Printer Friendly

Does iron concentration in a liver needle biopsy accurately reflect hepatic iron burden in [beta]-thalassemia?

[beta]-thalassemia major is an autosomal recessive disease characterized by absent or decreased synthesis of the [beta]-globin gene (1). Thalassemic children, estimated at 100 000 worldwide, are affected by chronic anemia and need regular blood transfusion (2). Because of the limited capacity of iron excretion in humans, the iron in transfused red cells accumulates in the body. The liver, heart, and pancreas are the target organs of iron-induced injury; therefore, the major pathological manifestations observed in [beta]-thalassemia major are chronic liver disease, evolving to cirrhosis, and dilative cardiomyopathy, both characterized by severe iron deposition (3, 4). The dangerous effects of iron excess can be managed by administration of chelators capable of removing iron from transferrin, ferritin, and other iron stores (5-7). Determination of the hepatic iron concentration (HIC) is one of the most valid procedures in assessing real body iron burden (8-10), which is important for adjusting each patient's chelation therapy over the years. HIC usually is measured on one part of a needle biopsy core, and the measured value is considered representative of the iron concentration in the whole liver (11 ). In 1995, a study by our group (3 ) first showed that iron is unevenly distributed in the livers of [beta]-thalassemic patients, and therefore, the iron content determined in a small liver fragment should be interpreted with caution because it cannot be considered a true representation of the mean HIC. These data were confirmed by us (12-14) and by other authors (15), who reported differences in HIC measurements on liver biopsy specimens.

Recently, a striking difference in HIC was found in biopsy samples from cirrhotic livers (range, 60-2851 [micro]g/g dry weight) (16). It therefore seemed of interest to measure iron concentrations in a large number of needle biopsies with the following objectives: (a) to establish whether the determination of iron concentration in a needle biopsy is representative of the mean iron content of the whole liver; (b) to determine whether HIC measured in only a portion of the needle biopsy is indeed representative of the mean liver iron content; and (c) to determine the minimum weight of liver parenchyma needed to obtain a HIC value that would be useful to evaluate the body iron burden. To this end we performed 54 needle biopsies from an autopsy liver of a thalassemic patient and evaluated the effect of each factor by use of analysis of variance (17).

The clinical data of the 29-year-old male affected by homozygous [beta]-0-thalassemia major were reported previously (3).

At autopsy, the right lobe of the liver was divided into 18 areas with diameters of ~1 cm (sampling sites), and three needle biopsies were performed in each area. The three biopsies were subdivided into two, three, and four parts (subsamples), respectively: the first subsample, marked with index 1, corresponded to the subcapsular part, whereas the indices for the other parts increased with depth. Fig. 1A shows the sampling map for these 162 [18 X (2 + 3 + 4)] subsamples.

The digestion procedure to obtain a solution suitable for inductively coupled plasma atomic emission spectroscopy (ICP-AES), the working calibration, and the ICP-AES conditions have been described previously (18).

All samples were weighed on a balance accurate to the sixth digit. It should be highlighted that weighing is fundamental to the precision of the entire procedure, e.g., a four-digit balance could lead to a 10% error when samples of ~1 mg are weighted. The sequence in analyzing the 162 samples was completely randomized to avoid any possible systematic error related to an ordered sequence. The same type of operation on all samples (e.g., weighing, diluting to the mark, spectral recording) was carried out by the same operator to avoid operator-generated errors. All procedures were tested by analyzing NIST bovine liver samples. During spectrophotometric analyses, one calibration out of five solutions was measured to check the reliability of the measurements.

HN[O.sub.3] and the ICP calibrator for iron (10 g/L) were from Aldrich; Triton X was from Merck. NIST Standard Reference Material 1577b, Bovine Liver, was used to validate the measurements.

