SA MEDIESE TYDSKRIF DEEL 65 2 JUNIE 1984 875
Serological techniques for the diagnosis
M. D. PAMMENTER, E. J. ROSSOUW
Serological tests for cysticercosis have long been used as a
diagnostic aid,s-l2 and because of the lack of specific clinical
symptoms they play a major role in its detection. In an attempt to
f!Dd the most efficient diagnostic technique we recently estab-
Three techniques - an indirect haemagglutination
lished an enzyme-linked immunosorbent assay (ELISA) and an
test, a fluorescent antibody test and enzyme-linked
indirect fluorescent antibody test (FAT), and have compared
immunosorbent assay (EUSA) - have been estab- their efficiency with that of a modification of the indirect
lished for the detection of cysticercosis antibodies in haemagglutination test (HAT) originally developed by Proctor er
1he serum or cerebrospinal fluid of patients with al. 9 The results obtained using these three assay systems are
cysticercosis. Results obtained using these techniques
reponed in this article.
have been compared todeterminethe most successful
serodiagnostic method. In each test there was a
marked difference between thedelection of active and
of calcified cysts. The results obtained using the three Material and methods
tests were remarkably similar. and none had a statis-
. ticaIIy significant advantage over the others with Antigen
reprd to the detllclion of cysticercosis antibodies.
Cysts were removed from infected pork and stored at -86°C
until required. The FAT antigen consisted of 6 J-lm thick sections
of 2 - 4 cysts cut by cryomicrotomy. The same antigenic
S . ....,J11M;_ 815-I1a preparation was used in the HAT and the ELISA. Cysts were
homogenized in 0,02M tris, 1,0 mM ethylenediamine tetra-
acetic acid (pH 7,4), using a ground-glass Potter-Elvehjem
homogenizer, until the outer membranous material of the cysts
was completely homogenized. The homogenate was centrifuged
Cysticercosis, a disease caused by the cystic stage of the pork at 105000 g to yield a supematant which was used as the antigen.
tapeworm (Taenia solium), has been described as the most
common parasitic disease affecting the central nervous system in
the world. l Under normal circumstances the adult worm lives in
the intestine ofman and the cystic stage takes place in pigs, which The HAT
become infected by consuming eggs excreted in human faeces. Cells were prepared as described by Hoq and Das. 13 Human
The eggs hatch and release oncospheres which migrate through O-negative or O-positive cells were fixed with 2% glutaraldehyde
the tissues, where they develop into cysts. Humans occasionally and tanned with a 1/40 000 tannic acid solution. Antigen was
ingest the eggs and become the intermediate host. While cysts attached to the cells at a concentration previously determined by
may be found in various tissues, they appear to have a titration. The HAT was performed in a total of 50 J-ll using
predilection for the central nervous system. 2 standard procedures. Cerebrospinal fluid (CSF) or serum
Despite repons showing a high prevalence3 ofcysticercosis in (diluted 1/10) was consecutively double-diluted with 2% bovine
Black population grou£s and indicating the imponant role it serum albumin (BSA) and mixed with an equal volume of a 1%
plays in mental illness, -6 the disease has never been regarded as sensitized cell suspension.
a serious problem by the South African medical profession.
Recently cysticercosis has received some prominent publicity,
and the advent of new drugs to combat the disease has brought it The FAT
to the fore again. 7
The condition is now treatable early and accurate diagnosis The FAT was performed as recommended by Kawamura. 14
may be crucial, but diagnosis has always been problematic. Sections of cysts were incubated at 37°C with CSF or a 1/10
Clinical symptoms include nausea, headache with disturbed dilution of serum in phosphate-buffered saline (0,066M sodium
vision, seizures, hemiparesis, raised intracranial pressure and onhophosphate, 0,85% NaCI, pH 7,2). After washing the
papilloedema, but none of these is specific for cysticercosis. The sections were incubated with a 1/30 dilution of fluorescein-
presence ofcysts in subcutaneous nodules confirms the diagnosis, conjugated anti-human IgG plus IgM (Hoechst OTKG 05),
but this feature is reponed to be relatively rare. s with 0,05% Evans blue as a counterstain.
