Serum Creatine Kinase Isoenzyme BB Is a Poor ... - Clinical Chemistry

2 downloads 0 Views 809KB Size Report
infarctions, and (C) brain contusion(s). We analyzed each patient's sera for creatine kinase isoenzyme BB (CK-BB), using a monoclonal antibody kit (impres-BB; ...
The most effective of the adulterants I tested is sodium chloride, which will be a concern only for laboratories that use the EMIT technology. This and other studies indicate that the minimum amount of sodium chloride that must be added to produce a negative result varies with different

assays, but it is substantial. The effective amounts used in this study would be difficult to store (e.g., under fingernails) and require time and stirring for solution to be complete. Others have reported that amounts from 50 to 75 gIL are effective in producing false negatives, depending upon the assay and drug concentration used (3-5). I found that 50 g/L was insufficient to affect the EMIT cannabinoid assay. Sufficient sodium chloride to produce falsely negative results will result in a residue (which can be noted by the collection-site person), a high relative-density reading, and a delta absorbance value less than the negative calibrator. Other adulterants that might be problematic include NaHC1O4, which should be readily recognized by its smell (even one adulterated sample in a group is easily detected) and its reaction with pH paper. Although NaHC1O4 is basic and a urine treated with it will give a pH reading of -‘-10 with a pH meter, if pH paper is used, a bright-red (but rapidly fading) color indicative of an acid pH of --1 is

Of the two assays currently of most interest, cocaine and cannabinoids, the cocaine assay was found to be a robust one, with only NaC1 producing a decreased result with the EMIT assay. The cannabinoid assay appears to be very sensitive to adulterants, yielding both decreased and increased results, depending upon the adulterant and immunoassay method used; however, most of these effects were in the positive rather than the negative direction. These results indicate that specimen adulteration is complicated for the subject by the fact that some adulterants shown to cause falsely lowered results can be readily detected by either trained collection-site personnel or by simple laboratory procedures such as temperature, pH, relative density, residue checks, and shake and sniff tests. In addition, the undesired result of an enhanced or false positive, produced by a number of potential adulterants, makes their use less attractive as a mechanism for producing a false-negative result. The laboratory needs to assess, based upon the methods used for screening, what preanalytical tests for detection of adulterants are necessary. This study was designed to serve as a starting point in making

such decisions.

produced.

Other false negatives of concern are those caused by dilute Joy and NaHCO3. Sodium bicarbonate in the concentration tested will not go completely into solution and will result in a pH of 8-9, which should be considered abnormal by the laboratory and should result in a request for a fresh sample. Joy did not cause any changes in appearance, pH, or relative density, but can be detected by vigorously shaking a small amount of the urine. More copious, longer-lasting bubbles are formed compared with normal urine, and when held to the light they refract it to give the typical rainbow appearance of soap bubbles. A major drawback, for the subject, to the use of Joy or NaHCO3 is the fact that these compounds also cause falsepositive results in several assays, hardly the result desired by the subject adding adulterants to ensure a negative

I gratefully acknowledge the gift of reagents by Roche Diagnostic Systems, and thank Dainien Brandeis, George Wadih, Tom Mertens, and Lori Hindenlang for their technical assistance.

result.

Washington,

References 1. Hoffman A. Steal this urine test. New York: Viking Penguin Books, 1987;240-2. 2. Andersen 0, Eriksen PB. Specific conductivity of urine and sensitivity of enzyme immunoassay methods of analysis for drugsof

abuse. Cliii Chem 1977:23;751-3. 3. Kim HJ, Cerceo E. Interference by NaCI with the EMIT method of analysis for drugs of abuse. Clin Chem 197622;1935-6. 4. Mikkelsen SL, Ash KO. Adulterants causing false negatives in illicit drug testing. Clin Chem 1988:342333-6. 5. Warner A. Opiates: use, abuse and current methods of detection and measurement. In-service training and continuing education. DC: Am Assoc Clin Chem, Inc., 1988.

