common indications were unresectable advanced colorectal cancer (n=20), pelvic malignant cancer (e. g. ovarian, cervix and prostate cancer, n=16), and perianal Crohn's disease with complex fistulas (n=16). Only in one female patient with pelvic malignant disease was the procedure converted to laparotomy due to obesity (conversion rate, 1.3%). 79 patients underwent laparoscopic stoma creation (completion rate, 98.7%) including loop ileostomy (n=30), loop sigmoid colostomy (n=40) and end sigmoid colostomy (n=9). Postoperative complications were documented in 9 patients (overall morbidity rate, 11.4%), including 4 minor complications treated conservatively (2 cases of prolonged atonia and 1 case each of pneumonia and urinary tract infection) and 5 major complications requiring reoperation (reoperation rate, 6.3%): one parastomal abscess (drainage), one stoma retraction following rod dislocation (laparoscopic stoma recreation), small bowel obstruction in two patients (small bowel resection), one port-site hernia (fascial closure), and hemorrhage (managed by re-laparoscopy). Mean operation time was 74 min (range, 30-245 min). Mean blood loss volume was 80 ml (range, 30-400 ml). Patients were discharged from hospital after a mean of 10.3 days (range, 3-47). Within a 1-year follow-up, no further stoma complications were documented. CONCLUSIONS: The advantages of laparoscopic stoma creation are low morbidity and reoperation rates, and no procedure-related mortality; our results suggest that laparoscopic stoma creation for fecal diversion is safe, feasible and effective. Therefore, at our institution, laparoscopic stoma creation is the method of choice for fecal diversion.

Rectal prolapse: which surgical option is appropriate?

Numerous surgical procedures have been suggested to treat rectal prolapse. In elderly and high-risk patients, perineal approaches such as Delorme's procedure and perineal rectosigmoidectomy (Altemeier's procedure) have been preferred, although the incidence of recurrence and the rate of persistent incontinence seem to be high when compared with transabdominal procedures. Functional results of transabdominal procedures, including mesh or suture rectopexy and resection-rectopexy, are thought to be associated with low recurrence rates and improved continence. Transabdominal procedures, however, usually imply rectal mobilization and fixation, colonic resection, or both, and some concern is voiced that morbidity, in terms of infection or leakage, and mortality could be increased. If we focus on surgical outcome, our own experience of laparoscopic resection-rectopexy for rectal prolapse shows that the laparoscopic approach is safe and effective, and functional results with respect to recurrence are favorable. However, the controversy "which operation is appropriate?" cannot be answered definitely, as a clear definition of rectal prolapse, the extent of a standardized diagnostic assessment, and the type of surgical procedure have not been identified in published series. Randomized trials are needed to improve the evidence with which the optimal surgical treatment of rectal prolapse can be defined.

Laparoscopic colectomy for diverticulitis is not associated with increased morbidity when compared with non-diverticular disease.

BACKGROUND AND AIMS: It was the aim of this prospective study to compare the outcome of laparoscopic sigmoid and anterior resection for diverticulitis and non-diverticular disease. PATIENTS AND METHODS: All patients who underwent laparoscopic colectomy for benign and malignant disease within a 10-year period were entered into the prospective PC database registry. For outcome analysis, patients who underwent laparoscopic sigmoid and anterior resection for diverticular disease were compared with patients who underwent the same operation for non-inflammatory (non-diverticular) disease. The parameters analyzed included age, gender, co-morbid conditions, diagnosis, procedure, duration of surgery, transfusion requirements, conversion, morbidity including major (requiring reoperation), minor (conservative treatment) and late-onset (postdischarge) complications, stay in the ICU, hospitalization, and mortality. For objective evaluation, only laparoscopically completed procedures were analyzed. Statistics included Student's t-test and chi-square analysis (p<0.05 was considered statistically significant). RESULTS: A total of 676 patients were evaluated including 363 with diverticular disease and 313 with non-inflammatory disease. There were no significant differences in conversion rates (6.6 vs. 7.3%, p>0.05), so that the laparoscopic completion rate was 93.4% (n=339) in the diverticulitis group and 92.7% (n=290) in the non-diverticulitis group. The two groups did not differ significantly in age or presence of co-morbid conditions (p>0.05). In the diverticulitis group, recurrent diverticulitis (58.4%), and complicated diverticulitis (27.7%) were the most common indications, whereas in the non-diverticulitis group, outlet obstruction by sigmoidoceles (30.0%) and cancer (32.4%) were the main indications. The most common procedure was laparoscopic sigmoid resection, followed by sigmoid resection with rectopexy and anterior resection. No significant differences were documented for major complications (7.4 vs. 7.9%), minor complications (11.5 vs. 14.5%), late-onset complications (3.0 vs. 3.5), reoperation (8.6 vs. 9.3%) or mortality (0.6 vs. 0.7%) between the two groups (p>0.05). In the postoperative course, no differences were noted in terms of stay in the ICU, postoperative ileus, parenteral analgesics, oral feeding, and length of hospitalization (p>0.05). CONCLUSION: These data indicate that laparoscopic sigmoid and anterior resection can be performed with acceptable morbidity and mortality for both diverticular disease and non-diverticular disease. The results show in particular that laparoscopic resection for inflammation is not associated with increased morbidity.

