Scandinavian Journal of Infectious Diseases
ISSN: 0036-5548 (Print) 1651-1980 (Online) Journal homepage: http://www.tandfonline.com/loi/infd19
Nitroblue Tetrazolium Test in Bacterial and Viral Infections Kjell B. Hellum To cite this article: Kjell B. Hellum (1977) Nitroblue Tetrazolium Test in Bacterial and Viral Infections, Scandinavian Journal of Infectious Diseases, 9:4, 269-276, DOI: 10.3109/ inf.1977.9.issue-4.03 To link to this article: http://dx.doi.org/10.3109/inf.1977.9.issue-4.03
Published online: 02 Jan 2015.
Submit your article to this journal
View related articles
Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=infd19 Download by: [University of Saskatchewan Library]
Date: 31 March 2016, At: 00:49
Scand J Infect Dis 9: 269-276, 1977
Nitroblue Tetrazolium Test in Bacterial and Viral Infections KJELL B. HELLUM
Scandinavian Journal of Infectious Diseases 1977.9:269-276.
From the School of Medicine, Medical Department B, University of Bergen, Bergen, Norway
ABSTRACT. The reduction of nitroblue tetrazolium (NBT) dye by neutrophils from 379 patients with infectious diseases and 268 controls has been examined. The mean NBT score was 29.8 % (72.3 % positive tests) in the 231 patients with non-tuberculous bacterial infections, 9.7 % (28.1 % positive tests) in the 135 patients with viral infections and 5.3% (1.5% positive tests) in the controls. Positive tests were demonstrated in 1 of 7 patients with tuberculosis and in 4 of 6 with mycoplasma pneumonia. Patients with urinary tract infections or septicemia had the highest percentage of positive tests, particularly when the infections were caused by gram-negative bacteria. In acute bacterial infection, the 176 patients who had not received any antibacterial therapy prior to testing had a significantly higher mean NBT score and proportion (77.8%) of Recent antibiotic treatment positive tests than the remaining 55 pretreated patients (54.5 70). seriously invalidates the NBT test results. In acute viral infection, 29 of the 38 positive tests were obtained from patients with acute hepatitis (mean score 20.0 %) or infectious mononucleosis (mean score 9.3 %). When evaluating the test results, special attention should be paid to patients with hepatitis. Endotoxin stimulated NBT tests disclosed normal enhancement of NBT reduction by neutrophils from the patients and the controls. Cautiously interpreted, the NBT test results may be useful as an adjunct in the differential diagnosis of major bacterial and viral infections.
INTRODUCTION Common leucocyte characteristics used in the differential diagnosis of infectious diseases include total and differential peripheral white blood cell counts, presence of toxic granulation, Dohle bodies, vacuolisation of the cytoplasm and alkaline phosphatase level (3 1). Unfortunately, these parameters often fail to prove the presence or absence of bacterial infection. The application of the histochemical nitroblue tetrazolium (NBT) test as an adjunct in the diagnosis of acute bacterial infection was reported in 1968 (22), when an increased spontaneous in vitro reduction of the NBT dye to formazan by human peripheral neutrophils was demonstrated in bacterial but not in viral infections or non-bacterial febrile disorders. The exact mechanisms involved with the internalisation and reduction of the dye are not entirely clear (2, 20, 28), but depend on phagocytic and oxidative potentials of the cells, thus making the reduced dye an indirect marker of phagocytic activity which seems to be stimulated during bacterial infections (19, 30). The diagnostic relevance of
the NBT test has, however, been questioned (4, 7, 20, 28, 29), and controversy exists as to its application in the evaluation of pyrexial syndromes of unknown etiology ( 1 , 3 , 10, 11, 18,27). Moreover, due to major influence of technical variables on the test results, carefully standardized and controlled test procedures seem necessary (1,4, 13, 14, 15). The aim of the present study was to evaluate a standardized histochemical NBT test in the differential diagnosis of major bacterial and viral infections. MATERIALS AND METHODS Patients and controls Originally, the study included 610 patients admitted to Medical Department B, Haukeland Hospital for febrile disorders consistent with various infections and 380 healthy members of the hospital staff, medical students and student nurses of the department who served as controls. The NBT test was performed within 1-48 h after patient admission, at a time when an etiologic microbiological diagnosis had been made in only 84 patients. In the remaining 526 patients, an etiologic infectious disease Scand J Infect Dis 9
270
K . B . Hellurn
Table I. Patient subjected to the NBT test No. of patients"
Scandinavian Journal of Infectious Diseases 1977.9:269-276.
Diagnosis Bacterial infections Lower respiratory tract infection Urinary tract infection Septicemia and bacterial endocarditis Acute bacterial meningitis Deep abscesses, empyema or peritonitis Acute arthritis or osteomyelitis Tuberculous infection Erysipelas Acute sinusitis Mycoplasma infections Viral infections Acute meningoencephalitis Acute hepatitis Infectious mononucleosis Exanthematous infections Measles Varicella Herpes zoster Herpes simplex Rubella 2 1 Vaccinia Mumps, uncomplicated Respiratory or gastrointestinal tract infections Influenza A Adenovirus 4 RS virus 1 Coxsackievirus B4
ij
r)
3 (ij 6 51 28 27
19
5
5
The parentheses indicate the number of patients with bacterial infections who had received systemic antibiotics within 48 h prior to the NBT test.