CLIN. CHEM. 35/4, 651-654 (1989)

Serum Creatine Kinase Isoenzyme BB Is a Poor Index to the Size of Various Brain Lesions Joyce G. Schwartz,’ Caries Bazan, 111,2Carols L Gage,3Thomas J. Pilhoda,1and Sheni

L

GilIham’

We dMded patients with brain lesions into three groups: (a) patients with primary or metastatic brain cancer, (b) brain infarctions, and (C) brain contusion(s). We analyzed each patient’s sera for creatine kinase isoenzyme BB (CK-BB), using a monoclonal antibody kit (impres-BB; International Immunoassay Laboratories). Computerized axial tomography (CAT) scans were performed on each patient. The size

of the various lesions was measured from the CAT scan and recorded in milliliters. Total 6K, CK-BB, and their ratios were compared with the volume of damaged brain tissue. We found no correlation between any of the variables and the various brain lesions. We attribute this lack of correlation to an intact blood-brain barrier, the rapid elimination or inactivation of CK-BB, or some combination of these factors.

‘Departments of Pathology and2 Radiology, University of Texas Health Science Center at San Antonio, 7703 Floyd Curl Drive, San Antonio, TX 78284-7750. 3Medical Center Hospital, San Antonio, TX 78284. Received December 13, 1988; acceptedJanuary 19, 1989.

Biochemical diagnosis of brain injury has traditionally been confined to analysis of cerebrospinal fluid. No specific blood test has been available, and there has been uncertainty whether such a test could be devised because of the bloodbrain barrier. CLINICALCHEMISTRY, Vol. 35, No. 4, 1989 651

We used a new assay that is specific for creatine kinase (EC 2.7.3.2) isoenzyme BB (CK-BB), the “Impres-BB” (International Immunoassay Laboratories Inc., Santa Clara, CA 95054-1529), to determine CK-BB activity in aera of patients with various brain lesions. Hitherto, determination of CK-BB has been carried out mainly by electrophoresis, ion-exchange chromatography, and immunoprecipitation or immunoinhibition with a specific antibody, followed by identification of functional enzymatic activity. More recent methods include radioimmunoassays for quantification of CK-BB isoenzymes by use of specific antibodies. The “Aburia-CK” RIA test kit (International Immunoassay Laboratories, Inc.) measures the CK-B subunit of both CK-MB and CK-BB isoenzymes, requiring a separate radioimmunoassay to “subtract” the CK-MB isoenzyme to obtain the quantitative value for CK-BB. Most of the literature on measurement of CK-BB has been based on electrophoretic methods. Guided by their electrophoretic data, many investigators have assumed there is no CK-BB in serum. There is increasing evidence that CK-BB was often missed in electrophoresis procedures (1) and (or) often overstated because of an albumin artifact (2). These factors have contributed to the disagreement among results obtained by various investigators. The Impres-BB kit, sold for experimental use only, provides monoclonal antibodies that are specific for CK-BB. High proportions of the CK-MM and CK-MB isoenzymes do not interfere in the method (3). The present study was carried out to ascertain the frequency of increased CK-BB in serum of patients with various brain injuries, and to see if we could demonstrate a relation between the concentration of CK-BB and lesion sine. MaterIals and Methods Patients and Samples We studied 51 consecutive patients admitted through the room with either brain contusion (n = 15), infarction (n = 20), or primary or metastatic brain cancer (n 16). Blood was sampled and serum was stored frozen at -20 #{176}C at the time the patients were admitted to the emergency room. Excess sera from 20 patients’ samples negative for human choriogonadotropin, beta subunit, were used as normal controls. emergency

Bland infarcts were measured to the farthest of the hypodensity. Lesion size was then measured from the printed images, with Kern Swiss Hartchrom calipers. Both the largest diameter of each lesion and the diameter perpendicular to the largest diameter were measured on each image that demonstrated the lesion (Figure 1). The cross-sectional area of the lesion was calculated by using an ellipse to approximate the shape of each lesion. Volume was determined for each image by multiplying the calculated area times the slice thickness (4). Total lesion volume was determined by adding the individual image volumes for each lesion. When multiple lesions were present, their individual volumes were summed to yield a total volume. hematoma. extent

Degree of Injuries The extent of brain tissue involved varied considerably in our 51 patients. The smallest lesion, 0.50 mL, was from a breast cancer metastatic to the brain; the largest lesion, 319.51 mL, involved a right middle-cerebral infarction.

