Alberto Maringhini

Biliary Sludge and Gallstones in Pregnancy: Incidence, Risk Factors, and Natural History

Cholesterol cholelithiasis is more common in women than in men [1-5]. This difference begins at puberty and continues throughout the childbearing years. Most of the epidemiologic studies have shown a positive association between the development of gallstones and previous pregnancies [3, 4, 6-8], but this association has not been confirmed in other studies [9-13]. Biliary sludge, a well-known ultrasonographic diagnosis, is a mixture of granules of calcium bilirubinate and cholesterol crystals within viscous bile [14, 15]. Sludge is frequently found in the gallbladders of women during pregnancy and after delivery [16, 17]. It usually disappears during the first months after delivery [17, 18]. To clarify the influence of pregnancy on gallbladder disease, we did a prospective study to determine the incidence of and the risk factors for biliary sludge and gallstones in pregnant women. We also assessed the natural history of these conditions in the first year after delivery. Methods Study Sample From 1986 to 1988, we enrolled all women (n = 272) admitted to the obstetrics department of our hospital who were in the first trimester of pregnancy. All participants volunteered and gave written informed consent. Six women were excluded: three because of a previous cholecystectomy and three because of technically inadequate ultrasonograms. All women had clinical interviews and ultrasonographic examinations of the gallbladder at entry (month 3 of pregnancy), later during pregnancy (month 6), and 2 to 4 weeks after delivery. We examined the following factors as possible predictors for the development of new sludge and new stones: age, previous use of oral contraceptives, previous pregnancy, number of previous deliveries, and percentage of body weight excess. The presence of sludge at the first-trimester examination (old sludge) was also considered as a predictor for the development of new gallstones. Age, breast-feeding, previous deliveries, and the diameter of stones found after delivery were considered as predictors for the disappearance of biliary sludge or gallstones. Body weight excess was defined as the percentage by which body weight exceeded ideal body weight [19]. Because of the high prevalence of -thalassemia in Sicily and its association with gallstones [20], we evaluated women for the presence of heterozygous -thalassemia by measuring hemoglobin A2 concentrations. These measurements were done using column chromatography. During follow-up, we assessed patients for vomiting, itching, biliary pain (colic pain in epigastrium or upper right abdomen that lasted more than 60 minutes and was not relieved by bowel movement), and dyspepsia (nausea, upper abdominal discomfort, and pyrosis). All clinical data were obtained by the examining physician before the ultrasonographic examination was done. All women with biliary sludge or gallstones had follow-up after delivery. A clinical interview and ultrasonographic examination were done every 3 months for 1 year or until an empty gallbladder was found on two consecutive ultrasonographic examinations. During the study, no patient received medication to dissolve gallstones. Ultrasonographic Examinations The ultrasonographic examinations were done using a high-resolution real-time scanner with a 3.5-MHz linear array transducer. Two of us (AM and AO) served as examiners. After an overnight fast, patients were examined in the supine position, in the oblique position right side up, during the change from one position to another, and in the standing position. All patients were on an unrestricted diet. An ultrasonic diagnosis of sludge was made when we found diffuse, low-amplitude echoes forming a fluid-fluid level [21]. Biliary sludge was characterized by homogeneous echoes or heterogeneous echoes of 2 to 5 mm with nonshadowing echogenic foci [22-24]. When heterogeneous echoes were present, sludge balls could not be differentiated from those gallstones [23] that do not completely deflect the ultrasonographic beam [25]. Consequently, the diagnosis of gallstones was made when shadowing, gravity-dependent, echogenic structures with a diameter of more than 2 mm were observed in the gallbladder [26]. We excluded slice-thickness artifact echoes, which can mimic sludge, using the criteria of Goldstein and Madrazo [27]. Before the study, the examiners had more than 5 years of experience with diagnostic ultrasonography and both had done more than 5000 examinations. For the first 100 women who enrolled in the study, the overall agreement between the observers was 97% and 100% for the diagnosis of sludge and stones, respectively. The value for chance-corrected agreement in the diagnosis of sludge was 91% (95% CI, 81% to 99%). Statistical Analysis The test for linear trend in proportions [28] was used to determine whether the frequency of variables at entry steadily increases or decreases as patients move from the best (empty gallbladder) to