diagnosis could be established in only 295 patients, and the remaining 231 patients together with 112 controls had to be excluded from further evaluation. Accordingly, the final study included 379 patients and 268 controls. 238 patients (13 1 males and 107 females aged 9-87 years, mean 56 years) had bacterial infections, 6 (3 males and 3 females aged 18-43 years, mean 33 years) had Mycoplasma pneumoniae infections and 135 patients (70 males and 65 females aged 4-82 years, mean 28 years) had acute viral infections (Table I). In the control group, 79 were males and 189 females, and the age ranged from 16 to 68 years, mean age being 29 years. Criteria of infection The diagnosis of lower respiratory tract infection was based on typical clinical and X-ray findings together with positive cultures of respiratory pathogens from transtracheal aspirates. 10 I patients 1 also1 had /positive blood cultures. Two patients had pure cultures of M. pneumoniae from the transtracheal aspirates, and 4 more patients had a significant rise in antibody titer to this agent, using the complement fixation (CF) test. Of the 94 patients with lower respiratory tract infections (Table I), Scand J Infect Dis 9
88 caused by bacteria and 6 by M. pneumoniae, 19 patients died. Autopsy performed in 15 cases confirmed the diagnosis. Acute urinary tract infection was diagnosed by positive bacterial cultures from suprapubic bladder aspirates together with clinical, radiographic and/or other laboratory findings. The majority of the 48 patients in this group had severe symptoms, and 8 patients died. Autopsy confirmed the diagnosis in each of these cases. A diagnosis of septicemia was made only when blood cultures grew out pathogens in the presence of systemic severe illness with spiking fever and/or shock symptoms. Altogether 41 patients had septicemia. Two more patients who had several positive blood cultures, but also clinical and laboratory evidence of valvular heart disease, were considered to have bacterial endocarditis. Eight patients died, including 1 patient with bacterial endocarditis, and autopsy substantiated the diagnosis in each case. Acute bacterial meningitis was diagnosed by positive bacteriological cultures from the cerebrospinal fluid (CSF) in patients with characteristic clinical and laboratory findings. In patients with deep abscesses, empyema, peritonitis, acute sinusitis and septic arthritis or osteomyelitis, the diagnoses were based upon positive bacteriological cultures of purulent specimens obtained aseptically by needle aspiration at surgery or at autopsy (4 cases). In 7 patients, tuberculous infection was demonstrated by positive cultures of Mycobacterium tuberculosis. Five patients had pulmonary involvement, 1 a tuberculous meningitis and the remaining patient had a tuberculous lymphadenitis. Erysipelas was diagnosed from the clinical picture together with a significant rise in antistreptolysin-0 titer and/or positive cultures of P-hemolytic streptococci from the site of infection. Viral meningoencephalitis was diagnosed from the clinical picture and the results of CSF examination (30-2 OOO predominantly mononuclear leukocyteslpl, protein increase up to 100 mg/dl, glucose ratio CSF:blood>O.S, absence of bacteria in stained smears and routine bacteriologic cultures). In 20 patients, the diagnosis was confirmed serologically by a 4-fold or greater rise in antibody titer to a specific virus: in 10 patients mumps virus, in 6 echovirus, in 2 coxsackievirus B and finally, coxsackievirus A and adenovirus in 1 patient each. Virus isolation from the CSF and/or faeces succeeded in 11 of these cases. 13 other patients had mumps diagnosed on clinical grounds alone. In the remaining 18 patients the evidence for a specific viral etiology was inconclusive. However, serologic tests for M. pneumoniae and Toxoplasma gondii also turned out negative, as did the search for extra- or intracranial processes capable of mimicking the CSF changes of viral meningitis. These patients all made a complete recovery without any administration of antibiotics prior to admission or in hospital. A diagnosis of acute viral hepatitis was based upon epidemiological, clinical and laboratory examination, including liver biopsies. HBsAg was demonstrated in 8 patients. Infectious mononucleosis was diagnosed on the basis of clinical symptoms, results of peripheral blood smear ex-
Nitroblue tetrazolium test in infections
27 1
Table 11. Unstimulated and endotoxin stimulated NBT tests in 379 patients with bacterial, mycoplasma or viral infections and in 268 normal controls Unstimulated test neutrophils
Positive testn Negative test
Endotoxin stimulated test % NBT-positive neutrophils
Meank1S.D. Range
No.
%
No.
%
Meanf1S.D. Range
% NBT-positive
Scandinavian Journal of Infectious Diseases 1977.9:269-276.
Diagnoses Normal controls Acute bacterial infections Untreated Pretreated Tuberculosis Pneumonia caused by M. pneumoniae Viral infections
No. of subjects 268
5.3k 3.1
0-18
4
1.5
264
98.5
46.8k13.2
22-79
176 55 7
32.6k21.6 21.1514.8 10.65 10.1
1-93 0-68 1-3 1
137 30 1
77.8 54.5
39 25 6
22.2 45.5
69.9k18.6 61.7k17.0 53.5k18.1
18-98 20-89 23-76
6 135
29.8f24.1 9.75 9.6
3-61 0-51
4 38
60.2k25.4 51.9k16.0
28-88 21-94
28.1
2 97
71.9
Positive test defined as >15% NBT-positive neutrophils (see text). amination and a significant rise in heterophile antibody titer. Patients with exanthematous viral infections or mumps all had readily recognizable diseases. In 5 patients with predominant respiratory or gastrointestinal symptoms, a diagnosis of viral infection was based on a significant rise in virus antibody titer together with virus isolation. 176 patients with bacterial non-tuberculous infections had not received any antibiotic therapy prior to the NBT test, whereas the remaining 55 patients in this group had received systemic antibiotics within the last 48 h (Table I). Histochemical NBT test
The test was performed by a standardized technique using heparinized whole blood, slightly modified from the test of Park et al. (22) as previously described (14, 15). Unstimulated and endotoxin stimulated tests were performed contemporaneously on the same blood sample (14,23). In each test, 300 neutrophils in each of 2 slide smears were classified as NBT-negative (without bluish-black formazan) or NBT-positive (containing deposits of formazan). The percentage of NBT-positive neutrophils was recorded as the NBT score and without knowledge of patient diagnosis. A control test was always performed in parallel with one'or more patient tests.
RESULTS Healthy individuals In the 268 normal controls, the percentage of NBT-positive neutrophils in the unstimulated test ranged from 0-18, mean 5.3 (Table 11). Scores above 15% were regarded as positive NBT tests. This upper level is well beyond the mean plus 2 standard deviations for the present controls and in accordance with previous results using an identical test procedure (16, 17). Thus, only 4 normal indi-
viduals (1.5 %) had a positive test. Endotoxin stimulated tests demonstrated a 3-fold or greater increase in the number of NBT-positive neutrophils compared to unstimulated test scores. However, the results ranged widely (22-79 %), mean score being 46.8 %.
Bacterial and mycoplasma infections In the 23 1 patients with non-tuberculous bacterial infections, the mean NBT score was 29.8% (range 0-93%) and 167 patients (72.3%) had a positive test. However, in the group of 176 patients who received no systemic antibacterial treatment prior to testing, the proportion of positive tests was significantly higher (77.8 %) than in the 55 pretreated patients (54.5%) and the mean scores in these groups were also significantly different (t-test,PO. l), indicating no functional impairment of neutrophil reducing capacity during systemic antimicrobial therapy. The highest number of positive tests in the nontreated patients was demonstrated in those with acute urinary tract infections (83.3 % positive tests, mean score 39.3%), closely followed by the septicemia group (80.0 % positive tests, mean score 38.1 %) and the group with lower respiratory tract infections (77.9 % positive tests, mean score 29.4%) (Fig. 1). The mean scores in patients with Scand J Infect Dis 9
272
K . B . Hellurn
D O
90
0
80
0,
70
a I
.
0
I
1
tW 3
60 W
2
8t
50
‘
40
z
30
a
Scandinavian Journal of Infectious Diseases 1977.9:269-276.
2 5
20
kw2 a
10
: 0
mer resFxrabi raci infections
Urinary tract infections
Septicemia and ndocarditis
I :
1
Menlngltls
Unstimulated NET test 0
Endotwin stimulated NET test
- mean xom
- - - -upper limit for r q a t i w NBT tesls
meningitis or other types of major non-tuberculous bacterial infections were markedly lower than in the former groups, the percentage of positive tests being only 70.6 and 70.0, respectively. In the 7 patients with active tuberculous infections, only 1 patient with a severe lymphadenitis had a positive test. Endotoxin stimulated tests in each of these patients demonstrated a normal enhancement of NBT dye reduction. In contrast, 4 of the 6 patients with mycoplasma infections had markedly elevated NBT scores. Thus, the results in this patient group were similar to those obtained in the patients with bacterial pneumonia. In 168 patients with untreated bacterial infec-
(see text
)
Fig. 1 . Unstimulated and endotoxin stimulated NBT tests in 176 patients with untreated acute bacterial infections.
tions, tuberculosis excluded, the causative organisms were recovered as monocultures (Table 111). 94 patients (56%) had infections with grampositive bacteria and 69 (73 %) had positive tests. In the 74 patients with infections due to gram-negative organisms, the percentage of positive tests was 81. The differences between the numbers of positive tests in infections caused by gram-positive versus gram-negative species were more prominent in the patients with septicemia or urinary tract infections than in the other disease groups. However, these differences did not reach significant levels (chi square test, P>O. I). Severe underlying disease (viz. malignancies, un-
Table 111. NBT tests according to gram stain properties of the etiologic agents recovered as monocultures from 168 patients with bacterial infections Percentage of patients with positive tests” (no. of patients tested)
Causative microorganisms
Lower respiratory tract infections
Urinary tract infections
Septicemia and endocarditis
Meningitis
Other acute infections
Total
Gram-positive species Gram-negative species
78 (SO) 81 (16)
67 (6) 89 (26)
67 (18) 88 (16)
60 ( 5 ) 67 (12)
73 (15) 50 (4)
73 (94) 81 (74)
Positive test defined as >15% NBT-positive neutrophils (see text). Scand J Infect Dis 9
Nitroblue tetrazolium test in infections
273
PP 0
-8-
.