Other Procedures We followed the protocol sheet provided by the kit manufacturer. A single lot of reagent was used for the entire study. Total CK. We measured total CK in a Cobas-Bio centrifugal analyzer (Roche Analytical Instruments, Inc., Nutley, NJ 07110). For each assay we used 5 uL of serum, 35 L of diluent, and 165 pL of reagent (Roche Reagent for “CKNAC”; Roche Diagnostic Systems, Nutley, NJ 07110). This procedure is based on the Oliver-Rosalki method in which the NADPH concentration is monitored at 340 nm and 37#{176}C (5, 6). CK-BB. We measured CK-BB by using the Impres-BB assay. Any patient’s sample having total CK activity >1000 UIL was diluted to bring the activity below 1000 U/L. The Impres-BB reagents include monoclonal antibodies specific for CK-BB to determine CK-BB activity. This assay also was performed in the Cobas-Bio centrifugal analyzer.

Measurement of Involved brain Tissue Patients with tumors were injected intravenously with “Reno-M-60” (Diatrizoate Meglumine Injection; E.R. Squibb and Sons, Inc., Princeton, NJ 08540), a radio-opaque contrast agent, before their CAT scans. For all brain CAT scans we used a Seimens DR3 or a Philips Tomoscan 6OITX scanner, with 4- or 5-mm slice thickness for the posterior fossa and 8- or 10-mm slice thickness for the remainder of the examination. The images were printed by either a Matrix M-10 Auto Loader or a Multispot FA printer, both using a 15-image format. Tumor lesions were measured to the largest extent of their contrast enhancement. Contusions were measured to include areas of both hemorrhage and edema. Hypertensive hemorrhages were measured to the maximum extent of the Nonstandard abbreviations: CK-BB, creatine kinase-BB; Vol BT, volume (mL) of damaged brain tissue; CK, total creatine kinase; CAT scan, computerized axial tomographic scan. 652

CLINICAL CHEMISTRY, Vol. 35, No.4, 1989

Fig. 1. Computerizedaidal tomography showing primaiy lymphoma of the brain A,row indicates largest diameters measured

Table 1. Total CK, CK-BB, and Volume of Damaged Brain Tissue among Various Psttentgroup Nermal

Canc.r

n=20

n=16

Contusion

65.40 50.92 Total CK range, U/L 32-108.00 17.00-261.00 Vol B, mL mean 29.41 Vol BT, mL, range 0.50-104.26 CK-BBmean, U/L 0.99 1.59 CK-BBrange, U/L 0.50-1.80 0-4.7 Vol B is the volume ofdamagedbraintissueas measuredby computerized axialtomography.

1074.43 88.00-7971.00 39.68 2.49-201.40 15.00 0-144.00

-

-

Statistics Because the distribution of all variables was skewed, we used the log (x + 1) transformation (7). The four groups of subjects were compared by using analysis of variance. Simple and partial correlations were calculated for all pairs of variables. All P-values are two-tailed. Total CK, CK-BB, and their ratios were compared with the volume of damaged brain tissue (Vol BT). The statistical analysis comparing CK-BB vs Vol BT and CK-BB/total CK vs Vol BT were excluded from the group of normal patients owing to lack of CAT scans within this group. Among the groups with brain lesions, total CK was highest in patients with contusions (mean 1074.43 UIL) and lowest in patients with cancer (mean 50.92 UIL). Both the mean and range of Vol BT were lower in the cancer group and greatest in the group of patients with infarcts (Table 1). CK-BB was markedly higher in the patients with contusions, although the difference was not statistically significant because of the wide range of values (Figure 2). The comparison of CK-BB vs Vol BT among all groups showed a correlation of r = -0.15 (P = 0.32), which was nonsignificant (Figure 3). The correlation of CK-BB with Vol BT was most nearly linear in the contusion group (r = -0.30, P = 0.30), although it was a negative or inverse correlation-i.e., as the amount of damaged brain tissue increased, the value for CK-BB decreased! The analysis of CK-BB/total CK vs CK-BB showed a moderate corrrelation and a linear relationship (r = 0.46, P = 0.0001) (Figure 4), indicating that, overall, as the total CK increased, so did CK-BB. Similarly, there was a moderate and inversely linear relationship (r = -0.40, P =