the worst (gallstones) ultrasonographic classes. When appropriate, the t-test for continuous variables and the chi-square test for categorical variables were used to determine the statistical significance of differences. Crude rates for events are given with 95% CIs, which are reported according to Simon [29]. The cumulative incidence rates of biliary sludge and gallstones in pregnancy were calculated using an actuarial life-table method and are given with 95% CIs [30]. Because biliary sludge or gallstones were still present in the immediate postpartum period, we assumed that the examination at this time indicated the status of the gallbladder at the end of pregnancy. Actuarial analysis was also applied to the incidence rates for disappearance of sludge and stones after delivery. A stepwise logistic regression analysis was done to identify the variables associated with the development of new sludge and new stones during pregnancy or with their disappearance after delivery, after adjusting for the effect of covariates [31]. Results Findings in the First Trimester We enrolled 272 women in the study during their first trimester of pregnancy (mean age SD, 27 5 years). Biliary sludge was found in 42 women (15%; CI, 11% to 20%) and gallstones were found in 17 women (6%; CI, 3% to 9%). Four women with gallstones also had biliary sludge. Eighty-one women (30%) had an excess body weight of more than 20%, 121 (44%) had had previous pregnancies, and 103 (38%) had used oral contraceptives. Only the presence of two or more deliveries before admission was associated with higher prevalence of biliary sludge or gallstones on the ultrasound examination done in the first trimester (Table 1). Table 1. Clinical Characteristics of 272 Pregnant Women at the Time of the First Ultrasonographic Examination* Findings Ultrasonographic Events and Symptoms At the end of the second trimester the actuarial incidence rates for new sludge and new stones were 14% (CI, 9% to 18%) and 2% (CI, 0.1% to 3%), respectively. Between the first-trimester examination to the control examination done 2 to 4 weeks after delivery, new sludge and new stones developed in 67 (29.6%) and 6 (2.4%) women, respectively. Thirty-four women (12.5%) were lost to follow-up (26 with empty gallbladders and 8 with sludge) (Table 2). The actuarial cumulative incidence rate was 31% (CI, 25% to 37%) for new sludge and 2% (CI, 0.2% to 4%) for new stones. Table 2. Ultrasonographic Gallbladder Findings in 272 Women Who Were Followed during Pregnancy and in the First Year after Delivery Vomiting (32%) and dyspepsia (28%) were the most frequent symptoms. Twenty-eight women had itching (10%). No relation between vomiting, dyspepsia or itching and biliary sludge or gallstones was found. Eight patients (3%) experienced biliary pain. These patients had no relapses of pain during pregnancy, and no treatment was necessary. Biliary pain was significantly more frequent among women with gallstones (5 of 17) than among women with biliary sludge (2 of 42) or an empty gallbladder (1 of 213) (P < 0.0001). Biliary pain did not occur in patients with new stones or new sludge. Vomiting occurred more often during the first trimester (95%) than during the second (6%) or the third (2%), but dyspepsia was more prevalent during the third trimester (70%). No difference was observed in the incidence of itching and biliary pain in the different trimesters. Predictors of Sludge and Gallstones Women with new sludge were less obese than women without sludge (mean body weight excess [SD], 12.4% 7.2% compared with 17.3% 11.2%; P = 0.016). This finding was confirmed by stepwise logistic regression analysis; the detection rate for new sludge decreased by 2% for each additional percentage point of body weight excess (CI, 1% to 5%). The number of patients who definitely formed gallstones during pregnancy was too small to assess predictive factors. However, the incidence of stones in women with formerly empty gallbladders (2 of 213 participants, 0.9%) was lower than in women with sludge (4 of 42 participants, 9.5%) (P = 0.0005). Findings in the Postpartum Period After delivery, 115 women were studied (92 with biliary sludge, 11 with gallstones and 12 with both sludge and gallstones) (see Table 2). Gallstones disappeared in seven women [30%, 3 women with stones only and 4 with both stones and sludge] after a mean follow-up of 9.7 4.6 months (range, 1 to 19 months). At this time, the rate of disappearance was 28% (CI, 10% to 46%). Four of these seven patients had old stones (stones detected at entry), whereas the other three women had new stones. The rate of sludge disappearance was higher. After a mean follow-up of 5.0 4.6 months (range, 1 to 23 months), sludge disappeared in 71 of 104 women (68.3%): in 65 of 92 with sludge only (70.7%) and in 6 of 12 with both stones and sludge (50%). At this time, the actuarial rate of disappearance was 61% (CI, 50% to 73%). Ten patients with sludge were lost to follow-up. We failed to find any relation between the time at whic