All data are represented as three 18 * p matrices, where 18 is the number of sampling sites, and p = 2, 3, or 4, the number of subsamples into which each needle biopsy was subdivided. The matrix elements [x.sub.ij] represent iron concentrations at sampling sites i (1-18) and depths j (1 to p, where 1 is the index of the superficial sample and p is the index of the deepest sample). For each matrix, the total mean value over the 18 * p data points is defined as:

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]

For each matrix, p mean values ([bar.x.sub.j]) could be defined among samples at the same depth in different sampling sites, and 18 mean values ([bar.x.sub.i]) could be defined over the p subsamples in each needle biopsy:

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]

The total sum of squares, [SS.sub.T] = [[summation].sup.p.sub.j] = 1 [[summation].sup.18.sub.i] = 1 [([x.sub.ij] - [bar.x.]).sup.2], with 18 * p - 1 degrees of freedom, can be divided into three contributions:

[SS.sub.S] = p * [[summation].sup.18.sub.i] = 1 [([bar.x.sub.i] - [bar.x.]).sup.2] has 17 degrees of freedom, and measures the part of the total sum attributable to the differences among sampling sites;

[SS.sub.D] = 18 * [[summation].sup.p.sub.j] = 1 [([bar.x.j] - [bar.x.]).sup.2] has p - 1 degrees of freedom, and measures the variation attributable to the depth of sampling;

[SS.sub.R] = [[summation].sup.p.sub.j] = 1 [[summation].sup.18.sub.i] = 1 [([x.sub.ij] - [bar.x.sub.i] - [bar.x.sub.j] + [bar.x.]).sup.2] has (p - 1) X 17 degrees of freedom, and is the part that cannot be explained by a given factor of variation and can be considered an estimate of experimental error.

[FIGURE 1 OMITTED]

Analysis of variance can be performed, using the F-test (17), by comparing the variances attributable to the various factors with the variance attributable to the experimental error.

The results presented in Table 1 show some characteristics that need some comment:

* The global mean value (20 954) on all 162 data almost coincided with the total mean values over the 2 X 18 (20 878), the 3 X 18 (21012), and the 4 X 18 (21 018) data for the needle biopsies divided into two, three, and four parts, respectively.

* The mean values of data at the same depth showed a steady decrease in iron concentrations from the subcapsular region to the inner part of liver parenchyma (Fig. 1B).

* The global mean value coincided with the mean calculated on 51 samples of the same liver (~1 g dry weight) presented in a previous work (3 ).

* The relative standard deviation associated with experimental error (~10%) was approximately twice the value we found when testing the precision and accuracy of the procedure on much heavier samples (18). We measured the concentrations of various elements other than iron on sets of seven samples of NIST bovine liver, each set characterized by a definite weight. For samples weighing 60-250 mg, the standard deviations did not vary significantly, whereas they increased dramatically for samples weighing 4-60 mg. This behavior, observed for all of the elements, cannot be ascribed to a loss of instrumental precision (ICP instrumental percentage error never exceeded 1%) but to real composition differences among the smallest samples.

* The total sum of the squares can be divided into the contributions attributable to experimental error, sampling site, and depth. The F-test at 5% probability applied to sampling and error variance gave a significant value only for needle biopsies divided into four parts (95th percentiles of the reference F distributions, [F.sub.17,17] = 2.29; [F.sub.17,34 ] = 1.94; and [F.sub.17,51 ] = 1.87). On the other hand, the F-test relating to the effect of depth afforded clearly significant values (95th percentiles of the reference F distributions, [F.sub.7,17] = 4.46; [F.sub.2,34] = 3.28; and [F.sub.3,51] = 2.80), highlighting that depth is the main factor determining variability among samples, as pointed out by the remarkable decrease in iron concentrations from the surface to the interior of the liver shown in Fig. 1B.

Our data clearly show that the iron concentration in a single liver needle biopsy, weighing ~5 mg and 2 cm long, may be considered representative of the mean HIC with a relative standard deviation of 15%. This value depends on the analytical procedure and the uneven iron distribution. Therefore, when only one portion of the needle biopsy core is used, the reduction in weight from 5 mg to 1 mg does not significantly increase the error attributable to the analytical procedure, although the measure is far less significant because the subcapsular portion of the needle biopsy contains much more iron than the inner part. In cases where the determination of HIC is important for monitoring iron chelation therapy, a random subdivision of the liver biopsy is to be avoided; if an entire liver biopsy is not available for chemical analyses, we suggest that the subcapsular and the deepest part of the biopsy be used as a unit to minimize the errors attributable to a casual choice of sample. Finally, we propose a consensus conference on the methods of trace element determination in needle biopsies. In our opinion, the standardization of these procedures may lead to measured HIC values more representative of the true HIC and much more useful for clinical purposes.