Diagnosis by computed tomography (CT) is becoming in-
creasingly popular and, with the advent ofimproved equipment,
increasingly accurate. However, its use is restricted and in TheELISA
cenain cases positive identification of cysts is still impossible. Flat-bottomed PVC immunoassay plates (Flow Laboratories
77-173-05) were coated with 50 J-ll antigen solution at room
temperature for 2 hours in a moist container. The antigen
Research Institute for Diseases in a Tropical Environment of solution contained 10 J-lg protein per millilitre ofO,05M sodium
the South African Medical Research Council, Durban carbonate buffer, pH 9,6. After coating, the plates were washed
M. D. PAMMENTER, PH.D., Senior Chief Research Officer with TST (0,05M tris, pH 8,0; 0,75M NaCl; 0,05% Tween 20)
E. J. ROSSOUW, DIP. MED. TECH., Technical Officer
and blocked with 2% BSA in a sodium carbonate buffer.
876 SA MEDICAL JOURNAL VOLUME 65 2 JUNE 1984
Blocking was accomplished after 1 hour, following which plates
were washed, dried and stored at -86°C. Plates were stored for ELlSA HAT FAT
up to 3 months without any apparent loss of activity. > • 0
The assay was conducted as described by Conradie and
Mbhele l5 with the following modifications: serum samples were
tested at four dilutions (1/50, 1/100, 1/200 and 1/400) made 32
with 4% normal pig serum in TST (the use ofnormal pig serum is
critical; without it a variable backgrolind colour develops). An
alkaline phosphatase-conjugated anti-human IgG (Sigma A
3150) diluted 1/500 with 4% pig serum was used and colour was
developed using 'Sigma 104' substrate tablets (Sigma 104-105) 16
in 1,OM diethanolamine and 0,5 mM MgC1 2, pH 9,8, after •
which the absorbance at 405 nm (OD 405) was recorded. Two
positive and negative controls were included in each assay plate.
The OD 405 of the negative controls was regarded as
background and subtracted from the OD 405 of the appropriate
• g8 • 0
dilutions of the positive controls and unknown samples to give a 0
'true' OD 405 reading. The positive control was deemed to have N
1 280 units per microlitre and a standard curve of units versus 0
'true' OD 405 was constructed for each plate. From this curve x ""
- - •
the number of units in unknown samples was calculated. 0
>- • C<
• C< :::>
Fig. 1 is a graphic representation of the results of the three tests
2 - 0 ....
performed on sera obtained from patients with cysticercosis (the • • MM
diagnosis being confirmed by either biopsy or a CT scan). The
control group comprised 79 patients suffering from amoebic
liver abscess. A complete clinical examination was performed on • • 000 MM
each patient in this group and all were found to be free of 1
symptomatic neurological disorders or palpable nodules. •
Although admittedly this could not exclude asymptomatic
cerebral or muscular cysticercosis, only 1 of the 79 sera tested
gave a positive reaction to more than one test, and to all intents A Ca -ve A Ca -ye A Ca -ye
and purposes this group could therefore be regarded as free of
cysticercosis. Fig. 1. Results of the three tests on serum from patients with
Fig. 1 shows that, as is the case with all tests of this type, cysticercosis and control subjects. The dotted line represents the
sensitivity must be weighed against specificity. We have selected point above which results were regarded as positive (A = active
cysticercosis; Ca = cranial calcifications; -ye = control subjects).
the following cut-off points for positivity: (I) ELISA - 80 units
and above; (il) HAT - a positive reaction at a dilution of 1/40 or
more; and (iil) FAT - 1 unit of fluorescence at a 1/10 dilution of Similar results were obtained with CSF (Table I), but the
serum. Using these values we were able to calculate the following sensitivity of the FAT was considerably less than with serum.
