Hepatic fibrogenic activity in chronic alcoholic pancreatitis.

The prevalence and the mechanisms of hepatic fibrosis in chronic alcoholic pancreatitis remain uncertain. The aim of this study was to investigate the fibrogenic activity of the liver in patients with chronic pancreatitis and its relation with either the alcohol or cholestasis. Liver biopsies were obtained from 16 patients with chronic pancreatitis at the time of surgery and from 10 organ donors. Samples were processed for histologic examination to assess the presence and extent of fibrosis, inflammatory reactions, and cholestasis- and alcohol-related lesions. In other samples, the collagen content was measured by morphometry, and prolylhydroxylase activity was determined. Liver-function tests, ultrasonography, and endoscopic retrograde cholangiopancreatography were performed before surgery in all the patients. Of patients with chronic pancreatitis, 75% had significantly greater hepatic fibrosis and prolylhydroxylase activity than the control group. Moreover, prolylhydroxylase activity in patients with chronic pancreatitis was higher in those with cholestasis or partial obstruction of the common bile duct than in those without cholestasis or partial obstruction of the common bile duct. Both the fibrogenic activity and the collagen content in the livers of patients with chronic alcoholic pancreatitis are significantly increased, even in those without histologic lesions, and these alterations may be secondary to a partial occlusion of the common bile duct.

Prospective evaluation of a minimally invasive approach for treatment of bile-duct calculi in the high-risk patient.

BACKGROUND: The best approach to bile duct stones in high-risk patients is controversial. We showed in a randomized trial that open surgery had a morbi-mortality similar to that of endoscopic sphincterotomy alone (ES) and less late biliary complications. The aim of this study was to evaluate a minimally invasive approach to duct stones in high-risk patients compared with open surgery or ES alone. METHODS: Sixty high-risk patients (mean age 80 years) suspected of duct stones were treated by ES + laparoscopic cholecystectomy (LC). High-risk factors were: age > 70 years, Goldman cardiac index > 13, chronic pulmonary disease, liver cirrhosis, neurologic deficit, and severe obesity. RESULTS: ERCP success was 87%. Duct stones were found in 75%. LC succeeded in 92%. Post-LC stay was 4 days. Overall morbidity was 19% and mortality was 3%. Recurrent symptoms (mean follow-up: 9 months) was 3.6%. When compared with open surgery or ES alone, ES + LC had a similar morbi-mortality, but shorter postop stay (p < 0.001). Late symptoms appeared in 20% after ES alone vs 4% after open surgery or ES plus LC (p < 0.04). CONCLUSIONS: Combined ES + LC is an effective alternative to open surgery or ES alone for treatment of duct stones in high-risk patients.

Apolipoprotein E polymorphism and gallstones.

BACKGROUND & AIMS: Apolipoprotein (apo) E is a genetically polymorphic protein influencing lipoprotein metabolism and the risk of both atherosclerosis and Alzheimer's disease. As opposed to common apo E3, apo E2 decreases and apo E4 increases hepatic lipoprotein uptake; hence, apo E4 could promote gallstone formation by increasing hepatic and biliary cholesterol concentrations. This study was designed to evaluate whether apo E polymorphism is related to gallstone risk. METHODS: apo E phenotype was determined in subjects older than 40 years of age (160 with and 125 without gallstones) and in 61 patients with cholesterol gallstones who underwent cholecystectomy. Bile composition, nucleation time, and gallstone features were analyzed in surgical patients. RESULTS: The E4/3 phenotype was enriched in both patients with gallstones and those who underwent cholecystectomy, with significantly (P < 0.006) higher epsilon 4 allele frequencies than in gallstone-free subjects (odds ratio, 2.67 [95% confidence limits, 1.23-5.93] and 3.62 [95% confidence limits, 1.49-8.91], respectively); women, but not men, accounted for these differences. The prevalence of the epsilon 4 allele increased with age in patients with gallstones, whereas the opposite occurred in gallstone-free subjects. Biliary lipid and gallstone cholesterol content tended to increase in the sequence E4 > E3 > E2 in patients who underwent cholecystectomy. CONCLUSIONS: Carrying the apo E4 isoform is a genetic risk factor for cholelithiasis in humans, thus adding another adverse effect of apo E polymorphism on health.

Laparoscopic treatment of acute biliary pancreatitis.