Inlrct

n=20 183.70 13.00-1038.00 71.08 1.57-319.51 2.32 0.10-7.00

n=15

Total CK mean, U/L

Results

Brain Lesions

0.0008) between CK-BB/total CK and total CK. CK-BB vs total CK also showed a moderate and linear relation, which was significant owing to one data point with a high total CK and a high CK-BB (r = 0.43, P = 0.0003). The relation between CK-BB/total CK and Vol BT was neither linear or significant (r = -0.12, P = 0.42). DIscussIon CK-BB is present in various human tissues, predominantly the central nervous system, the gastrointestinal tract, and the pregnant uterus. However, it is rapidly eliminated and (or) inactivated compared with other CK isoenzymes -J 000

.

.32 n=45 p =

E

I

.

. 00

#{149},

.

I.

#{149}

.#{149}t.I.

.

ID

. S

.

S

10

100

cK-89

1000

044.)

Fig. 3. Clinical correlation between the amount of damaged brain tissue and CK-B8 isoenzyme measured in serum u0

bOO 0 cotu

Ms

Vj.



r0.46 p = 0.0001

n=68 S

ID

.u0

.

.2 . 1.00

S

#{149} . .

#{149} 10 -

Can.r

ConIslon

,forot

Fig.2. Rangeof CK-B8 (U/L) in serum among differentpatient groups and normal controls, based on log transformationanalysis

100 O(-S8

ioOo

044.)

Fig. 4. Ratio of CK-B&total Cl( vs CK-88 Asthe total CK increases,so dose Ihe seam CK-8B

antvtty

CLINICAL CHEMISTRY, Vol. 35, No.4,1989663

(8). The half life of CK-BB in human serum was estimated to be 5 h by Harwood et al. (9), 1 h by Vladutiu et al. (10). Therefore, measurable CK-BB activity is rarely seen, and in the serum of normal adults it is very low and below the detection limit of most methods. When tissue containing CK-BB is diseased or damaged, increased CK-BB activity can appear in the serum. CK-BB has been described as a marker for adenocarcinoma of the prostate, breast, ovary, colon, other adenocarcinomas of the gastrointestinal tract, and for small-cell anaplastic carcinoma of the lung (11). One of the most important causes of CK-BB in serum is brain damage in caseswith concomitant breakdown of the bloodbrain barrier (8). An intact blood-brain

barrier seems to block the penetration of CK-BB activity into the serum effectively. Bayer et al. (12) measured 118 U of CK-BB per liter in the cerebrospinal fluid of a patient suffering from brain damage without being able to detect CK-BB in the serum. On the other hand, a few cases have been reported of serum CK-BB activity, apparently of brain origin, in serum (13,14). In some as-yetundelineated cases, therefore, the blood-brain barrier seems to break down. The cancer patients in our study exhibited some damage to their blood-brain barrier, as judged from the amount of contrast dye present on their CAT scan. All intra-axial brain tumors with enhancement must have some disruption of the blood-brain barrier if the dye is to be seen within the brain parenchyma. We have no logical explanation for the small amount of CK-BB activity in our cancer patients’ serum as compared with those patients with contusions and infarcts. We saw no significant correlation between CK-BB as measured with the Impres-BB kit and the volume of damaged brain tissue in patients with either brain contusions, infarcts, or primary or metastatic brain cancer. We attribute this lack of correlation to an intact blood-brain barrier (which is more likely in infarcts and contusions), rapid elimination or inactivation of CK-BB, or some combination of these. Statistically, we also note that with an n value equal to ours En = 71(20 were controls)], we have a power of 0.80 for detecting an r of 0.34 with a two-tailed t-test with 0.05 significance level. That is, we had a reasonable chance of detecting a correlation as small as 0.34.

More work is needed to clarify the conditions and the transport mechanisms that can lead to the leakage of brain isoenzymes into the blood (8). Our study shows that CK-BB is a poor index of actual damaged brain tissue.

