College of American Pathologists
Printable Version

  Clinical Abstracts





cap today

September 2001

Acute respiratory distress syndrome as a manifestation of thrombotic thrombocytopenic purpura
Thrombotic thrombocytopenic purpura classically presents with thrombocytopenia and microangiopathic hemolytic anemia and some combination of neurologic abnormalities, renal dysfunction, and fever. Other forms of organ system dysfunction—for example, cardiac or hepatic—also may be a part of TTP, but respiratory dysfunction is not generally recognized as a major manifestation. Fifty-six cases of TTP presented at the authors' institution between 1981 and 1998. For seven (12.5 percent) of these cases, acute respiratory distress syndrome was a major clinical manifestation of TTP (age range, 43 to 85 years; all female). In four patients, ARDS developed as a surgical complication, and three patients presented to the hospital with respiratory distress. The diagnosis of ARDS was based on the guidelines of the American-European Consensus Conference on ARDS. Six patients required ventilatory support. TTP was not the initial consideration in any of the seven patients. As soon as TTP was diagnosed, the patients were treated with plasma exchange (six patients) or fresh frozen plasma infusion (one patient). Four patients showed significant clinical improvement in their ARDS within 48 hours, and these four patients eventually achieved complete remission of TTP and ARDS. Three patients died because of delays in recognizing TTP and initiating plasma exchange therapy. The four patients who responded to therapy have remained in complete remission during more than three years of followup. The authors concluded that ARDS may occur as a clinical manifestation of TTP and that all patients with ARDS should be evaluated for unrecognized TTP.

Chang JC, El-Sayed MA. Acute respiratory distress syndrome as a major clinical manifestation of thrombotic thrombocytopenic purpura. Am J Med Sci. 2001;321:124-128.

Correspondence: Jae C. Chang, MD, Good Samaritan Hospital, 2222 Philadelphia Drive, Dayton, OH 45406;

Point-of-care lactate testing in the intensive care unit
Blood lactic acid levels, which normally do not exceed 1.0 mmol/L, may be significantly elevated in patients with critical illness. Blood lactate levels have shown clinical usefulness for the prognosis of critical illness and assessing the effects of therapeutic maneuvers. There are two types of lactic acidosis: type A lactic acidosis, in which the finding is the result of poor tissue perfusion, and type B lactic acidosis, in which there is no clinical evidence of poor perfusion. Several different mechanisms may account for increased blood lactate. While blood lactate has been reported to be a good correlate of clinical outcomes, the reliability of a single value as a marker of adequate resuscitation or as a predictor of patient outcome has been questioned. Sequential determinations of lactate apparently have higher predictive value than single lactate measurements. The authors prospectively studied the reliability and costs of two different point-of-care lactate testing devices and compared them with central laboratory lactate determinations. They studied 40 patients from a surgical ICU in Germany and included only patients who had a blood lactate concentration of more than 2.0 mmol/L and for whom the measurement was performed in the central laboratory. Patients receiving lactated ringers solution as an IV were excluded from the study population. The authors performed 197 lactate measurements with two systems—the battery-powered Accusport handheld lactate analyzer and the benchtop blood gas analyzer Chiron 865 series. Both measurements were compared with the lactate measured by the hospital central laboratory, which was considered the reference method. Measurements were performed on the day of inclusion in the study, which was considered to be baseline, eight hours later, and in the morning of the first, second, and third days thereafter. The time elapsed from blood sampling to availability of data was significantly longer for the central laboratory—85 minutes on average—than for both POC systems, which ranged from one to 10 minutes each. Excellent agreement was noted between the lactate measurements from arterial blood using the Accusport method and the central laboratory, with a relative error of -2.74 percent and a bias of -0.15 mmol/L. The correlation between the blood gas measurements using the blood gas analyzer also agreed well with the reference method bias: 0.09 mmol/L and relative error of 2.24 percent. The total cost for measuring blood lactate was lower using the Accusport device than using the blood gas analyzer. The authors concluded that blood lactate could be rapidly, easily, accurately, and economically measured at the bedside using the Accusport analyzer as well as benchtop blood gas analyzers.