UNLABELLED: Laparoscopic cholecystectomy (LC) has become the standard treatment of gallstones. Application of LC in certain complications of biliary stones such as acute biliary pancreatitis (ABP) is not well defined. 10-30% of patients with ABP present associated bile duct stones, and the realization of a preoperative ERCP has been routinely proposed. Nevertheless, this examination may be unnecessary in most patients. AIM: To investigate the applicability of laparoscopic surgery for treatment of ABP. MATERIALS AND METHODS: Between Jan-1992 and June-1995, 368 patients were prospectively evaluated for LC, 274 for indications other than ABP. (Group I, LC) and 91 as a consequence of ABP. (Group II, ABPxL). ERCP was indicated when ultrasonography showed a dilated bile duct (> 8 mm) or when the liver function test (LFT) presented high scores. Age, sex, operative time, incidence of bile duct stones, postoperative stay and morbimortality were evaluated. RESULTS: The two groups were well matched for age, sex and associated medical risk factors. There were no differences in the operative time, conversion rate or postoperative morbidity (10% vs 10%). ERCP was performed in 25 patients in Group II and bile duct stones were found in 12 cases. In all cases an intraoperative cholangiography was performed, and in 6 patients, bile duct stones were removed by laparoscopic means. Three patients were converted to open surgery on finding duct stones which could not be treated by laparoscopic means. Mean postoperative stay was significantly longer in Group II than in Group I. In two cases, pancreatic pseudocyst was attempted with a laparoscopic approach. CONCLUSIONS: Definitive treatment of ABP could be accomplished effectively by laparoscopy, with selective indication of ERCP.

Effects of stimulation and inhibition of acid secretion on gastric damage induced by platelet-activating factor.

Intravenous administration of platelet-activating factor (PAF) induces extensive damage in rat gastric mucosa. The aim of the present study was to examine whether the presence or absence of acid in the gastric lumen could modify the PAF-induced gastric damage. The effects of inhibition of basal and pentagastrin-stimulated acid secretion by ranitidine on the deep histological gastric damage induced by 30 min of infusion of PAF (100 ng/kg/min) were assessed by using a histological score. Inhibition of basal gastric acid secretion did not prevent the histological gastric damage induced by PAF. Stimulation of gastric acid secretion by pentagastrin significantly increased PAF-induced gastric damage, and this effect was reversed by a dose of ranitidine that returns acid secretion to baseline levels. This acid-related damage was confined to the deep mucosa, since scanning electron microscope analysis ruled out an additional surface damage in PAF-infused rats when gastric acid was stimulated. The data indicate that a certain amount of acid may worsen the deep gastric mucosal damage induced by PAF.

Propranolol in prevention of recurrent bleeding from severe portal hypertensive gastropathy in cirrhosis.

The two main causes of gastrointestinal bleeding in cirrhosis are oesophageal varices and portal hypertensive gastropathy (PHG). Rebleeding from varices can be prevented by beta-blockers, but it is not clear whether these drugs effectively reduce rebleeding from PHG. 54 cirrhotic patients with acute or chronic bleeding from severe PHG took part in a randomised, controlled trial to investigate the efficacy of propranolol in prevention of rebleeding from PHG. 26 patients were randomised to receive propranolol daily at a dose that reduced the resting heart rate by 25% or to 55 bpm (20-160 mg twice daily), throughout mean follow-up of 21 (SD 11) months. 28 untreated controls were followed-up, with the same examinations, for 18 (13) months. The actuarial percentages of patients free of rebleeding from PHG were significantly higher in the propranolol-treated patients than in the untreated controls at 12 months (65% vs 38%; p less than 0.05) and at 30 months of follow-up (52% vs 7%; p less than 0.05). Propranolol-treated patients had fewer episodes of acute bleeding than controls (0.010 [0.004] vs 0.120 [0.040] per patient per month). Multivariate analysis showed that absence of propranolol treatment was the only predictive variable for rebleeding. Actuarial survival was slightly higher in the propranolol group than in the controls, but the difference was not significant. Thus, long-term propranolol treatment significantly reduces the frequency of rebleeding from severe PHG, and may improve the prognosis of cirrhotic patients with this disorder.

Effects of propranolol on gastric microcirculation and acid secretion in portal hypertensive rats.