sensitivities for the tests: ELISA 45,5%; HAT 33,3%; and FAT From the results presented in Table II iris clear that there was a
48,9% (Table I). close correlation between the effectiveness oftests. The apparent
TABLE I. DETECTION OF ANTICYST ANTIBODIES IN THE SERUM AND CSF OF
PATIENTS WITH CYSTICERCOSIS
ELlSA HAT FAT
No. % No. % No. %
Overall 20/44 45,5 17/51 33,3 22145 48,9
Calcified cysts only 3/18 16,7 5/24 20,8 1119 5,3
Active cysticercosis 12117 70,6 11120 55,0 12115 80,0
Overall 25/45 55,6 16/51 31,4 12/50 24,0
Calcified cysts only 8/23 34,8 4/24 16,7 3/24 12,5
Active cysticercosis 12115 80,0 8/16 50,0 5/14 35,7
Combined CSF and serum
Overall 45/89 50,6 33/102 32,4 33/94 35,1
Calcified cysts only 11/41 26,8 21175 28,0 4/43 9,3
Active cysticercosis 24/32 75,0 19/36 52,8 17/29 58,6
(control group) 1179 1,3 1/79 1,3 5/79 6,3
SA MEDIESE TYDSKRIF DEEL 65 2 JUNIE 1984 877
TABLE 11. DEGREE OF AGREEMENT BETWEEN THE RESULTS OF
THE VARIOUS TESTS·
Tests Agreement K P
ELlSA + HAT 38/44 (86~) 0,720 <0,0001
ELlSA + FAT 35/42 (83~) 0,667 < 0,0001
HAT + FAT 39/52 (75~) 0,447 < 0,0005
ELlSA + HAT 23/47 (49%) 0,078 <0,05
ELlSA + FAT 32/47 (68%) 0,396 < 0,0007
HAT + FAT 42/50 (84%) 0,585 <0,0001
• K and P were calculated using the McNemar lest of symmetry_
low sensitivity of the tests (Table I) can be explained by the inability of serological tests to detect calcified cysts in no way
status of the disease. The patients diagnosed as having cysticer- detracts from their usefulness.
cosis could be divided into two groups, those with active cysts The value of serological tests is not uniformly accepted,18,19,21
and those with calcified lesions only. There was a high sensitivity and the clinician must therefore know the implications of a
for active cysticercosis, but calcified cysts were difficult to detect laboratory result. In our tests a positive result was highly
serologically (Fig. I, Table I). indicative of active cysticercosis requiring treatment. With the
Sera from patients with other parasitic diseases were also ELISA 75% of cases of active cysticercosis can be detected; this
tested; no significant difference from the control subjects was can be increased to 88% if all three tests are used. In about 13% of
found in the rate of positive titres in serum from patients infected active cases a positive result was obtained in only one of the three
with Schistosoma species (10/49, FAT) or Hymenolepis (1/9, tests, and these are the cases causing .concern. If the test in
FAT). However, the ELISA gave positive results for 7 of a total question is the ELISA or the HAT the result is still highly
of 10 sera from patients with hydatid disease. suggestive o~ the presence of the disease, since false-positive
results are rare in both these tests, However, the FAT would
seem to have the highest false-positive rate (5 out of79 results in
the control group and 10 out of 49 in patients with schisto-
Discussion somiasis), so that a single positive FAT result may be regarded
Several tests for cysticercosis, with sensitivities ranging from with some suspicion. Unfortunately we were unable to determine
25% to 100% and false-positivity rates of2 - 25%, have been de- the statistical significance of these results because of the small
veloped. 16 Those results compare favourably with the results ob- number ofsamples and the impossibility ofabsolute exclusion of
tained during the present study. While the overall sensitivity ofthe cysticercosis.
ELISA (50%) appears lower than has been reported by Diwan er Our results show that, on the whole, no test has a statistical
al. ll (60%) and Arambulo er alP (75%), its sensitivity in the advantage over the others. However, the ELISA is significantly
detection of active cysticercosis is at least equal (75%). The more effective than the HAT or the FAT in detecting active
ELISA system used during this study had the advantage over cysticercosis when CSF samples are tested (P = 0,02). This
that used by Diwan er al. ll that only one test was necessary. difference may be related to the small number of samples,
Those workers used a control ELISA with porcine soluble although with an agreement ofonly 33% between the ELISA and
proteins as antigen; this was necessary to eliminate false-positive the FAT this seems unlikely.