;I*
.
i
O
9-
8 0
Scandinavian Journal of Infectious Diseases 1977.9:269-276.
D
- - - --- - -
I 0
MenlKj.3encephalitis
nfectious noncmucleosis
Acute hepatits
Various infections
Unstimulated NBT test Endotoxin stimulated NBT test
- mean
score ----upper
iimt tor negatw NBT tests (see text)
controlled diabetes mellitus, heart failure, uremia, collagen diseases, liver cirrhosis, bowel infarction and gangrene) was present in 37 patients without antibiotic pretreatment and 14 patients on antibiotics. Furthermore, 12 patients were on corticosteroid therapy (prednisone 5-40 mg per day), 8 in the non-treatment group and 4 in the treatment group. Two more patients in the latter group received treatment with azathioprine (100 mg per day). Exclusion of these patient categories did neither significantly alter the proportion of positive tests, nor the mean scores in the various disease groups. Finally, these patients's neutrophils had a normal response to endotoxin stimulation (Fig. 1). Viral infections In the 135 patients with viral infections, the mean NBT score was 9.7% in the unstimulated test and only 38 patients (28.1 %) had positive tests (Table 11). 22 of the 28 patients (78.6%) with acute viral
hepatitis A or B had positive tests during the early icteric stage, mean score being 20.0% NBTpositive neutrophils (Fig. 2). The 27 patients with infectious mononucleosis also had a relatively high number (25.9%) of positive tests (mean score 9.3 %). However, neither clinical picture nor laboratory test results in the patients with hepatitis 18-77 1959
Fig. 2. Unstimulated and endotoxin stimulated NBT tests in 135 patients with acute viral infections.
or infectious mononucleosis disclosed any significant difference between the severity of the disease process in patients with positive and negative tests. In the 51 patients with viral meningoencephalitis, the mean score was 6.4 % and only 6 (1 1.8 %) had positive tests. Similar results were obtained in the patients with exanthematic viral syndromes (mean score 6.1 %, positive tests in 2 patients with measles) and in the remaining 10 patients with other acute viral infections (mean score 5.8%, positive test in 1 patient with uncomplicated mumps). Endotoxin stimulated tests in patients with acute viral infections disclosed normal or supranormal NBT dye reducing capacity. DISCUSSION The present study was undertaken to evaluate prospectively the histochemical NBT test as an indicator of bacterial versus viral infections in patients admitted to an infectious disease department with syndromes compatible with major bacterial infections or viral diseases. Patients with minor localized illness, e.g. sore throat, upper respiratory tract infection, mild gastrointestinal syndromes, frequency/dysuria syndrome or superficial cutaneous infection were deliberately omitted because the Scand J Infect Dis 9
Scandinavian Journal of Infectious Diseases 1977.9:269-276.
274
K . B . Hellurn
etiology will often not be established and neither would a rapid differential laboratory test be of major importance in these instances. However, even with these reservations, more than 1/3 of the patients with suspected infectious diseases had to be excluded, partly because of lack of etiologic diagnosis and partly because the febrile disorders were eventually diagnosed as non-infectious. The patients and controls differed substantially as to sex and age distribution. In the control group, a female preponderance was marked whereas the patient groups included more males than females. Furthermore, the patients with bacterial infections were significantly older than the other patients or controls. However, these incompatibilities would seem unimportant since neither sex nor age (excluding newborns) previously have been shown to influence the NBT test (9, 10, 18, 24). The overall results in the patients with nontuberculous bacterial infections, viral infections and healthy controls differed significantly, mean scores being 29.8, 9.7 and 5.3%, respectively. More important, however, was the demonstration of strikingly different proportions of positive tests in untreated bacterial infections (77.8 %) and viral infections (28.1 %), a finding which is also supported by the results of previous studies (3, 10, 12, 18, 22). Furthermore, acute hepatitis and infectious mononucleosis accounted for 29 of the 38 positive tests in the 135 patients with viral infections. In the remaining 80 patients with other viral infections, only 9 had positive tests. Accordingly, the specificity of the test was markedly influenced by the substantial number of patients with hepatic involvement (which is also a characteristic feature of infectious mononucleosis), and special attention should be paid to patients with hepatitis when evaluating NBT test results in patients with infections. The rapid differentiation of bacterial and viral meningitis is of vital importance and the NBT test in this setting is, therefore, of interest. The results in untreated bacterial and viral meningitis differed significantly as positive NBT tests were demonstrated in 70.6% and 11.8%, respectively. However, the 5 false negative tests obtained in the 17 patients with bacterial meningitis is not compatible with appropriate diagnostics and calls for more accurate laboratory examinations. These discouraging results in bacterial meningitis are in accordance with previous studies (3, 8, 12), and might reflect Scund J Infect Dis 9
the localization of the disease process. The importance of systemic infection for the test to become positive has been stressed (24), and the inflammatory process confined to the cerebrospinal tract might in some patients not impinge on a sufficient number of circulating neutrophils. In patients with septicemia, febrile urinary or lower respiratory tract infection, the proportion of positive tests were closely similar, only about 1/5 of the untreated patients having a false negative test. Bacteremia occurs in at least 50% of febrile upper urinary tract infections (5) and also often in bacterial pneumonia, notably pneumococcal pneumonia as was demonstrated in the present patients. The widespread and often blood-borne infection might bear a close relation to the markedly higher percentage of positive tests demonstrated in these conditions compared to the results in the more localized infections grouped as “various acute infections” (Fig. 1). The elevated NJ3T response in pneumonia caused by M. pneumoniae did not differ from that of untreated bacterial pneumonia, and should not be grouped as “false positive” tests since the clinical picture and the pathologic process in mycoplasma infection often mimics bacterial pneumonia and is also amenable to antibiotic treatment. It should also be remembered that pulmonary embolism which may be confused with lower respiratory tract infection, only exceptionally generates a positive NBT test (16, 25). Thus, in patients with pneumonia-like syndromes, positive NBT tests suggest a bacterial or mycoplasma etiology. Appropriate antibacterial therapy significantly influenced the NBT dye reduction in patients with bacterial infections. The reason for this may be eradication of a substantial number of causative microorganisms. However, previous reports (10, 18) and personal experience have shown that the percentage of NBT-positive neutrophils may drop from significantly elevated to normal levels within less than 1 day’s treatment which in most cases would be a short time for the eradication of the infection. A direct effect of the antibiotics on neutrophil NBT dye reduction has been proposed (10, 26), but therapeutic levels do not seem to inhibit this reduction in vitro (26). Furthermore, inhibition of endotoxin stimulated tests in patients treated with antibiotics could not be demonstrated in the present study. Whatever the reasons, recent antibiotic treatment will seriously invalidate the test
Scandinavian Journal of Infectious Diseases 1977.9:269-276.