Boldt J, Kumle B, Suttner S, et al. Point-of-care (POC) testing of lactate in the intensive care patient. Acta Anaesthesiol Scand. 2001;45:194-199.

Reprints: Dr. Joachim Boldt, Dept. of Anesthesiology and Intensive Care Medicine, Klinikum der Stadt Ludwigshafen, Bremserstr. 79 D-67063 Ludwigshafen, Germany

CMV in bronchoalveolar lavage fluid
Infection with cytomegalovirus is, in many cases, associated with increased morbidity and mortality, particularly in lung allograft recipients, for whom it is associated with chronic rejection. Rapid diagnosis of CMV infection is of great clinical relevance, and the most commonly used methods for diagnosis are direct-antigen detection and molecular techniques, the most promising of which is polymerase-chain reaction. A positive CMV PCR test in transbronchial lung biopsy specimens has been shown to precede manifest CMV disease by two weeks in lung transplant recipients. Likewise, quantitative PCR assessments of CMV DNA load in urine and blood samples from kidney transplant recipients have shown promise in diagnosing patients at risk of infection. A recent study has shown that quantitative PCR analysis is capable of accurately measuring the amount of CMV load in tissue and bronchoalveolar lavage specimens in lung transplant recipients. The authors undertook a longitudinal evaluation to determine whether CMV DNA quantities in BAL fluid were associated with clinically manifest CMV disease. They studied 35 consecutive lung transplant recipients at a university hospital in Sweden. Sixteen of the patients were studied for two years, seven for 18 months, and 10 patients for 12 months after surgery. Two patients died within the seven months after surgery, one of multi-organ failure and the other of malignant lymphoma. A total of 340 protocol and diagnostic bronchoscopies were performed, with an average of eight episodes per patient within the first postoperative year. Quantitative CMV DNA levels were determined in these 340 BAL samples. Seventeen of the patients developed CMV disease with pneumonitis, and 27 CMV episodes were diagnosed. The patients with CMV disease had a significantly higher mean level of CMV copies per milliliter of BAL fluid (around 1,100) compared with those without (180 copies on average). Viral load as well as acute rejection requiring treatment were independent risk factors associated with CMV disease. No significant differences were found between the groups with regard to HLA typing, basic immunosuppressive therapy, and CMV serologic status. Using the mean level of all episodes without CMV disease could discriminate only nine of the 27 CMV episodes. The authors concluded that viral load is increased during periods of episodic CMV disease in lung transplant patients, but the quantitative PCR assessment of CMV DNA in BAL fluid is not discriminative enough to be useful as a diagnostic tool for CMV disease.

Riise GC, Anderson R, Berström T, et al. Quantification of cytomegalovirus DNA in BAL fluid: a longitudinal study in lung transplant recipients. Chest. 2000;118:1653-1660.

Reprints: Gerdt C. Riise, MD, PhD, Dept. of Respiratory Medicine, Sahlgrenska University Hospital, SE-413 45 Göteborg, Sweden