results caused by the development of high levels of background When selecting a test system there are other aspects to be
colour which it was suggested were due to host-related proteins considered; for example, although the HAT does not require
being carried over with the cyst-antigen preparation. We have sophisticated equipment, interpretation of the results is sub-
overcome this problem by using 4% normal pig serum as the jective and requires a skilled operator. The F AT has the
diluent in the test. This resulted in an almost 50% reduction in disadvantages of both requiring expensive equipment and
the OD 405 values and reduced background colour to low levels needing an experienced technician to interpret the result. An
(0 - 50 units), obviating the use of a control antigen. However, examination of Fig. I shows that with both the HAT and the
this procedure may decrease the potential sensitivity of the test. FAT a number of results fall within the area which could easily
It seems unlikely that the vast differences in test sensitivities be misinterpreted. Although the ELISA may not be any more
reported by various workers 8,9-12,17 are due solely to differences accurate than the other tests, the result is always unequivocal and
in laboratory technique. The selection of patients for assessment cannot be misinterpreted through inexperience.
of the tests plays an important role. ll ,17 The results of this study It was not the intention ofthis study to promote any particular
clearly indicate that patients with calcified cysts only have low technique but rather to assess the currently available tests. It can
antibody levels and that inclusion of their sera lowers the be concluded that the tests are highly effective in detecting active
calculated sensitivity. With modem equipment, calcified cysts cysticercosis (except when the FAT is used on a CSF sample)
are readily detected and many cases ofcysticercosis which would and that a positive result, especially if produced by more than
previously have remained undiagnosed are now being detected. one test, can be regarded with a high degree of certainty as
Because of the low antibody titres in these patients, serological indicating cysticercosis.
tests appear to be less effective than has previously been
reported. 9,10,18 We wish to thank the staffofthe Cato Ridge abattoir for supplying
the infected pork, Professor J. van Dellen and Dr M. Walshman of
Natural calcification of cysts is regarded as denoting spon- the Departments of Neurosurgery and Radiology at Wentworth
taneous regression of the disease with return to normal and Hospital for providing the serum and CSF samples and CT scan
disappearance of symptoms. 2,19 Subjects with calcified cysts results, and the South African Medical Research Council for
would not be considered for treatment,20 and for this reason the permission to publish.
878 SA MEDICAL JOURNAL VOLUME 65 2 JUNE 1984
12. Arambulo PV, Wals KW, Bullock S, Kagan IG. Serodiagnosis of human
1. Wiederholt WC, Grisolia J. Cysticercosis. Arch Neuro11982; 39: 533. cysticercosis by microplate enzyme-linked immunospecific assay. Acta Trap
2. Dixon HBF, Lipscomb FM. Cyscicercosis: An Analysis and Folluw-up of 450 (Basel) 1978; 35: 63-67.
Cases (British Medical Research Council Special Report Series No. 299). 13. Hoq MS, Das PC. Preparation ofhuman cells for the assay of serum fibrinogen
London: HMSO, 1961. degradation products using haemagglutination inhibition. S Afr] Haemacol
3. Heinz HJ, MacNab GM. Cysticercosis in the Bantu of southern Africa. S Afr] 1971; 3: 101-105.
Med Sci 1965; 30: 19-31. 14. Kawamura A. Fluorescent Antibody Techniques and their Applicarions. Tokyo:
4. Powell SJ, Proctor EM, Wilmot AJ, MacLeod IN. Cysticercosis and epilepsy Universiry of Tokyo Press, 1969: 70.
in Africans: a clinical and serological study. Ann Trap Med Parasico/l966; 60: 15. Conradie JD, Mbhele BEL. Quantitation of serum ferritin by enzyme-linkel1
152-/58. immunosorbent assay (ELISA). S Afr Med] 1980; 57: 282-287.
5. Heinz HJ, Klinrworth GK. Cysticercosis and epilepsy. S Afr] Med Sci 1965; 16. Flisser A, Perez-Montford R, Larralde C. The immunology of human and
30: 32-36. . animal cysticercosis. Bull WHO 1979; 57: 839-856.