We thank the Assessorato alla Sanita della Regione Autonoma della Sardegna for financial support.

References

(1.) Olivieri NF. The R-thalassemias. N Engl J Med 1999;341:99-109.

(2.) Cao A. 1993 William Allan Award Address. Am J Hum Genet 1994:54:397-402.

(3.) Ambu R, Crisponi G, Sciot R, Silvagni R, Nurchi VM, Faa G, et al. Uneven hepatic iron and phosphorous distribution in R-thalassemia. J Hepatol 1995;23:544-9.

(4.) Faa G, Sau F, Abbruzzese P, Silvagni R, Nurchi VM, Crisponi G, et al. Uneven cardiac iron distribution in R-thalassemia. Cardiovasc Pathobiol 1997;3/4: 1-7.

(5.) Faa G, Crisponi G. Iron chelating agents in clinical practice. Coord Chem Rev 1999;184:291-310.

(6.) Crisponi G, Nurchi VM, Silvagni R, Faa G. Oral iron chelators for clinical use. Polyhedron 1999;18:3219-26.

(7.) Brittenham GM, Cohen AR, McLaren C, Martin MB, Griffth PM, Nienhuis AW, et al. Hepatic iron stores and plasma ferritin concentration in patients with sickle cell anemia and thalassemia major. Am J Hematol 1993;42:81-91.

(8.) Angelucci E, Baronciani D, Lucarelli G, Baldassarri M, Galimberti M, Giardini C, et al. Needle liver biopsy in thalassemia analyses of diagnostic accuracy and safety in 1184 consecutive biopsies. Br J Haematol 1995;89:757-61.

(9.) Tricta F, Spino M. Assessment of iron chelation in thalassemia: experience with Deferiprone. 9th International Conference on Oral Chelation, March 25-28, Hamburg, Germany, 1999:22.

(10.) Galanello R, De Virgiliis S, Giagu N, Cao A, Mancosu MG, Faa G. Evaluation of iron overload in thalassemia intermedia. 9th International Conference on Oral Chelation, March 25-28, Hamburg, Germany, 1999:29.

(11.) Bassett ML, Halliday JW, Powell LW. Value of hepatic iron measurements in early hemochromatosis and determination of the critical iron level associated with fibrosis. Hepatology 1986;6:24-9.

(12.) Faa G, Nurchi V, Demelia L, Ambu R, Parodo G, Congiu T, et al. Uneven hepatic copper distribution in Wilson's disease. J Hepatol 1995;22:303-8.

(13.) Aragoni MC, Crisponi G, Nurchi VM, Sciot R, Ambu R, Faa G, et al. Chemometric methods applied to an ICP-AES study of chemical element distributions in autopsy livers from subjects affected by Wilson and /3-thalassemia. J Trace Elem Med Biol 1995;9:215-21.

(14.) Faa G, Sciot R, Farci AMG, Callea F, Ambu R, Congiu T, et al. Iron concentration and distribution in the newborn liver. Liver 1994;14:193-9.

(15.) Villeneuve JP, Bilodeau M, Lepage R, Cote J, Lefebvre M. Variability in hepatic iron concentration measurement from needle biopsy specimens. J Hepatol 1996;25:172-7.

(16.) Emond MJ, Bronner MP, Carlson TH, Lin M, Labbe RF, Kowdley KV. Quantitative study of the variability of hepatic iron concentrations. Clin Chem 1999;45:340-6.

(17.) Massart DL, Vandeginste BGM, Deming SN, Michette Y, Kaufman L. Evaluation of sources of variation in data. Analysis of variance. In: Chemometrics: a textbook. Amsterdam: Elsevier, 1988:59-74.

(18.) Crisponi G, Nurchi VM, Silvagni R, Lubinu G, Ambu R, Faa G, et al. Critical evaluation of analytical procedures for trace-element determinations in human liver using ICP-AES. Atom Spectrosc 1995;16:73-8.