Influence of thymidylate synthase and p53 protein expression on clinical outcome in patients with colorectal cancer.

AIMS: Thymidylate synthase (TS) and tumor suppressor p53 are two proteins with an influence on tumor resistance to radio-chemotherapy that is well known. For this reason we tested the effect of TS and p53 expression on clinical outcome (tumor recurrence and survival) in patients after curative tumor resection, especially in patients who received adjuvant radio-chemotherapy. PATIENTS AND METHODS: A total of 120 patients with colorectal cancer were included in the study. A curative resection was possible in 83 patients, and 30 of this group received adjuvant therapy. For the immunohistochemical staining of tumor specimens, monoclonal antibody (mAb) TS 106 against TS and mAb DO-1 against p53 protein were used. TS positivity was defined as a moderate to high staining intensity in the cytoplasma of cells and p53 positivity as nuclear staining of tumor cells in >10% of these cells. RESULTS: Thymidylate synthase immunoreactivity was found in 59% of all cases and p53 staining in 51%. No relation between clinicopathological features and p53 expression was found in contrast to TS expression, where a highly significant association of TS-positive cases with tumor invasion (pT) was observed. Curatively resected patients with a TS-positive tumor developed tumor recurrence/distant metastases significantly more often than TS negative tumors. The same result was found when comparing p53-positive with p53-negative tumors and TS+/p53+ with TS-/p53- tumors. TS expression was highly significantly associated with poor survival and was the strongest independent prognostic factor in multivariate analysis, followed by lymph node status. CONCLUSION: Thymidylate synthase expression seems to be an independent prognostic factor and a possible predictor of tumor recurrence in patients with colorectal cancer.

[Typical and avoidable doctor's errors in surgical gastroenterology]. / Arztliche Erfahrung, Kenntnis der Leitlinien, interdisziplinäres Gespräch. Wie man Kunstfehler in der Gastroenterologie vermeidet.

Everyone makes mistakes, and even the greatest efforts may fail to eliminate them completely. In contrast, typical errors and miscalculations--which by the very fact of being typical are predictable--can and must be avoided. The reasons for such mistakes may be lack of experience, inadequate interdisciplinary cooperation, and ignorance of the current state of the art. This applies equally, and in particular, to the field of surgical gastroenterology. Intensive interdisciplinary discussions, a healthy surgical hierarchy, a knowledge of pertinent guidelines and the latest literature and, last, but not least--against the present background of discussions about working hour and diagnosis-related groups--a rested and motivated physician, are the pillars of error-free patient management.

Laparoscopic colectomy for recurrent and complicated diverticulitis: a prospective study of 396 patients.

