Study of Forensic and Clinical Source Hemoglobin Interference with the dupont aca Ethanol Method

Transcription

1 ANNALS O F CLINICAL AND LABORATORY SCIEN CE, Vol. 16, No. 5 Copyright 1986, Institute for Clinical Science, Inc. Study of Forensic and Clinical Source Hemoglobin Interference with the dupont aca Ethanol Method RICHARD H. GADSDEN, SR., P h.d. Department of Laboratory Medicine, Medical University of South Carolina, Charleston, SC ABSTRACT The enzymatic alcohol (ethanol, ALC) analytical pack for the dupont aca is intended primarily for the analysis of serum or plasma samples containing no visibly detectable free hemoglobin (non-hemolyzed samples). It has becom e apparent, through clinical and medicolegal consultation, that this methodology has been applied for assay of ethanol content in hemolyzed samples (sample sources: clinical and forensic cases). The influence of sample free hemoglobin on the ace ethanol m ethod was investigated. W ith hem olyzed clinical samples (containing prim arily oxyhemoglobin) a significant decrease was found in known ethanol concentration at 0.5 g per dl and greater of hemoglobin w ith the influence being greatest at lower ethanol concentration. W ith forensic samples (containing varying amounts of oxy-, reduced, and m ethemoglobins), the same observation was made with m easured decrease in ethanol concentration beginning at 0.1 g per dl hemoglobin and greater. In essence, the higher the sample hemoglobin concentration the greater would be the effect in decreasing the apparent ethanol concentration. Preparation of protein-free trichloroacetic acid supernatants and their assay on the aca and by head-space gas chromatography resulted in excellent recovery of the sample ethanol. It is concluded that the effect of sample free hemoglobin is spectral in nature, and its presence results in potential and real overblanking for the aca ethanol enzymatic reaction sequence. Introduction Blood samples from living traum atized patients often show visible evidence of hemolysis. This is due primarily to skeletal m uscle crush and/or internal organ damage with resultant red cell hem o lysis. W ith the prospect of surgical interv en tio n and necessity for th e use of an esth etic agents, an assessm ent of in to x ican ts b eco m es necessary. T he resp o n sib ility of th e lab o rato ry is to determ ine and confirm8 w ith confidence the blood ethanol concentration of the victim. The dupont aca (automated clinical analyzer) is often available to the lab * dupont Instrum ents-aca, Barley Mill Plaza, Wilmington, DE /86/ $01.20 Institute for Clinical Science, Inc.

2 4 0 0 GADSDEN oratory for this rapid assessm ent. The hem olyzed clinical serum sam ple contains primarily oxyhemoglobin. Blood sam ples taken at autopsy are always hemolyzed. The degree of hem o lysis varies with respect to circum stances of death (extensive trauma) and to tim e of death and tim e to harvesting of the sample(s). These samples contain varying amounts of oxy-, m et-, and reduced hem oglobins. It has been our experience, through medicolegal consultations, th at forensic and clinical h em olyzed serum samples have been analyzed for ethanol c o n ten t using th e dupont aca alcohol (ALC) m ethod and without sample pretreatm ent. Alcohol dehydrogenase (ADH; alcoh o l: N A D + o x id o r e d u c ta s e, E C ) has been shown to be relatively specific for enzymatic reduction of ethanol (EtO H )1,7 using the m easurem ent of form ed N A D H (nicotinam ide-adeninedinucleotide, reduced) for estim ating the sam ple E to H concentration. Several published methods dem onstrate the utilization of this principle for clinical purpose Buchner and Redtzki3 recomm ended the inclusion of semicarbazide in the reaction m ixture in order to favor maximum oxidation of the EtO H substrate, w hereas in our laboratory5 tris (hydroxymethyl) am inom ethane (TRIS) is used as a trapping agent9 for the acetaldehyde and as the reaction buffer. The dupont ALC m ethod employs the following reaction sequence for the measurem ent of serum EtOH: c le o tid e (N A D H ) [o n e m o le E to H re s u lts in fo rm atio n o f one m ole of N A D H ]. T he re a c tio n is m o n ito re d bichromatically at 383 nm (the minimal spectral absorbance of N A D H ) and at 340 nm (the maximal spectral absorbance of N A D H ). P rio r to in itia tio n of th e enzym atic reaction sequence, biochromatic m easurem ent of the sample-buffer mixture serves as the blank absorbance. M aterials and Methods P r e p a r a t io n o f S e r u m S a m p l e s C o n t a in in g F r e e H e m o g l o b in a n d E to H 1. A pool of hum an serum was prepared from patient sam ples rem aining from analyses in our clinical laboratory. C riteria for individual sam ple acceptance were: clear (non-turbid), nonicteric, and have no visible presence of hemoglobin (red chromogen). The total free hem o globin of the serum pool was 25 mg per dl, as determ ined by the spectrophotometric method of H erboe6. 2. W hole blood sam ples with ethylenediam inetetraacetic acid (EDTA) as th e anticoagulant w ere obtained from healthy individuals. An hem olysate was p re p a re d by th e a d d itio n o f w h ite saponin, centrifuged to rem ove red cell ghosts, white blood cells, and any other solid. The total hemoglobin of the hemolysate was determ ined by the cyanmethemoglobin using the Coulter S-Plus *. EtO H + NAD - Acetaldehyde + N A D H +(H +) Acetaldehyde + T R IS Complex The TRIS serves as the buffer for the enzym atic reaction as well as reacting This hem olysate was used for prepara- ( trap p in g ) w ith th e form ed acetalde- tion of serum containing free hemoglohyde in order to drive the reaction to bin w hich is referred to hereafter as completion; i.e., to maximize formation of reduced nicotinamide-adenine dinu- * Coulter Electronics, Hialeah, FL.