Percutaneous blind biopsy versus laparoscopy with guided biopsy in diagnosis of cirrhosis

A prospective controlled study of the diagnostic accuracy of blind percutaneous liver biopsy in comparison to laparoscopy plus guided biopsy for the recognition or exclusion of cirrhosis has been performed. One hundred twenty-six patients with a clinical diagnosis of chronic, diffuse, well-compensated liver disease were randomized into two groups and submitted either to percutaneous blind liver biopsy (PB: 64 patients) or to laparoscopy with guided biopsy (LB: 62 patients), in order to assess the accuracy of either procedure in diagnosing cirrhosis. PB correctly recognized or ruled out cirrhosis in 52 patients (82%). Inconclusive results were mostly false negative, as demonstrated by the presence on endoscopy of esophageal varices or by subsequent LB. LB demonstrated presence or absence of cirrhosis in all patients. The difference between the rate of accurate results of the two procedures is statistically significant. It is concluded that in patients without esophageal varices, LB should be the investigation of choice for the assessment of liver structure since the presence of cirrhosis can be missed in up to 20% of cases by PB.

Pseudocysts in Acute Nonalcoholic Pancreatitis (Incidence and Natural History)

Epidemiological studies on pancreaticpseudocysts are retrospective analyses on alcoholicpatients. The aims of this study were to investigate theincidence, natural history, and predictors of theappearance and disappearance of pancreatic fluidcollections and pseudocysts after nonalcoholic acutepancreatitis. We carried out a prospective cohort studyin a series of 926 patients with acute pancreatitis.Pancreatic fluid collections or pseudocysts were treatedonly after complications. We studied pancreatic fluidcollections from 83 patients (8.9%): 48 of whomdeveloped pseudocysts (5.1%). Both were less frequent after biliary pancreatitis (P < 0.0001). Inthe first 60 days of follow-up, patients with fluidcollections or pseudocysts showed more complicationsthan spontaneous disappearance; two of them died. After the 60th day, spontaneous disappearancewas more frequent, and at one year the cumulativeincidence of complications and spontaneous disappearancewas 36% and 56%, respectively. A total of 33 patients with fluid collection needed interventionaltreatment (surgery or percutaneous or endoscopicdrainage). Pseudocysts that were small (<5 cm) ordeveloped in the tail had a higher incidence ofspontaneous disappearance: 22/24 (91.7%) and 11/12 (91.7%),respectively. In conclusion, fluid collections andpseudocysts after nonalcoholic pancreatitis have a lowincidence of complications and mortality with a high rate of spontaneous disappearance. Wesuggest treating them only aftercomplications.