Rats with chronic portal hypertension have an increased gastric mucosal blood flow and an impaired acid secretory response to pentagastrin stimulation. This study investigated the effects of propranolol, a drug widely used in the treatment of portal hypertension, on gastric mucosal blood flow and acid secretion in portal hypertensive rats with partial portal vein occlusion. In study I, gastric mucosal blood flow, measured by hydrogen gas clearance, and acid output were determined in basal conditions and after the administration of propranolol (0.3 mg/kg/min) or vehicle infusion in portal hypertensive and sham-operated rats. Gastric mucosal blood flow was significantly higher (p less than 0.005) in portal hypertensive rats than in sham-operated rats, and it was significantly reduced (p less than 0.05) after propranolol infusion in portal hypertensive rats (39 +/- 2 vs. 35 +/- 3 ml/min/100 gm). Basal acid output was not significantly modified with propranolol infusion, either in portal hypertensive or sham-operated animals. In study II, acid output was measured in portal hypertensive rats in basal conditions, during pentagastrin (20 micrograms/kg/hr) infusion and after the combined infusion of pentagastrin and propranolol (or vehicle). Pentagastrin infusion significantly increased (p less than 0.001) acid output, but propranolol did not modify stimulated gastric acid secretion. In conclusion, propranolol induced a slight but significant reduction in gastric mucosal blood flow and did not change basal or stimulated gastric acid secretion.

Increased gastric PGE2 biosynthesis in cirrhotic patients with gastric vascular ectasia.

Plasma levels of glucagon, secretin, norepinephrine, arginine-vasopressin, and prostaglandin biosynthesis in the gastric mucosa were determined in cirrhotic patients with gastric vascular ectasia associated with hypoacidity, in cirrhotics without this lesion, and in healthy controls. Plasma concentrations of glucagon, secretin, and norepinephrine were similar in cirrhotics with gastric vascular ectasia and cirrhotics without this lesion, these concentrations being significantly higher (p less than 0.05) than in healthy controls. However, there was no significant difference between plasma levels of arginine-vasopressin in patients with cirrhosis (with or without gastric vascular ectasia) and those in healthy controls. The biosynthesis of prostaglandin E2 in the antrum of the gastric mucosa was significantly higher in cirrhotics with gastric vascular ectasia than in cirrhotics without this lesion (p less than 0.05) and healthy controls (p less than 0.005). Prostaglandin E2 in the corpus was significantly higher (p less than 0.05) in cirrhotics with gastric vascular ectasia than in healthy controls. The biosynthesis of 6-keto PGF1 alpha (a stable metabolite of prostacyclin) and PGF2 alpha in the corpus and antrum of gastric mucosa was not significantly different in cirrhotics with gastric vascular ectasia, cirrhotics without this lesion and healthy controls. Since prostaglandin E2 has a vasodilator and acid-inhibitory effect, we speculate that high content of this prostanoid in the gastric mucosa may play a role in the pathogenesis of ectatic capillaries and acid inhibition present in some cirrhotic patients.

Gastric vascular ectasias in cirrhosis: association with hypoacidity not related to gastric atrophy.

To investigate whether hypergastrinemia and low serum levels of pepsinogen I are associated with gastric hypoacidity in cirrhosis with capillary ectasia of gastric mucosa and whether this alteration is secondary to the presence of atrophic gastritis, two groups of patients were studied: 1) 12 cirrhotic patients with diffuse gastric red spots at the endoscopic examination, and 2) 12 cirrhotic patients with endoscopically normal mucosa. Vascular ectasia of the gastric mucosa was histologically confirmed in all patients with gastric red spots. The study of base-line and stimulated acid gastric secretion showed that 9 of 12 (75%) cirrhotics with gastric vascular ectasia had achlorhydria and that 8 of these 9 patients had high base-line gastrin serum levels (greater than 130 pg/ml) and low base-line pepsinogen I serum levels (less than 20 ng/ml). Base-line gastrin and pepsinogen I serum levels were significantly greater and lower, respectively, in patients with gastric vascular ectasias than in cirrhotics without these lesions. None of the patients of either group had complete atrophy in the corpus of the stomach, and only 4 of the 9 cirrhotics with gastric vascular ectasia and achlorhydria had moderate atrophy. These results show that achlorhydria is frequently associated with hypergastrinemia and low pepsinogen I serum levels in patients with cirrhosis and gastric vascular ectasias and suggest that this disturbance is not secondary to a morphologic abnormality of the gastric mucosa.

