Clinical chemistry - ACS Publications - American Chemical Society

---

Anal. Chem. 1985, 57, 114R-130R (599) Xu, X.; LI, R.;Wang, G., Zhongnan Kuangye Xueyuan Xuebao, (3), 63 (1983); Chem. Abstr., 700, 1 6 7 2 3 1 ~ . (600) Yakovieva, E. F.; Rybakova, E. V.; Zharkova, d. N.. Meto& Khlm. Podgot. Anal. Meter., Mater. Semln. 143-7 (1982);Chem. Abstr., 707, 58534x. (601) Yakovieva, V. S.; Kondakova, S. V., Ogneupory: Issled. Proizvod. 66-9 (1983); Chem. Abstr., 700, 90214h. (602) Yamagushl, I., SumnomO Tokushu Klnzoku Giho, 6 , 71 (1982): Chem. Abstr., 9 7 . 2 2 9 2 8 4 ~ . (603) Yamaguchi, S.; Fujino, Y.; Nagata, S.; Kaneko, H.; Hashimoto, K.; Hirabayashi, M., Nucl. Instrum. Methods Phys. Res., 278 (I-3), 598 (1983); Chem. Abstr., 700, 1 3 1 6 8 0 ~ . (604) Yamakawa, K.; Yonezawa, S.; Yoshizawa, S., Anal. Chem. Symp. Ser. 1983, 77 (Chem. Sens.), 758-61; Chem. Abstr., 700, 1431656. (605) Yamamoto, A.; Watahiki, S.; Shimizu, M.; Konishi, M.,Nippon Kinzoku Gakkal Kaiho, 22 (7), 658 (1983); Chem. Abstr., 9 9 , 198814d. (606) Yamamoto, Y.; Hasebe, K.; Kambara, T., Anal. Chem., 55 (12), 1942 (1983); Chem. Abstr., 9 9 , 1 1 5 1 0 8 ~ . (607) Yang, S.; Wang, H., Fenxi Huaxue, 70 ( I l ) , 701 (1982); Chem. Abstr., 9 9 , 32316t. (608) Yasuda, H.; Nishisaka, K.; Ito, R.; Yasube, T., Totubetswu HokokushoNippon Tekko Kyokal, 3 4 , 375 (1982); Chem. Abstr., 9 8 , 3 8 6 3 4 ~ . (609) Yoshimori, T.; Taniuchi, T., Bunsekl Kagaku, 32 (g), 547 (1983); Chem. Abstr., 9 9 , 2051756. (610) Yu, R.; Wen, 2.; Fei, C., Huaxue Xuebao (Zengkan), 136-40 (1981); Chem. Abstr., 9 8 , 2092069. (611) Yu, X.; Cal, R.; Liang, H., Wuhan Daxue Xuebao, Zkan Kexueban, (I), 105 (1984); Chem. Abstr., 707, 32511n. (612) Zamaraeev, V. P.; Popovskaya, N. A., Zavod. Lab., 49(6), 21 (1983); Chem. Abstr., 9 9 , 98239b.

(613) Zatka. V. J.; Zeiding, N., Anal. Chem., 56 (9), 1734 (1984); Chem. Abstr., 707, 47793.9. (614) Zeng, S.; Cai, R.; Yu, X.; Zeng, Y., GaWeng Xuexiao Huaxue Xuebao, 4 (4), 444 (1983); Chem. Abstr., 9 9 , 2 0 5 1 4 8 ~ . (815) Zhan, C., Huaxue ShJle, 25 (I), 12 (1984); Chem. Abstr., 707, 103125~. (616) Zhang, C. P.; Qi, D. Y.; Zhou, T. Z., Talanta, 29 (12)8 I l l 9 (1982); Chem. Abstr., 9 8 , 10043517. (617) Zhang, 0.;Wang, S.; Wang, H., Huaxue Xuebao, 47 (I), 90 (1983); Chem. Abstr., 9 8 , 1 5 4 5 6 4 ~ . (618) Zhang, L.; Lai, R.; Zhang, Z . , Fenxi Huaxue, 72 (5),398 (1984); Chem Abstr .. 70 7 . 031830. (819) Zhang, X.;’Xu, L.;Xu, 6,Huaxue ShJle, 24 ( l o ) ,298 (1983); Chem. Abstr ., 707, 6 5 0 7 6 ~ . (820) Zhang, Y., Fenxi Huaxue, 70 (S), 555 (1982); Chem. Abstr., 9 9 , 132874s. (621) Zheng, S., Fenxl Huaxue, 7 7 (9), 884 (1983); Chem. Abstr., 700, 61025s. (622) Zheng, Y.; Wang, Z., FenxiHuaxue, 7 7 (E), 586 (1983); Chem. Abstr., 700, 7 9 0 1 9 ~ . (623) Zheng, Y; Wang, Z.,Huaxue Shy/, 5(2), 70 (1983);Chem. Abstr., 9 9 , 47178~. (824) Zhou, N.; Frech, W.; Lundberg, E., Anal. Chlm. Acta, 753, 23 (1983); Chem. Abstr.. 9 9 , 224324r. (625) Zhou, T.; Qi, D.; Zhang, C., Huaxue Xuebao, 47 (3), 237 (1983); Chem Abstr ., 9 9 , 32393r. (826) Zhukova, M. P.,Zh. Anal. Khim., 3 7 ( 9 ) , 1629 (1982);Chem. Abstr., 98, 1 0 8 0 4 ~ . (627) Zhukova, M. P.; Zueva, N. A., Zavod. Lab., 48(12), 30 (1982); Chem. Abstr., 9 8 , 82993m.

.

.

Clinical Chemistry K. E. Stinshoff,* J. W. Freytag, P. F. Laska, and Lori Gill-Pazaris Diagnostic & Bioresearch Systems Division, Biomedical Products Department, E. I. du Pont de Nemours & Company, Wilmington, Delaware 19898

“Clinical chemistry is the application of the science of chemistry to the understanding of the human in health and disease” (1). To fulfill this mission, clinical chemistry is, of necessity, an applied science which must be able to support medical decisions by means of reliable, quick, and economical analyses. Equally important is its involvement with research into the development of the biochemical bases of medical diagnosis and patient management. In both areas the subjects of clinical chemistry overlap with those of other disciplines. It would exceed the scope of this review to detail all developments that have been published over the last 2 years. Instead, we have concentrated our efforts on those areas that we believe have the greatest impact on the routine use of clinical laboratories and on the improvement of medical care in the future. The routine laboratory will profit most from recent advances in instrumentation and data management. These advances are driven by both technological progress and economic considerations. The future role of the clinical laboratory, however, will also be determined by progress in molecular biology that leads to the identification of new analytes of medical interest. Establishment of the utility and reliable analysis of these markers will be the greatest challenge in this area of clinical chemistry. Developments in the field of immunological assays have contributed much to meet this challenge. Immunoassays, therefore, will be an additional major concern of this review. INSTRUMENTATION AND DATA MANAGEMENT To evaluate developments in instrumentation and data mana ement for the clinical laboratory, important economic consi erations have to be taken into account. Progress in sensor technology will also be considered, followed by a review of new instrument technology. Finally, new developments in data management will be discussed. Economics. As of October 1984 virtually all U.S. hospitals receiving Medicare payments have come under a new pro-

d

ll4R

spective payment system known as Diagnostic Related Groups or DRGs (2). Clinical laboratories are now treated as cost rather than profit centers, which makes justification of new analytical instrumentation more elaborate and difficult ( 3 , 4 ) . Currently, only operating costs are covered under DRGs while capital expenses are reimbursed at cost. This is due to change by October of 1986 when capital costs are to become prospectively reimbursed as well (5). These changes leave a narrow time window for hospitals to easily acquire advanced technology. A result of this is a current rush to market of systems vying to be the most cost effective general purpose analyzer. These have generallytaken the form of highly reliable, random access, discrete sample processors with throughputs in the range of 100-700 tests per hour and possessing extended calibration intervals (6). Another result is the worldwide development of lost cost instruments suitable for use in private and group medical practices which are not yet constrained by DRGs (7). Nearly all of the new analytical systems take advantage of low cost computing power in order to reduce operator time and the required skill level. Personnel cosb are as important in most cases as the cost of reagents and hardware (8-11). Sensors. Recent general reviews of relevant biomedical sensor technology are available (12,13). Linear photodiode arrays continue to be used in various spectrophotometers (14) and LC detectors (15). They have been applied in high-resolution densitometry (16) and reviewed with respect to multiwavelength derivative spectroscopy for clinical use (I7). Multicomponent infrared analysis of serum lipids has also been shown (28).Multiwavelength analysis of hemoglobin fractions is of continued interest (19-22). New co-oximeters (23) use five or more wavelengths to resolve and quantitate hemoglobin fractions and thereby accurately compute oxygen saturation of arterial blood. Polychromatic analysis continues to be a useful technique for improving the accuracy of other clinical assays (24,%), including some reflectance applications (26). Clinical applications of ion selective and potentiometric electrodes have been extensively reviewed in recent literature

0003-2700/85/0357-114R3$06.50/0 0 1985 American Chemical Society

CLINICAL CHEMISTRY

.

.-

I

d

(27-30). ISE systems for measurement of sodium and potassium have received widespread clinical acceptance, frequently replacing flame photometry for routine analysis. These systems have become easy to operate and maintain. They are ideally suited for use in small, remote or STAT laboratories especially since whole blood systems are now available from a t least five manufacturers (31). Electrodes for chloride, total carbon dioxide, total calcium, and ionized calcium are also in routine use. Ionized calcium measurement and interpretation remain a source of controversy and have been extensivelv reviewed (32). Immovements in selectivitv w t h neutral &er-based membrank electrodes have spark& renewed interest in K' (33). Na' (34). and Li' ISE's A differentd pH technique has provlded c l m i d y useful results for glucose, urea, and creatinine (35,36). Interferences with K+determination in urine continue to be discussed (37) and some significant improvements have been made using silicone rubber membranes (38). Little progress is evident in the commercialization of electrode systems for enzyme analysis. An amylase electrode instrument has been described (39).A laasy carbon electrode has amperometrically determined LDfl enzyme activity (40) by kinetidy monitoring the concentration of NADH cofactor.