Asymptomatic hepatocellular carcinoma in Child's A cirrhosis

The present study deals with the natural history of 37 asymptomatic patients with cirrhosis and hepatocellular carcinoma, 25 with 2-9-cm tumors who were not surgically treated (first group) and 12 with tumors smaller than 4 cm who underwent resection (second group). All patients were in Childs A class. Two-year survival (according to life-table analysis by the Kaplan-Meier method) was 50% in the first group and 39% in the second group. This difference was not significant. In the first group no relation was found between survival and initial tumor size or alpha-fetoprotein levels. Ultrasound examinations at 3-mo intervals revealed the following patterns of tumor growth: (a) no significant growth during the follow-up (9 patients); (b) significant growth (tumor size at least doubling) only in the final stage of the disease (11 patients); (c) initial significant growth followed by a period of no increase in size (5 patients). These findings show that in our geographical area (a) 2-yr survival of untreated asymptomatic patients with hepatocellular carcinoma associated with cirrhosis does not differ from that of similar patients undergoing resection and (b) the tumor can exhibit long periods of no growth alternating with periods of exponential growth.

Ultrasonography and alpha-fetoprotein in diagnosis of hepatocellular carcinoma in cirrhosis.

The accuracy of ultrasound (US) and alpha-fetoprotein (AFP) in the diagnosis of hepatocellular carcinoma (HCC) in 363 patients with cirrhosis (C) and a clinical suspicion of HCC was assessed. The ultrasonographic patterns of HCC and their relationship with AFP values were analyzed. Echographic patterns were distributed as follows: 47 patients had sonodense lesions; 30 patients had hypoechoic lesions; 47 had mixed-pattern lesions, and in four patients focal dilated intrahepatic bile ducts were demonstrated. The sensitivity of US was 90%; specificity was 93.3%. Serum AFP level ⩾500 ng/ml (RIA) was the first clue to the diagnosis in 71 patients (48.6%); specificity was 100%. In 28 patients AFP levels became significantly elevated during follow-up after US detection of HCC. No relationship between echo pattern and serum AFP levels was demonstrated. An algorithm for diagnosis of HCC is proposed.

Sludge and stones in gallbladder after pregnancy: Prevalence and risk factors

The prevalence of sludge and stones in the gallbladder of 298 women in the immediate post-partum period was ultrasonographically assessed. We have investigated some risk factors for the development of sludge or stones in these patients and followed up most of these patients by ultrasonography to detect the presence of sludge and/or stones in the year following their discovery. We found sludge in 80 (26.2%) and gallstones in 16 (5.2%) of these patients. Age, obesity and months of oral contraceptive use were risk factors only for the presence of gallstones. After 1 year of follow-up only 2 out of 45 patients with sludge but 13 out of 15 patients with gallstones still had abnormal ultrasonographic findings.

Ascites, pleural, and pericardial effusions in acute pancreatitis. A prospective study of incidence, natural history, and prognostic role.

Ascites and pleural and pericardial effusions can be observed during acute pancreatitis. The aims of this study were to evaluate their incidence, natural history, and prognostic role in patients with acute pancreatitis. One hundred patients consecutively admitted with a diagnosis of acute pancreatitis were prospectively submitted to abdominal, pleural, and cardiac ultrasonography at admission and during follow-up. Ascites was found in 18 patients, pleural effusion in 20, and pericardial effusion in 17. Twenty-four patients of this series had severe pancreatitis; three of them died. All effusions disappeared spontaneously in patients who survived pancreatitis up to two months after dismissal. At multivariate analysis ascites and pleural effusion were demonstrated to be accurate independent predictors of severity. The respective odds ratios were 5.9 [95% confidence interval (CI), 1.5–23.0%) and 8.6 (95% CI, 2.3–32.5%). Furthermore the presence of pleural effusion, ascites, and pericardial effusion were associated with an increased incidence of pseudocyst during follow-up. Ascites and pleural and pericardial effusions are frequent during acute pancreatitis. Pleural effusion and ascites are accurate predictors of severity in these patients.

Clinical presentation and ultrasonography in the diagnosis of pancreatic cancer.