Temporal relationship between the impairment of free water excretion and antidiuretic hormone hypersecretion in rats with experimental cirrhosis.

To investigate the temporal relationship between the impairment of water excretion, sodium retention, and antidiuretic hormone hypersecretion in cirrhosis, free water excretion (estimated by the minimum urinary osmolality) and urinary antidiuretic hormone excretion (which correlates with the plasma levels of this hormone) were measured weekly after an oral water load in 18 rats with carbon tetrachloride-induced cirrhosis and in 20 control animals. The onset of ascites (as an index of sodium retention) in cirrhotic rats was estimated by sequential paracentesis. Thirteen cirrhotic animals developed an impairment of water excretion 2-5 wk after the onset of ascites. The urinary excretion of antidiuretic hormone in these animals, which was normal before the impairment of water excretion, increased markedly within the week in which this abnormality was first detected and remained high thereafter. The remaining 5 cirrhotic rats did not experience an impairment of free water excretion in spite of developing ascites. The urinary excretion of antidiuretic hormone in these animals was similar to that of control rats during the entire study. In all urine samples obtained from cirrhotic rats, there was a highly significant direct linear correlation between the urinary excretion of antidiuretic hormone and the minimum urinary osmolality. Our results show the following: in rats with carbon tetrachloride-induced cirrhosis, sodium retention preceded the impairment of water excretion; and in these animals, the defect in water metabolism correlated chronologically and quantitatively with antidiuretic hormone hypersecretion. These findings are consistent with the concept that antidiuretic hormone is a major determinant of the impaired water metabolism in cirrhosis.

[Technics for the study of the renal excretion of free water in normal rats and those with experimental liver cirrhosis]. / Técnicas para el estudio de la excreción renal de agua libre en ratas normales y con cirrosis hepática experimental.

Different techniques to measure free water excretion in rats, administered an oral water overload and with measurement of its ability to excrete it into the urine have been studied. When 30 or 50 ml/kg b.wt. were administered and the urine excreted in 3 h was collected, a decrease on the urinary osmolality (UOSM) was observed with respect to the baseline UOSM, which was similar in both overloads, although the percentage of the overload excreted was significantly greater with 50 ml/kg. However, the UOSM obtained was hypertonic as compared to plasma osmolality (POSM) indicating that this determination was not useful to study free water excretion. In a further study it was investigated if there was any period of time in which all the animals excreted hypotonic urine. However, results indicated that the period for excreting a maximally diluted urine was very variable in time. The best technique to study free water excretion in these animals was the collection of each spontaneously voided urine independently, to measure the minimal UOSM. When a 50 ml/kg water load was administered and the minimal UOSM was determined it was observed to be lower than POSM in all the animals indicating that this technique was useful to study this derangement in these animals.

Urinary excretion of 6-keto-prostaglandin F1 alpha, thromboxane B2 and prostaglandin E2 in cirrhosis with ascites. Relationship to functional renal failure (hepatorenal syndrome).

The aim of the study was to investigate the urinary excretion of 6-keto-PGF1 alpha (a stable metabolite of PGI2), thromboxane B2 (TxB2; a stable metabolite of TxA2), and PGE2 in 18 normal subjects, 49 cirrhotics with ascites without renal failure (GFR = 90 +/- 4 ml/min, means +/- S.E.M.) and 20 cirrhotics with functional renal failure (FRF) (GFR = 36 +/- 3). The study was made after 5 days on a 50 mEq sodium diet and without diuretics. Plasma renin activity (PRA), plasma norepinephrine concentration (NE) and plasma antidiuretic hormone concentration (ADH) were also measured. Cirrhotics without FRF showed a significantly higher urinary excretion of 6-keto-PGF1 alpha, TxB2 and PGE, (15.9 +/- 1.7 ng/h, 3.0 +/- 0.3 ng/h, and 6.2 +/- 1.0 ng/h) than did normal subjects (9.2 +/- 0.9, 1.3 +/- 0.1 and 2.3 +/- 0.4). On the contrary, the urinary excretion of these prostaglandins was normal or reduced in patients with FRF (5.3 +/- 0.8, 1.3 +/- 0.2 and 1.9 +/- 0.4). PRA, NE and ADH were significantly increased in cirrhotics with FRF (15.2 +/- 3.9 ng/ml/h, 1026 +/- 149 pg/ml and 4.1 +/- 0.3 pg/ml) and in patients without FRF (8.0 +/- 1.4, 667 +/- 67 and 3.9 +/- 0.3) as compared to normal controls (1.3 +/- 0.2, 275 +/- 46 and 2.4 +/- 0.2). These results suggest that renal hemodynamics in cirrhosis depends upon a critical equilibrium between the activity of endogenous vasoconstrictor systems and the renal production of the vasodilator prostaglandins PGI2 and PGE2. In addition, they do not support FRF in cirrhosis being related to an increased renal production of the vasoconstrictor prostaglandin TxA2.