3 STUDY OF FORENSIC AND CLINICAL SOURCE HEMOGLOBIN INTERFERENCE 40 1 Clinical source samples. The oxyhemoglobin content was m aintained at 95 to 98 p ercen t of th e total hem oglobin as monitored by the IL Co-Oximeter. 3. O pen heart, whole blood samples w ere obtained at autopsy from traum a victim s (vehicular accidents, gunshot wounds) who otherw ise w ere deem ed to be h e alth y p rio r to d eath. A utopsies w ere perform ed 20 to 24 hours postd e a th. T h e se sa m p le s w e re g ro ssly hem olysed. T hey w ere centrifuged to rem ove detritus and the total hemoglobin m easured as in # 2. The hemolysate (supernatant) was used for preparation of serum containing free hemoglobin which is re fe rre d to h e re a fte r as Forensic so u rc e sa m p le s. H e m o ly sa te s w e re pooled and analyzed with the IL Co-Oxim eter and found to have the following characteristics: 12 percent methem oglobin, 55 p ercen t oxyhem oglobin and 33 percent reduced hemoglobin. 4. D uplicate sam ples of th e pooled serum was pipeted into 5 ml aliquots. A ppropriate aliquots of the Clinical and Forensic hem olysates w ere pip eted to result in serum hem oglobin concentration of 0.05, 0.1, 0.2, 0.3, 0.5, 1.0, 2.5, t Instrum entation Labs., Lexington, MA. 5.0 and 10.0 g p er dl, respectively. The final concentration of hem oglobin and chemical forms of hemoglobin was verified by the spectrophom etric m ethods as stated in the foregoing. 5. F o r each C linical and F o ren sic source of h em oglobin c o n c en tratio n, E to H was added to approxim ate final concentrations of 50, 100, 150, and 200 mg per dl, respectively. The efficiency of p ip etin g and final concentration of E to H as determ ined by head-space gas chromatography (GC) are noted in table I. This does dem onstrate the technical pipeting proficiency in support of validation of the subsequent aca and GC analytical findings. A l c o h o l A n a l y s e s E ach of th e p re p a re d sam ples was analyzed in duplicate using the aca ALC m ethod10 and by head-space G C.4 The m ethods w ere standardized using standards prepared as described by Gadsden and Terry.4 Excellent agreem ent of the m ethods is d em o n strated in table II. These data are from results obtained in our laboratory on a routine request basis. None of these clinical serum and plasma TABLE I E th a n o l '5 0 '* E th a n o l '1 0 0 ' E th a n o l '1 5 0 ' E th a n o l C L I N I C A L X SD % CV Range F O R E N S I C X SD % CV Range *Protocol ethanol (EtOH) concentration, mg per dl. +X of determined EtOH concentration, mg per dl. These data represent proficiency of individual sample pipeting based upon EtOH recovery as determined by gas chromatography analysis from serum samples containing varying amounts of Clinical and Forensic source hemoglobin(s). See table III for the experimental matrix and table IV for the aca analysis of protein-free supernatants. (At each EtOH level, N = 36 [total N = 144]).