Benign causes of an elevated serum CA-125 concentration
CA-125 performs with low sensitivity as a screening test for ovarian cancer. It is better suited to confirming diagnosis, monitoring response to therapy, and monitoring recurrence. CA-125 is associated with normal and neoplastic derivatives of the coelomic epithelium. It is produced by normal endometrium and decidua and is present in normal ovarian epithelia and that of the colon, lung, pancreas, kidney, and gall bladder. CA-125 is increased during menstruation, in the first trimester of pregnancy, and in endometriosis. It also may be seen in inflammation of the peritoneal and pleural epithelia, pleural effusions, and ascites. Large increases in CA-125 may be seen in benign cysts or fibromata when these are associated with ascites or rupture of the tumor. A clinical chemistry group in England undertook a study, via a retrospective audit, to establish the occurrence of elevated CA-125 values in benign conditions during a 42-month period that went from May 1996 through October 1999. During this period, 19 study patients who ranged in age from 21 to 84 years (median age, 52) showed increased serum concentrations of CA-125 without evidence of ovarian cancer. The levels of the analyte in these patients ranged from 39 to 1,464 kU/L (median value, 119 kU/L). Six of the patients had nonmalignant gynecological disease, six had abdominal disorders, and seven had lung disease. The gynecological disorders included endometriosis, uterine fibroid, salpingitis, and pelvic abscess. The abdominal conditions included ascites due to bowel obstruction and edema, appendicular abscess, tuberculous peritonitis, urinary tract infection, and ascites due to cirrhosis. The lung conditions included pleural effusion, pulmonary embolism, pneumonia, and Dressler's syndrome. The author concluded that one should be cautious when attributing elevated CA-125 as an indicator of ovarian malignancy.

Buamah P. Benign conditions associated with raised serum CA-125 concentration. J Surg Oncol. 2000;75: 264-265.

Reprints: P.K. Buamah, MD, Unit of Cancer Studies, Dept. of Clinical Biochemistry, Queen Elizabeth the Queen Mother Hospital, Margate, Kent CT9 4AN, England

Endothelin 1 in cystic fibrosis
Endothelin 1 is produced mainly by cells of the endothelium and by such cells as the epithelial cells of the airways. It has autocrine and paracrine functions, however, it also may function as a hormone. ET1 causes vasospasm and promotes bronchospasm, proliferation of smooth muscle fibers, and inflammation of the airways. The authors examined ET1 levels in the context of cystic fibrosis. They studied 31 cystic fibrosis patients who were a mean age of 12 years and 28 control subjects who were a mean age of 10.6 years. The diagnosis in all cases was made on the basis of clinical symptoms and sweat chloride levels. The patients were prospectively selected and classified into groups according to the severity of their pulmonary disease. Group A comprised 16 patients who were an average age of 13 years and had impaired lung function. Group B consisted of 15 patients who were an average age of 11.2 years and had unimpaired lung function. Group C was made up of 28 healthy control subjects. Criteria for this classification included the grade of finger-clubbing, Brasfield chest radiograph score, spirometric values, and arterial blood PO2 values. Plasma levels of ET1 and atrial natriuretic peptide were performed by radioimmunoassay. The authors also measured renin, serum aldosterone, and serum and urine electrolytes. The plasma ET1 levels were significantly higher among those in group A than among those in groups B and C; the levels did not differ between groups B and C. Median levels were 3.2 pg/mL in group A, 2.0 pg/mL in group B, and 2.5 pg/mL in group C. These levels correlated positively with the severity of finger-clubbing, heart rate, arterial blood PCO2, plasma IR-ET levels, and serum aldosterone levels. It correlated negatively with arterial blood PO2, forced vital capacity, forced expiratory volume at one second, and Brasfield chest radiograph score. PO2 was the only independent factor found to affect plasma endothelin levels in multivariate regression analysis. The authors concluded that plasma ET1 levels are increased in cystic fibrosis and related to the severity of impairment of pulmonary function in that disease.

Siahanidou T, Nikolaidou P, Doudounakis S, et al. Plasma immunoreactive endothelin levels in children with cystic fibrosis. Acta Paediatr. 2000;89:915-920.