One thousand twenty patients consecutively admitted because of a clinical suspicion of pancreatic cancer were investigated to evaluate the accuracy of simple clinical, laboratory, and ultrasonographic data in the diagnosis of pancreatic cancer. Age, weight loss, recent-onset diabetes mellitus, palpable abdominal mass or gallbladder, elevated serum bilirubin or alkaline phosphatase levels, and ultrasonography were significant criteria in discriminating 80 pancreatic cancers from 940 controls. The most sensitive criteria were ultrasonography (83%). weight loss (66%), and bilirubin level of >3 mg/dl(61%); the most specific were ultrasonography (99%), recent-onset diabetes (97%), and a distended palpable gallbladder (94%). Only ultrasonography demonstrated an elevated positive predictive value (86%), while weight loss, elevated bilirubin and alkaline phosphatase, besides ultrasonography had an elevated negative predictive value (95%). These results show that advanced pancreatic cancer may be excluded with simple clinical and laboratory data; ultrasonography can confirm the diagnosis with a high degree of accuracy. We suggest that the results of any new diagnostic tests for pancreatic cancer be compared with these clinical findings.

Treatment of Small Hepatocellular Carcinoma Associated with Cirrhosis by Percutaneous Ethanol Injection: A Trial with a Comparison Group

BACKGROUND Ethanol injection has been reported to be effective in the treatment of hepatocellular carcinoma, but no controlled randomized trials have been performed. We therefore performed a trial comparing ethanol injection with an untreated, matched historical comparison group in the treatment of hepatocellular carcinoma. METHODS From 1992 to 1993, 35 patients (14 Childs A and 21 Childs B cirrhosis) with small (< 4 cm) hepatocellular carcinoma associated with cirrhosis were treated by ethanol injection. Each patient was matched with an untreated case (followed up during the period 1984-89) for variables known to have independent prognostic value (age, Childs classification, number of lesions, alpha-fetoprotein, and modality of diagnosis). RESULTS The 1-, 2-, and 3-year survival rates of ethanol-treated patients were 86% (95% confidence interval (CI), 69-94), 53% (95% CI, 34-68), and 33% (95% CI, 15-52), whereas the survival rates of the comparison group were 75% (95% CI, 56-85), 26% (95% CI, 13-41), and 14% (95% CI, 5-27) (P = 0.01). The 1-, 2-, and 3-year survival rates of Childs A were 100%. 87% (95% CI, 30-97), 71% (95 CI, 33-90), 71% (95% CI, 33-90) in the ethanol-treated patients and 92 (95% CI, 59-99), 43% (95% CI, 23-73), and 21% (95% CI, 23-72) in untreated patients. The 1-, 2-, and 3-year survival of Childs B patients were 76% (95% CI, 59-97), 32% (95% CI, 13-53), and 9% (95% CI, 0.8-33) in the treated group and 61% (95% CI, 40-83), 14% (95% CI, 3-32), and 9% (95% CI, 1-26) in the treated group. CONCLUSIONS These data suggest that ethanol injection prolongs the life of patients with hepatocellular carcinoma associated with Childs A cirrhosis but seems not to influence the survival of Childs B patients.

Ultrasonographic and radionuclide detection of hepatocellular carcinoma in cirrhotics with low alpha-fetoprotein levels

A total of 67 cirrhotic patients with clinically suspected neoplastic degeneration and low alpha‐fetoprotein levels were assessed prospectively with ultrasound and gold (198Au) scintigraphy. Ultrasound showed space‐occupying lesions in 22 of the 24 patients who had a final diagnosis of hepatocellular carcinoma (HCC) (sensitivity, 95.8%) and excluded the presence of HCC in 37 of the 43 patients with cirrhosis only (specificity, 86.0%; efficiency, 90.8%). Scintigraphy demonstrated a cold defect in 22 of the 24 patients who had a final diagnosis of HCC (sensitivity, 95.8%) and excluded the presence of HCC in 22 of the 43 patients with cirrhosis only (specificity, 51.1%; efficiency, 69.8%). It was concluded that the most accurate screening plain in cirrhotic patients suspected of having HCC with alpha‐fetoprotein values below 500 mg/ml would consist of ultrasonography followed, as clinically indicated, by ultrasonographic or laparoscopic guided biopsy.