BACKGROUND: It was the aim of this prospective study to evaluate the outcome of laparoscopic surgery for diverticular disease. METHODS: All patients who underwent elective laparoscopic colectomy for diverticular disease within a 10-year period were prospectively entered into a PC database registry. Indications for laparoscopic surgery were acute complicated diverticulitis (Hinchey stages I and IIa), chronically recurrent diverticulitis, sigmoid stenosis or outlet obstruction caused by chronic diverticulitis. Surgical procedures (sigmoid and anterior resection, left colectomy and resection rectopexy) included intracorporeal dissection and colorectal anastomosis. Parameters studied included age, gender, stage of disease, procedure, duration of surgery, intraoperative technical variables, transfusion requirements, conversion rate, total complication rate including major (requiring re-operation), minor (conservative treatment) and late-onset (post-discharge) complication rates, stay on ICU, hospitalisation, mortality, and recurrence. For objective evaluation, only laparoscopically completed procedures were analysed. Comparative outcome analysis was performed with respect to stage of disease and experience. RESULTS: A total of 396 patients underwent laparoscopic colectomy. Conversion rate was 6.8% ( n=27), so that laparoscopic completion rate was 93.2% ( n=369). Most common reasons for conversion were directly related to the inflammatory process, abscess or fistulas. The most common procedure was sigmoid resection ( n=279), followed by anterior resection ( n=36) and left colectomy ( n=29). Total complication rate was 18.4% ( n=68). Major complication rate was 7.6% ( n=28), whereas the most common complication requiring re-operation was haemorrhage in 3.3% ( n=12). Anastomotic leakage occurred in 1.6% ( n=6). Minor complications were noted in 10.7% ( n=40), late-onset complications occurred in 2.7% ( n=10). Mortality was 0.5% ( n=2). Mean duration of surgery was 193 (range 75-400) min, return to normal diet was completed after 6.8 (range 3-19) days. Mean hospital stay was 11.8 (range 4-71) days. No recurrence of diverticulitis occurred. CONCLUSION: Laparoscopic surgery for diverticular disease is safe, feasible and effective. Therefore, laparoscopic colectomy has replaced open resection as standard surgery for recurrent and complicated diverticulitis at our institution.

Laparoscopic colorectal surgery in obese and nonobese patients: do differences in body mass indices lead to different outcomes?

BACKGROUND: The aim of this prospective study was to compare the outcome of laparoscopic colorectal surgery in obese and nonobese patients. METHODS: All patients who underwent laparoscopic surgery for both benign and malignant disease within the past 5 years were entered into the prospective database registry. Body mass index (BMI; kg/m(2)) was used as the objective measure to indicate morbid obesity. Patients with a BMI >30 were defined as obese, and patients with a BMI <30 were defined as nonobese. The parameters analyzed included age, gender, comorbid conditions, diagnosis, procedure, duration of surgery, transfusion requirements, conversion rate, overall morbidity rate including major complications (requiring reoperation), minor complications (conservative treatment) and late-onset complications (postdischarge), stay on intensive case unit, hospitalization, and mortality. For objective evaluation, only laparoscopically completed procedures were analyzed. Statistics included Student's t test and chi-square analysis. Statistical significance was assessed at the 5% level (p < 0. 05 statistically significant). RESULTS: A total of 589 patients were evaluated, including 95 patients in the obese group and 494 patients in the nonobese group. There was no significant difference in conversion rate (7.3% in the obese group vs 9.5% in the nonobese group, p > 0.05) so that the laparoscopic completion rate was 90.5% (n = 86) in the obese and 92.7% (n = 458) in the nonobese group. The rate of females was significantly lower among obese patients (55.8% in the obese group vs 74.2% in the nonobese group, p = 0.001). No significant differences were observed with respect to age, diagnosis, procedure, duration of surgery, and transfusion requirements (p > 0.05). In terms of morbidity, there were no significant differences related to overall complication rates with respect to BMI (23.3% in the obese group vs 24.5% in the nonobese group, p > 0.05). Major complications were more common in the obese group without showing statistical significance (12.8% in the obese group vs 6.6% in the nonobese group, p = 0.078). Conversely, minor complications were more frequently documented in the nonobese group (8.1% in the obese group vs 15.5% in the nonobese group, p = 0.080). In the postoperative course, no differences were documented in terms of return of bowel function, duration of analgesics required, oral feeding, and length of hospitalization (p > 0.05). CONCLUSION: These data indicate that laparoscopic colorectal surgery is feasible and effective in both obese and nonobese patients. Obese patients who are thought to be at increased risk of postoperative morbidity have the similar benefit of laparoscopic surgery as nonobese patients with colorectal disease.

