Clinical Chemistry 49, No. 2, 2003
3. Clarke R, Stansbie D. Assessment of homocysteine as a cardiovascular risk factor in clinical practice. Ann Clin Biochem 2001;38:624 –32. 4. Andersson A, Isaksson A, Hultberg B. Homocysteine export from erythrocytes and its implication for plasma sampling. Clin Chem 1992;38:1311–5. 5. Willems HPJ, Bos GMJ, Gerrits WBJ, den Heijer M, Vloet S, Blom HJ. Acidic citrate stabilizes blood samples for assay of total homocysteine. Clin Chem 1998;44:343–5. 6. Palmer-Toy DE, Szczepiorkowski ZM, Shih V, Van Cott EM. Compatibility of the Abbott IMx homocysteine assay with citrate-anticoagulated plasma and stability of homocysteine in citrated whole blood. Clin Chem 2001;47: 1704 –7. 7. Al-Khafaji F, Bowron A, Day A, Scott J, Stansbie D. Stabilization of blood homocysteine by 3-deazaadenosine. Ann Clin Biochem 1998;35:780 –2. 8. Ubbink J, Vermaak H, van der Merwe A, Becker P. The effect of blood sample aging and food consumption on plasma total homocysteine levels. Clin Chim Acta 1992;207:119 –28. 9. Moller J, Rasmussen K. Homocysteine in plasma: stabilization of blood samples with fluoride. Clin Chem 1995;41:758 –9. 10. Hughes P, Carlson TH, McLaughin Kathleen, Bankson DD. Addition of sodium fluoride to whole blood does not stabilize plasma homocysteine but produces dilution effect on plasma constituents and hematocrit. Clin Chem 1998;44:2204 – 6. 11. Nauck M, Bisse E, Nauck M, Wieland H. Pre-analytical conditions affecting the determination of the plasma homocysteine concentration. Clin Chem Lab Med 2001;39:675– 80. 12. Duarte NL, Wang XL, Wilcken DEL. Effects of anticoagulant and time of plasma separation on measurement of homocysteine. Clin Chem 2002;48: 665– 8. 13. Fuchs D, Jaeger M, Widner B, Wirleitner B, Artner-Dworzak E, Lebhuber F. Is hyperhomocysteinemia due to the oxidative depletion of folate rather than to insufficient dietary intake? Clin Chem Lab Med 2001;39:691– 4. 14. Naurath HJ, Riezler R, Pu¨ tter S, Ubbink JB. Does a single vitamin Bsupplementation induce functional vitamin B-deficiency ? Clin Chem Lab Med 2001;39:768 –71. 15. Candito M, Causse´ E, Couderc R, Demuth K, Diop ME, Drai J, et al. Motivations et conditions de demande d’un dosage d’homocyste´ ine. Enque ˆte transversale dans huit laboratoires hospitaliers. Ann Biol Clin 2002; 60:321– 4. 16. Rauh M, Verwied S, Knerr I, Do¨ rr HG, So¨ nnichsen A, Koletzko B. Homocysteine concentrations in a German cohort of 500 individuals: reference ranges and determinants of plasma levels in healthy children and their parents. Amino Acids 2001;20:409 –18.
Characterization of Biochemical and Clinical Correlates of Hypocholesterolemia after Hepatectomy, Ivo Giovannini,1,2* Carlo Chiarla,2 Francesco Greco,1 Giuseppe Boldrini,1 and Gennaro Nuzzo1 (1Department of Surgery, Hepatobiliary Unit, Surgical Intensive Care, and 2IASI-CNR Center for Pathophysiology of Shock, Catholic University School of Medicine, Via Alessandro VII, 45 I-00167 Rome, Italy; * author for correspondence: fax 39-06-3051343, e-mail
[email protected]) Although hypocholesterolemia in posttraumatic states is considered a generic expression of acute-phase response, a relationship between severity of hypocholesterolemia and bad prognosis has also been found (1– 6 ). The issue is still poorly understood and is complicated by the circumstance that plasma cholesterol is affected simultaneously by multiple factors whose impacts have never been distinctly characterized. The postoperative state after hepatectomy is an interesting condition in which most of these factors may be simultaneously present. We have assessed in detail the main biochemical and clinical correlates of hypocholesterolemia in a large group of patients undergoing hepatectomy.