Nitroblue tetrazolium test in infections
results as demonstrated by the 45.5 % false negative tests in pretreated patients with major bacterial infections, and should be taken into consideration when evaluating the test results. Infections caused by gram-negative species caused positive NBT tests more often than did those with a gram-positive etiology. Although not statistically significant, this difference was demonstrated in septicemia, lower respiratory tract infection, urinary tract infection and in bacterial meningitis. The reasons for this discrepancy are speculative, but might be associated with endotoxemia in gramnegative infections (6). In vitro, the concentration of enterobacterial endotoxin, which is also used as a standardized challenging factor in the stimulated NBT test, correlates with the amount of neutrophil NBT reduction (21). However, a number of other factors may also cause an increased dye reduction by the neutrophils, viz. bacteria and particulate and soluble products, immune complexes, serum complement and acute phase proteins (1, 9, 28). For practical purposes, it should be remembered that acute hepatitis and infectious mononucleosis are often associated with false positive tests. In patients receiving antimicrobial treatment, the test results may also be misleading. Obviously, the test can be no substitute for carefully performed cultures of body fluids and tissues and serological confirmation. In conclusion, however, the present results suggest a justification for the introduction of a standardized histochemical NBT test as an adjunct to clinical and other laboratory parameters in the differential diagnosis of major bacterial and viral infections. ACKNOWLEDGEMENT The study was supported by grants from The Norwegian Research.Council for Science and Humanities.
REFERENCES I . Baehner, R. L.: Use of the nitroblue tetrazolium test in clinical pediatrics. Am J Dis Child 128:449, 1974. 2. - Subcellular distribution of nitroblue tetrazolium reductase (NBT-R) in human polymorphonuclear leukocytes (PMN). J Lab Clin Med 86: 785, 1975. 3. Bittner, S. J., Kieff, E., Windhorst, D. & Meier, P.: The use of the unstimulated nitroblue tetrazolium test as a routine screening test for bacterial infection in an adult population: a reassessment. Am J Clin Pathol 60:843, 1973. 4. Bjorksten, B.: The nitroblue tetrazolium (NBT) test. A methodological and clinical study. Ume% University Medical Dissertations, No. 15, 1974.
275
5 . Brumfitt, W.: Urinary infection. Lettsomian Lectures, February 11 and 25, 1974. Trans Med SOC
Lond 90: 105, 1974. 6. Caridis, D. T., Reinhold, R. B., Woodruff, P. W. H. & Fine, J.: Endotoxemia in man. Lancet 1: 1381, 1972. 7. Editorial: Another look at the N.B.T. test. Lancet 1:664, 1974. 8. Esposito, R. & DeLalla, F.: N.B.T. test in bacterial meningitis. Lancet 1: 747, 1972. 9. Feigin, R. D.: NBT test in the diagnosis of febrile patients. N Engl J Med 285: 347, 1971. 10. Feigin, R. D., Shackelford, P. G., Choi, S. C., Flake, K. K., Franklin, F. A. & Eisenberg, C. S.: Nitroblue tetrazolium dye test as an aid in the differential diagnosis of febrile disorders. J Pediatr 78: 230, 1971. 1. Freeman, R. & King, B.: Nitroblue-tetrazolium tests. Lancet 1: 104, 1975. 2. Gordon, A. M., Rowan, R. M., Brown, T. & Carson, H. G.: Routine application of the nitroblue tetrazolium test in the clinical laboratory. J Clin Pathol26: 52, 1973. 3. Gordon, P. A., Stuart, J., Lee, T. R., Breeze, G. R. & Pugh, R. N. H.: The cytocentrifuge NBT test. J Clin Pathol28: 674, 1975. 4. Hellum, K. B. & Solberg, C. 0.: Influence of anticoagulants on the nitroblue tetrazolium test. Scand J Infect Dis 5 : 67, 1973. 5 . Hellum, K. B.: Standardization of the nitroblue tetrazolium test. Influence of pH, dye concentration and sample storage. Scand J Infect Dis 9: 125, 1977. 6. - Nitroblue tetrazolium test in pulmonary thromboembolism and pneumonia. Scand J Infect Dis 9: 131, 1977. 17. Hellum, K. B. & Solberg, C. 0.:Granulocyte function in bacterial infections in man. Acta Pathol Microbiol Scand (B) 85: 1, 1977. 18. Matula, G. & Paterson, P. Y.: Spontaneous in vitro reduction of nitroblue tetrazolium by neutrophils of adult patients with bacterial infection. N Engl J Med 285:311, 1971. 19. McCall, C. E., DeChatelet, L. R., Butler, R. & Brown, D.: Enhanced phagocytic capacity, the biologic basis for the elevated histochemical nitroblue tetrazolium reaction. J Clin Invest 54: 1227, 1974. 20. Nathan, D. G.: NBT reduction by human phagocytes. N Engl J Med 290: 280, 1974. 21. Nydegger, U. E., Miescher, A., Anner, R. M., Creighton, D. W., Lambert, P. H. & Miescher, P. A.: Serum and cellular factor involvement in nitroblue tetrazolium (NBT) reduction by human neutrophils. Klin Wochenschr 54: 377, 1973. 22. Park, B. H., Fikrig, S. M. & Smithwick, E. M.: Infection and nitroblue tetrazolium reduction by neutrophils. A diagnostic aid. Lancet 2: 532, 1968. 23. Park, B. H. & Good, R. A.: N.B.T. test stimulated. Lancet 2: 616, 1970. 24. Park, B. H.: The use and limitations of the nitroblue tetrazolium test as a diagnostic aid. J Pediatr 78: 376, 1971. 25. Rowan, R. M., Gordon, A. M., Chaudhuri, A. K. R. & Moran, F.: A further application of the nitroblue tetrazolium test. Br Med J 3: 317. 1974. Scand J Infect Dis 9
Scandinavian Journal of Infectious Diseases 1977.9:269-276.
276
K . B . Hellurn
26. Rubinstein, A. & Pelet, B.: False-negative N.B.T. tests due to transient malfunction of neutrophils. Lancet 1: 382, 1973. 27. Segal, A. W., Trustey, S. F. & Levi, A . J.: Reevaluation of nitroblue-tetrazolium test. Lancet 2:879, 1973. 28. Segal, A . W.: Nitroblue-tetrazolium tests. Lancet 2: 1248, 1974. 29. Steigbigel, R. T., Johnson, P. K. & Remington, J. S.: The nitroblue tetrazolium reduction test versus conventional hematology in the diagnosis of bacterial infection. N Engl J Med 290: 235, 1974. 30. Stossel, T. P.: Evaluation of opsonic and leukocyte function with a spectrophotometric test in patients with infection and with phagocyte disorders. Blood 42: 121, 1973. 31. Wintrobe, M. M.: Clinical hematology, pp. 1266-1312. Lea & Febiger, Philadelphia 1974.