6. Powell SJ, Proctor EM, Wilmot AJ, Bamen AM. Neurological complications 17. Rydzewski AK, Chisholm ES, Kagan IG. Comparison of serologic tests for
of cysticercosis in Africans. Ann Trap Med ParasitoI1966; 60: 159-164. human cysticercosis by indisecr hemagglutination, indisect immunofluo-
7. Groll E. Cysticercosis humana y praziquantel. Bol Chil Parasicol 1981; 36: rescent antibody and agar gell precipitin tests.] Parasiro/1975; 61: 154-155.
29-37. . 18. Byrd SE, Locke GE, Biggers S, Percy AK. The computed tomographic
8. Nieto D. Cysticercosis of the nervous system. Neurology 1956; 6: 725-738. appearance of cerebral cysticercosis in adults and children. Radiology 1982;
9. Proctor EM, Powell SJ, Elsdon-Dew R. The serological diagnosis of cysticer- 144: 819-823. :
cosis. Ann Trap Med ParasitoI1966; 60: 146-151. 19. McCormick GF, Zee CS, Heiden J. Cysticercosis cerebri. Arch NeU,-011982;
10. Flisser A, Woodhouse E, Larralde C. Human cysticercosis: antigens, antibodies 39: 534-539. .
and non responders. ClinExplmmuno/1980; 39: 27. 20. Boteto D, Castano S. Treatment of cysticercosis with praziqtiantel in
11. Diwan AR, Coker-Vann M, Brown P et al. Enzyme-linked immunosorbent Columbia. Am] Trop Med Hyg 1982; 31: 810-821.
assay (ELISA) for the detection ofantibody to cysticerci of Taenia solium. Am] 21. Loo L, Braude A. Cerebral cysticercosis in San Diego. Medicine (Baltimore)
Trop Med Hyg 1982; 31: 364-369. 1982; 61: 341-359.
Endogenous immunoreactive digitalis-like
substance in neonatal serum and
A. D. SEVERS, . L. L. SPRUYT, H. I. SEIFART, A. KRIEGLER, D. P. PARKIN,
P. P. VAN JAARSVELD
Summary statistically highly significant. In the case of infants
with DLS values of 1 - 1,5 ng/ml in terms of digoxin,
Therapeutic levels of digoxin in the serum of untreated approximately 1 week was required to reach non-
neonates delivered to mothers who had not received therapeutic digoxin levels, Le. below 0,5 ng/ml.
the drug prenatally were detected by radio-immuno- Gel chromatography showed thatthe DLS in neonatal
assay. Digoxin lev~s in neonates should be interpreted serum was more closely associated with protein than
with care because of the unknown contribution by the is authentic digoxin. In placental extracts it followed
endogenous digitalis-like substance (DLS) to the level the elution profile of the protein completely, but it
of the drug. shifted to fractions with a lower molecular weight than
Three commercially available radio-immunoassay haemoglobin after trypsinization. The level of DLS in
kits were compared with regard to thei r sensitivity and neonatal serum was also increased by more than half
reproducibility in detecting the endogenous DLS. The its original value by trypsinization. Proteolysis therefore
kit from Clinical Assays (Cambridge, Mass., USA) was seems to have a releasing effect on DLS. The molecular
selected for further investigations. In a series of 35 size of this substance is probably in the same range as
paired samples of maternal and cord blood the average that of polypeptides, since it was not dialysable from
DLS values in terms of digoxin were 0,52 ± 0,07 and trypsinized and untreated samples through a mem-
0,81 ± 0,27 ng/ml respectively. This difference is brane with a 22000 dalton molecular weight cut-off
S Air Med J 1984; 65: 878-882.
Departments of Pharmacology and Internal Medicine,
University of Stellenbosch, Parowvallei, CP
A. D. BEYERS, M.B. CH.B.
L. L. SPRUYT, B.sc. (pHARM.) HONS, M.B. CH.B. The existence of an endogenous digitalis-like factor has received
H. I. SEIFART, DR. RER. NAT. much anention in the recent literature. l The cardiac glycosides
A. KRIEGLER, TECH. DIPL. lend themselves to these investigations, since their concentrations
D. P. PARKIN, B.sc. (FARMAKOL.) HONS, M.B. CH.B. can be determined with great specificity by radio-immunoassay
P. P. VAN JAARSVELD, PH.D. and their high-affinity inhibitory action on Na+-K'"-adenosine