1277
Diagnosis
and
Prognosis Meningitis
in
Pyogenic
pyogenic meningitis accurate bacteriological diagnosis is essential, but the customary testsgram stain and culture of cerebrospinal fluid (c.s.F.)-have disadvantages. When done by inexperienced staff on emergency call, a gram stain may be misleading. Culture is slow and may be negative in patients who have received antibiotics IN
before
coming to hospital. In view of these shortcomings, interest has revived in diagnostic techniques which depend on the presence in c.s.F. of bacterial products rather than whole, viable organisms. Several facets of this topic are covered in the April issue of the Journal of Pediatrics. 1-4 The two main methods are detection of specific polysaccharide antigens and detection of endotoxin. Observations of free capsular polysaccharide in the c.s.F. of patients with meningococcal meningitis go back to the beginning of the century, and in the 1930s infections were successfully diagnosed by a capillary-tube agglutination test. 5-7 After a brief vogue, interest lapsed and was not reawakened until 1971, when EDWARDS showed that small amounts of meningococcal polysaccharide could be detected in serum and c.s.F. by countercurrent immunoelectrophoresis (c .I.E.).8 The diagnostic value of C.LE. in meningococcal meningitis has since been amply confirmed.9-11 Countercurrent immunoelectrophoresis has also been used successfully to diagnose pneumococcal, 12-14 haemoD., Stechenberg, B. W., Chang, M. J., Dunkle, L. M., Wong, M. L., Palkes, H., Dodge, P. R., Davis, H. J. Pediat. 1976, 88, 542. 2. Feldman, W. E. ibid. p. 549. 3 Berman, N. S., Siegel, S. E., Nachum, R., Lipsey, A., Leedom, J. ibid. p. 553.
14-16
and Escherichia coli meningitis!’ with success-rates of 70% to over 90%. Both c.s.F. and serum should be examined, since the serum is sometimes positive even when c.s.F. is negative. Previous antibiotic treatment has little effect on the success-rate of C.I.E.-an important advantage over culture. Meningococcal and pneumococcal meningitis can be rapidly typed by direct C.I.E. of c.s.F. against specific antisera. Antibody-coated latex particles have also been used to detect polysaccharide antigens in C.S.F.14 18 19 This test is simple, gives an answer within three minutes, and, for meningococcal and haemophilus infections, is more sensitive than c.I.E. Experience with the latex test is less extensive than withC.I.E., but the latex test may prove useful in clinical side-rooms and particularly so in rural hospitals of developing countries where meningitis is rife and where laboratory facilities are scarce. The success of bothC.I.E. and the latex test depends heavily on the quality of the antisera employed. Some commercial antisera, produced for use in agglutination tests, do not produce a precipitation reaction and are thus unsuitable for use in c.I.E. Particular difficulty has been experienced with antisera to the group-B meningococcal polysaccharide antigen, but antisera have now been prepared which give a good precipitation reaction with this antigen and with the very similar antigen of the Kl strain of E. coli, which is responsible for many cases of neonatal meningitis. 17 Gram-negative bacteria multiplying within the C.S.F. can produce enough endotoxin to cause gelation of limulus lysate, so the limulus test can be used to diagnose this form of meningitis. 20 21 A diagnostic success-rate of nearly 100% has been obtained in meningococcal and haemophilus meningitis, whilst tests were negative in meningitis due to gram-positive organisms and in other neurological conditions. The limulus test has been adapted for use as a side-room investigation21 but great care is needed to protect the reagents and containers from endotoxin contamination since this will cause false-positive reactions. The limulus test is a very sensitive indicator of the presence of gram-negative meningitis but it cannot distinguish between different gram-negative organisms. It is thus less helpful clinically than tests which provide an exact bac-
philus,l ’2
teriological diagnosis. The prognosis of pyogenic meningitis varies,
1. Feigin, R.
4. McCracken, G. H. ibid. p. 706. 5. Rake, G. J. exp. Med. 1933, 58, 375. 6. Maegraith, B. G. Lancet, 1935, 1, 545. 7. Alexander, H. E. J. clin. Invest. 1937, 16, 207. 8. Edwards, E. A. J. Immun. 1971, 106, 314. 9. Greenwood, B. M., Whittle, H. C., Dominic-Rajkovic, O. Lancet, 1971, ii, 519. 10. Tobin, B. M., Jones, D. M. J. clin. Path. 1972, 25, 583. 11. Higashi, G. I., Sippel, J. E., Girgis, N. I., Hassan, H. Scand. J. infect. Dis. 1974, 6, 233. 12. Coonrod, J. D., Rytel, M. W. Lancet, 1972, i, 1154.
13. 14.