Address for reprints: K . B. Hellum. M.D., Medisinsk avdeling B , Haukeland Sykehus, N-5000 Bergen, Norway
Scand J Infect Dis 9
ISSN: 0036-5548 (Print) 1651-1980 (Online) Journal homepage: http://www.tandfonline.com/loi/infd19
Nitroblue Tetrazolium Test in Bacterial and Viral Infections Kjell B. Hellum To cite this article: Kjell B. Hellum (1977) Nitroblue Tetrazolium Test in Bacterial and Viral Infections, Scandinavian Journal of Infectious Diseases, 9:4, 269-276, DOI: 10.3109/ inf.1977.9.issue-4.03 To link to this article: http://dx.doi.org/10.3109/inf.1977.9.issue-4.03
Published online: 02 Jan 2015.
Submit your article to this journal
View related articles
Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=infd19 Download by: [University of Saskatchewan Library]
Date: 31 March 2016, At: 00:49
Scand J Infect Dis 9: 269-276, 1977
Nitroblue Tetrazolium Test in Bacterial and Viral Infections KJELL B. HELLUM
Scandinavian Journal of Infectious Diseases 1977.9:269-276.
From the School of Medicine, Medical Department B, University of Bergen, Bergen, Norway
ABSTRACT. The reduction of nitroblue tetrazolium (NBT) dye by neutrophils from 379 patients with infectious diseases and 268 controls has been examined. The mean NBT score was 29.8 % (72.3 % positive tests) in the 231 patients with non-tuberculous bacterial infections, 9.7 % (28.1 % positive tests) in the 135 patients with viral infections and 5.3% (1.5% positive tests) in the controls. Positive tests were demonstrated in 1 of 7 patients with tuberculosis and in 4 of 6 with mycoplasma pneumonia. Patients with urinary tract infections or septicemia had the highest percentage of positive tests, particularly when the infections were caused by gram-negative bacteria. In acute bacterial infection, the 176 patients who had not received any antibacterial therapy prior to testing had a significantly higher mean NBT score and proportion (77.8%) of Recent antibiotic treatment positive tests than the remaining 55 pretreated patients (54.5 70). seriously invalidates the NBT test results. In acute viral infection, 29 of the 38 positive tests were obtained from patients with acute hepatitis (mean score 20.0 %) or infectious mononucleosis (mean score 9.3 %). When evaluating the test results, special attention should be paid to patients with hepatitis. Endotoxin stimulated NBT tests disclosed normal enhancement of NBT reduction by neutrophils from the patients and the controls. Cautiously interpreted, the NBT test results may be useful as an adjunct in the differential diagnosis of major bacterial and viral infections.
INTRODUCTION Common leucocyte characteristics used in the differential diagnosis of infectious diseases include total and differential peripheral white blood cell counts, presence of toxic granulation, Dohle bodies, vacuolisation of the cytoplasm and alkaline phosphatase level (3 1). Unfortunately, these parameters often fail to prove the presence or absence of bacterial infection. The application of the histochemical nitroblue tetrazolium (NBT) test as an adjunct in the diagnosis of acute bacterial infection was reported in 1968 (22), when an increased spontaneous in vitro reduction of the NBT dye to formazan by human peripheral neutrophils was demonstrated in bacterial but not in viral infections or non-bacterial febrile disorders. The exact mechanisms involved with the internalisation and reduction of the dye are not entirely clear (2, 20, 28), but depend on phagocytic and oxidative potentials of the cells, thus making the reduced dye an indirect marker of phagocytic activity which seems to be stimulated during bacterial infections (19, 30). The diagnostic relevance of
the NBT test has, however, been questioned (4, 7, 20, 28, 29), and controversy exists as to its application in the evaluation of pyrexial syndromes of unknown etiology ( 1 , 3 , 10, 11, 18,27). Moreover, due to major influence of technical variables on the test results, carefully standardized and controlled test procedures seem necessary (1,4, 13, 14, 15). The aim of the present study was to evaluate a standardized histochemical NBT test in the differential diagnosis of major bacterial and viral infections. MATERIALS AND METHODS Patients and controls Originally, the study included 610 patients admitted to Medical Department B, Haukeland Hospital for febrile disorders consistent with various infections and 380 healthy members of the hospital staff, medical students and student nurses of the department who served as controls. The NBT test was performed within 1-48 h after patient admission, at a time when an etiologic microbiological diagnosis had been made in only 84 patients. In the remaining 526 patients, an etiologic infectious disease Scand J Infect Dis 9
270
K . B . Hellurn
Table I. Patient subjected to the NBT test No. of patients"
Scandinavian Journal of Infectious Diseases 1977.9:269-276.
Diagnosis Bacterial infections Lower respiratory tract infection Urinary tract infection Septicemia and bacterial endocarditis Acute bacterial meningitis Deep abscesses, empyema or peritonitis Acute arthritis or osteomyelitis Tuberculous infection Erysipelas Acute sinusitis Mycoplasma infections Viral infections Acute meningoencephalitis Acute hepatitis Infectious mononucleosis Exanthematous infections Measles Varicella Herpes zoster Herpes simplex Rubella 2 1 Vaccinia Mumps, uncomplicated Respiratory or gastrointestinal tract infections Influenza A Adenovirus 4 RS virus 1 Coxsackievirus B4
ij
r)
3 (ij 6 51 28 27
19
5
5
The parentheses indicate the number of patients with bacterial infections who had received systemic antibiotics within 48 h prior to the NBT test.