Immobilized enzyme-electrode systems have been ap lied to the measurement of creatine and creatinine (41)andlto the immunoassay of hepatitis B surface antigen (42). Chemfet sensors appear most likely to succeed a t in vivo or bedside applications (43, 44). Fiberoptic probes with spectrophotometric, fluorescence and reflectance capabilities are available,particularly for in vivo measurement of pH, pOP. and hemoglobin. Although efforts continue to perfect these technologies, alternate approaches for clinical lahoratory in vitro instrumentation appear more practical a t this time. Instrument Technology. The current economic climate may be seen as driving clinical instrument technology in three directions. First, the traditional clinical labs need low operating cost systems with comprehensive analyte flexibility and adaptability to changing workloads. Second, the emerging physician group and solo practice labs need low purchase cost, easy to operate devices with overall performance equivalent to that provided by the traditional lab. These latter instruments need a smaller test menu which ideally integrates blood cell counting, hemoglobin, common coagulation tests, traditional chemistries such as glucose, urea, etc., and selected enzymes. Third, biochemical diagnostic testing, in general, may be well served by eliminating instrumentation and s u p port personnel wherever possible. High performance 'dipstick" type assays capable of confirming or disallowing a hypothesis such as "disease X is present" or 'this woman is pregnant" have always been desirable. It is only recently that biotechnology is available to produce such systems. Perhaps the most active area of innovative instrument technology has been that of dry 'slide" or multilayer film chemistries in combination with reflectance, fluorescence, and potentiometric analysis (45). The principal instruments in the clinical marketplace and their performance have been widely reviewed (46-55). The last 2 years have seen the commercial success of fluorescence polarization via the Abbott TDx (56). Used principally for therapeutic drug assays the technology has recently been modified for random access, discrete sample processing with a unit dose packaging of reagents (57). Fluorescence is now widely used for therapeutic drug assays. Other analyzers (58-60)have been modified for fluorescence and optimized for random access, discrete analysis. This broadening of existing designs is also evident in the Beckman ASTRA which has been modified to assay six different enzymes (61) and to link two instruments to increase its test menu, throughput, and data handling capability. Random access to samples and reagents in modern analyzer designs has its roots in both computer and robotic technob gies. General discussions of robotics are readily available (6244). Two new analyzers (65,66) are noteworthy for their use of software to continuously optimize hardware resources to handle the complex scheduling of reagent and sample additions. Several general purpose robots suitable for sample handling have been demonstrated (67,68). The cost effectiveness and integration of this technology into clinical laboratory operations remain as challenges Some diverse developments are brought together a t this point where progress has been seen either in the integration of technologies or in new applications of older principles. Hematofluorometry or fluorescent analysis of whole blood samples has been accomplished at Bell Laboratories (69) and elsewhere. This front face fluorescence technique may he optimally suited for porphyrin analysis in the case of lead poisoning (70). Heterogeneous immunoassays are becoming increasingly automated. One system uses superparamagnetic particles, 1 pm in diameter, as the solid phase support. They stay suspended in solution until removed with a simple permanent magnet. Thyroid stimulating hormone, total T3, T4, and free T4 are readily assayed (71). A combination of technologies has been assembled in a radiolabeled heterogeneous immunoassay for allergen specific IgE testing (72).A parallel array of cellulose threads forms the solid phase support in a way which allows large panels of tests to be run on a single sample without volume dependent pipetting. Polaroid film is used for the radiograph, which is scanned with a small infrared densitometer. Another novel separation technique has been described (73). usine evanescent wave detection of totallv internally reflected cght w kinetically mnniwr the reactiok of an antigen bnund t o a glass surface. ANALYTICAL CHEMISTRY. VOL. 57. NO. 5. APRIL 1985

11SR

CLINICAL CHEMISTRY

Particle counting immunoassay (PACIA) continues to be developed (74, 75). This technology offers instrumentation simplicity and low net reagent costs. Like any other new approach, however, it must displace widely used existing technology with a equivalent range of analytes. Other nonisotopic immunoassay techniques recently discussed include time-resolved fluorescence (76) and chemiluminescence (77). Flow injection analysis has been recently reviewed (78,79) and although capable of many clinical assays there is little reason to expect economic success of the technology in clinical chemistry at this time. Quite distinct from flow injection analysis is flow cytometry which brings together many analytical techniques to sort and/or classify cells and particles (80-82). This technology has been extended to the automation of microscopic urinalysis by International Remote Imaging Systems in the yellow IRIS system (83). Two-dimensional gel electrophoresis is a technique for separating and visualizing proteins in biological specimens (84-86). The separation is achieved first with respect to charge by an isoelectric focusing step in one dimension and then with respect to size and charge by electrophoresis in a second dimension. Visualization is often accomplished with direct staining or by autoradiography. Each gel may produce hundreds of spots, presenting significant problems in quantitation and interpretation. The high cost of automation currently precludes the technique from routine clinical use. Recent efforts however, to apply laser (87) and video scanners (88) and software for image enhancement and pattern recognition to electrophoresis strips (89)may point to a new direction in diagnostics (90). Other techniques that currently are used in medical research and may gain acceptance in the clinical chemical laboratory are nuclear magnetic resonance, particularly 31PNMR (91), and laser-microprobe mass analysis (LAMMA) (92). Data Management. Over 60% of those hospital laboratories serving more than 300 beds now have computerized data management systems. These are generally minicomputer based and are highly customized to manage the data flow within the laboratory and between the laboratory and physician or hospital information systems. About 50% of smaller laboratories have access to some form of computerization. These laboratories are presently experiencin some of the greatest change with respect to com uters, as jocumented in two recent surveys (93, 94). Unfer the new prospective reimbursement system, the clinical laboratory is accounted as a cost center. This means that the ability to control laboratory use, automate manual procedures, and to measure and document productivity have become more important goals in the clinical laboratory (95,96). The introduction of DRG’s and the rapid expansion of computer technology are helping to change the mix of hardware and software in clinical laboratories. The large, minicomputer based, multitasking, multiuser systems have not witnessed dramatic change in the last 2 years relative to microcomputer systems in the laboratory. Microor personal computer applications have not changed a great deal either, but what has changed is the price to performance ratio of the hardware. Useful application software has proliferated and computer literacy has increased dramatically. Current pro ess reports in medical and laboratory computing are accessib e in the annual proceedings of major conferences (97-101). A summary list of microcomputer applications is provided in Table I with references to recent work in many of the areas. The single user microcomputer has become an important management tool (102) as general purpose business tools are frequently adapted for use by the laboratory supervisor. The flexibility of the microcomputer is often cited as a key to its success in the laboratory. Its low cost, portability, ease of programming, and reconfigurability have solved many problems that would simply take too long or cost too much to solve by any other means. The most profitable applications have been in quality control and other data capture, text editing, and reporting. The expectations of the users are generally that microcomputers will play a bigger role in cost control and assist with complex problems of planning and scheduling (103, 104). There are very limited data documenting the cost effectiveness of clinical laboratory computerization (189-191). Justification notwithstanding, computers are often viewed as a means to improve the cost effectiveness of test result com-

ff

116R 0 ANALYTICAL CHEMISTRY, VOL. 57, NO. 5, APRIL 1985

Table I. Miorocolpputer Applications application calibrations communica-

tions data base instrument interfacing interpretive reporting

general functions standard curves nonlinear assay calculations data base searching remote system diagnostics networking record keeping (all types) data acquisition control and operatioe quality control and monitoring protein analysis electrolytes and blood gases organ profiles biochemical profiles therapeutic drug msnitoring graphic presentations diagnostic coding (SNOP, SNOMED)

inventory control preventive maintenance

stock level maintenance cost optimization maintenance logs reminder services

planning and accounting

staff and outpatient scheduling budgeting, forecasting cost accounting utilization review sample labeling and reading

production

quality control Levey-Jennings plots multirule procedures trend analysis cumulative sum procedures average of normals patient data checks statistics product evaluations instrument correlation simulations reference intervals text editing workload recording

manuals, protocols result reporting productivity analysis

references

105-107 108, 109

110-116 117-120 121-124

125-127 128-134 135-137 138-140

141-144 145, 146 147

148-191 152-154 155 156

157 158 159-161 95 162-164

165, 166 167-170 171 172, 173 174, 175 176-178 179 180

181, 182 183

184 185-188

munications, data collection, record keeping, and management support systems. Given appropriate systems anaiysis of laboratory functions (192),networks of microcomputers are becoming a major element in the computerization of small to mid-sized laboratories (193-195). This systems analysis is often missing in smaller institutions particularly with respect to stand-alone microcomputers. Trial and error with offthe-shelf software and BASIC language programming is frequently the order of the day (196). It is not clear at this time that the inherent power of the microcomputer is anywhere near optimally used, given the high costs of development, installation, and maintenance of useful software. There are some emerging, directly profitable, data management applications which can be understood economically in terms of the Tax Equity and Fiscal Responsibility Act of 1982 (TEFRA) (197). A part of this legislation eliminated separate billing for hospital laboratory physician services,thus bundling all reimbursement through the hospital at a fixed rate per diagnosis. An effective strategy then, for laboratory physicians to increase their revenue, is to become involved in consultations which are reimbursable (198,199). Therapeutic drug monitoring programs for consultations are now available which function at the near expert level for optimizing drug dosage (200-202). Interpretative reporting of laboratory results is well suited to the microcomputer where the domain of knowledge is relatively complete and accepted (203-206).

CLINICAL CHEMISTRY

Table 11. Instrument Based Data Management Functions real time mode

batch mode

instrument data acquisition data filing system status monitoring result calculation, reporting QC limit checking inventory, workload counting on line help system diagnostics multiinstrument management

host computer communications log sheets and data entry file review, edit, maintenance collated result reporting QC analysis and reporting productivity analysis instrument calibration maintenance recorda remote “factorv” diaanostics

Artificial intelligence approaches are only just beginning to have an impact on laboratory data interpretation (207-209). The Bell Laboratories UNIX operating system is turning up in personal computers (210,211) and recently in laboratory systems. This is a mature, multiuser, multitasking operating system. Although it is relatively hardware intensive, its high performance allows sophisticated application environments to be created easily. Software written for UNIX and the UCSD p-system is almost completely portable between widely differing hardware systems (212). This can enhance the success of a commercial software enterprise. The use of off-the-shelf software for laboratory data management is increasing. Several instrument vendors are offering the BBN Research Systems RS/ 1package for use on high end personal computer accessories (213,214). Much of the pubIished applications work is also being done with powerful spreadsheets like Lotus 1-2-3 and dbase 11. There are no standards as yet for laboratory data management systems (215,216). Most instruments and data management systems have low level (RS-232) communication capability, but at higher levels of protocol each system is unique. Microcomputers are frequently attached to those systems a~ data buffers and protocol converters for result reporting or communication with other systems. Biovation, Creative Computer Applications, and EDMAC among other corporations provide highly specialized data acquisition terminals and keypads for workstation environments like urinalysis and hematology where a large fraction of the data is acquired by humans. These special terminals then store, format and forwatd their data to other computer devices. Biovation Corp. and Dawning Technologies Corp. have developed specialized workstation terminals which are well adapted to convert instrument protocols to host computer protocols and vice versa, while presenting a uniform screen format a t each workstation. Comprehensive data management systems are offered by most instrument system manufacturers as adjuncts to the instrument or as internal components of the system. Table I1 lists some of the data management functions of computer systems found in or attached to instruments today. The table distinguishes between those functions which are generally carried out while the analyzer is operating (real time) and those functions carried out more or less on operator demand (batch mode). It is interesting to observe that former “batch mode“ functions such as quality control interpretation and instrument calibration are becoming ”real time” functions of many systems.