Fossieck, B., Craig, R., Paterson, P. Y. J. infect. Dis. 1973, 127, 106. Whittle, H. C., Tugwell, P., Egler, L. J., Greenwood, B. M. Lancet, 1974, ii, 619. 15. Ingram, D. L., Anderson, P., Smith, D. H. J. Pediat. 1972, 81, 1156. 16. O’Reilly, R. J., Anderson, P., Ingram, D. L., Peter, G., Smith, D. H. J. clin. Invest. 1975, 56, 1012. 17. McCracken, G. H., Sarff, L. D., Glode, M. P., Mize, S. G., Schiffer, M. S., Robbins, J. B., Gotschlich, E. C., Orskov, I., Orskov, E. Lancet, 1974, ii, 246. 18. Newman, R. B., Stevens, R. W., Gaafar, H. A. J. Lab. clin. Med. 1970, 76, 107. 19. Severin, W. P. J. J. clin. Path. 1972, 25, 1079. 20. Nachum, R., Lipsey, A., Siegel, S. E. New Engl. J. Med. 1973, 289, 931. 21. Ross, S., Rodriguez, W., Controni, G., Korengold, G., Watson, S., Khan, W. J. Am. med. Ass. 1975, 233, 1366.
1278
patients dying within a few hours of admishospital, others recovering rapidly. Certain are of prognostic value-in pneumococcal signs meningitis, for example, coma on presentation and a low c.s.F. white-cell count indicates a gloomy prognosis-but on clinical grounds alone the outcome often cannot be foretold. Bacterial antigen some
sion
to
and endotoxin levels in c.s.F. have therefore been quantitated to see if they could be of help in prognosis. Antigen concentration in c.s.F. is closely related to the c.s.F. bacterial count2 and a relationship between antigen titre and prognosis might therefore be expected. This has proved to be the case, for in meningococcal, 22 21 pneumococcal,2a haemophilus,l 12 16 and E. coli17 meningitis those patients with severe brain damage and late sequelae had a higher concentration of antigen in c.s.F. than patients who recovered uneventfully. However, in most investigations there was considerable overlap between the two groups. Antigen usually disappears from the c.s.F. 24-48 hours from the start of treatment. Longer persistence of antigen is ominous. Persistence of antigen for more than 48 hours after the start of treatment, especially if accompanied by a poor clinical response, may be an indication for a change of therapy. Antigen in the serum, particularly if present in high titre or if present for more than 2 or 3 days, is another bad sign in haemophilus,t6 E. coli,t7 and meningococcal
meningitis.22 Detection of bacterial products in c.s.F. and serum is proving a useful adjunct to the management of patients with pyogenic meningitis. It allows rapid bacteriological diagnosis and, early in the course of their illness, it singles out the patients with a poor prognosis. It should not be regarded as a replacement for routine bacteriological investigations but as a supplement, for the diagnostic success-rate of antigen detection and culture combined is greater than that of either technique alone. In addition, culture may identify one of the more un-usual causes of bacterial meningitis and can provide valuable information on antibiotic sensitivities.
Irradiation of the Thyroid Gland THE people of Hiroshima and Nagasaki were exposed to external gamma and neutron radiation from the nuclear bombs "Fat Man" and "Little Boy", but not to any significant fallout. An accident during a thermonuclear explosion at Bikini in 1954 unfortunately subjected a group of Marshall 22.
Whittle, H. C., Greenwood, B. M., Davidson, N. McD., Tomkins, A., Tugwell, P., Warrell, D. A., Zalin, A., Bryceson, A. D. M., Parry, E. H. O., Brueton, M., Duggan, M., Oomen, J. M. V., Rajkovic, A. D. Am. J. Med. 1975, 58, 823. 23. Hoffman, T. A., Edwards, E. A. J. infect. Dis. 1972, 126, 636. 24. Tugwell, P., Greenwood, B. M., Warrell, D. A. Q. Jl Med. (in the press).