diagnosis could be established in only 295 patients, and the remaining 231 patients together with 112 controls had to be excluded from further evaluation. Accordingly, the final study included 379 patients and 268 controls. 238 patients (13 1 males and 107 females aged 9-87 years, mean 56 years) had bacterial infections, 6 (3 males and 3 females aged 18-43 years, mean 33 years) had Mycoplasma pneumoniae infections and 135 patients (70 males and 65 females aged 4-82 years, mean 28 years) had acute viral infections (Table I). In the control group, 79 were males and 189 females, and the age ranged from 16 to 68 years, mean age being 29 years. Criteria of infection The diagnosis of lower respiratory tract infection was based on typical clinical and X-ray findings together with positive cultures of respiratory pathogens from transtracheal aspirates. 10 I patients 1 also1 had /positive blood cultures. Two patients had pure cultures of M. pneumoniae from the transtracheal aspirates, and 4 more patients had a significant rise in antibody titer to this agent, using the complement fixation (CF) test. Of the 94 patients with lower respiratory tract infections (Table I), Scand J Infect Dis 9
88 caused by bacteria and 6 by M. pneumoniae, 19 patients died. Autopsy performed in 15 cases confirmed the diagnosis. Acute urinary tract infection was diagnosed by positive bacterial cultures from suprapubic bladder aspirates together with clinical, radiographic and/or other laboratory findings. The majority of the 48 patients in this group had severe symptoms, and 8 patients died. Autopsy confirmed the diagnosis in each of these cases. A diagnosis of septicemia was made only when blood cultures grew out pathogens in the presence of systemic severe illness with spiking fever and/or shock symptoms. Altogether 41 patients had septicemia. Two more patients who had several positive blood cultures, but also clinical and laboratory evidence of valvular heart disease, were considered to have bacterial endocarditis. Eight patients died, including 1 patient with bacterial endocarditis, and autopsy substantiated the diagnosis in each case. Acute bacterial meningitis was diagnosed by positive bacteriological cultures from the cerebrospinal fluid (CSF) in patients with characteristic clinical and laboratory findings. In patients with deep abscesses, empyema, peritonitis, acute sinusitis and septic arthritis or osteomyelitis, the diagnoses were based upon positive bacteriological cultures of purulent specimens obtained aseptically by needle aspiration at surgery or at autopsy (4 cases). In 7 patients, tuberculous infection was demonstrated by positive cultures of Mycobacterium tuberculosis. Five patients had pulmonary involvement, 1 a tuberculous meningitis and the remaining patient had a tuberculous lymphadenitis. Erysipelas was diagnosed from the clinical picture together with a significant rise in antistreptolysin-0 titer and/or positive cultures of P-hemolytic streptococci from the site of infection. Viral meningoencephalitis was diagnosed from the clinical picture and the results of CSF examination (30-2 OOO predominantly mononuclear leukocyteslpl, protein increase up to 100 mg/dl, glucose ratio CSF:blood>O.S, absence of bacteria in stained smears and routine bacteriologic cultures). In 20 patients, the diagnosis was confirmed serologically by a 4-fold or greater rise in antibody titer to a specific virus: in 10 patients mumps virus, in 6 echovirus, in 2 coxsackievirus B and finally, coxsackievirus A and adenovirus in 1 patient each. Virus isolation from the CSF and/or faeces succeeded in 11 of these cases. 13 other patients had mumps diagnosed on clinical grounds alone. In the remaining 18 patients the evidence for a specific viral etiology was inconclusive. However, serologic tests for M. pneumoniae and Toxoplasma gondii also turned out negative, as did the search for extra- or intracranial processes capable of mimicking the CSF changes of viral meningitis. These patients all made a complete recovery without any administration of antibiotics prior to admission or in hospital. A diagnosis of acute viral hepatitis was based upon epidemiological, clinical and laboratory examination, including liver biopsies. HBsAg was demonstrated in 8 patients. Infectious mononucleosis was diagnosed on the basis of clinical symptoms, results of peripheral blood smear ex-
Nitroblue tetrazolium test in infections
27 1
Table 11. Unstimulated and endotoxin stimulated NBT tests in 379 patients with bacterial, mycoplasma or viral infections and in 268 normal controls Unstimulated test neutrophils
Positive testn Negative test
Endotoxin stimulated test % NBT-positive neutrophils
Meank1S.D. Range
No.
%
No.
%
Meanf1S.D. Range
% NBT-positive
Scandinavian Journal of Infectious Diseases 1977.9:269-276.
Diagnoses Normal controls Acute bacterial infections Untreated Pretreated Tuberculosis Pneumonia caused by M. pneumoniae Viral infections
No. of subjects 268
5.3k 3.1
0-18
4
1.5
264
98.5
46.8k13.2
22-79
176 55 7
32.6k21.6 21.1514.8 10.65 10.1
1-93 0-68 1-3 1
137 30 1
77.8 54.5
39 25 6
22.2 45.5
69.9k18.6 61.7k17.0 53.5k18.1
18-98 20-89 23-76
6 135
29.8f24.1 9.75 9.6
3-61 0-51
4 38
60.2k25.4 51.9k16.0
28-88 21-94
28.1
2 97
71.9
Positive test defined as >15% NBT-positive neutrophils (see text). amination and a significant rise in heterophile antibody titer. Patients with exanthematous viral infections or mumps all had readily recognizable diseases. In 5 patients with predominant respiratory or gastrointestinal symptoms, a diagnosis of viral infection was based on a significant rise in virus antibody titer together with virus isolation. 176 patients with bacterial non-tuberculous infections had not received any antibiotic therapy prior to the NBT test, whereas the remaining 55 patients in this group had received systemic antibiotics within the last 48 h (Table I). Histochemical NBT test
The test was performed by a standardized technique using heparinized whole blood, slightly modified from the test of Park et al. (22) as previously described (14, 15). Unstimulated and endotoxin stimulated tests were performed contemporaneously on the same blood sample (14,23). In each test, 300 neutrophils in each of 2 slide smears were classified as NBT-negative (without bluish-black formazan) or NBT-positive (containing deposits of formazan). The percentage of NBT-positive neutrophils was recorded as the NBT score and without knowledge of patient diagnosis. A control test was always performed in parallel with one'or more patient tests.
RESULTS Healthy individuals In the 268 normal controls, the percentage of NBT-positive neutrophils in the unstimulated test ranged from 0-18, mean 5.3 (Table 11). Scores above 15% were regarded as positive NBT tests. This upper level is well beyond the mean plus 2 standard deviations for the present controls and in accordance with previous results using an identical test procedure (16, 17). Thus, only 4 normal indi-
viduals (1.5 %) had a positive test. Endotoxin stimulated tests demonstrated a 3-fold or greater increase in the number of NBT-positive neutrophils compared to unstimulated test scores. However, the results ranged widely (22-79 %), mean score being 46.8 %.
Bacterial and mycoplasma infections In the 23 1 patients with non-tuberculous bacterial infections, the mean NBT score was 29.8% (range 0-93%) and 167 patients (72.3%) had a positive test. However, in the group of 176 patients who received no systemic antibacterial treatment prior to testing, the proportion of positive tests was significantly higher (77.8 %) than in the 55 pretreated patients (54.5%) and the mean scores in these groups were also significantly different (t-test,PO. l), indicating no functional impairment of neutrophil reducing capacity during systemic antimicrobial therapy. The highest number of positive tests in the nontreated patients was demonstrated in those with acute urinary tract infections (83.3 % positive tests, mean score 39.3%), closely followed by the septicemia group (80.0 % positive tests, mean score 38.1 %) and the group with lower respiratory tract infections (77.9 % positive tests, mean score 29.4%) (Fig. 1). The mean scores in patients with Scand J Infect Dis 9
272
K . B . Hellurn
D O
90
0
80
0,
70
a I
.
0
I
1
tW 3
60 W
2
8t
50
‘
40
z
30
a
Scandinavian Journal of Infectious Diseases 1977.9:269-276.
2 5
20
kw2 a
10
: 0
mer resFxrabi raci infections
Urinary tract infections
Septicemia and ndocarditis
I :
1
Menlngltls
Unstimulated NET test 0
Endotwin stimulated NET test
- mean xom
- - - -upper limit for r q a t i w NBT tesls
meningitis or other types of major non-tuberculous bacterial infections were markedly lower than in the former groups, the percentage of positive tests being only 70.6 and 70.0, respectively. In the 7 patients with active tuberculous infections, only 1 patient with a severe lymphadenitis had a positive test. Endotoxin stimulated tests in each of these patients demonstrated a normal enhancement of NBT dye reduction. In contrast, 4 of the 6 patients with mycoplasma infections had markedly elevated NBT scores. Thus, the results in this patient group were similar to those obtained in the patients with bacterial pneumonia. In 168 patients with untreated bacterial infec-
(see text
)
Fig. 1 . Unstimulated and endotoxin stimulated NBT tests in 176 patients with untreated acute bacterial infections.
tions, tuberculosis excluded, the causative organisms were recovered as monocultures (Table 111). 94 patients (56%) had infections with grampositive bacteria and 69 (73 %) had positive tests. In the 74 patients with infections due to gram-negative organisms, the percentage of positive tests was 81. The differences between the numbers of positive tests in infections caused by gram-positive versus gram-negative species were more prominent in the patients with septicemia or urinary tract infections than in the other disease groups. However, these differences did not reach significant levels (chi square test, P>O. I). Severe underlying disease (viz. malignancies, un-
Table 111. NBT tests according to gram stain properties of the etiologic agents recovered as monocultures from 168 patients with bacterial infections Percentage of patients with positive tests” (no. of patients tested)
Causative microorganisms
Lower respiratory tract infections
Urinary tract infections
Septicemia and endocarditis
Meningitis
Other acute infections
Total
Gram-positive species Gram-negative species
78 (SO) 81 (16)
67 (6) 89 (26)
67 (18) 88 (16)
60 ( 5 ) 67 (12)
73 (15) 50 (4)
73 (94) 81 (74)
Positive test defined as >15% NBT-positive neutrophils (see text). Scand J Infect Dis 9
Nitroblue tetrazolium test in infections
273
PP 0
-8-
.