317
Hepatectomies were performed in 92 patients (47 women, 45 men). The mean (⫾ SD) age was 57 ⫾ 12 years, body weight was 70 ⫾ 11 kg, the ratio of actual to ideal body weight was 1.13 ⫾ 0.16 (1983 Metropolitan Tables), body surface area was 1.78 ⫾ 0.15 m2, and body mass index (weight/height2) was 25.0 ⫾ 4.0 kg/m2. Thirty-six patients had primary liver malignancy (23 with hepatocarcinoma, 10 with cholangiocarcinoma, 3 with other neoplasms), 35 had secondary hepatic malignancies (23 from colorectal cancer, 12 from other sources), and 21 had benign lesions. Eighteen patients had liver cirrhosis. Fiftyfour patients were in ASA class I (7 ), 7 in class II, 30 in class III, and 1 in class IV. No patient was on cholesterollowering medication. Hepatectomies consisted of 43 minor (⬍3 liver segments) and 49 major resections (3– 6 segments). The mean number of resected segments was 3 ⫾ 1. There were 17 associated bowel operations (resections for primary malignancy or Roux-en-Y biliary reconstructions). The duration of the operations was 390 ⫾ 149 min, and the duration of normothermic liver ischemia (used in 61 patients) was 49 ⫾ 28 min. Seventy-one patients recovered without complications, whereas 15 had nonlethal complications: 9 had intraabdominal or pulmonary sepsis, 5 had transient liver insufficiency, and 1 had a biliary fistula without sepsis. Diagnosis of sepsis was based on previously defined criteria (5 ). Six patients died. The study was carried out prospectively except for the inclusion of three nonsurvivors observed outside the prospective period; this improved the significance of results in nonsurvivors without bias because the pattern of death was similar in all cases (systemic sepsis with liver and/or respiratory insufficiency, progressing to multiple organ dysfunction syndrome). This patient population provided a continuous distribution of observations from minor to extreme surgical procedures (and degrees of postoperative illness) suited to assess correlates of hypocholesterolemia over a wide range of pathophysiologic abnormalities. The database included 478 venous blood measurements. These were performed according to the clinical routine, without the need for consent, preoperatively and on postoperative days 1, 3, and 7 in all patients and thereafter only in those with complications until recovery or death. The following variables were considered: plasma cholesterol concentration, albumin, total protein, fibrinogen, creatinine, urate, alkaline phosphatase, ␥-glutamyltranspeptidase, total and indirect bilirubin, prothrombin activity, hematocrit, hemoglobin, blood cell counts, number of resected liver segments, duration of the operation and of eventual liver ischemia, occurrence of cirrhosis, neoplastic disease, previous chemotherapy, associated bowel operations, sepsis, cholestasis, and substrate doses in patients on parenteral nutrition. Statistical analysis was based on least-squares regressions and “best fit” procedures selecting the simplest possible regressions controlling the largest possible variability of cholesterol, based on Mallows’ Cp criteria (8 ). Plasma cholesterol decreased on postoperative days 1
318
Technical Briefs
and 3 in all patients. In those recovering without complications, at postoperative day 7 cholesterol tended to return to preoperative values. In survivors who developed complications, it remained low, increasing later; in nonsurvivors it decreased further until death (Fig. 1). The trend for more severe hypocholesterolemia in these groups became significant at postoperative days 3 and 7 (Regression 1 in Fig. 1). Analysis of all postoperative measurements showed that cholesterol was correlated directly with the preoperative concentration, albumin, total protein, fibrinogen, urate, alkaline phosphatase and ␥-glutamyltranspeptidase (accounting for cholestasis), prothrombin activity, hematocrit, hemoglobin, and red blood cell and platelet counts, and inversely with total and indirect bilirubin, white blood cell count, number of resected segments, and duration of operation and hepatic ischemia. Cholesterol was also lower in patients with simultaneous bowel operations, cirrhosis, and/or sepsis and in nonsurvivors. Correlations were similar in the whole sample and in subgroups of measurements (r2 ⫽ 0.12– 0.40; P ⬍0.001 for all). Multiple regression analysis selected the best simultaneous correlates of postoperative cholesterol that explained the maximum possible portion of its variability (Regression 2 in Fig. 1). In nonsurvivors, from postoperative day 3 onward, cholesterol and its ratio