Cefotaxime is more effective than is ampicillin-tobramycin in cirrhotics with severe infections.

We compared the effectiveness and incidence of nephrotoxicity of ampicillin-tobramycin and cefotaxime in 73 cirrhotics who had severe bacterial infection. Most of these patients had spontaneous peritonitis and/or bacteremia. Patients were randomly allocated into two groups. Group I included 36 patients treated with ampicillin-tobramycin and Group II comprised 37 patients treated with cefotaxime. Patients from both groups were similar with respect to clinical data, standard liver and renal function tests, types of infection and isolated organisms. Ninety-two per cent of bacteria isolated in Group I and 98% of those isolated in Group II were susceptible in vitro to ampicillin-tobramycin and to cefotaxime, respectively. Ampicillin-tobramycin cured the infection in 56% of Group I patients, and cefotaxime in 85% of Group II patients (p less than 0.02). Five patients treated with ampicillin-tobramycin, and none treated with cefotaxime developed superinfections (p = 0.024). Nephrotoxicity (impairment of renal function associated with an increase of urinary beta 2-microglobulin to over 2,000 micrograms per liter) occurred in two patients in Group I and none in Group II. These results suggest that broad-spectrum cephalosporins should be considered as first choice antibiotics in cirrhotic patients with severe infections.

Oral, nonabsorbable antibiotics prevent infection in cirrhotics with gastrointestinal hemorrhage.

To investigate if oral, nonabsorbable antibiotics prevent bacterial infections in cirrhotics with gastrointestinal hemorrhage, 140 consecutive patients were randomly allocated into two groups: 68 patients (Group I) were given oral, non absorbable antibiotics (gentamicin + vancomycin + nystatin or neomycin + colistin + nystatin) from the inclusion into the trial up to 48 hr after cessation of the hemorrhage, or until emergency surgery or death in those cases who continued bleeding; and 72 patients (Group II) did not receive oral, nonabsorbable antibiotics. Both groups were similar in relation to clinical and laboratory data and characteristics of the hemorrhage. The incidence of infection was significantly lower in Group I than in Group II (11 patients in Group I and 25 in Group II developed proved infections; p less than 0.025). This difference was due to the fact that spontaneous bacteremia and peritonitis and urinary tract infection caused by enteric bacteria occurred almost exclusively in Group II. Two patients of Group I and 10 of Group II developed spontaneous bacteremia and/or peritonitis caused by enteric bacteria (p less than 0.025). These results indicate that prophylactic administration of oral, nonabsorbable antibiotics markedly reduces the incidence of infections caused by enteric bacteria in cirrhotic patients with gastrointestinal hemorrhage.

Evidence that renal prostaglandins are involved in renal water metabolism in cirrhosis.

Plasma antidiuretic hormone (ADH) and urinary prostaglandin E2 excretion (UPGE2V) were measured in basal conditions, after water restriction, and after water-loading in 10 normal subjects (free water clearance after the water load, CH2O, 9.6 +/- 0.8 ml/min) and in 27 patients with cirrhosis and ascites (13 with a positive CH2O: 3.6 +/- 0.5; 14 with a negative CH2O: -0.37 +/- 0.007). Plasma ADH and UPGE2V were significantly increased in patients with a positive CH2O as compared with normal subjects. Patients with a negative CH2O showed a significantly higher plasma ADH and a lower UPGE2V and GFR than did normal subjects and patients with the positive CH2O. In 18 additional subjects (6 normal and 12 with cirrhosis, ascites, and a positive CH2O) submitted to a sustained water overload, the i.v. administration of 450 mg of lysine acetylsalicylate (LAS) induced a marked reduction of UPGE2V, but it had no effect on plasma ADH. LAS did not alter GFR and CH2O in normal subjects; however, it reduced CH2O in all the 12 patients (from 5.1 +/- 0.4 to 0.6 +/- 0.3) and the GFR in only 6 of these patients. These results suggest (a) that renal PGE2 plays an important role in the maintenance of water excretion in cirrhosis with ascites, and (b) that impaired ability to dilute the urine in cirrhosis may be a consequence of the simultaneous occurrence of impaired renal hemodynamics, nonostomic hypersecretion of ADH, and reduced renal production of PGE2.