;I*
.
i
O
9-
8 0
Scandinavian Journal of Infectious Diseases 1977.9:269-276.
D
- - - --- - -
I 0
MenlKj.3encephalitis
nfectious noncmucleosis
Acute hepatits
Various infections
Unstimulated NBT test Endotoxin stimulated NBT test
- mean
score ----upper
iimt tor negatw NBT tests (see text)
controlled diabetes mellitus, heart failure, uremia, collagen diseases, liver cirrhosis, bowel infarction and gangrene) was present in 37 patients without antibiotic pretreatment and 14 patients on antibiotics. Furthermore, 12 patients were on corticosteroid therapy (prednisone 5-40 mg per day), 8 in the non-treatment group and 4 in the treatment group. Two more patients in the latter group received treatment with azathioprine (100 mg per day). Exclusion of these patient categories did neither significantly alter the proportion of positive tests, nor the mean scores in the various disease groups. Finally, these patients's neutrophils had a normal response to endotoxin stimulation (Fig. 1). Viral infections In the 135 patients with viral infections, the mean NBT score was 9.7% in the unstimulated test and only 38 patients (28.1 %) had positive tests (Table 11). 22 of the 28 patients (78.6%) with acute viral
hepatitis A or B had positive tests during the early icteric stage, mean score being 20.0% NBTpositive neutrophils (Fig. 2). The 27 patients with infectious mononucleosis also had a relatively high number (25.9%) of positive tests (mean score 9.3 %). However, neither clinical picture nor laboratory test results in the patients with hepatitis 18-77 1959
Fig. 2. Unstimulated and endotoxin stimulated NBT tests in 135 patients with acute viral infections.
or infectious mononucleosis disclosed any significant difference between the severity of the disease process in patients with positive and negative tests. In the 51 patients with viral meningoencephalitis, the mean score was 6.4 % and only 6 (1 1.8 %) had positive tests. Similar results were obtained in the patients with exanthematic viral syndromes (mean score 6.1 %, positive tests in 2 patients with measles) and in the remaining 10 patients with other acute viral infections (mean score 5.8%, positive test in 1 patient with uncomplicated mumps). Endotoxin stimulated tests in patients with acute viral infections disclosed normal or supranormal NBT dye reducing capacity. DISCUSSION The present study was undertaken to evaluate prospectively the histochemical NBT test as an indicator of bacterial versus viral infections in patients admitted to an infectious disease department with syndromes compatible with major bacterial infections or viral diseases. Patients with minor localized illness, e.g. sore throat, upper respiratory tract infection, mild gastrointestinal syndromes, frequency/dysuria syndrome or superficial cutaneous infection were deliberately omitted because the Scand J Infect Dis 9
Scandinavian Journal of Infectious Diseases 1977.9:269-276.
274
K . B . Hellurn
etiology will often not be established and neither would a rapid differential laboratory test be of major importance in these instances. However, even with these reservations, more than 1/3 of the patients with suspected infectious diseases had to be excluded, partly because of lack of etiologic diagnosis and partly because the febrile disorders were eventually diagnosed as non-infectious. The patients and controls differed substantially as to sex and age distribution. In the control group, a female preponderance was marked whereas the patient groups included more males than females. Furthermore, the patients with bacterial infections were significantly older than the other patients or controls. However, these incompatibilities would seem unimportant since neither sex nor age (excluding newborns) previously have been shown to influence the NBT test (9, 10, 18, 24). The overall results in the patients with nontuberculous bacterial infections, viral infections and healthy controls differed significantly, mean scores being 29.8, 9.7 and 5.3%, respectively. More important, however, was the demonstration of strikingly different proportions of positive tests in untreated bacterial infections (77.8 %) and viral infections (28.1 %), a finding which is also supported by the results of previous studies (3, 10, 12, 18, 22). Furthermore, acute hepatitis and infectious mononucleosis accounted for 29 of the 38 positive tests in the 135 patients with viral infections. In the remaining 80 patients with other viral infections, only 9 had positive tests. Accordingly, the specificity of the test was markedly influenced by the substantial number of patients with hepatic involvement (which is also a characteristic feature of infectious mononucleosis), and special attention should be paid to patients with hepatitis when evaluating NBT test results in patients with infections. The rapid differentiation of bacterial and viral meningitis is of vital importance and the NBT test in this setting is, therefore, of interest. The results in untreated bacterial and viral meningitis differed significantly as positive NBT tests were demonstrated in 70.6% and 11.8%, respectively. However, the 5 false negative tests obtained in the 17 patients with bacterial meningitis is not compatible with appropriate diagnostics and calls for more accurate laboratory examinations. These discouraging results in bacterial meningitis are in accordance with previous studies (3, 8, 12), and might reflect Scund J Infect Dis 9
the localization of the disease process. The importance of systemic infection for the test to become positive has been stressed (24), and the inflammatory process confined to the cerebrospinal tract might in some patients not impinge on a sufficient number of circulating neutrophils. In patients with septicemia, febrile urinary or lower respiratory tract infection, the proportion of positive tests were closely similar, only about 1/5 of the untreated patients having a false negative test. Bacteremia occurs in at least 50% of febrile upper urinary tract infections (5) and also often in bacterial pneumonia, notably pneumococcal pneumonia as was demonstrated in the present patients. The widespread and often blood-borne infection might bear a close relation to the markedly higher percentage of positive tests demonstrated in these conditions compared to the results in the more localized infections grouped as “various acute infections” (Fig. 1). The elevated NJ3T response in pneumonia caused by M. pneumoniae did not differ from that of untreated bacterial pneumonia, and should not be grouped as “false positive” tests since the clinical picture and the pathologic process in mycoplasma infection often mimics bacterial pneumonia and is also amenable to antibiotic treatment. It should also be remembered that pulmonary embolism which may be confused with lower respiratory tract infection, only exceptionally generates a positive NBT test (16, 25). Thus, in patients with pneumonia-like syndromes, positive NBT tests suggest a bacterial or mycoplasma etiology. Appropriate antibacterial therapy significantly influenced the NBT dye reduction in patients with bacterial infections. The reason for this may be eradication of a substantial number of causative microorganisms. However, previous reports (10, 18) and personal experience have shown that the percentage of NBT-positive neutrophils may drop from significantly elevated to normal levels within less than 1 day’s treatment which in most cases would be a short time for the eradication of the infection. A direct effect of the antibiotics on neutrophil NBT dye reduction has been proposed (10, 26), but therapeutic levels do not seem to inhibit this reduction in vitro (26). Furthermore, inhibition of endotoxin stimulated tests in patients treated with antibiotics could not be demonstrated in the present study. Whatever the reasons, recent antibiotic treatment will seriously invalidate the test
Scandinavian Journal of Infectious Diseases 1977.9:269-276.