Fig. 1. Postoperative plasma cholesterol as a fraction of preoperative concentration (top) and results of regression analysis (bottom). (Top), mean ⫾ SD values are in bold; absolute values are in italics. (Bottom), regressions: postoperative plasma cholesterol (CHOL, mmol/L) as a function of preoperative concentration (PRECHOL, mmol/L), number of resected liver segments (NSEG), occurrence of complication (COMPL), death (DEATH), simultaneous bowel operations (SIMOP), presence of liver cirrhosis (CIR), sepsis (SEP), hematocrit (HCT, %), plasma albumin concentration (ALB, g/L), alkaline phosphatase concentration (ALKP, U/L; reference interval, 79 –279 U/L). Regressions combine continuous and discontinuous variables. Coefficients for continuous variables (variables in parentheses) are regression slopes and estimate the mean change in postoperative cholesterol per unit change in each variable. The coefficients for discontinuous variables (variables in subscripts) are partial regression intercepts and estimate the mean decrease in postoperative cholesterol that is associated with the event indicated in the subscript; if the event does not occur, the coefficient is zero (P ⬍0.001 for whole regression and for each coefficient).
with the preoperative value were correlated directly with time to death (r2 ⫽ 0.60 – 0.85; P ⬍0.001), but the discriminant power of single measurements in predicting outcome was poor because the pattern of death was characterized by severe and persistent decreases in cholesterol and not by isolated decreases, which also occurred in survivors. No patient in our study survived after having cholesterol ⬍1.5 mmol/L and a postoperative/preoperative ratio ⬍0.4 for more than 6 days. These results show that hypocholesterolemia is not a simple expression of acute-phase response but a more complex phenomenon quantifiably related to severity of illness. Regression 2 in Fig. 1 quantifies the simultaneous impact of “adverse” factors on decreasing cholesterol. It shows that postoperative cholesterol, in addition to being related to the preoperative value, decreases with increasing magnitude of operation and decreases further if cirrhosis is present, if sepsis occurs, and with decreasing hematocrit and albumin. Indeed, each one of these factors may be associated with accelerated clearance or diminished synthesis of cholesterol-containing lipoproteins. The magnitude of the operation increases the demand for cholesterol in cell repair and new cell synthesis (9 ), while limiting the synthetic capability of the residual liver. This is aggravated by liver cirrhosis. Sepsis involves a stronger proinflammatory cytokine response, with a reduction in hepatic lipoprotein synthesis, a stimulation of lipoprotein receptor activity, and further impairment of liver function (4, 6, 10 –12 ). The relationship with hematocrit involves hemodilution from blood loss, and the relationship with albumin involves the parallel impact of acute-phase response, altered synthetic adequacy of the residual liver, and hemodilution on both cholesterol and albumin. Among the adverse factors affecting cholesterol in our study, cholestasis is unique because it stimulates release of cholesterol-rich lipoprotein-X from the liver (13 ) and is thus expected to increase plasma cholesterol. This is consistent with the mean increase of 0.002 mmol/L per unit increase in alkaline phosphatase, as quantified in Regression 2 of Fig. 1, which is similar to that estimated in a previous study (5 ), and may explain the failure of patients with cholestasis to manifest extreme hypocholesterolemia. These findings help to clarify how hypocholesterolemia in postoperative and critically ill patients becomes a cumulative index of severity of illness and its relationship with poor prognosis. The pattern of death is not characterized by low cholesterol at one single time; rather it is characterized by persistent hypocholesterolemia, as supported by separate findings showing recovery after transient postoperative decreases in cholesterol to ⬍0.5 mmol/L followed by a steady increase (5 ). Additional adverse implications of hypocholesterolemia may be related to impaired protection against proinflammatory mediators (4, 14 ), including impaired antioxidant capacity. As we also found, hypocholesterolemia is associated with low concentrations of albumin and urate
Clinical Chemistry 49, No. 2, 2003
(which account for one-half of the antioxidant capacity of plasma) (15 ), of vitamin E (carried in lipoproteins together with cholesterol) and other antioxidants (16 ), and in sepsis, with reduced antioxidant protection by sulfur amino acids (17 ). In sepsis, it is also related to impaired energy and amino acid disposal, which is partly reversed by increasing the amino acid supply (2, 18 ). Recent studies also suggest that cholesterol becomes an essential substance in extreme illness (19 ). At present, however, the main clinical implication of severe hypocholesterolemia in acute states is the need for rapid treatment and resolution of the underlying illness.