[Surgical therapy of rectal carcinoma]. / Chirurgische Therapie des Rektumkarzinoms.

Total mesorectal excision (TME) has gained a revolutionary impact on the surgical therapy of rectal cancer within the last 2 decades, providing superior local tumor control in comparison to conventional resection. Consequently, 85% of rectal carcinomas can be resected by sphincter-preserving surgery without compromising either oncologic radicality or continence. With the introduction of TME, local recurrence rates have been reliably decreased below 10% after curative resection. Surgical dissection along the connective tissue space between rectal and parietal pelvic fascia with complete mesorectal excision results in reliable excision of all relevant lymphatic pathways with preservation of continence and sexual function. Complete removal of a TME specimen is mandatory in carcinomas of the middle and lower third of the rectum. Both removal of the complete TME specimen and careful pathologic examination of the circumferential resection margin have decisive significance. An additional pelvic lymphadenectomy with the potential risk of increased morbidity does not improve prognosis. As a spread of tumor distally along the bowel wall rarely exceeds a few centimeters, a distal resection margin of 1-2 cm is oncologically sufficient in sphincter-saving procedures without compromising prognosis. Taken together, the convincing results of TME provide a rationale for using TME as the dissection policy of choice to resect rectal cancers in the distal two-thirds of the rectum, despite the absence of direct evidence from prospective randomized trials. The question whether laparoscopic curative resection for rectal cancer is oncologically adequate cannot be definitely answered to date, as results of randomized studies are currently missing. However, the preliminary results of laparoscopic resection for rectal cancer provided by centers are promising.

Pouch reconstruction in the pelvis.

ILEAL POUCH RECONSTRUCTION: Proctocolectomy with ileal pouch-anal anastomosis (IPAA) is the procedure of choice in mucosal ulcerative colitis (MUC) and familial adenomatous polyposis (FAP). Because the disease is cured by surgical resection, functional results, pouch survival prognosis, and disease or dysplasia control are the major determinants of success. There is controversy as to whether the IPAA should be handsewn with mucosectomy or stapled, preserving the mucosa of the anal transitional zone. Crohn's disease is a contraindication for IPAA, but long-term outcome after IPAA is similar to that for MUC in patients with indeterminate colitis who do not develop Crohn's disease. As development of dysplasia and cancer in the ileal pouch have been reported, a standardized surveillance program is mandatory in cases of MUC, FAP, and chronic pouchitis. COLONIC POUCH RECONSTRUCTION: Construction of a colonic pouch is a widely accepted technique to improve functional outcome after low or intersphincteric resection for rectal cancer. Several randomized studies comparing colo-pouch-anal anastomosis (CPA) with straight coloanal anastomosis (CAA) have found the pouch functionally superior. Most controlled studies cover only 1-year follow-up, but randomized studies with 2-year follow-up show similar functional results of CPA and CAA. Evacuation difficulty as initially observed was related to pouch size, and the results with smaller pouches (5-6 cm) are more favorable, showing adequate reservoir function without compromising neorectal evacuation. The transverse coloplasty pouch may offer several advantages to J-pouch reconstruction. Current series question whether the neorectal reservoir is the physiological key of the pouch, but rather the decreased motility. The major advantage reported with colonic pouch reconstruction is the lower incidence of anastomotic complications.

Intraoperative suspicion of gallbladder carcinoma in laparoscopic surgery: what to do?

The frequency of gallbladder cancer in Europe is less than 1% of all gallstone operations. With the introduction of laparoscopic surgery and the higher acceptance of this technique, patients with gallstones have gallbladder removal much earlier in their gallstone history. So the percentage of gallbladder carcinomas will decrease in the future. We report on our surgical procedures in patients with suspicious gallbladders having laparoscopic gallbladder removal, and how to proceed after the diagnosis of gallbladder carcinoma. From June 1990 to December 2001, we have performed 7,130 cholecystectomies in a single department. 47 of these patients (0.66%) were identified as having carcinoma. There were 40 females and 7 males, with a mean age of 70.6 years. In 17 cases (36%) there was a preoperative suspicion of malignancy. Most commonly, in 30 cases (64%), malignancy was suspected intraoperatively or diagnosed postoperatively after pathological examination of the resected gallbladder. We recommend removal with a bag for all gallbladders with a suspected wall or scleroatrophic calcified gallbladder area. In stage Tis or T1 laparoscopy + cholecystectomy is sufficient. For T2 and T3 we perform reoperation with liver bed resection and lymphadenectomy.