4 4 0 2 GADSDEN T A B L E II P r e c is io n Corre 1a tio n aca GC aca(y) v s. GC(X) cc s i X X SD X X Y SD Y These data are results of duplicate ethanol (EtOH) analysis of serum and plasma samples (N = 100) by the aca and gas chromatography methods. These were routine clinical samples having no visibly detectable evidence of free hemoglobin (hemolysis). samples contained visibly detectable evidence of hemolysis. Results and Discussion In table III are docum ented typical analytical results for serum containing varying amounts of Clinical and Forensic source of hemoglobins and for each level o f h e m o g lo b in v a ry in g a m o u n ts of E to H. T he se d a ta are e x p re sse d in term s of p ercen t recovery of E to H as determ ined by the aca m ethod com pared to that determ ined on the same sam ple by gas chrom atographic (GC) analysis (aca E to H, mg p er dl h- GC E to H, mg p er dl X 100). It is noted that, regardless of th e case source of hemoglobin, the effect is to decrease the a p p a re n t specim en E to H c o n te n t as determ ined by the aca m ethod. H ere 94 to 96 percent or less recovery of E to H is considered to be of significance. For Clinical source serum free hem o globin, th is d e p re ssio n o f a p p a re n t EtO H content is significant at 0.5 g per dl of h em o g lo b in an d g re a te r. T he reduction is greatest at the lower sample E to H concentration and w ith increasing reduction at all EtO H levels as the sample free hem oglobin content increases (example: at 1.0 g p er dl Clinical source hemoglobin, the E to H recovery is 84.7 percent at an approximate 50 mg p er dl E to H, w hereas, there is a 94.0 percent recovery at an approxim ate 200 mg per dl of EtOH). For samples containing Forensic case source of hem oglobin, the decrease in Hgb gm/dl EtOH mg/dl T A B L E III P ercent R ec o ve ry, F o r e n s ic \ aca/gc* C lin ica l% o u> Percent recovery of EtOH (aca EtOH, mg/dl -i GC EtOH, mg/dl x 100). fsera containing Forensic source hemoglobin. $Sera containing Clinical source hemoglobin. Protocol EtOH concentration, mg/dl (see table I) for determined EtOH concentration. These data represent percent recovery of E t O H, as determined by the aca method and gas chromatography analysis, from serum samples containing varying amounts of Forensic and Clinical source hemoglobin(s).

5 STUDY O F FORENSIC AND CLINICAL SOURCE HEMOGLOBIN INTERFERENCE 40 3 m easured E to H recovery by th e aca m ethod begins at a low er hem oglobin concentration than for the Clinical case so u rc e o f h e m o g lo b in. S ig n ific a n t decrease in the m easured EtO H recovery begins at 0.1 g per dl of hemoglobin w ith greater decreases noted at higher hem oglobin concentrations and at the levels of E to H co n ten t studied. The effect with Forensic source hemoglobin is th e sam e as for th e C linical source h e m o g lo b in,-th e m a g n itu d e o f th e decrease differs. In figure 1 are dem onstrated typical absorption spectra for hem oglobin from C linical source (I) and from F orensic source (II). The Clinical source sample contains prim arily oxyhem oglobin (95 p ercent of total hem oglobin) while the Forensic source sample contains a mixture of met-(12 percent), oxy-(55 percent), and reduced (33 percent) hem o g lo b in s as d e te r m in e d w ith th e IL Co-Oximeter. The wavelengths 340 and 383 nm are n o ted on th e resp ectiv e spectra. The absorptivity at these wavelengths for Clinical source hem oglobin are relatively equal; thus, blanking for hem olyzed clinical sam ples (sera) does not result in significant interference to the aca EtO H m ethod until higher concentrations of free oxyhem oglobin are p re s e n t. W ith th e F o re n sic so u rc e hemoglobin, it is noted that the absorptivity at 383 nm exceeds that at 340 nm. This significant difference in absorptivities results in sam ple overblanking at lower mixed hem oglobin concentrations; thus, the m easured decrease in the aca E to H co n ten t of F orensic sam ples at lower sample hem oglobin concentration is noted. In o rd er to validate th e concept of overblanking, protein-free supernatants (PFS) of the Clinical and Forensic hemoglobin- and EtO H -containing sera (aliquots of samples shown in table III) were prepared. The PFS was prepared by the addition of one ml of the sample to two F i g u r e 1. Absorption spectra scans of Clinical (I) and Forensic (II) source hemoglobin(s). The Clinical source sample contains 95% oxyhemoglobin; the Forensic source sample contains 12 percent met-, 55 percent oxy-, and 33 percent reduced hemoglobins. The wavelengths of 340 nm and 383 nm are noted on each spectra. (See Results and Discussion for pertinence.) ml of six percent freshly prepared trichloroacetic acid (TCA; 1:5 dilution of 30 percent TCA; stable when refrigerated), covered and mixed well, centrifuged for five m inutes at 2000 X g, the supernatant rem oved and analyzed11 by the aca m ethod. Results are shown in table IV. These data may be com pared to those show n in ta b le I w h ich are re s u lts o btained by GC analysis of non-tca tre a te d aliquots. T he results indicate excellent absolute recovery of E to H from TCA tr e a te d sa m p le P F S s as

