blood and plasma. Ann Clin Biochem. 1987;24:161-6. Frank Kyne'. Sean Maguire'. Sean O'Broin2. Peadar McGing'. Shaun McCann2. Edwin Wright1. 'Biochem.
range) for the main groups were as
available kits for RIA are the Sandimmune kit (Sandoz,
commercially
Basle, Switzerland),
follows: heart (38%, -15% to 96%), liver (54%, -7% to 135%), and renal (55%, -4% to 89%). The correlation coefficients and regression equations obtained for the main types of organ transplants were as follows: Heart,TDx = 1.26HPLC + 26(r= 0.95, n = 77); TDx = 1.14 Sandoz + 22 (r 0.97, n = 15). Liver, TDx = 1.25 HPLC +57 (r = 0.90, n = 52); TDx 0.88 Sandoz +115 (r = 0.85, n on 12). Renal (same 17 patients analyzed by all three assays), TDx on 1.52 HPLC - 8 (r = 0.93); TDx = 1.09 Sandoz + 63 (r = 0.89). The overall regression equations for all samples assayed were as follows: = 1.22 HPLC + 46 (r = 0.94, n = 166); TDx = 1.06 Sandoz + 66 (r on 0.96, n = 55). For the renal patients, our regression equation is markedly different from that of Yatscoff et al. (3): TDx = 1.14 HPLC +6 (r = 0.967, n on 44; Yatscoifetal.);TDx = 1.57 HPLC -10 (r = 0.95, n = 22; this report). Results for the Abbott TDx kit compared with those for the RIA were as follows: TDx = 1.03 Sandoz + 5 (r on 0.99, n = 44; Yatseoff et al.); TDx = 1.06 Sandoz + 66 (r on 0.96, n = 55; this report, Figure 1B). In our study,
and the Cyclo-
tracSp kit (Incatar Corp., Stillwater, MN). Both of these kits use the same monoclonal antibody and give similar results. This monoclonal antibody is produced by Sandoz to be selective for
the parent
However, be-
compound.
cause of cross-reactivity with some CsA metabolites, the Sandixmnune kit yielded results -12% higher than by HPLC in 44 renal patients (1). Recently, an assay of CsA has been developed for the Abbott TDx (Abbott Diagnostics, N. Chicago, IL 60064) in which another Sandoz antibody specific for the parent compound is used (2). This new assay was evaluated by Yatscoff et al. (3), who reported good correlation between results by TDx, HPLC, and RIA (Sandimmune kit). Their study of CsA in whole blood from 44 renal-transplant recipients indicated that the TDx values averaged 24% higher than the HPLC values. We evaluated the new TDx assay, controls and also trough blood samples from the following transplant recipients: heart (n = 77), liver (n = 52), renal (n = 22), and “others” (n = 15), which included bone marrow transplant recipients and patients with autoimmune disease. We assayed all of the patients’ samples (n = 166) by HPLC (4) and a portion (n = 55) by RIA (Sandimmune kit); both of these assays are used routinely for CsA monitoring in our laboratories and both perform well in both internal quality-control schemes and in the Cyclosporine Quality Assurance Scheme, London, U.K. Whole-blood CsA calibrators provided with the TDx kit were assayed by HPLC and RIA, and yielded values identical to those assigned to them by the manufacturers. Whole-blood CsA values by the TDx in 166 patients were, on average, 48% higher than those by HPLC (Figure 1A). Results were higher for all transplant types, and the extent ofthe overestimation varied greatly. The percentage of overestimations (mean, using
U,.
0
the 55 samples assayed by the TDx gave values 19% higher than when assayed by RIA (Sandiminune). This compares with a 6% assay difference reported by Yatacoffet al. (3). We found results (n on 39) 12% higher by Sandimmune RIA than by HPLC, in agreement with the findings of Johnston et al. (1). We conclude that, in our hands, the monoclonal antibody used in this TDx assay is not as specific for CsA as claimed by Yatscoff et al. (3), and that the extent of overestimation varies broadly in all transplant types. Although the assay is rapid, precise, and has probable clinical utility, it is not as accurate as either the HPLC reference method or the Sandoz specific monoclonal assay.
a
B
0
0
o
0
200
aoo
600
Hea,1
oy
#{149}Live,
o
a
#{149}Line,
Renal
*
Bone
a
Autoininome
eon
400
Menow
Dineen.
p
200
Heat
a
Renal
*
Bone
Menow
#{149}Autoinmur.