IMMUNOASSAYS The need to measure very specific parameters in low concentrations has triggered a high interest in immunoassay techniques. Overviews of the rapid development in this area of clinical chemistry are given in ref 217-219. Most of the advances over the past 2 years in the field of immunoassay involve applications or modifications of existing detection systems or the development of monoclonal antibodies for use in assays. The most common detection systems, including those employingradioisotopes, enzymes, fluorescence markers, chemiluminescence, and light-scattering techniques, will be reviewed. Developments in use of monoclonal antibodies for immunoassay will also be discussed. In addition, it should be mentioned that a number of articles on solid phase coating technologies have ap eared recently (220-224). Radioimmunoassay. Devegpments in radioimmunoassays are less frequent than before due to restrictive laws realating the use of radioactivity and the increased emphasis placed

on the search for alternatives to radiolabels. Radioimmunoassays are heterogeneous systems in that the signal cannot usually be stopped or started and separation of free from bound radioisotope is required. However, an article on the use of internal quenching agents such as bismuth oxide in homogeneous radioimmunoassay systems (225) appeared during an earlier review period. Many different radioimmunoassay types have been used for measurement of a variety of analytes including proteins, haptens and drugs. Radioimmunoassaysare being increasingly adapted for direct measurement of analytes previously requiring lengthy extraction or column separation procedures to eliminate interference from cross-reacting substances (226-228). Other developments also aim at simplifying RIA procedures while retaining their high sensitivity. Centrifugation was avoided by the use of magnetic particles coupled to antibody or antigen (229). Centrifugation is still necessary but less critical in a soluble phase system using microencapsulated antibodies (230). A novel automated RIA for the measurement of haptens, which is based on a solid phase technology, is discussed in ref 231. Enzyme Immunoassays. Enzymes were the first markers to successfully replace radioisotopes in immunoassays. Since this use was first described, literature on enzyme immunoassays has far outgrown that on radioimmunoassays. For a comprehensive overview refer to ref 232 and 233. Enzyme immunoassays also replace other antibody based techniques such as hema glutination, fluorescent antibody staining techniques, an diffusion in gel (234). Enzyme based assays have been shown to be less sensitive than equivalent radioisotope based assays (235),although the enzyme signal may potentially be amplified to increase sensitivity of the assay, reduce interferences, or both. A number of recent reports have discussed methods for accomplishing this including use of the biotin/avidin system (236-240), antibody bridging technique (241,242),or antibody chimeras (243). Increases in sensitivity over standard EIA of greater than 50-fold have been observed in some cases. Enzyme immunoassayscan be grouped into three categories; homogeneous, heterogeneous soluble phase, and heterogeneous. Homogeneous assays as utilized in the EMIT technology have been limited to the measurement of drugs and other haptens of low molecular weight at relatively high concentrations. Recent articles describe the application of the EMIT technology to new haptens (244)and the performance of the technology on automated systems (245-247). A variation of the EMIT technology, enzyme modulator immunoassay (EMMIA), has also been recently reviewed (248). Heterogeneous enzyme immunoassays are characterized by the fact that no change in enzyme activity occurs when the enzyme conjugated ligand binds to the antibody. A meaningful enzyme signal, therefore, can only be determined after separation of free from bound enzyme conjugate. Variations of this system include coupling of antibody to particles. In one commercially available competitive system, free and labeled analyte compete for binding sites on antibody coupled to bacterial cell wall. A recent report describes the adaptation of this method for semiautomated measurement on a continuous flow analyzer fitted with a microfilter (249). Another heterogeneous enzyme immunoassay, the ELISA technique, has also been extensively described (250-252). A special application of this method is the measurement of analytes in serum by amperometric detefmination of the product of an enzyme reaction using a PO, sensor (253). Other antibody coated solid phase ELISA assays have been discussed recently, including one in which antibody in serum is detected by its ability to bind enzyme labeled antigen (254). Antigen coated solid phase ELISA systems have also been described. One novel technique utilizes antigen (e.g., immune complex) absorbed to special cells (255, 256). In another technique an antigen coated column is used to remove free enzyme-antibody conjugate remaining after addition of free analyte. The signal mediated by the remaining enzyme-antibody conjugate is detected in the eluant (257). Quantities as low as mol of analyte/L can be detected and the technique is adaptable to automation. Finally, systems have been described in which the analyte to be measured is itself absorbed to the solid phase (258). In one such system the anti en is collected on swabs and then absorbed onto treated beafs (259). This method has been applied to the measure-

6i

ANALYTICAL CHEMISTRY, VOL. 57,

NO. 5,

APRIL 1985

* 117R

CLINICAL CHEMISTRY

ment of bacterial antigens and has demonstrated good sensitivity and specificity. Fluorescent Immunoassay. FIA traditionally has suffered from interference problems in serum and the need for specific instrumentation. Both problems now seem to have been overcome and recent reviews (260-262) demonstrate the current interest. Among the publications to be mentioned are those describing the fluorescent energy transfer assays for automated fluorescence analyzers. In this technique the reaction of a fluorescent-labeledantigen with a quencher-labeled antibody is utilized (260). A related homogeneous technique is the substrate labeled fluorescent immunoassay (SLFIA) (263). This is based on the enzymatic cleavage of a fluorogenic substrate coupled to the antigen. Antigen-antibody reaction reduces the availability of the substrate. This method has been applied to the quantitation of drugs (264-269), haptens (270), and proteins (271), and the system can be used in conjunction with automated clinical analyzers (272, 273). An interesting adaptation of the SLFIA technique describes the simultaneous measurement of two drugs by use of different fluorogenic enzyme substrates whose hydrolysis products can be excited by different wavelengths (274). In another article the use of solid-phase reagent strips for measurement of drugs by the SLFIA technique is reported (275). All components of the assay are incorporated in one of two possible dry reagent formats, and the assay is run in either a “reversible complex formation” or conventional competitive binding mode. Automated equipment has been developed to measure fluorescence and reflectance from these strips (276). The fluorescent polarization immunoassay (FPIA) has been applied to the quantitation of a number of drugs (277-279). Recent reports have shown these assays to be free of significant interference from bilirubin, hemoglobin, and triglycerides (280, 281). As mentioned above, background interference can limit sensitivity and detection levels, particularly in homogeneous assays. There are a number of methods for reducing or avoiding background interference, including use of rate determinations or fluorescent probes which fluoresce outside the range of natural interferences. Reports of alternative fluorescent probes such as porphyrins and chlorophylls have appeared (282). These substances have large Stokes’ shifts and are less susceptible to interference caused by autofluorescing substances and light scattering. In addition, sample pretreatment methods such as laser photobleaching have been described. In this method, an intense narrow band of laser light is used to photolyze unwanted chromophores in human serum and leave other serum components intact (283). The technique is not applicable to some methods such as FPIA and may reduce assay sensitivity in others. Another method of eliminating background interference, described in earlier reports, is the principle of time-resolved fluorescent immunoassay (TR-FIA) (284). In this heterogeneous technique, compounds such as the lanthanide chelates, which have exceptionally large Stokes’ shifts and high quantum yields, are used. This method is based on the fact that lanthanide chelates have longer decay times than naturally occurring fluorophores. Recent articles describe applications of the technique (285, 286), as well as automated systems employing it (287). Sensitivities equal to or better than RIA and EIA have been observed. Also, wide ranges of detection and good counting precision have been reported. Heterogeneous assay systems also allow the reduction of interferences in patient samples. A number of commercial assays are available and noncommercial prototypes or applications for clinical (288,289) and research use (290,291) have been described. Use of heterogeneous systems alone, however, may not be sufficient to totally eliminate interference as demonstrated by one recent report describing a competitive FIA system for measurement of drugs in serum (292). Although performance characteristicssimilar to EIA or RIA have been observed, high sample dilution factors and frequent corrections with sample blanks are required. Light Scattering, Light scattering techniques have been reviewed in recent articles (293-295). In two common light scattering measurement methods, turbidimetry and particle counting, particle based reagents are frequently used to increase sensitivity. The principle of turbidimetric measurement 118R

ANALYTICAL CHEMISTRY, VOL. 57, NO. 5, APRIL 1985

is analogous to measurement of absorbance and may be performed by spectrophotometry. Latex is the polymer of choice for these particles and in many protein and hapten assays nonspecific agglutination is reduced by coating the particles with f(AB’)2fragments of antibody. The particle enhanced turbidimetric inhibition assays (PETINIA) are based on the ag lutination of hapten-coated latex particles by antibody a n t inhibition of this agglutination by free analyte. This methodology has been developed for a fully automated analyzer (296, 297) and has also been adapted to centrifugal analyzers (298). Centrifugal analyzers have also been used to measure a number of drugs and haptens by homogeneous immunoprecipitation inhibition (299). In this turbidimetric technique, analyte and analyte-conjugate compete for binding sites on antibody. No particle or sample pretreatment are needed. The method gives comparable results to the EMIT system. PACIA, a light scattering immunoassay based on particle counting, has already been mentioned in the section on instrumentation (300-303). In this technique which requires specialized particle counting equipment, light scattering is used to detect individual latex particles and a discriminator used to differentiate particle sizes. Electronic signals from particles smaller than 0.6 pm are ignored thus minimizing interferences from dust and chylomicrons. Impulses corresponding to particles greater than 1bm are also ignored thereby ensuring measurement of unaggregated particles. Both conventional competitive and direct noncompetitive PACIA assays have been described. In the latter type assay, the agglutination of coated latex particles by antigen or antibody is compared to that of standard preparations. Comparison of turbidimetry to some particle counting systems had demonstrated the latter to be twice as sensitive. In one recent article (304) detection limits of to mol/L were reported for direct measurement of proteins using antibody-coated latex particles in an automated continuous-flow system. When sera were diluted greater than 400-fold, interferences were eliminated and no sample pretreatment was required. In nephelometry, the scattered light rather than the decrease in intensity due to scattering is measured. This technique offers a high sensitivity, and a number of nephelometric assays have been recently decribed (305-309). Like other light scattering assays, however, they may be limited by serum intereferences (310-312) and sample pretreatment requirement. Chemiluminescence. Chemiluminescent molecules can be detected with a higher sensitivity than their colored and fluorescent counterparts. Concentrations in the and potentially down to mol/L range can be measured. These observations have encouraged the development of immunoassays combining the sensitivity of chemiluminescencewith the specificity of immunologic reactions (313-315). A number of different approaches have been verified: In the luminescence immunoassay (LIA), luminol or its derivatives are utilized as labels. Heterogeneous LIA’s have been applied to measure haptens in body fluids. Their performance was similar to RIA’S (316-319). Homogeneous LIA’s have successfullybeen used to analyze haptens in urine (320). Measurement of proteins has been hampered by poor analytical performance (321). However, the use of acridinium salts as labels has rendered promising results (322-323). In luminescence enzyme immunoassay (LEIA) an enzyme is coupled to an antigen. It serves to produce or degrade a substrate that is itself used in a luminescent reaction. In a recent development (324) horseradish peroxidase and luciferin were utilized, resulting in a signal that was high enough and of sufficient duration to be measured by photographic film (325). The combination of luminescence and fluorescence immunoassay (LFIA) has yielded interesting results (326,327). The combination of aminobutylethylisoluminol and fluorescin as labels offers the possibility of a very sensitive and widely applicable homogeneous assay (328-330). Monoclonal Antibodies. Production and applications of monoclonal antibodies have been the focus of many recent publications (331, 332). One of the greatest advantages of monoclonals is their ability to recognize defined antigenic determinants and therefore to differentiate between molecules which share a high degree of homology. This factor has been successfully applied to the measurement of proteins from different tissues (333) and different species (334). The high