6 40 4 GADSDEN T A B L E IV E th a n o l '5 0 * E th a n o l '1 0 0 ' E th a n o l '150' E th a n o l '2 0 0 ' C L I N I C A L X SD % CV Range F O R E N S I C X SD % CV Range Represents the approzimate EtOH concentration. trepresents the absolute EtOH concentration, mg per dl. These data represent EtOH content of protein-free supernatants prepared from serum samples containing varying concentration of Clinical and Forensic source hemoglobin(s) (see table II) as determined b y the aca method. See table I for correlation with gas chromatography analysis of the hemoglobin-containing samples. (At each EtOH level, N = 36 [total N - 144]). d e te r m in e d b y th e aca e n z y m a tic m ethod co m p ared to E to H ab so lu te recovery from identical hemoglobin-containing sam ple aliquots as analyzed by GC analysis. Extended studies of Clinical and Forensic source hemoglobin(s) have verified these findings. D ata resu ltin g from absolute E to H analysis of hem oglobin-containing sera as in table III by the aca and GC m ethods w e re c o m p a re d sta tistic a lly. T h e se results are show n in figures 2 and 3. W ith th e Clinical source hem oglobincontaining samples, the slope was with an overall bias of 5.9. W ith the Forensic source hemoglobin-containing sam ples, th e slope was 1.00 w ith an over-all bias of These data confirm the more profound effect of Forensic source hem oglobin on the aca m ethod for E to H analysis. The slopes are linear in n atu re, thus dem o n stratin g an aca o v e rb la n k in g e q u iv a le n t to sam p le hemoglobin concentration. However, it is not infered that aca E to H results may be corrected if serum free hem oglobin content is known. There would be variation of oxyhemoglobin content in Clinical source hem olyzed serum sam ples and, most importantly, variation in quantitative admixtures of chemical hemoglobin species in Forensic source hem o lyzed samples. Sum m ary T h e s e s tu d ie s d e m o n s tr a te th a t patient serum containing free hem oglobin has a negative bias (i.e., there is a decrease in the m easured level) for F i g u r e 2. Correlation of aca EtOH content vs. GC EtOH content of serum samples containing varying concentrations of Clinical source hemoglobin. At each hemoglobin level the samples contained varying levels of EtOH (see table III for matrix). At increasing hemoglobin concentration recovery of EtO H decreases as determ ined by the aca method when com pared to the definitive GC m ethod. (At each EtOH level, n = 1 8 (total n = 7 2 ).