Nitroblue tetrazolium test in infections
results as demonstrated by the 45.5 % false negative tests in pretreated patients with major bacterial infections, and should be taken into consideration when evaluating the test results. Infections caused by gram-negative species caused positive NBT tests more often than did those with a gram-positive etiology. Although not statistically significant, this difference was demonstrated in septicemia, lower respiratory tract infection, urinary tract infection and in bacterial meningitis. The reasons for this discrepancy are speculative, but might be associated with endotoxemia in gramnegative infections (6). In vitro, the concentration of enterobacterial endotoxin, which is also used as a standardized challenging factor in the stimulated NBT test, correlates with the amount of neutrophil NBT reduction (21). However, a number of other factors may also cause an increased dye reduction by the neutrophils, viz. bacteria and particulate and soluble products, immune complexes, serum complement and acute phase proteins (1, 9, 28). For practical purposes, it should be remembered that acute hepatitis and infectious mononucleosis are often associated with false positive tests. In patients receiving antimicrobial treatment, the test results may also be misleading. Obviously, the test can be no substitute for carefully performed cultures of body fluids and tissues and serological confirmation. In conclusion, however, the present results suggest a justification for the introduction of a standardized histochemical NBT test as an adjunct to clinical and other laboratory parameters in the differential diagnosis of major bacterial and viral infections. ACKNOWLEDGEMENT The study was supported by grants from The Norwegian Research.Council for Science and Humanities.
REFERENCES I . Baehner, R. L.: Use of the nitroblue tetrazolium test in clinical pediatrics. Am J Dis Child 128:449, 1974. 2. - Subcellular distribution of nitroblue tetrazolium reductase (NBT-R) in human polymorphonuclear leukocytes (PMN). J Lab Clin Med 86: 785, 1975. 3. Bittner, S. J., Kieff, E., Windhorst, D. & Meier, P.: The use of the unstimulated nitroblue tetrazolium test as a routine screening test for bacterial infection in an adult population: a reassessment. Am J Clin Pathol 60:843, 1973. 4. Bjorksten, B.: The nitroblue tetrazolium (NBT) test. A methodological and clinical study. Ume% University Medical Dissertations, No. 15, 1974.
275
5 . Brumfitt, W.: Urinary infection. Lettsomian Lectures, February 11 and 25, 1974. Trans Med SOC
Lond 90: 105, 1974. 6. Caridis, D. T., Reinhold, R. B., Woodruff, P. W. H. & Fine, J.: Endotoxemia in man. Lancet 1: 1381, 1972. 7. Editorial: Another look at the N.B.T. test. Lancet 1:664, 1974. 8. Esposito, R. & DeLalla, F.: N.B.T. test in bacterial meningitis. Lancet 1: 747, 1972. 9. Feigin, R. D.: NBT test in the diagnosis of febrile patients. N Engl J Med 285: 347, 1971. 10. Feigin, R. D., Shackelford, P. G., Choi, S. C., Flake, K. K., Franklin, F. A. & Eisenberg, C. S.: Nitroblue tetrazolium dye test as an aid in the differential diagnosis of febrile disorders. J Pediatr 78: 230, 1971. 1. Freeman, R. & King, B.: Nitroblue-tetrazolium tests. Lancet 1: 104, 1975. 2. Gordon, A. M., Rowan, R. M., Brown, T. & Carson, H. G.: Routine application of the nitroblue tetrazolium test in the clinical laboratory. J Clin Pathol26: 52, 1973. 3. Gordon, P. A., Stuart, J., Lee, T. R., Breeze, G. R. & Pugh, R. N. H.: The cytocentrifuge NBT test. J Clin Pathol28: 674, 1975. 4. Hellum, K. B. & Solberg, C. 0.: Influence of anticoagulants on the nitroblue tetrazolium test. Scand J Infect Dis 5 : 67, 1973. 5 . Hellum, K. B.: Standardization of the nitroblue tetrazolium test. Influence of pH, dye concentration and sample storage. Scand J Infect Dis 9: 125, 1977. 6. - Nitroblue tetrazolium test in pulmonary thromboembolism and pneumonia. Scand J Infect Dis 9: 131, 1977. 17. Hellum, K. B. & Solberg, C. 0.:Granulocyte function in bacterial infections in man. Acta Pathol Microbiol Scand (B) 85: 1, 1977. 18. Matula, G. & Paterson, P. Y.: Spontaneous in vitro reduction of nitroblue tetrazolium by neutrophils of adult patients with bacterial infection. N Engl J Med 285:311, 1971. 19. McCall, C. E., DeChatelet, L. R., Butler, R. & Brown, D.: Enhanced phagocytic capacity, the biologic basis for the elevated histochemical nitroblue tetrazolium reaction. J Clin Invest 54: 1227, 1974. 20. Nathan, D. G.: NBT reduction by human phagocytes. N Engl J Med 290: 280, 1974. 21. Nydegger, U. E., Miescher, A., Anner, R. M., Creighton, D. W., Lambert, P. H. & Miescher, P. A.: Serum and cellular factor involvement in nitroblue tetrazolium (NBT) reduction by human neutrophils. Klin Wochenschr 54: 377, 1973. 22. Park, B. H., Fikrig, S. M. & Smithwick, E. M.: Infection and nitroblue tetrazolium reduction by neutrophils. A diagnostic aid. Lancet 2: 532, 1968. 23. Park, B. H. & Good, R. A.: N.B.T. test stimulated. Lancet 2: 616, 1970. 24. Park, B. H.: The use and limitations of the nitroblue tetrazolium test as a diagnostic aid. J Pediatr 78: 376, 1971. 25. Rowan, R. M., Gordon, A. M., Chaudhuri, A. K. R. & Moran, F.: A further application of the nitroblue tetrazolium test. Br Med J 3: 317. 1974. Scand J Infect Dis 9
Scandinavian Journal of Infectious Diseases 1977.9:269-276.
276
K . B . Hellurn
26. Rubinstein, A. & Pelet, B.: False-negative N.B.T. tests due to transient malfunction of neutrophils. Lancet 1: 382, 1973. 27. Segal, A. W., Trustey, S. F. & Levi, A . J.: Reevaluation of nitroblue-tetrazolium test. Lancet 2:879, 1973. 28. Segal, A . W.: Nitroblue-tetrazolium tests. Lancet 2: 1248, 1974. 29. Steigbigel, R. T., Johnson, P. K. & Remington, J. S.: The nitroblue tetrazolium reduction test versus conventional hematology in the diagnosis of bacterial infection. N Engl J Med 290: 235, 1974. 30. Stossel, T. P.: Evaluation of opsonic and leukocyte function with a spectrophotometric test in patients with infection and with phagocyte disorders. Blood 42: 121, 1973. 31. Wintrobe, M. M.: Clinical hematology, pp. 1266-1312. Lea & Febiger, Philadelphia 1974.
Address for reprints: K . B. Hellum. M.D., Medisinsk avdeling B , Haukeland Sykehus, N-5000 Bergen, Norway
Scand J Infect Dis 9