CLINICAL CHEMISTRY

specificity of some monoclonal antibodies also can have negative consequences. Monoclonal antibodies can be selective for phenotypic variations of macromolecules which, for diagnostic purposes, should not be differentiated (335). The use of monoclonal antibodies in commercial and noncommercial two-site immunoradiometric or immunoenzymatic assays has been well documented. In this type of assay the specificityof monoclonals can be combined with the sensitivity of a two-site binding assay. The system has proven valuable for the measurement of proteins showing large degrees of homology with other related proteins. Many combinations of monoclonal and polyclonal antibodies have been used in attempts to obtain the best sensitivity, but results are not consistent. In some commercial (336,337)and noncommercial (338,339)assays, monoclonal antibodies are used for both the solid phase and the label, and excellent sensitivities have been observed. In other reports (340,341)better sensitivities are achieved if monoclonal antibodies are used for the solid phase and polyclonal antibodies as tracers. In these cases, the multiple attachment of heterogeneous antibody populations to the antigen may lead to increased avidity and overall to improved sensitivity. Some systems have achieved optimal performance using a polyclonal solid phase and labeled monoclonals (342,343). The signal produced by labeled monoclonal has been reported in some of these systems to be superior to labeled polyclonal antibodies. Reports of applications of monoclonal antibodies in cometitive assays have been fewer in number. This may in part l e due to incomplete or variable precipitation of antigenmonoclonal antibody complexes in soluble phase double antibody systems as well as the generally lower affinity of many monoclonal antibodies (344). Competitive systems have been adapted successfully in recent indirect sequential assays using solid phase antigen to separate the unbound monoclonal antibody from the antigen-antibody complex in a second, separate incubation (345). In another approach a novel precipitation system was designed to overcome the problem of the separation of bound from free tracer in a soluble phase assay (346). The monoclonal antibody was labeled with biotin and avidin, rather than a second antibody being- added to effect separation. Combinations of monoclonal antibodies have been used to increase the average affinity for antigen over that demonstrated by any single monoclonal (347). The mixtures of monoclonal antibodies appear in this case to bind analyte in a cooperative manner, forming circular complexes. It is clear that monoclonals will continue to play an important role in immunoassay but it is equally apparent that use of monoclonals will sometimes require new strategies to achieve the performance characteristics expected. ANALYTES OF CLINICAL INTEREST The list of molecules of clinical interest today is expansive and evergrowing. It includes metabolic and catabolic intermediates, transport proteins, lipoproteins, enzyme/isoenzymes, nutrients and vitamins, coagulation proteins, complement proteins, receptors, cellular antigens, antibodies, acute phase proteins, immune complexes, hormones, steroids, immunoregulatory factors, infectious agents, therapeutic agents, etc. Those areas that have received the greatest attention over the past 2 years have been the areas of cancer tumor markers, infectious agents, endocrine hormones, mar era of heart and vascular disese, and biological response modifiers of the immune system. Cancer and Tumor Markers. The clinical value of analytical tests for cancer has been a subject of intense debate over the past decade. Most of the controversy is a result of claims that some of these tests provide the ca acity to diagnose early cancer, and although this aspect is hig ly desirable, few if any such tests have withstood the scrutiny of carefully controlled studies. Greater promise, however, has been forthcoming from biochemical testa that monitor the efficacy of treatment of malignant disease. Here a few specific tumor markers have made a valuable contribution to patient care. It is beyond the scope of this review to cover all the important advances made in the area of cancer and tumor markers. Rather, we have chosen to select certain interesting areas that have been im acted by recent advances in immunoassay development anfthat might be representative of state of the art in this field.

i,

R

Carcinoembr onic antigen (CEA) is an oncofetal antigen synthesized by tze cells lining the normal fetal gastrointestinal tract. This protein is also synthesized in small amounts by the gastrointestinal glycocalyx of adults, but since most of it is secreted into the lumen of the gut under normal conditions, only low concentrations of CEA are found in the circulation of the healthy adult. The development of adenocarcinomas of the astrointestinal tissues, however, cause CEA to be secretel into the circulatory system. Gold and Freeman, in 1965 (348),were first to describe CEA as a result of its isolation from both embryonic gastrointestinal and neoplastic tissues. It has since been found to be a 180000 molecular weight glycoprotein possessing 4040% carbohydrate. It is a single polypeptide containing about 40 asparagine-linked complex oligosaccharide chains of heterogeneous nature (349). The high degree of glycosylation imparts unusual solubility characteristicsin that CEA remains in solution in the presence of 1M perchloric acid, a property which has previously been exploited as a tool in its urification and in the development of diagnostic assays. C E l is known to have a complex antigen system in that a number of related antigens, such as nonspecific cross-reacting antigen (NCA), nonspecific cross-reacting antigen 2 (NEA), and normal fetal antigen have been described (350). At first it was believed that CEA, in the sera of adults, was specific for the presence of colorectal carcinoma. Further examination, however, has revealed that CEA is also present in the sera of other types of malignant as well as benign conditions, includin carcinomas of the colon, rectum, pancreas, lun , stomach, reast, ovary, thyroid, and nonmalignant con8itions such as pulmonary emphysema, alcoholic cirrhosis, cholecytitis, rectal polyps, and benign breast disease (351-354). The first generation of diagnostic tests for CEA were RIA's developed with polyclonal antiserum. Because of their inability to be highly discriminating, these RIA's were shown to be not useful in the early diagnosis of cancer but were useful as prognostic indicators or in indicating recmence of tumors following treatment (355). A newer generation of CEA tests developed in the last 2 years, particularly those employing monoclonal antibodies (356),show increased disease-related specificity for carcinomas and are clearly useful for diagnosis, assessment of therapy response, and detection of recurrent disease for colorectal cancer, malignant breast cancer, and medullary carcinoma of the thyroid. These assays employ the two-site sandwich immunoassay configuration using two different monoclonals to two distinct antigenic sites and either an enzyme or a radioisotope label. Furthermore, the high degree of specificity offered by the monoclonals has eliminated the need of a pretreatment step designed to remove nonspecific cross-reacting antigens. The assays are sensitive to 0.5 ng mL CEA and have a measuring range of 0.5-200 ng/mL EA. a-Fetoprotein (AFP), much like CEA, is also a oncofetal antigen whose biosynthetic origin is not only the fetal gastrointestinal tract but also the fetal liver and yolk sac. AFP has been extensively characterized as a biological molecule in the past 2 years (357-359). Its amino acid sequence has been determined and shown to possess 39% sequence identify with serum albumin. AFP contains 590 amino acids and one N-glycosylation site. AFP is present in the circulation of a newborn at a concentration of approximately 30 pg/mL after having reached peak levels as high as 4 mg/mL in the 12-15th week of gestation. By 1year postpartum, however, AFP levels decrease to the normal adult concentration of 1-16 ng/mL. The reappearance of AFP in the adult circulation is often taken as a signal of the presence of some malignancy, particularly primary hepatocellular carcinoma, a germ-cell tumor containing yolk sac cell elements or a nonseminomatosis germ-cell tumor of the testis (360). The first generation immunoassay developed for AFP, utilizing polyclonal antibodies, was found useful for monitoring the treatment of AFP-producing tumors, however, the finding of elevated serum AFP levels in some patients with nonmalignant liver disease, particularly acute and chronic viral hepatitis and cirrhosis, has limited the value of these assays as independent specific tests to establish the diagnosis of cancer. A more recent generation of immunoassays that employ two high affinity monoclonal antibodies directed at two separate epitopes of human AFP exhibits extraordinary specificity and appears to be suitable for screening high-risk populations for AFPproducing tumors (361). The assay is configured in a two-site,

%

c!

ANALYTICAL CHEMISTRY, VOL. 57, NO. 5, APRIL 1985

llQR

CLINICAL CHEMISTRY

one-step simultaneous sandwich immunoradiometric assay. The assay is capable of quantitiating AFP levels as low as 0.5 ng/mL with a 1-h incubation. Unlike the prior generation of immunoassays that employ polyclonal antiserum this assay is capable of distinguishing AFP-producing carcinomas and various benign liver diseases. A possible mechanism for this enhanced assay sensitivity and specificity, put forth by Bellet et al. (361),is that benign liver disease may result in the secretion of “AFP-like”proteins that cross react with the less specific antibodies present in the prior polyclonal-mediated immunoassays. A murine monoclonal antibody has been created that, when put into an immunoassay, is useful in monitoring the serum levels of a tumor-associated antigen of epithelial ovarian origin [CA 1251. The monoclonal was developed by somatic hybridization of spleen cells from mice that were immunized with an epithelial ovarian carcinoma cell line [OVCA 4331 (362). The OV 125 monoclonal was screened and defined against 60 selected ovarian tumors and normal adult cells, by indirect immunofluorescence. The spectrum of reactivity included only ovarian carcinomas of epithelial origin, such as, endometroid tumors, clear cell carcinoma, and undifferentiated epithelial tumors. No reactivity was found with any benign or malignant tumors of stromal or germ-cell origin. The antigenic determinant defined by the antibody is a high molecular weight cell-surface glycoprotein that is expressed in coelomic epithelium during embryonic development (363). The immunoradiometric assay developed for CA 125 utilized a polystyrene bead coated with OC 125 antibody as a solid-phase immunoabsorbent. Quantities of CA 125 were expressed in arbitrary units based on comparison with a primary reference standard prepared from purified antigen. The minimum detection level of CA 125 was 1.4 U/mL (364). A clinical trial of 888 normal patients, 143 patients with benign disease, 200 patients with nongynecological cancer, and 101 patients with ovarian carcinomas showed less than 1% of the healthy persons and 6% of the patients with nonmalignant disease has serum CA 125 levels above 35 U/mL. In contrast, 82% with surgically demonstrated ovarian cancer had elevated levels of antigen. Furthermore, rising and falling levels of CA 125 correlated with progression and regression of the disease in 93% of the cases. Even more recently Niloff et al. (365), found elevated levels (>35 U/mL) of CA 125 in patients with recurrent fallopian tube carcinoma, late stage (V) endometrial carcinoma, and adenocarcinoma of the cervix. This led to the speculation, due to the tissue distribution, that CA 125 may be associated with Muellerian duct differentiation. Although additional proof is needed, this early work suggests that the monitoring of CA 125 may have significant value in monitoring tumor burden and disease progression and regression in various gynecologic cancers. Much like CA 125, a murine monoclonal antibody (1116 NS 19-9) has been developed which, when used in an immunoassay, proves useful in monitoring patients with advancing colorectal carcinoma, as well as providing a diagnostic adjunct for adenocarcinoma of the upper gastrointestinal tract (e.g., pancreatic cancer, gastric cancer, and hepatobiliary cancer) (366). The monoclonal antibody was developed by immunizing mice with the human colorectal carcinoma line SW 1116 (367). The antibody was shown to react with a carbohydrate antigenic determinant, a sialylated lacto-N-fucopentaose I1 related to Lewis blood group substance (368). The same antigenic determinant has been identified on a glycolipid (a sialylated lacto-N-fucopentaose I1 ceramide) extracted from the SW 1116 cell line and from meconium. A “forward” two-site immunoradiometric (sandwich) assay was developed, with the 19-9 monoclonal antibody both immobilized on a polystyrene bead and iodinated in the tracer solution (369). Quantities of CA 19-9 were expressed in units based on comparison with frozen primary reference standards, prepared by spiking partially purified antigen, isolated from cell-culture supernatant fluids, into pooled human serum. The CA 19-9 unit is an arbitrary activity corresponding to approximately 0.8 ng of purified antigenic material. The average CV for the assay was approximately 10% in the range of 5-120 units/mL. The minimum detectable dose was 1.4 units mL. The average concentration of CA 19-9 in sera from 102 putative healthy individuals was 8.4 (standard deviation 7.4) units/mL. A reference level of 40 units/mL was chosen to discriminate between positive and negative samples. The initial clinical