[Laparoscopic cholecystectomy--surgical standard in cholelithiasis]. / Laparoskopische Cholecystektomie--Chirurgischer Standard bei Cholelithiasis.

Laparoscopic surgery showed a dramatic development in the last years of the 20th century. From the beginning laparoscopic cholecystectomy (LCCE) has been the pacemaker of this development. Today laparoscopic cholecystectomy is the first choice for treatment of cholecystolithiasis in nearly all surgical clinics. Therefore laparoscopic cholecystectomy is the most common part of minimal invasive technique. LCCE is the golden standard in therapy of gallstones, more than 90% of cholecystectomies in specialized clinics are done laparoscopically. It is an established, evidence based operation today. Open cholecystectomy is left for special indications only. A problem of LCCE is the occult carcinoma of the gallbladder. In histological proven carcinoma of the gallbladder LCCE is the adequate operation only for Tis and T1 carcinoma. In T2 and T3 carcinoma a radical oncologic resection with lymph node dissection should be performed. Due to the poor prognosis T4 tumors should be left with laparoscopic biopsy only.

Urinary excretion of thromboxane and markers for renal injury in patients undergoing cardiopulmonary bypass.

Urinary excretion of selected markers for renal injury, as well as urinary excretion rates of the thromboxane metabolite, 11-keto-thromboxane B2 (11k-TXB2), was studied in 36 male patients undergoing coronary bypass surgery using cardiopulmonary bypass (CPB). In all patients, excretion of both tubular (N-acetyl-beta-D-glucosaminidase [beta NAG]; alpha 1-microglobulin [alpha 1-MG]) and glomerular markers (albumin [Alb]; transferrin [Trf]; immunoglobulin G [IgG]) sharply increased on Day 1 after CPB, and they remained elevated throughout the observation period of 5 days. Urinary excretion rates of 11k-TXB2 markedly increased on Day 1 after surgery, and they rapidly decreased thereafter. In 12 of the 36 patients, a temporary increase of serum creatinine levels (> 1.30 mg/dl) was noted following surgery. A positive correlation was found between serum creatinine levels and excretion of the tubular enzyme beta NAG (r = 0.36; p < 0.05), but not between creatinine levels and alpha 1-MG or the glomerular markers. Furthermore, no correlation between urinary excretion of 11k-TXB2 and any of the urinary markers for renal injury could be detected. Our data do not strengthen the hypothesis that acute renal injury observed during CPB is related to exaggerated thromboxane biosynthesis in these patients. Monitoring of urinary markers for incipient renal damage, particularly excretion of beta NAG, might be of additional diagnostic value for detection of otherwise subclinical renal injury in patients undergoing CPB.

Thromboxane B2 urinary metabolites in patients undergoing cardiopulmonary bypass.

The urinary excretion of selected markers for renal injury and thromboxane metabolites was studied in 16 patients undergoing cardiopulmonary bypass (CPB). Excretion of both tubular and glomerular markers sharply increased on day 1 after CPB and remained elevated throughout the observation period (five days). Immunoreactive thromboxane B2 (i-TXB2, mainly reflecting 2,3-dinor-TXB2) and immunoreactive 11-keto-thromboxane B2 (i-11-keto-TXB2) were measured by direct enzyme immunoassays. TXB2, 2,3-dinor-TXB2 and 11-keto-TXB2 were also measured in selected samples by GC-MS. Urinary excretion rates of both i-TXB2 and i-11-keto-TXB2 markedly increased on day 1 after surgery and decreased thereafter. Following CPB, excretion rates of 2,3-dinor-TXB2 and TXB2 displayed parallel changes, suggesting that in these patients most urinary TXB2 derives from blood platelets rather than the kidney. Taken together, our observations do not support the hypothesis that acute renal injury observed after CPB is caused by exaggerated thromboxane biosynthesis in the kidney.