We thank International Immunoassay Laboratories for the Impres-BB reagents, and Margaret Darnell for her assistance in obtaining

specimens.

References 1. Gajda AT, Turley isoenzyme

C, McDaniel RC. Frequent occurrence ofCK-1 in coronary care patient sera [Abstracti. Clin Chem

1981;27:1027. 2. Massey TH, Barta JS. Creatine kinase isoenzymes in neonate plasma by cellulose acetate electrophoresis: albumin and adenylate kinase artifacts. Clin Chem 1982;28:1174-6. 3. Package insert for !mpres-BB. Santa Clara, CA: International Immunoassay Laboratories, Inc., 1988. 4. Selby SM, ed. Standard mathematical tables, 18th ed. Cleveland, OH: Chemical Rubber Co., 1970:13. 5. Oliver iT. A spectrophotometric method for the determination of creatine phosphokinase and myokinase. Biochem J 1955;61:116-22. 6. Rosalki SB. An improved procedure for serum creatine phosphokinase determination. J Lab Clin Med 1967;69:696-705. 7. Cohen J. Statistical power analysis for the behavioral sciences. 2nd ed. Hilladale, NJ: Lawrence Erlbaum Associates, 1988:93. 8. Lang H. The creatine kinase BE isoenzyme. hi: Lang H, ed. Creatine kinase isoenzymes: pathophysiology and clinical application. New York: Springer-Verlag, 1981:243-69. 9. Harwood SJ, Catrou PG, Cole GW. Creatine phosphokinase isoenzyme fractions in the serum of a patient struck by lightning. -

Arch Intern Med 1978;138:645-6. 10. Vladutiu AO, Schachner A, Schaefer PA, et al. Detection of creatine kinase BE isoenzyme in sera of patients undergoing aortocoronary bypass surgery. Clin Chirn Acts 1977;75:467-73. 11. Coolen RB, Pragay DA, Nosanchuk JS, et al. Elevation of brain-type creatine kinase in serum from patientswith carcinoma. Cancer 1979;44:1414-8. 12. Bayer PM, Gabi F, Granditach G, et al. Creatine kinase isoenzymes in cerebrospinal fluid in a case of brain damage. Clin

Chem 1976;22:1405-7. 13. Phillips J, Homer B, Horgan J, et al. Brain type CPK-BB and cerebral ischaemia after subarachnoid haemorrhage. Lancet 982;ii:662-3. 14. Kaste M, Hernesniemi J, Somer H, et al. Creatine kinase isoenzymesin acute brain injury. J Neurosurg 1981;55:511-5.

CLIN. CHEM. 35/4, 654-658 (1989)

Enzymic Urea Assay: a New ColorimetricMethod Based on HydrogenPeroxide Measurement FrancIs Lespinas,’ Georges Dupuy,

Francolse Revol,2 and Claude Aubry2

We describe a new enzymic colorimetric method in which urea is measured in serum by use of a single reagent mixture. Ammonia produced by urea hydrolysis, catalyzed by urease, reacts with glutamate and ATP in the presence of glutamine synthetase. The ADP so produced is assayed in reactions catalyzed sequentially by pyruvate kinase and

‘Laboratoire d’Analyses Medicales, 4, place Gainbetta, 24700 Montpon-Menesterol, France. 2bioM#{233}neux, Marcy I’Etoile, 69260 Charbonni#{232}res-les-Bains, France. 3To whom correspondence should be addressed. Received August 31, 1988; accepted January 24, 1989. 654

CLINICALCHEMISTRY, Vol. 35, No. 4, 1989

pyruvate

oxidase

in a system

that generates

hydrogen

peroxide. The hydrogen peroxide is measured at 500 or 550 nm in a reaction catalyzed by horseradish peroxidase, with phenol/4-aminophenazone as the chromogen. The reaction is complete in 15 mm at 37 #{176}C. The standard curve is linear

up to a urea concentration of 40 mmol/L. Precision is good; CVs ranged from 2.5% to 3.1%. Results by the present method compared well with those by a candidate Reference Method and are not subject to interferences from commonly used drugs and anticoagulents. Several sensitive colorimetric assays have been developed for metabolites since Trinder (1) described the coupled