6

120R

ANALYTICAL CHEMISTRY, VOL. 57, NO. 5, APRIL 1985

evaluation conducted on more than 1600 coded sera obtained from blood donors at the NCI/Mayo Clinic Serum Bank indicated that CA 19-9 was elevated in a large fraction of sera (67%)from patients with advanced adenocarcinomas of the upper gastrointestinal tract, including those with pancreatic, hepatobiliary and gastric carcinomas. A smaller fraction (18%) of patients with carcinomas of the large bowel had elevated serum CA 19-9 levels, the majority among patients with metastatic disease. In contrast, none of the healthy donors from the serum bank and only 4 of 1023 of the blood donor specimens (0.4%) had CA 19-9 levels greater than or equal to 40 units/mL. These early data appear very encouraging, and at the moment, CA 19-9 appears to be potentially useful as a laboratory adjunct in following patients with pancreatic, hepatobiliary, and gastric cancer, particularly in combination with CEA determinations. The test also appears able to identify some patients with early resectable malignancies of the upper gastrointestinal tract. Rigorous prospective clinical trials will be necessary to verify these hypotheses. The measurement of prostate-specific acid phosphatase concentrations in the serum of males has been commonplace since its activity was first shown, 45 years ago (370),to be elevated in patients with prostatic adenocarcinoma. Acid phosphatase now is generally used as a marker in the diagnosis of metastatic carcinoma of the prostate and to follow the effectiveness of therapy in this disease. The problem, however, is that the sensitivity and specificity of previous assays have been inadequate to diagnose prostate cancers prior to their metastatic involvement of other tissues, predominantly in the axial skeleton. The first generation of assays were directed toward establishing a specific enzymatic assay for the prostatic acid phosphatase isoenzyme using specific substrates or inhibitors. These assays, although quite useful for prognostic applications, lack the sensitivity to be of any true diagnostic value. The newer assays being developed employ antibodies specific to the prostatic acid phosphatase enzymes of the serum. These assays are configured either as competitive radioimmunoassay that quantitate the mass of the enzyme (371),two-site immunoradiometric assays (372),or as endogenous capture immunoassays that quantitate PAP enzyme activity (373). The sensitivity of these assays are now in the ran e of 0.1 ng/mL. A stable reference material for prostatic acidg phosphatase, derived from human tissue and human seminal fluid, has been developed which is compatible for use with both kinetic measurements and immunoassays. Inspite of all these advances, the measurement of prostate-specific acid phosphatase is still of little value in the detection of occult (stage I) adenocarcinoma of the prostate. In addition to PAP, another potential marker for the diagnosis and monitoring of prostate cancer is a serum protein called prostate-specific antigen [PA] (374, 375). This 33000 molecular weight glycoprotein is found in the serum of healthy males at low levels (0-2.7 ng/mL) and is present at elevated levels in the sera of patients with enlarged prostate glands, either benign or malignant. The principal assays to date are either two-site (sandwich) immunoradiometric assays or competitive radioimmunoassays. The lower end sensitivity is 0.25 ng/mL protein. Although the presence of this antigen tends to be rather nonspecific and cannot be used to differentiate patients with prostatic tumors from benign prostatic hypertrophy, it is useful in detecting all patients with prostate abnormalities. The monitoring of PA has been shown to be a very sensitive measure of the changes in prostatic tumor patients following diagnosis. The precise role that an assay for P A will play in the future in the diagnosis and treatment of prostatic cancer is not clear. Perhaps a combined measurement of PA and PAP will improve the clinical sensitivity with which prostatic cancer can be detected. The hormone called human chorionic gonadotropin is secreted by the trophoblast and is believed to be responsible for maintainin the corpus luteum during pregnancy. A glycoprotein,h8G is composed of two subunits called a (14500 mol wt) and ,tl (22000 mol wt). The hCG alpha subunit is identical with the a subunits of the human pituitary hormones lutropin (LH), follitropin (FSH), and thyrotropin (TSH). The p subunits are different but share a great deal of homology and provide the specificity for interaction with the particular target tissue (376). Secretion of hCG increases rapidly in the first few weeks of pregnancy and its presence in serum or urine

CLINICAL CHEMISTRY

is used for the early detection of pregnancy. Since hCG is also secreted by neoplastic forms of trophoblast tissue, the quantitation of serum levels of hCG is also very important in the diagnosis and followup of gestational trophoblastic neoplasia-cancers that originate from placental tissue. The most malignant form, choriocarcinomas,may become apparent following hydatidiform mole or after delivery of a normal fetus. hCG assays are also important in the detection of threatened abortion, ectopic pregnancy, and in staging and followup of nonseminomatous germ-cell testicular tumors. A number of immunoassays are currently available for the quantitation of hCG, but only the more sensitive and specific ones are of any utility in the diagnosis of trophoblastic cancers (0.5-3 mIU/mL). The latest improvements in the specificity of hCG assays involve the preparation of antisera or monoclonal antibodies that recognize the unique carboxy terminus of the hCG p subunit which is absent from the LH molecule (377-379). The assay configuration commonly employed is the two-site (sandwich) immunometric assay with either an enzyme or a radioisotope as the label. These ultrasensitive assays provide the physician with a very useful tool for diagnosing and monitoring gestational trophoblastic neoplasia. In addition to the more classic view of cancer and tumor markers described above, there are a host of less specific substances that are proving to be useful in the diagnosis and management of cancer. For example, some of the enzymes that are routinely measured in serum such as lactate dehydrogenase, alkaline phosphatase, and aldolase have been found to be useful as nonspecific tumor markers that complement other clinical information. Unique isoenzyme forms for enzymes like creatine kinase (CK-1 is elevated in patients with small cell carcinoma of the lung), y-glutamyl transferase (isoenzyme I and I1 are present in hepatocellular carcinoma), 5’-nucleotidase (elevated in patients with ovarian cystadenocarcinoma), and amylase (a unique isoenzyme of amylase called serous ovarian neoplastic amylase is present in cystadenocarcinoma of the ovary) are used in the diagnosis of cancers (380-382). Several newer enzymes, whose methodological assays have been improved through the development of immunological assays over the last couple of years, may also prove to be potential indicators of neoplastic transformation. These include the following: cytidine 5’-monophospho-Nacetylneuraminic acid hydrolase for ovarian carcinoma; acid and alkaline deoxyribonuclease for the breast cancer; pancreatic elastase for pancreatic adenocarcinoma; 5’-nucleotide phosphodiesterase for hepatic metastic disease; and glycoprotein glycosyltransferases like galactosyltransferase isoenzyme form I1 for lung and ovarian cancer. The frontier of cancer markers, however, is the field of oncogenes and oncogene products. Oncogenes are genes present in normal human cells whose overexpression or mutation results in malignant transformation of cells. There are more than 20 oncogenes identified to date and most have been found to be associated in some way with the regulation of normal cell division and differentiation (383). Of particular interest is a family of oncogenes that encode for tyrosine protein kinases. Thus far, the link between aberrant expression of an oncogene and cancer is strongest for Burkitt’s lymphoma and certain bladder cancers (384). In Burkitt’s lymphoma, the “c-myc”oncogene is found translocated to an unnatural chromosomal position, next to an immunoglobulin gene, that results in the overexpression of this gene and causes abnormal cellular proliferation. In bladder carcinoma, cancer has been linked to mutation of a member of the “ras” oncogene family (385). These tumor cells carry an altered form of the “ras” oncogene which leads to the expression of a “ras” protein containing a single amino acid change. Both DNA probe technolo of the gene itself as well as monoclonal antibody mediategmmunologic assays of oncogene products serve as approaches to using oncogenes as diagnostic indicators of cancer. Infectious Diseases. Any review covering the advances and status of analytes of clinical interest for the 1982-1984 period would be totally remiss without a section on infectious diseases-particularly acquired immune deficiency syndrome. In addition to AIDS, hepatitis and sexually transmitted diseases were areas which saw significant advances. Traditionally, the diagnosis of infectious diseases has been based on the isolation of the infecting organism in pure culture. Although this culture procedure is very accurate and highly reliable, in