D.e
1000 600
HPLC CA Ug/L
t000
SAJ’COZ CSA pg / I.
Fig. 1. Cyclosporlne concentrations in whole blood determinedby the AbbottTDx (monoclonalantibody)vs (A) HPLC and ( the Sandimmune selectiveRIA
1658 CLINICAL CHEMISTRY, Vol. 37, No. 9, 1991
Some caution should be exercised in the interpretation of results, and the assay requires further investigation. References 1. JohnstonA, MarsdenJT, Holt DW. The continuing need for quality assessment of cyclosporine measurement. Clin Chem
1989;35:1309-12. 2. Wang P, Meucci V, Simpson et al. A
antibody fluorescent polarisation immunoassay for cyclosporine. Transplant Proc 1990;22:1186-8. 3. Yatacoff RW, Copeland KR, Faraci CJ. Abbott Thx monoclonal antibody assay evaluated for measuring cyclosporine in whole blood.Cliii Chem 1990;36:1969-73. 4. Maguire S, Kyne F, UaConaill D. An improved extraction procedure for an HPLC method for cycloaporin A in whole blood and plasma. Ann Clin Biochem monoclonal
1987;24:161-6. Frank Kyne’
Sean Maguire’ Sean O’Broin2
Peadar McGing’ Shaun McCann2 Edwin Wright1 ‘Biochem. Lab. Mater Misericordiae Hosp. Dublin 7, Ireland
2Dept. of Haematol. St. James’s Hosp. Dublin
8, Ireland
Effect of Storage Time on Peptic Activity in Gastric Biopsies
To the Editor: Pepsin appears to play a crucial role in the development of acute and chronic ulceration in the upper gastrointestinal tract (1,2). However, which method to use to evaluate peptic secretion is still debated (3). Several factors may compromise the accuracy of pepsin measurement in gastric juice; in particular, the storage time appears crucial to a reliable determination (4, 5). A different approach involves the direct evaluation of pepsin-pepsinogen concentrations in gastric mucosa. As yet, no data are available on the influence of storage time on peptic activity in mucosal biopsies. Here we report the results obtained for peptic activity measured in mucosal biopsies stored for different times. Twelve biopsy specimens were obtained from six duodenal ulcer patients (five men, one woman, mean age 48 years, range 28-60 years) during upper gastrointestinal endoscopy: one specimen each from the fundus, another from the corpus of each subject’s stomach. The specimens were
treated as previously described (6). The zymogen in the homogenate was transformed into pepsin by acidifying the medium (HC1, 0.01 mol/L). We divided the supernate of each sample into five aliquots and measured the pepticactivity in each by the method of Berstad (7), with human hemoglobin as a substrata. The activity was determined by comparing the results with a standard curve obtained by using purified pig pepsin as a standard. Peptic activity was expressed as pg/mg oftotal homogenate proteins, the latter being determined by the method of Bradford (8), with reagents from Bio-Rad Labs. (Milan, Italy). To evaluate the influence of the storage time, we measured the pepsin activity in the 12 homogenate specimens from the six patients at time 0 (immediately after biopsy collection) and after seven, 14,30, and 90 days of storage at -20 #{176}C. All determinations were made in triplicate.