Renal kallikrein excretion in cirrhotics with ascites: relationship to renal hemodynamics.

To investigate if the renal kallikrein-kinin system may be involved in the homeostasis of renal perfusion in cirrhosis, urinary kallikrein activity was measured in 11 normal subjects, 31 cirrhotics with ascites and preserved renal plasma flow (548.2 +/- 32.2 ml per min) and glomerular filtration rate (85.8 +/- 3.4 ml per min), and 18 cirrhotics with functional renal failure (renal plasma flow: 229.9 +/- 23.4 ml per min; glomerular filtration rate: 34.9 +/- 3.3 ml per min). Plasma renin activity, plasma norepinephrine concentration and the urinary excretion of prostaglandin E2 were also measured in these subjects. Cirrhotics without renal failure showed a significantly higher renin (4.9 +/- 1.1 ng per ml per hr), norepinephrine (458.2 +/- 50.4 pg per ml), urinary kallikrein (15.4 +/- 1.8 pkat per min) and urinary prostaglandin E2 (0.52 +/- 0.08 ng per min) than did normal subjects (1.08 +/- 0.1 ng per ml per hr, 218.1 +/- 18.2 pg per ml; 8.4 +/- 1.4 pkat per min and 0.24 +/- 0.02 ng per min, respectively). Cirrhotics with renal failure showed a significantly higher renin (16.1 +/- 3.4 ng per ml per hr) and norepinephrine (739.4 +/- 79.2 pg per ml), and a significantly lower urinary kallikrein (5.2 +/- 0.6 pkat per min) and urinary prostaglandin E2 (0.15 +/- 0.02 ng per min) than did normal subjects and cirrhotics without renal failure. Glomerular filtration rate correlated (p less than 0.001) with urinary kallikrein (r = 0.53), urinary prostaglandin E2 (r = 0.55), plasma renin (r = -0.41) and norepinephrine (r = -0.44).(ABSTRACT TRUNCATED AT 250 WORDS)

Effect of demeclocycline on renal function and urinary prostaglandin E2 and kallikrein in hyponatremic cirrhotics.

8 cirrhotics with hyponatremia were given demeclocycline (DMC) 900 mg/day to investigate its effect on renal function, plasma renin activity, aldosterone and urinary excretion of prostaglandin E2 and kallikrein. In 7 patients DMC induced an increase of free water clearance (from -0.36 +/- 0.06 to 0.13 +/- 0.06 ml/min) and serum sodium concentration (from 125.4 +/- 0.09 to 131.1 +/- 1.0 mEq/l, mmol/l). In 5 of these patients DMC also induced a marked reduction of glomerular filtration rate (from 72.2 +/- 6.2 to 31,2 +/- 4.7 ml/min) and renal plasma flow (from 468 +/- 98 to 195 +/- 55 ml/min) which could not be explained on the basis of hypovolemia. In each case this renal impairment was not associated with changes in urinary concentration of beta 2-microglobulin, urinary casts excretion, fresh urine sediment or urine protein content and disappeared after discontinuation of the drug. DMC induced a marked increase in the urinary excretion of prostaglandin E2 (from 0.82 +/- 0.27 to 6.16 +/- 1.91 ng/min) in 6 out of the 7 patients who responded to DMC and a marked reduction in urinary kallikrein (from 16.1 +/- 4.4 to 4.2 +/- 1.6 pkat/min) in the 5 patients who developed renal insufficiency. The serum DMC concentration was greater than 5 micrograms/ml in all patients who responded to DMC, greater than 8 micrograms/ml in all cases who developed renal insufficiency and of 3 micrograms/ml in the case not responding to DMC. (ABSTRACT TRUNCATED AT 250 WORDS)