many cases the procedure is so time-consumingthat the results are unavailable to the clinician facing the crisis of an acutely ill patient. This is especially the case for slow-growing bacteria and viruses that must be isolated on cultured cells. Equally important is that many medically significant microorganisms, most notably the hepatitis A, hepatitis B, and human rotoviruses are not cultivable in most widely available tissueculture systems. The goal over the past few years has been, and will continue to be for sometime into the future, the development of either rapid immunodiagnostic tests for detecting the presence of antigens from these organisms or a serological test for detecting circulating antibodies signifying previous exposure and potential immunity. Acquired Immune Deficiency Syndrome (AIDS): The acquired immune deficiency syndrome was first recognized in 1981 (386) as a generally fatal disorder of cell mediated immunity manifested clinically by opportunistic infections (predominantly Pneumocystis carnii pneumonia) or Kaposi’s sarcoma. The underlying disorder affects the patient’s cellmediated immunity, resulting in T-cell lymphoma, especially of the helper/inducer subset bearing the OKT4+ surface marker. To date, more than 6000 cases of AIDS have been reported to the Center for Disease Control (387). The incidence of AIDS in homosexual men with multiple sexual partners, intravenous drug abusers, hemophiliacs, blood transfusion recipients, and close heterosexual contacts of members of the high risk groups strongly suggested that the disease was spread by an infectious agent. Very recently, several converging lines of research, conducted predominantly in two independent laboratories, Robert Gallo of the NCI (388-390) and Luc Montagnier of Institut Pasteur (391-393), have linked a human T-cell lymphocytotropic retrovirus HTLV I11 or LAV (for lymphoadenopathy associated virus) to the pathogenesis of AIDS. HTLV I11 is the newest member of the HTLV family. The first human T-cell lymphotropic virus, HTLV I, was discovered in 1980 by Poiesz et al. (394), and has since been shown to be the etiological agent of certain adult T-cell leukemia-lymphomas. HTLV 11, discovered in 1980 by Kalyanaraman et al. (395) was isolated in a patient with leukemia and in an intravenous drug abuser but has not been definitively linked to human disease by seroepidemiologic results to date. The genes of HTLV I and HTLV I1 have been cloned and sequenced and shown to contain a great deal of homology (396,397). These viruses do not possess known “onc” genes but contain the characteristic genes of retroviruses such as the core proteins, reverse transcriptase, and envelope glycoproteins. HTLV I11 shares many properties with HTLV I and 11, including T-cell tropism and a propensity to infect T-cells with OKT4+ (helper inducer) phenotype (398). All three viruses also have simi ar reverse transcriptases, at least some cross-reactive antigens, major structural proteins of similar size, and the capacity to induce multinucleated giant cells. Recent nucleic acid hybridization studies have revealed some homologies among all three members of the HTLV family. HTLV 111, unlike HTLV I and 11, is characterized by substantial genetic polymorphism. At this point, HTLV I11 has been isolated from the blood and lymph nodes of over 95 persons and >90% of the patients with early stage AIDS have been tested positive for circulating antibodies for HTLV 111. HTLV I11 has never been isolated from healthy heterosexual donors (115 unsuccessful attempts). Most recently, it was possible to detect circulating antibodies reactive against HTLV I11 in every patient with AIDS, using a combination of ELISA and Western electroblot assay for detecting antibodies (399). Knowledge of HTLV I11 and the practical application of convenient diagnostic testing will clearly have profound clinical implications on the spread and management of AIDS. A rapid test for circulating antibodies to HTLV I11 would prevent donation of blood products by persons infected by HTLV 111. The availability of sensitive tests for HTLV I11 antigenemia or circulating antigen-antibody complexes would permit the identification of early cases of AIDS and assessing the risks of the disease. In addition, such testing may make it possible to begin formulating rational interventions directed against a course and ultimately a cure or vaccine. The same can be said for HTLV I and 11. Unlike AIDS, viral hepatitis has been described for more than 70 years. Early in this century it was suspected that a microorganism, acting on the liver, was the etiologic agent for

i

ANALYTICAL CHEMISTRY, VOL. 57, NO. 5, APRIL 1985

121 R

CLINICAL CHEMISTRY

what was described as epidemic jaundice. About the same time it was also noticed that a similar liver malady was associated with vaccinations. It took another 40 years, however, before these two clinical entities were identified as distinct disorders which differed clinically, epidemiologically, and serologically: infectious hepatitis (hepatitis A) corresponding to epidemic jaundice and serum hepatitis (hepatitis B) correspondin to the disease spread by vaccination. In 1974 (400, 4011, anotier infectious agent was recognized as causing acute chronic hepatitis. Although originally named hepatitis C, it is now referred to as non-A, non-B (NANB) hepatitis principally because its diagnosis is based on the exclusion of exposure to hepatitis A or B. Even more recently (402, 403), a forth viral hepatitis, delta hepatitis, has been described. This form of hepatitis is caused by the simultaneous infection of an individual with hepatitis B virus and the newly described delta virus. Delta hepatitis is more fulminant and deadly than hepatitis B alone. In addition to causing hepatitis, hepatitis B virus has also been linked casually to liver cancer. Epidemiological studies have revealed a striking correspondence between the areas of high incidence hepatitis B infection and the frequency of primary hepatocellular carcinoma (404). Most recently, Rutter et al. (405) have isolated and analyzed, by molecular cloning, integrated hepatitis B virus sequences from a human hepatoma cell line. Additional studies, however, will be required before such a link can be definitively established. In 1975 Gravelle et al. (406) recovered viruslike particles from the stool of acutely ill patients infected in a food-borne outbreak of hepatitis A. These virus particles were characterized as spherical particles with variable diameters ranging from 24 to 29 nm. The virion, a picornovirus, contains single-stranded RNA surrounded by a protein capsid of 32 capsomers. The presence of either an outer protective envelope or virus subunit components has not been reported. Viral hepatitis A is a highly contagious disease transmitted by the fecal-oral route, primarily by person-to-person contact and by exposure to contaminated food or water. Following infection through oral ingestion, hepatitis A virus multiplies in the gastrointestinal endothelium and then spreads to the liver by way of the blood and lymphatic vessels. Characteristic increases in serum alanine aminotransferase activities are observed 17-38 days after the initial infection. Diagnosis is made by performing an antibody-based serological assay for circulating antibodies for hepatitis A virus. Since viremia in hepatitis A is usually associated with low levels of circulating virus which do not persist into the acute phase of the disease, the laboratory diagnosis of hepatitis A by the analysis of serum for whole virions or components therefrom is not readily possible. Diagnosis, on the other hand, is made by examining the patient sera for antihepatitis A titers, as either an IgM or an IgG. Several techniques are available for distinguishing these two antibody populations: (i) treatment of the serum with a mild reducing agent to dissociate the pentameric IgM, thus eliminating its activity, (ii) use of specific anti-IgM or anti-IgG as blocking agents, (iii) addition of protein-A/staphylococcus aureaus to absorb out the IgG population, or (iv) the addition of immobilized anti-IgM (p-chain specific) to absorb out the IgM population. Commercial test kits using either radioisotopes or enzymes as labels are now available for the rapid and simple detection of IgM-specific antihepatitis A in patient serum (404,407). Hepatitis B or serum hepatitis is caused by the hepatitis B virus, originallytermed the Dane particle upon its discovery (408). The mature hepatitis B virion is much more complex than the hepatitis A virion. Inside a double-shelled spherical particle, measuring 42 nm in diameter, is a double-stranded DNA and a virus-specific DNA polymerase. The entire nucleic acid sequence of the DNA genome has been determined and cloned (405). The outer envelope is a complex mixture of polypeptides which manifest an antigenic activity previously described as the Australia antigen or hepatitis associated antigen but now is termed as the hepatitis B surface antigen (HBsAg). The number and size of the constituent polypeptides are still debated; however it is clear that there are two major polypeptides designed PI (p23) and PI1 (gp28). The PI1 peptide appears to be the glycosylated version of the PI peptide. It also has been shown, immunologically, that there are multiple subdeterminants or serotypes of the HBsAg, now specified as “adw”, “adr”, “ayw”,and “ayr”. It is thought that 122R

ANALYTICAL CHEMISTRY, VOL. 57, NO. 5, APRIL 1985

these polypeptides are present as disulfide-linked dimers of 50000 molecular weight, embedded in a typical lipid bilayer. All four of these major subtypes exhibit the same clinical picture, and since they are transmitted and replicated unchanged, they can be used in epidemiologicalstudies of viral transmission. More recently several reports (409) have identified additional subdeterminants; however, until they are more fdly characterized, their significance will remain unclear. The inner envelope is a 7 nm thick shell that makes up the boundary of the electron-dense nucleocapsid and manifests an antigenic activity that has been termed hepatitis B core antigen (HBcAg). Other than the fact that HBcAg is distinct (immunologically) from HBsAg, little else is known about its structure. An even more mysterious component is the serologically defined hepatitis B “e” antigen (HBeAg). The exact location of the HBeAg is unknown, but is believed to be present in the core and probably part of the HBcAg in cryptic form (410). Hepatitis B is a serious worldwide public health problem. Over 15% of the population in the USA show positive serological markers for hepatitis B virus. The route of infection is parenteral, primarily via contact with blood or other body fluids. As a result, multiply transfused patients, those receiving chronic dialysis, hemophiliacs, newborns, homosexual men, intravenous drug abusers, health-care professionals, and sexual partners of hepatitis B patients are at especially high risk. For a patient with acute hepatitis, who may be jaundiced and have elevated liver transaminase enzyme levels in the serum, the additional presence of HBsAg in the serum identifies him clearly as having a hepatitis B infection. The presence of anti-HBsAg, anti-HBcAg, or anti-HBeAg is also used to identify patients who have recovered from a hepatitis B infection. Since the discovery of HBsAg by Blumberg et al. in 1964 (411) there has been a continual evolution of the testing technology for HBsAg toward higher sensitivity and shorter turnaround times. In the United States, methods for the detection of HBsAg are regulated by the Bureau of Biologics, a division of the Food and Drug Administration. Methods for detecting HBsAg are classified into three categories (first, second, or third generation tests) based principally on their sensitivity. The first-generation tests, used primarily between 1967 and 1970, rely on radial immunodiffusion technology. These tests are very insensitive, requiring a 2-3 day development time. These tests evolved to the second-generation tests that utilize counter-immunoelectrophoresis procedures. The counterimmunoelectrophoresistests were clearly much faster (2-3 h) but provided little more sensitivity than the immunodiffusion technology. The quantum leap came with the next generation of tests which employ solid-phase immunoradio(and enzymo)metric assays of the sandwich type. These assays are extremely sensitive with the capability of measurin HBsAg at levels near 1 ng/mL. Very recently an immunorafiometric assay has been described that employs a mixture of high affinity monoclonal antibodies directed at specific antigenic surface determinants (409). This assay exhibits state-ofthe-art sensitivity and specificity, as well as the fact that it can detect all HBsAg even in the presence of human antihepatitis B immune complexes. Tests for HBcAg, antiHBsAg, anti-HBeAg, and anti-HBcAg in third-generation RIA formats are also available. Non-A, non-B hepatitis was first recognized in 1974 (400) and now is one of the most frequent forms of viral hepatitis seen in the United States. It accounts for up to 90% of the posttransfusion hepatitis cases and approximately 25% of all incidences of sporadic hepatitis. Non-A, non-B hepatitis can be transmitted both systemically and nonsystemically. Blood products such as fibrinogen and factor VI11 are frequent sources of NANB infectivity. Although NANB infections may produce fulminant liver disease, NANB hepatitis tends to be clinically less severe than hepatitis B and it is not known whether NANB hepatitis predisposes to hepatocellular carcinoma. Epidemiologic studies suggest that more than two forms of NANB exist (401). Attempts to identify the casual agent(s) by electromicroscopy are numerous and varied (412). Examinations of both serum and liver tissue have revealed viruslike particles ranging in size from 14 to 180 nm. None of these claims have withstood scrutiny. The most recent report to date is that of Prince et al. (413),who claim to have identified a virus from chimpanzee liver cell cultures innoculated with NANB infectious material. Morphological