The intra-assay reproducibility was determined by measuring repeatedly (11 times) the pepsin concentration in three homogenates with low, medium, and high enzyme activities. The between-assay reproducibifity was evaluated by measuring the same specimens in five different analytical runs. Variation of pepsin concentration over time was statistically assessed by analysis of variance (ANOvA) for repeated measures, with use of the program 2V of the BMDP statistical package (9). We considered as the grouping factor the gastric region (fundus and corpus)and as the trial factor the time of storage. The within- and between-assay coefficientsofvariation (CVs) demonstrated a good reproducibility. The within-assay CVs were 13.1% at 8.0 pg/mg of total proteins, 5.1% at 146.0 pg/mg, and 9.1% at 248.0 pg/mg; the between-assay CVs were 21.6% at 8.1 pg/mg, 11.2%at 142.9 pg/mg, and 13.3%at 240.0 pg/mg. The variation of pepsin concentration over time, as assessed by the ANOVA for repeated measures, wasnot statistically significant (F = 1.70). Moreover, there was no significant difference between the two gastric regions (F = 1.01) and no significant interaction between region and time (F = 1.73). Mean pepsin concentration (± SE) was 135.4(16.03) pg/mg at time 0; 133.3 (17.9) pg/mg after seven days; 132.4 (18.7) pg/mg after 14 days; 125.8 (14.4) pg/mg after 30 days;and 135.8(16.9) pg/mg after 90 days. In conclusion, our findings demonstrate that, contrary to data reported in the literature regarding the dependenceof gastric juice pepsin on storage time, there were no changes in peptic
activity in gastric homogenates
stored
at pH 7.4 until at least 90 days. This apparent discrepancy may dependon an interaction between the activated gastric juice pepsin and other factors present beforefreezing, e.g., bile salts, pepsinogen, pH, and ionic strength of the gastric juice. Further studies are in progress to evaluate the clinical usefulness of pepsin determination in gastric mucosa. References
1. SoIl AR. Pathogenesis of peptic ulcer andimplications for therapy. N Engi J Med 1990;322:909-16.
2. Venables CW. Mucus, pepsin andpeptic ulcer. Gut 1986;27:233-.8. 3. SamlofflM. Pepticulcer many proteinases of aggression. Gastroenterology
1989;96:586-.95. 4. Sivakumaran T, De Gara CJ, Walker WHC, Siletti C, Burget D, Hunt RH. Stability of gastric juice pepsin and a method for its preservation. Clin Chem 1984;30: 1582-3.
5. Pilotto A, VianelloF, Di Mario F, eta!. Effect of temperature and storage time on pepsin concentration in human gastric juice. IRCS Med Sci 1986;14:155-6. 6. PlebaniM, Di Mario F, Dal SantoPL,et al. Pepsinogen group I measurement in endoscopic gastroduodenal biopsies. Cliii Chem 1990;36,4:682-4. 7. Berstad A. A modified hemoglobin substrate method for the estimation of pepsin in gastric juice. Scand J Gastroenterol 1970;5:343-8. 8. Bradford MA. A rapid and sensitive method for the quantification ofmicrogram quantities of protein utilising the principle of protein-dye binding. Anal Biochem 1976;72:248.-51. 9. Dixon WJ, Brown MD, Engelman L, Jennrich RI. BMDP statistical software manual. Berkeley, CA: University of California, 1990:489-527. Mario
Plebani
Francesco Di Mario’ Bastianello GerrnanI’ Fabio Vianello’ Paola Dotto’ Remo Naccarato’
Angelo Burlina Depts. of Clin. Biochem. ‘Gastroenterol. Complesso Ospedale Univ. di Padova
tamicin
and vancomycin.
The patient was a premature infant in the neonatal intensive-care unit (estimated gestational age, 25 weeks) who hadbeenbornwith an abdominal wall defect through which a portion of liver protruded (omphalocele). This deformity was treated nonsurgically, because of the infant’s unstable respiratory status, by the topical applica-
tion of mercurochrome (merbromin), 5 g/L, each hour for the first 48 h and then daily for another six days. This therapy is suggested by the pediatric surgical literature to promote re-epithelialization of the omphalocele (1, 2). However, recent reports indicate a significant risk of mercury intoxication in these children, secondary to systemic absorption ofthe topical mercury preparation (3, 4). This infant’s concentration of serum mercury was measured during treatment and found to be >100 pglL (nontoxic adult value