Guido Crisponi, [1] * Rossano Ambu, [2] Franco Cristiani, [1] Gabriella Mancosu, [2] Valeria Marina Nurchi, [1] Rosalba Pinna, [1] and Gavino Faa [2] ([1] Dipartimento di Chimica Inorganica ed Analitica, Universita di Cagliari, Complesso Universitario di Monserrato, 09042 Monserrato-Cagliari, Italy; [2] Dipartimento di Citomorfologia, Divisione di Anatomia Patologica, Universita di Cagliari, Via Ospedale 60, 09124 Cagliari, Italy; * author for correspondence: fax 39-0706754478, e-mail crisponi@unica.it)
Table 1. Results of HIC determinations for needle biopsies.

No. of TM (a, b) MV1 (c) MV2 (c) MV3 (c)
parts

2 20 878 22 354 19 402

3 21 012 22 988 20 508 19 540

4 21 018 24 676 21 002 19 286

No. of MV4 (c) [SS.sub.T] [S.sup.2.sub.T] [SS.sub.R]
parts x [10.sup.-3] x [10.sup.-3] x
 [10.sup.-3]

2 260 705 7448 59 582
 35 (f)
3 503 372 9498 197 893
 53 (f)
4 19 110 862 902 12 154 252 581
 71 (f)

No. of [S.sup.2.sub.R] [SS.sub.S] [S.sup.2.sub.S] [SS.sub.D]
parts x [10.sup.-3] x [10.sup.-3] x [10.sup.-3] x
 [10.sup.-3]

2 3505 122 680 7216 78 443
 17 (f) 17 (f)
3 5820 191 611 11 271 113 867
 34 (f) 17 (f)
4 4952 250 018 14 707 360 302
 51 (f) 17 (f)

No. of [S.sup.2.sub.D] [S.sup.2.sub.S]/ [S.sup.2.sub.D]/
parts x [10.sup.-3] [S.sup.2.sub.R] [S.sup.2.sub.R]
 (d) (e)

2 78 443 2.06 22.4
 1 (f) (2.29) (g) (4.46) (g)
3 56 934 1.94 9.8
 2 (f) (1.94) (g) (3.28) (g)
4 120 101 2.97 24.3
 3 (f) (1.87) (g) (2.80) (g)

Global mean value (h) [MV.sub.162] = 20 954

(a) TM, total mean value; MV, mean value; [SS.sub.T], total sum;
[S.sup.2.sub.T], total variance; [SS.sub.R], sum attributable to
experimental error; [S.sup.2.sub.R], variance attributable to
experimental error; [SS.sub.S], sum attributable to variability
among sampling sites; [S.sup.2.sub.S] variance attributable to
variability among sampling sites; [SS.sub.D] sum attributable to
depth; [S.sup.2.sub.D], variance attributable to depth.

(b) Total mean values, calculated on 18 x p samples, are reported
for the three cases (needles subdivided into 2, 3, and 4 parts,
respectively).

(c) Mean values for subsamples at the same depth.

(d) Ratio between the variance attributable to the variability among
sampling sites and the variance attributable to experimental error.

(e) Ratio between the variance attributable to depth and variance
attributable to the experimental error.

(f) Degrees of freedom.

(g) 95th percentiles of the reference F distributions.

(h) Global mean value for all 162 [(2 + 3 + 4) x 18] subsamples.
COPYRIGHT 2000 American Association for Clinical Chemistry, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2000 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:Technical Briefs
Author:Crisponi, Guido; Ambu, Rossano; Cristiani, Franco; Mancosu, Gabriella; Nurchi, Valeria Marina; Pinna
Publication:Clinical Chemistry
Date:Aug 1, 2000
Words:2615
Previous Article:Identification of two LDL-receptor mutations causing familial hypercholesterolemia in Indian subjects by a simplified rapid PCR-heteroduplex method.
Next Article:Agreement study of methods based on the elimination principle for the measurement of LDL- and HDL-cholesterol Compared with ultracentrifugation in...
Topics:


Related Articles
Type IV glycogen storage disease. (Images in Pathology).
Liver biopsy may become ancillary prognostic test.
Liver dysfunction and steatosis in familial hypobetalipoproteinemia.
Biopsy or biomarkers: is there a gold standard for diagnosis of liver fibrosis?
Ferritin is not an indicator of available hepatic iron stores in anemia of copper deficiency in rats.
Assessment of liver function: pre- and peritransplant evaluation.
Clinical chemistry and post-liver-transplant monitoring.

Terms of use | Privacy policy | Copyright © 2021 Farlex, Inc. | Feedback | For webmasters