CLINICAL CHEMISTRY

characteristicssuggest that it belongs to 'a hitherto undescribed class of animal virus. Two weeks prior to Prince's report, Seto and co-workers (414)described a reverse transcriptase activity in association with NANB agent(& These latest findings are potentially very important; however, they will need to be confirmed and amplified. The diagnosis of NANB hepatitis, therefore, still remains to be by exclusion of hepatitis A, hepatitis B, cytomegalovirus,Epstein-Barr virus, herpes virus, Coxsackie B virus, and other liver dysfunctions. For the moment, there is no antigen-antibody system recognized by all research groups or the FDA as being specifically related to NANB infection. Delta hepatitis is a newly discovered deadly and mysterious form of hepatitis caused by the coinfection of two viruses: hepatitis B virus and delta agent. Two recently reported outbreaks (415),one in Venezuela and one in Worcester, MA, resulted in the death of almost 25% of the infected patients, The delta agent is an infectious particle, presumably a virus, which is found only in patients who are either acutely or chronically infected with hepatitis B virus. Although transmissible as an independent infectious agent, delta virus can only infect and cause illness in the presence of hepatitis B virus. Severe outbreaks, therefore, can occur anywhere hepatitis is common and the fact that over 800000 people in the United States are known to be carriers of hepatitis B, and thus vulnerable to delta infection, presents a rather ominous specter. The agent consists of a small piece of single-stranded RNA surrounded by a novel protein antigen "delta antigen" and an outercoat of hepatitis B surface antigen (HBsAg). It is one of the smallest viruses known, not much larger than the viroids that cause disease in plants. Several reports (416) exist describing the development of radioimmunoassays for the detection of delta antigen; however, these assays were setup using delta antigen purified from liver tissue of patients who succumbed to acute hepatitis infection. A very recent report (403) describes the development of a solid-phase RIA using delta antigen purified from the serum of individuals that were coinfected with hepatitis B virus and delta agent. Perhaps it will now be possible to establish assays in clinical settings that do not rely on the availability of autopsy tissue. Sexually Transmitted Diseases. There are three predominant venereal pathogens that have received a great deal of study over the past few years toward the development of rapid diagnostic methodologies: Chlamydia trachomatis, Herpesviridae, and Neisseria gonorrheae. Chlamydia trachomatis are gram-negative, obligate intracellular parasites. Although classified as a bacterium because of their cell wall structure, C. trachomatis lack important metabolic functions for the production of energy and thus require a cell culture host system for isolation. Chlamydia trachomatis predominantly infect columnar epithelial cells of the lumen of the human inguinal and ocular regions after recognizing specific cell-surface sialic acid baring receptors. Once phagocytosed, chlamydiae replicate within unique intracellular inclusion bodies with a cycle time of 24-48 h. Chlamydia infection can lead to urethritis, cervicitis, epididymitis, endometritis, pelvic inflammatory disease, perihepatitis, and cervical dysplasia. Chlamydia infection has become a serious public-health problem in the United States with as many as 5-10 million suspected new cases each year. Although chlamydial infection can be effectively treated with antibiotics, the inability of most physicians to obtain cultures, combined with the subtle and nonspecific clinical manisfestations, has made their control difficult. Currently, isolation in a cycloheximide-treated McCoy cell monolayer is the most common diagnostic procedure. This technique is based on the detection of specifically stained (immunofluorescence, iodine, or Giemsa) inclusion bodies which are visualized in the cell monolayer after 2-3 days of culture, with the aid of a light microscope. Recently a more rapid direct test for the detection of Chlamydia trachomatis has been developed (417, 418). This technique uses fluorescein-labeled monoclonal antibodies that stain individual elementary and reticulate bodies in direct smears. This procedure still requires the preparation of a slide and microscopic examination. Future research is aimed at the development of a rapid immunometric assay for the detection of Chlamydial antigens (419). Herpes simplex virus (HSV) infections are among the most common infections in humans. There are two serotypes of herpes virus: HSV 1 and HSV 2. HSV 1,primarily respon-

sible for recurring orolabial lesions and cold sores, is found in almost 80% of the population. HSV 2 is much less prevalent and is responsible for genital herpes and infections of neonates. More recently, it has been recognized that a significant percentage of genital herpes is also HSV 1 related. The differential diagnosis of these two viridae, therefore, has become important. Identification of herpes infection is particularly important during pregnancy, in that if sheeding occurs at the time of delivery, the newborn can die or have severe brain damage. If HSV infection is present at term, the delivery should be performed by Cesarean section to reduce the risk of transmission of the virus to the newborn. This element of time and risk to the newborn emphasizes the need for a rapid and sensitive diagnostic test for HSV. HSV 1and 2 are envelope virus similar to varicella-zoster virus, cytomegalovirus, and Epstein-Barr virus. Surrounding their DNA is a 162 capsomere icosahedron of approximately 100 nm diameter. This core is surrounded by a lipid bilayer membrane derived from the host cell nuclear membrane. The two serotypes of HSV share common antigens but also express unique determinants as well. The standard and most sensitive technique for the detection of HSV involves extensive tissue culture procedures on human foreskin fibroblast monolayers. This procedure can take up to 7-10 days for a final determination in specimens with low infectivity. Obviously this procedure is too lengthy to provide an early confirmation of active herpes infection prior to neonate delivery. More rapid procedures involving first a 24-h culture followed by immunofluorescence staining and fluoromicroscopy (biotin-avidin fluorescent-antibodytechnique) have proven to be equivalent in sensitivity (420,421). Direct immunofluorescence and direct immunoperoxidase staining of scraping of HVS lesions, although very rapid, have not proven to be as sensitive (83% sensitive) as virus isolation results (422-424). Most recently several sensitive enzyme-linked immunosorbent capture assays for the detection of HSV antigen have been reported (425-529). These procedures proved to be highly sensitive and specific when compared to tissue culture results. Total assay time was approximately 4.5 h. It should be noted, however, that such an ELISA test procedure will detect antigen in the absence of virus infectivity. Neisseria gonorrhoeae is a gram-negative bacterium that is venereally transmitted only in the human and is the cause of gonorrhea. Gonorrhea ranks first among reported communicative diseases in the United States. Infection in the male generally causes severe anterior urethritis. Infection in the female, on the other hand, is generally asymptomatic and presents few signs of the disease. There is a strong need and desire to control the spread of this disease. Diagnosis of gonorrhea in the male with overt urethritis is quite simple and is based on characteristic clinical signs and symptoms. Confirmation of the diagnosis involves a simple gram-stain of the urethral discharge. Diagnosis in the asymptomatic individual, particularly the female, is much more complicated. The recommended procedure for diagnosing the infection in women involves cultures of a cervical swab specimen on Thayer-Martin or similar media. This procedure is 80-90% sensitive and requires 2-3 days. Clearly, a more rapid immunological-based assay for the detection of Neisseria gonorrhoeae would provide a tool for the control of gonorrhea. Such a system has been described recently (430-433) and involves an enzyme-linked immunometric capture assay. Although this assay does provide adequate sensitivity for the proper diagnosis of culture-positive males, the sensitivity falls short of what is required for detecting cervical gonorrhea. Trichomoniasis is a venereal infection caused by the protozoon Trichomonas uaginalis. This organism is becoming a widespread problem, primarily because a large majority of the carriers (50% females and up to 80% males) are asymptomatic. The primary clinical symptoms are vaginal discharge, pain during intercourse, or symptoms of urinary tract infection or urethritis in men. The principal means of diagnosis in women is the microscopic examination of wet-mount of vaginal secretions. This protozoon is easily recognized by light microscopy. Although the approach is simple, fast, and specific, it lacks sensitivity in that only 50-70% of culture positives are identified by even the very skilled microbiologist (434). The problem with culture is that it is extremely complex involving a unique media formulation not readily available. The development of a rapid immunodiagnostic test would be ANALYTICAL CHEMISTRY, VOL. 57, NO. 5, APRIL 1985

123R

CLINICAL CHEMISTRY

of great utility in the diagnosis and control of trichomoniasis. Endocrine Function. The development of rapid and sensitive immunological assays for many hormones has both revolutionized the understanding of the endocrine systems as well as has provided an extremely useful clinical tool for the diagnosis and therapeutic monitoring of endocrine disease. Thyroid Function. The principal function of the thyroid gland is to regulate the metabolic activity of the whole organism through the controlled secretion, into the circulatory system, of the hormones L-thyroxin (T4) and 3,6,3’-triiodoL-thyroxin (T3). The release of T4 and T3 from the thryoid gland is in turn controlled by circulating levels of the anterior pituitary hormone, thyroid-stimulatinghormone (TSH), which is then under the control of thyrotropin-releasing hormone which is secreted from the hypothalamus. The major fraction of circulating T4 and T3 and bound to plasma proteins, viz., thyroxin-binding globulin (TBG), thyroxin-binding prealbumin, and albumin. The availability of a variety of laboratory tests permits evaluation of many aspects of thyroid homeostatic control. Of these, measurement of total T4, thyrotropin with and without thyroliberin test, as well as a “free-T4 parameter” such as the T4/TBG ratio or ffd-index are the most widely applied tests. Several recent studies (435-438) have dealt with the subject of whether or not 8 direct measurement of free circulating levels of T4 is more useful at interpreting the correct clinical status of cellular thyrometabolic function than is a measurement of the free T4-index. Many factors are known to influence circulating concentrations of the principal binding protein (TBG) and thus make difficult the interpretation of total thyroid hormone concentrations. Moreover, recent findings (439) on thyroid hormone action at the cellular level indicate that the hormone action proceeds by entry of the free hormone into the target cell, the conversion of the prohormone T4 to T3, the binding of T3 to an intracellular receptor protein, and the possible modulation of the normal signal at a postreceptor level. Considering this sequence of events, it is unlikely that the thyrometabolic state of a patient may be solely reflected by the concentration of free T4 in serum. The measurement of serum TSH has become an important tool for the diagnosis of hypothyroidism, hyperthyroidism, and diminished thyroid reserved. Up until just recent1 (400-442), the major problem has been the inability of t l e diagnostic tests for TSH (generally RIA) of providing adequate sensitivity for the accurate quantitation of TSH at levels much below normal (mean normal level is 1.5 f 0.2 MIU/mL). Such limitations of sensitivity and specificity restricted the clinical utility of basal TSH measurements to the diagnosis of primary hypothroidism. The availability of the improved techniques, which permit discrimination of normal from suppressed TSH concentrations, have now made it practical to utilize basal TSH measurements to diagnose hyperthyroidism. Hypertension is now viewed as having two different pathophysiologic etiologies: renal hypertension and essential hypertension. Both, however, involve the newly defined renin-angiotension-aldosterone system that serves as the long-term biochemical regulator of blood pressure and sodium/fluid-volume homeostsitis. When the juxtaglomerular cells surrounding the afferent arterioles of the cortical glomeruli sense a drop in the perfusion pressure, they secrete from granular stores, the proteolytic enzyme, renin. Upon entering the blood, the released renin proteolytically cleaves a circulatin a-2 globulin, angiatensinogen, made in the liver, to form the dgecapeptide angiotensin I. After one assage through the lung, the inert angiotensin I is converte by pulmonary enzymes to the physioactive octapeptide, angiotensin 11. Angiotensin I1 raises the blood pressure in several ways. An immediate change in blood pressure is achieved through its action as a vasoconstrictor, both by the energization of peripheral arteriole smooth muscle and by activatin hormones in the sympathetic nervous system. A slower anI f more long term effect on blood pressure is achieved by the stimulation of the zona glomerulosa cells of the adrenal cortex to secrete the steroid hormone, aldosterone. The major function of aldosterone is to facilitate the exchange of potassium for sodium in the distal renal tubule, causing sodihm reabsorption and potassium and hydrogen loss. There are many ways for disorders of this complex interplay of factors to result in a clinical manisfestation of hypertension. In primary hyperaldosteronism, elevated levels of aldosterone can be caused