We acknowledge the kind assistance of Maurizio Cianfanelli (Catholic University Medical School) and Rosaria Di Pasquale, nurse and sister of a nonsurviving patient.
References 1. Alvarez C, Ramos A. Lipids, lipoproteins, and apoproteins in serum during infection. Clin Chem 1986;32:142–5. 2. Chiarla C, Giovannini I, Siegel JH, Boldrini G, Coleman WP, Castagneto M. Relationship of plasma cholesterol level to doses of branch-chain amino acids in sepsis. Crit Care Med 1990;18:32– 6. 3. Windler E, Ewers-Grabow U, Thiery J, Walli A, Seidel D, Gretern H. The prognostic value of hypocholesterolemia in hospitalized patients. J Clin Invest 1994:72:939 – 43. 4. Fraunberger P, Pilz G, Cremer P, Werdan C, Walli AK. Association of serum tumor necrosis factor levels with decrease of cholesterol during septic shock. Shock 1998;10:359 – 63. 5. Giovannini I, Boldrini G, Chiarla C, Giuliante F, Vellone M, Nuzzo G. Pathophysiologic correlates of hypocholesterolemia in critically ill surgical patients. Intensive Care Med 1999;25:748 –51. 6. Fraunberger P, Nagel D, Walli AK, Seidel D. Serum cholesterol and mortality in patients with multiple organ failure. Crit Care Med 2000;28:3574 –5. 7. Djokovic JL, Hedley-White J. Prediction of outcome of surgery and anesthesia in patients over 80. JAMA 1979;242:2301– 6. 8. Seber GAF. Linear regression analysis. New York: Wiley, 1977:369 – 82. 9. Chijiiwa K, Kozaki N, Naito T, Okamoto S, Kuroki H Yamashita H, et al. Hepatic bile acid synthesis and DNA synthetic rate after partial hepatectomy. Br J Surg 1996;83:482–5. 10. Akgu¨ n S, Ertel NH, Mosenthal A, Oser W. Postsurgical reduction of serum lipoproteins: interleukin-6 and the acute-phase response. J Lab Clin Med 1998;131:103– 8. 11. Cerra FB, Siegel JH, Border JR, Wiles J, McMenamy RR. The hepatic failure of sepsis: cellular vs substrate. Surgery 1979;86:409 –22. 12. Siegel JH, Giovannini I, Coleman B, Cerra FB, Nespoli A. Pathologic synergy in cardiovascular and metabolic compensation with cirrhosis and sepsis. Arch Surg 1982;1117:225–38. 13. Cooper AD. Hepatic lipoprotein and cholesterol metabolism. In: Zakim D, Boyer TD, eds. Hepatology: a textbook of liver disease, 2nd ed. Philadelphia: WB Saunders, 1990:96 –123. 14. Feingold KR, Funk JL, Moser AH, Shigenaga JK, Rapp JH, Grunfeld C. Role for circulating lipoproteins in protection from endotoxin toxicity. Infect Immun 1995;63:2041– 6. 15. Miller NJ, Rice-Evans CA. Spectrophotometric determination of antioxidant activity. Redox Rep 1996;2:161–71. 16. Muldoon MF, Kritchevsky SB, Evans RW, Kagan VE. Serum total antioxidant activity in relative hypo and hypercholesterolemia. Free Radic Res 1996;25: 239 – 45. 17. Chiarla C, Giovannini I, Siegel JH, Boldrini G, Castagneto M. The relationship between plasma taurine and other amino acid levels in human sepsis. J Nutr 2000;130:2222–7. 18. Chiarla C, Giovannini I, Boldrini G, Castagneto M. The influence of amino acid load and energy expenditure on plasma cholesterol levels in sepsis. Acta Med Rom 1989;27:166 –70.