7 STUDY OF FORENSIC AND CLINICAL SOURCE HEMOGLOBIN INTERFERENCE F i g u r e 3. Correlation of the aca EtO H content vs. GC EtO H content of serum samples containing varying concentrations of Forensic source hemoglobin. At each hemoglobin level the samples contained varying levels of EtO H (see table III for matrix). At increasing hem oglobin concentration recovery of EtOH decreases as determ ined by the aca method when compared to the definitive GC method. (At each EtOH level, n = 18 (total N = 72). E to H analysis using th e d u P o n t aca m ethod. T he d eg ree of negative bias varies with respect to: the sample concen tratio n of hem oglobin; th e sam ple c o n c e n tra tio n o f E to H ; a n d, w ith respect to the chemical form(s) of hem o globin present. The aca m ethod for ethanol analysis is shown to be accurate and precise when nonhemolyzed serum /plasma is the sample an d for analysis o f p ro te in -fre e su p e rn a ta n ts p re p a re d from sam ples containing free hem oglobin (the hemolyzed sample). Conclusion H em olyzed sam ples (serum /plasm a showing visible evidence of free hemoglobin) from a Clinical source should be pretreated w ith TCA and the resultant clear PFS subjected to analysis for E to H by the aca m ethod. Results com pare well with the definitive GC method. All F orensic source serum /plasm a samples show evidence of free hemoglobin, and the chemical forms of hemoglobin present are m ultiple in nature. Thus, all Forensic source samples should be pretreated with TCA and the resultant clear PFS utilized for analysis of E to H by the aca m ethod. Results com pare well with the definitive GC method. T his stu d y focuses on th e sp ectral interference (overblanking) with respect to a single analyte (EtO H ) on a single analytical system (the dupont aca). The clinical laboratorian should be aware of the potential and/or real interference by any quantity of sample free hem oglobin using discrete analyzers. This study does validate th at bichrom atic blanking and analysis will result in inaccurate analytical results. Acknowledgments The author expresses recognition and appreciation to Mrs. Rachel M. Crews, MT (ASCP) and to Mrs. Sharon W. Callaway, BSMT (ASCP) for their diligent and expert technical assistance with the analytical portion of this study. Support was received from Du Pont in supplying adequate aca ALC packs and from Mr. John W. M cchristian Jr., attorney-at-law, El Paso, TX and the Southern Pacific Transportation Co. for impetus and fiscal aspects. Bibliography 1. B a r r o w, E. S. G. and L e v i n e, S.: Oxidation of alcohols by yeast alcohol dehydrogenase and by living cells. Arch. Biochem. Biophys. 47: , B r in k, N. G., B o n n i c h s e n, R. K., and T h e o r - e l l, H. A.: A modified method for the enzymatic microdetermination of ethanol. Arch. Pharmacol. Toxicol. 10: , B u c h e r, T. and R e d e t z k i, H.: Eine spezifische photometriche Bestimming von Athylalkohol auf fermentativen wege. Klin. Wochenschr. 29: , G a d s d e n, R. H., Sr., and T erry, C. S.: Alcohols in biogical fluids by gas chrom atography (autom ated head-sp ace m ethod). S elected Methods of Emergency Toxicology of the series Selected Methods of Clinical Chemistry, vol. 11. Frings, C. S. and Faulkner, W. R., eds. Washington, AACC Press, 1986, pp G a d s d e n, R. H., Sr., and T a y l o r, E. H.: E thanol in biological fluids by enzym ic analysis. Selected Methods of Emergency Toxicology of the series Selected Methods of Clinical Chemistry, vol. 11. Frings, C. S. and Faulkner, W. R., eds. W ashington, AACC P ress, 1986, pp

8 40 6 GADSDEN 6. H a r b o e, M.: A m ethod for determ ination of hemoglobin in plasma by near-ultraviolet spectrophotom etry. Scand. J. Lab. C lin. Invest. 11:66-70, J o n e s, D., G e r b e r, L. P., and D r e l l, W.: A rapid enzymatic method for estimating ethanol in body fluids. Clin. Chem. 16: , Pa p p a s, A. A., G a d s d e n, R. H., S r., and T a y l o r, E. H.: Serum osomolality in acute intoxication: A prospective clinical study. Am. J. Clin. Path. 84:74-79, Merck Index, 9th ed.: Rahway, NJ. Merck and Co., Inc., 1976, p Test Methodology for the aca; ALC, ethyl alcohol. Instrum ent Manual, vol. 3: Chemistry, Du Pont Co., Diag. and Bioresch. Systems Div., Wilmington, DE, June Technical Bulletin: Performance characteristics of the Du Pont ethyl alcohol (ALC) method with plasma and supernatants prepared from whole blood. Compì.: J. J. Dicht, Du Pont Co., Diag. and Bioresch. Systems Div., Wilmington, DE, Dec