Reprints: T. Siahanidou, First Dept. of Paediatrics, Athens University, Aghia Sophia Children's Hospital, GR-115 27, Athens, Greece

Limited clinical value of folate measurement in evaluating macrocytosis
Macrocytosis with or without anemia is a common finding in clinical practice, and folate levels are often obtained as part of clinical and laboratory evaluations. The clinical utility of serum and erythrocyte folate levels has been questioned. This study retrospectively reviewed all folate measurements at three hospitals over one year. Medical records and laboratory determinations were studied. The clinicians' responses were sought out in the medical record, and a cost analysis was performed. A total of 2,998 folate levels (84.2 percent serum; 15.8 percent erythrocyte) were measured, and 68 (2.3 percent) were low. Of the 68 patients, 66 (97 percent) had charts available for review. Thirty-one (47 percent) had macrocytosis without anemia and 35 (53 percent) were anemic (22 had macrocytic anemia, 10 were normocytic, and three were microcytic). Cobalamin levels were measured in all patients and were low or borderline-low in five patients with macrocytic anemia. Of the 66 low folate levels, only 35 (53 percent) were noted in the medical record, and it was noted that 16 of these patients received folic acid. In the other 19 patients (all of whom were inpatients), folate supplementation was not given. The authors calculated that $89,814 was spent on folate testing, which changed physician behavior in nine patients (a cost of $9,979 per change in management). Empiric folate supplements in lieu of testing for all patients would have cost a maximum of $6,914 and would have saved a minimum of $82,900 in direct charges. The authors concluded that folate testing has limited value and seldom changes physician behavior. In cases of macrocytosis, empirical supplementation with folic acid should be used in lieu of testing for deficiency.

Robinson AR, Mladenovic J. Lack of clinical utility of folate levels in the evaluation of macrocytosis or anemia. Am J Med. 2001;110:88-90

Correspondence: Andrew R. Robinson, MD, 1719 E. 19th Ave., 5C East, Denver, CO 80218

Reliability of pediatric urinalysis in young febrile children
A number of young children initially seen with "fever of unknown origin" are found to have urinary tract infection upon further workup. Reports in the literature indicate that standard urinalysis has relatively low sensitivity for detecting UTIs in young febrile infants. Urine culture typically is obtained in younger febrile infants who do not have an obvious source of infection, and urinalysis is used in older infants in the same situation. Nonetheless, proper urine culture in this age group requires bladder catheterization; thus one should know which patients may be adequately screened using a "bag" urinalysis. A study was undertaken to measure the sensitivity of standard urinalysis for detecting urinary tract infection to determine if sensitivity varies by age of patient and to make specific recommendations as to when a urine culture and not just a screening urinalysis should be obtained. The study was a retrospective medical chart review of 37,450 febrile children younger than two years who were seen at Children's Hospital, Boston, Mass. Forty-four percent of the patients were girls. Median age of the study subjects was 10.6 months, and median temperature on examination was 38.8°C. Urine cultures were obtained for 11,089 (30 percent) patients. The urinalysis on these patients was considered positive if any one of the following was positive: leukocyte esterase, nitrite, or pyuria of 5 blood cells per high-powered field or greater. The patients who had paired urinalysis and urine culture were used to calculate the sensitivity, specificity, and likelihood ratios of the urinalysis. The sensitivity of the urinalysis in this context was 82 percent and did not vary by age subgroups; specificity was 92 percent. The likelihood ratio for a positive urinalysis was 10.6, and the likelihood ratio for a negative urinalysis was 0.19. The prevalence of urinary tract infection in the patients examined was 2.1 percent overall (2.9 percent for girls; 1.5 percent for boys). When broken down by ethnic group, the prevalence of urinary tract infection in girls was 5.0 percent in white patients, 2.1 percent in Hispanic patients, and 1.0 percent in black patients. Among boys, the prevalence was 2.2 percent in Hispanic patients, 1.4 percent in white patients, and 0.8 in black patients. Higher prevalence was also seen among patients with a temperature of 39.0°C or higher. A 13 percent prevalence of urinary tract infection was found among white girls younger than six months who had a temperature of 39°C or greater. From the calculation of post-test probability of UTI, it was found that when the prevalence of UTI is 2.0 percent, one urinalysis among 250 will produce a false-negative test result.

Bachur R, Harper MB. Reliability of the urinalysis for predicting urinary tract infections in young febrile children. Arch Pediatr Adolesc Med. 2001;155:60-65.