319
19. Bakalar B, Zadak Z, Pachl J, Hyspler R, Crhova S. Influence of severe trauma on cholesterol synthesis. Intensive Care Med 2000;26(Suppl 3):S357.
Deletion of the C4-CYP21 Repeat Module Leading to the Formation of a Chimeric CYP21P/CYP21 Gene in a 9.3-kb Fragment as a Cause of Steroid 21-Hydroxylase Deficiency, Hsien-Hsiung Lee,1* Shwu-Fen Chang,2 YannJinn Lee,3 Salmo Raskin,4 Shio-Jean Lin,5 Mei-Chyn Chao,6 Fu-Sung Lo,7 and Ching-Yu Lin,1 (1 King Car Food Industrial Co., Ltd., Yuan-Shan Research Institute, No. 326, Yuan Shan Rd., Sec. 2, Yuan Shan, Ilan 264, Taiwan, Republic of China; 2 Graduate Institute of Cell and Molecular Biology, Taipei Medical University, Taipei 110, Taiwan, Republic of China; 3 Department of Pediatrics, Mackay Memorial Hospital, Taipei 104, and College of Medicine, Taipei Medical University, Taipei 110, Taiwan, Republic of China; 4 Genetika-Centro de Aconselhamento e Laboratorio de Genetica, Curitiba, Parana, Brazil 80410; 5 Department of Pediatrics, College of Medicine, National Cheng Kung University, Tainan 704, Taiwan, Republic of China; 6 Department of Pediatrics, Division of Genetics, Kaohsiung Medical University Hospital, Kaohsiung 807, Taiwan, Republic of China; 7 Division of Endocrinology, Department of Pediatrics, Chang Gung Children’s Hospital, Taoyuan 330, Taiwan, Republic of China; * author for correspondence: fax 886-3-9228030, e-mail hhlee@ms2. kingcar.com.tw) Gross gene deletions have been reported in 20% of alleles in patients with congenital adrenal hyperplasia (CAH) involving a 21-hydroxylase deficiency (1 ). This type of deletion occurs in the RCCX module, including the CYP21P, tenascin A (TNXA), RP2, C4B, CYP21, and tenascin B (TNXB) genes, as evidenced by a 30-kb deletion identified by pulse-field electrophoresis (2 ). Inactivation of the CYP21 gene may also occur through intergenic recombination with transfer of deleterious mutations from the neighboring CYP21P pseudogene. The frequency of gene deletions or conversions in CAH is controversial (3–5 ) and is dependent on the population studied. Evidence for gene deletions and/or conversions is traditionally obtained by Southern blot analysis. Multiple probes and separate restriction endonuclease digestions are used. TaqI generates 3.7-kb (functional) and 3.2-kb (pseudogene) fragments, and BglII produces 11-kb (functional) and 12-kb (pseudogene) fragments. These analyses have been used since 1984 (1, 3, 5–9 ). However, the method is indirect and time-consuming, and densitometry of fragments can be prone to error. To identify the interchange region and improve detection of gene deletions and conversions in the RCCX module (10 –13 ), we have developed a novel Southern blot analysis that uses two restriction endonucleases, AseI and NdeI, and requires only one probe. In addition, we use a PCR product amplified with locus-specific primers covering the TNXB gene to the 5⬘ end of CYP21P or CYP21