B

124R

ANALYTICAL CHEMISTRY, VOL. 57, NO. 5, APRIL 1985

by single or multiple adenomas of the adrenal cortex, by an adrenal carcinoma, or by biIateral adrenal hyperplasia. In renal hypertension, the malfunction lies somewhere in the bar0 receptor mechanism in the kidney. To identify the primary biochemical lesion, an assortment of diagnostic assays is required (443). Aside from the need to measure serum and/or urine levels of renin, angiotensin I and 11, aldosterone, assays for prorenin (the inactive precursor of renin), the putative natriuretic hormone, vasopressin, and adrenalcorticotropin harmone (ACTH) are also of importance (444,445). Prorenin is the inactive precursor of renin and is found in high levels in the circulation. The reason for its presence, however, is not well understood. The elusive “natriuretic factor” has been an enigma to physiologists for many years. Recent studies (446) have identified some atrial peptides which possess the potent natriuretic activity ascribed to this entity. ACTH, secreted from the anterior pituitary, is more directly involved with glucocorticoid (e.g., cortisol) secretion from the adrenal cortex than with aldosterone secretion. However, in the absence of ACTH, the cells of the zona glomerulosa partially atrophy, leading to aldosterone deficiency. Finally, a growing number of studies (447, 448) have indicated that the neurohypophyseal antidiuretic hormone (ar inine vasopressin), because of its involvement in the contro of renal urine concentration, is also involved in the pathogenesis of salt-retentive hypertension. Sensitive immunologic assays for all of these factors are useful for diagnosing and monitoring disorders of the endocrine system that manifest the broad clinical picture called hypertension. Gastrin/ Gastrinoma Syndrome. The release of acid into the stomach is under the complex control of both vagal innervation and hormone action. After the ingestion of protein or distension of the stomach, the cells of the antral pyloric glands release the hormone, gastrin. The hormone is absorbed into the blood stream and is carried to the parietal cells of the fundic region of the stomach causing them to produce and secrete hydrochloric acid. Gastrin is actually a family of polypeptides. The gene for gastrin codes for a 101 amino acid precursor. The active forms of gastrin found in the circulation, however, are a 17-amino acid polypeptide called “little-gastrin” or G-17 and a 34-amino acid polype tide called ”big gastrin” or G-34. G-17 is formed from G-34y! removal of a 17-amino acid NH,-terminal fragment (449). Approximately two-thirds of circulating gastrin consists of the larger molecular species, G-34. The relative concentration of these two forms of gastrin, on the other hand, appears to be indicative of different gastrinoma syndromes. Sensitive diagnostic tests, utilizing specific antibodies that recognize the COOH-terminal 17 amino acids separate from the NH2-terminalamino acids, have been reported and thus are capable of quantitating G-17 and G-34. The NH,-COOH ratio for healthy individuals is 0.39. In patients with atrophic gastritis, the ratio is 0.37. Patients with benign sporadic localized gastrinoma have a ratio of 0.17 whereas patients with a malignant variety of gastrinoma have a value often much greater than 1. The ability to accurately measure circulating levels of the various forms of gastrin, therefore, clearly provides an important tool for the management of gastrinoma syndrome (450). Growth hormone or somatotropin is an anterior pituitary hormone that is primarily responsible for the control of growth of organ systems such as skeleton, connective tissue, muscles, and viscera such as liver, intestine, and kidneys. The control, however, is mediated through a group of recently discovered peptides known as somatomedins (Sm) or insulin-like growth factors (451). The discovery of these factors and elucidation of their physiologic role, biochemical characterization, and alterations in disease have resulted in greater understanding of the mode of action of growth hormone and the development of specific laboratory tests that can be used to diagnose and evaluate growth-relateddisorders. Growth hormone is a linear polypeptide with 190 amino acids and contains two intrachain disulfides. It exerts its effects through the binding to specific cell surface receptors of the target organs. The somatomedins and/or insulin-like growth factors are much less characterized. Three somatomedins (Sm-A, Sm-B, and Sm-C) and two insulin-like growth factors (IGF I and IGF 11) have been described. The only two peptides whose molecular structure has been fully defined are IGF I and 11. They are single-chain peptides that resemble proinsulin, with which they share 50 percent homology in amino acid sequence. They are syn-

P

CLINICAL CHEMISTRY

thesized and secreted by the liver in response to growth hormone. More recent studies (452,453) have revealed that Sm-C and IGF I are probably identical as are IGF I1 and Sm-A. Sm-B, on the other hand, is probably not a somatomedin at all but rather a small molecular weight antiprotease contaminated with epidermal owth factor. In the circulation these growth hormone depengnt factors are bound to carrier proteins. Abnormal circulating levels of these growth factors can lead to a number of high growth or low growth syndromes. The etiology for such abnormalities ean range from undernutrition and craniopharyngiomas to pituitary igantism and receptor unresponsiveness. Alterations in Sm evels are also found in chronic hepatic and renal diseases. Several assay methodologies are available for measuring circulating levels of these growth factors including bioassays, receptor assays, and immunoassays (454,455). The immunoassays are most desirable from the point of view of speed and convenience and are just now making their way into the clinical laboratory. The pancreas is an organ which exhibits both exocrine and endocrine function. The exocrine function is manifest in the secretion of digestive (amylolytic, lipolytic, and proteolytic) es into the small intestine in response to both hormonal Z g u r a l control. The endocrine function is predominantly associated with the production of hormones and factors that regulate the utilization of blood glucose. This family of hormones includes insulin, proinsulin, C-peptide, glucagon, and pancreatic polypeptide. Insulin is synthesized as an inactive precursor polypeptide, proinsulin, by the @ cells of the pancreas. Proinsulin is made as a linear polypeptide chain of 86 amino acids with three intrachain disulfide bonds. After synthesis, proinsulin is transferred from the endoplasmic reticulum of the cell to the peripheral elements of the Golgi. Here, conversion of proinsulin to the active insulin hormone occurs by proteolytic removal of the “connecting”of C-peptide (amino acids 31-65) leaving two peptide fragments connected by disulfide bridges. The active hormone (51 amino acid dipeptide) and the C-peptide are packaged together into secretory granules that await blood glucose stimulation before release into the blood circulatory system via the pancreatic vein. Note, that the C-peptide is not degraded but secreted along with insulin in equimolar amounts. The predominant disease associated with insulin is diabetes mellitus. Several other clinical disorders like hypoglycemia and insulinoma also relate to abnormal plasma levels of this pancreatic hormone. There are two forms of diabetes mellitus; Type I or insulindependent diabetes mellitus which is represented by an absolute deficiency in the ability of the pancreas to produce and/or release insulin and Type I1 or non-insulin-dependent diabetes mellitus which is manifest by a relative insulin deficienc as we11 as an insulin-resistance most often associated with ogesity. Although the measurement of insulin levels is not routinely necessary for the characterization of Type I or Type I1 diabetes, its measurement can be useful for diagnosing unusual causes of diabetes such as unexplained insulin-resistant states due to insulin-receptor disorders or antireceptor antibodies. Pancreatic tumors (Insulinomas)that cause release of sporadic supraphysiological levels of insulin, causing transient hypoglycemia, can also be diagnosed b monitoring plasma insulin during a 72-h fasting state. Tge ability to measure both intact proinsulin, insulin, and C-peptide has significant clinical value, particuarly in differentiating the various etiologic mechanisms for what clinically is broadly categorized as diabetes mellitus (456-458). For example, in some cases, although adequate levels of proinsulin are synthesized, the mechatlism for conversion of proinsulin to active insulin is missing, or in other cases, the normal mechanism for hormone secretion is failing. Even the production of variant forms of insulin due to an abnormal insulin gene has been identified as the mechanism for the diabetic syndrome. Very sensitive assays for insulin and the C-peptide, using radioimmunoassay technology, have been developed (459). The radioimmunoassa for insulin is also useful and sensitive for proinsulin, althougi under normal conditions, the amount of circulating proinsulin is very low. There are, however, eases when proinsulin comprises a major component of circulating insulin, and in these cases, quantificiation of C-peptide is useful to allow interpolation of insulin/proinsulin ratios. Insulin does not cross-react in the C-peptide assay, whereas proinsulin does to a moderate extent (5-20%). Occassionally, in cases where insulin cannot be measured because of inter-

’i

fering antibodies produced in response to earlier insulin treatment are present, C-peptide must be measured to get a measure of insulin. Glucagon is synthesized and secreted by the a cells of the islets of Langerhans as a straight-chain polypeptide of 29 amino acids. Glucagon has an opposite effect on blood glucose as that of insulin; it causes hyperglycemia by stimulating hepatic glucose production via glycogenolysis. Low blood glucose levels result in the release of glucagon into the circulation and high blood glucose suppresses glucagon secretion. The primary utility of measuring serum levels of glucagon is in the differential dia nosis of glucagonoma, a tumor of the a cells. A sensitive ra ioimmunoassay for glucagon has been described (460). The key to a useful assay has been the development of very specific antisera which do not cross react with glucagon-like peptides of enteric origin. Human pancreatic polypeptide is a rather new pancreatic hormone, the function of which is still unclear. Some recent studies have suggested that pancreatic polypeptide possesses some regulatory functions on the gastrointestinal activity, such as pancreatic e x h e secretion and gall bladder motility (461). The avian analog has been shown to be a powerful gastric stimulant when injected into chickens. Sensitive radioimmunoassays for quantifying plasma levels of this 36 amino acid peptide have been developed and are presently used as research tools for providing insight into the role of this hormone in health and disease (462). Tumors containing and secreting pancreatic polypeptide have also been described. Atherosclerosis. Phospholipids, triglycerides, and cholesterol are relatively insoluble in aqueous solutions like plasma and therefore must be transported in the blood stream as supramolecular complexes with a series of transport proteins. These protein-lipid complexes, called lipoproteins, are discrete, generally spherical particles ranging in size from 8 to 800 nm. The lipoproteins can be distinguished according to the particular transport protein (apolipoproteins) present in the complex, the density of the complex, and the specific lipid content in the complex. There are five principle classes of lipoproteins: chylomicra, very low density lipoproteins (VLDL), intermediate density lipoproteins (IDL), low density lipoproteins (LDL), and high density lipoproteins (HDL). Chylomicra are of the lowest density and largest size and are composed primarily of exogenous triglycerides (>95% by weight) absorbed into the circulatory system from the intestinal tract after the ingestion of fatty meal. These lipids are destined for utilization of storage in peripheral tissues. The protein associated with chylomicra is a special intestinal form of apolipoprotein B called apo B-48. VLDL have a density of