Reprints: Richard Bachur, MD, Division of Emergency Medicine, Children's Hospital, 300 Longwood Ave., Boston, MA 02115;

Acquired activated protein C resistance in cancer patients with venous thromboembolism
Activated protein C resistance is associated with the factor V Leiden genotype, while acquired activated protein C resistance occurs in the absence of this mutation. The authors performed a prospective, controlled study to evaluate the prevalence of inherited and acquired activated protein C resistance in cancer patients with and without venous thromboembolism. Group one (55 patients) had cancer and thromboembolism, and group two (58 patients) had cancer without thromboembolism. Two groups of control patients also were studied. One control group (54 patients) had thromboembolism without cancer, and the other control group (56 patients) consisted of apparently healthy volunteers without a clinical history of venous thromboembolism or malignancy. The mean values of the activated protein C sensitivity ratio (the ratio between the activated partial thromboplastin times with and without activated protein C) were significantly lower in the groups with cancer and the control group with thromboembolism as compared to the healthy controls. Cancer patients with venous thromboembolism had lower activated protein C sensitivity ratios than did cancer patients without thromboembolism (P=0.03 after excluding those patients with the factor V Leiden mutation). Thirty-three percent of the patients who had venous thromboembolism but no cancer had activated protein C resistance due to the factor V Leiden mutation, compared with about four percent of the other patients. In contrast, acquired activated protein C resistance was found in 54 percent of the cancer patients with thromboembolism, 17 percent of the cancer patients without thromboembolism, and 19 percent of the patients with thromboembolism who did not have cancer. The authors concluded that acquired activated protein C resistance is a common finding in cancer patients with venous thromboembolism and that the presence of the factor V Leiden mutation in these patients is unusual.

Haim N, Lanir N, Hoffman R, et al. Acquired activated protein C resistance is common in cancer patients and is associated with venous thromboembolism. Am J Med. 2001;110:91-96.

Correspondence: N. Haim, MD, Rambam Medical Center, P.O.B. 9602, Haifa 31096, Israel

Estrogen and LDL oxidation
Estrogen probably exerts a beneficial influence on lipid metabolism, and postmenopausal women typically show increased levels of low-density lipoprotein, lipoprotein(a), and total cholesterol, as well as decreased high-density lipoprotein, especially HDL2. Potentially protective effects of estrogen include inhibition of endothelial hyperplasia, reduced arterial impedance, enhanced production of prostacyclin, and increased insulin sensitivity. Furthermore, estrogen has an anti-oxidant function. Free radical-mediated LDL oxidation is recognized to be an atherogenic condition because these molecules are rapidly taken up by scavenger receptors on macrophages, leading to the formation of foam cells. A study was undertaken to investigate the effect of physiological levels of female hormone estrogen on in vitro LDL oxidation. The comparison was between healthy pre- and postmenopausal women. Vitamin E, a major lipid soluble anti-oxidant in LDL, was also measured. A commonly employed copper ion- dependent method was used to determine the oxidizability of LDL in 20 premenopausal women who were a mean age of 27 years and in 23 postmenopausal women who were a mean age of 51 years. The mean serum estradiol level was 576 pmol/L in the premenopausal group and 64 pmol/L in the postmenopausal group. Oxidation of LDL in the two groups did not differ significantly by measuring the lag phase of conjugated dienes formation or the generation of thiobarbituric acid reactive substances over four hours of oxidation. Likewise, vitamin E levels were similar in the two groups. The authors concluded that physiological levels of estrogen might not affect in vitro LDL oxidation.

Wen Y, Doyle MCT, Cooke T, et al. Effect of menopause on low-density lipoprotein oxidation: Is estrogen an important determinant? Maturitas. 2000;24:233-238.

Reprints: Yichuan Wen, School of Health Sciences, University of Wolverhampton, 62-68, Lichfield St., Wolverhampton, WV1 1DJ, United Kingdom;