Islanders and Japanese fishermen to acute radiation from fallout-the only population so affected. They have been under surveillance ever since; weI have commented on some of the earlier reports,’ and a twenty-year review has now appeared.2 The fallout was deposited on people’s skin and surroundings and on food and water supplies. This resulted in very high epidermal doses, mostly from nonpenetrating beta rays, and whole-body gamma doses estimated at 175 rads on Rongelap and 14-69 rads in neighbouring islands. Because many radioisotopes of iodine were produced and ingested, thyroid doses were higher, calculated at a total amount of about 220-450 rads to adults and 700-1400 rads to children on Rongelap (because of many uncertainties these are rough estimates). Initially, transient nausea and depression of leucocyte-counts were common, and there were severe skin burns, all of which healed. In the succeeding years a few ill-effects were manifest, such as an increase in miscarriages and stillbirths, and chromosome abnormalities in blood-cells. (There has also been an increase in the incidence of non-thyroid cancer in the heavily exposed group, including a case of acute myeloid leukaemia, but the numbers are small; so far no skin cancers have appeared.) Then growth retardation was noted in 5 of the 19 children exposed before ten years of age. Soon evidence ofmyxoedema became apparent in 2 of them, and when satisfactory methods of measuring thyroxine were devised, low levels were found (the protein-bound iodine levels had, confusingly, been normal, apparently because of an iodoprotein in Marshallese unrelated to thyroid hormone). Thyrotrophin levels were normal. Nine years after exposure, thyroid nodules began to appear (but not in the 2 myxoedematous boys). By 1974, 27 of the 86 exposed people on Rongelap had nodules, with smaller numbers in the less heavily exposed people of the other islands. These nodules were all found clinically; scanning might have revealed larger numbers.3 The highest incidence of nodules has been in the heavily exposed group who were less than ten years old at the time of the accident. 1 of the 4 children exposed in utero developed a thyroid adenoma. 24 of these patients have had thyroid operations, revealing benign lesions in 21 and carcinoma in 3. Many glands had hairlike vessels on the surface, reminiscent of thyroids treated with radioiodine. Many of the adenomas were papillary; some patients were given tracer doses of 1311 before operation, and autoradiography showed diminished or absent function in the nodules. Most of the thyroids contained many minute encapsulated lesions 1. Lancet, 1966, ii, 580; ibid. 1968, i, 625. 2. A Twenty-Year Review of Medical Findings in a Marshallese Population Accidentally Exposed to Radioactive Fallout. By R. A. CONARD and others. Brookhaven National Laboratory, Upton, New York, 1975. 3. Favus, M. J., Schneider, A. B., Stachura, M. E., Arnold, J. E., Yun Ryo, U., Pinsky, S. M., Colman, M., Arnold, M. J., Frohman, L. A. New Engl. J. Med. 1976, 1294, 1019.
Diagnosis
and
Prognosis Meningitis
in
Pyogenic
pyogenic meningitis accurate bacteriological diagnosis is essential, but the customary testsgram stain and culture of cerebrospinal fluid (c.s.F.)-have disadvantages. When done by inexperienced staff on emergency call, a gram stain may be misleading. Culture is slow and may be negative in patients who have received antibiotics IN
before
coming to hospital. In view of these shortcomings, interest has revived in diagnostic techniques which depend on the presence in c.s.F. of bacterial products rather than whole, viable organisms. Several facets of this topic are covered in the April issue of the Journal of Pediatrics. 1-4 The two main methods are detection of specific polysaccharide antigens and detection of endotoxin. Observations of free capsular polysaccharide in the c.s.F. of patients with meningococcal meningitis go back to the beginning of the century, and in the 1930s infections were successfully diagnosed by a capillary-tube agglutination test. 5-7 After a brief vogue, interest lapsed and was not reawakened until 1971, when EDWARDS showed that small amounts of meningococcal polysaccharide could be detected in serum and c.s.F. by countercurrent immunoelectrophoresis (c .I.E.).8 The diagnostic value of C.LE. in meningococcal meningitis has since been amply confirmed.9-11 Countercurrent immunoelectrophoresis has also been used successfully to diagnose pneumococcal, 12-14 haemoD., Stechenberg, B. W., Chang, M. J., Dunkle, L. M., Wong, M. L., Palkes, H., Dodge, P. R., Davis, H. J. Pediat. 1976, 88, 542. 2. Feldman, W. E. ibid. p. 549. 3 Berman, N. S., Siegel, S. E., Nachum, R., Lipsey, A., Leedom, J. ibid. p. 553.
14-16
and Escherichia coli meningitis!’ with success-rates of 70% to over 90%. Both c.s.F. and serum should be examined, since the serum is sometimes positive even when c.s.F. is negative. Previous antibiotic treatment has little effect on the success-rate of C.I.E.-an important advantage over culture. Meningococcal and pneumococcal meningitis can be rapidly typed by direct C.I.E. of c.s.F. against specific antisera. Antibody-coated latex particles have also been used to detect polysaccharide antigens in C.S.F.14 18 19 This test is simple, gives an answer within three minutes, and, for meningococcal and haemophilus infections, is more sensitive than c.I.E. Experience with the latex test is less extensive than withC.I.E., but the latex test may prove useful in clinical side-rooms and particularly so in rural hospitals of developing countries where meningitis is rife and where laboratory facilities are scarce. The success of bothC.I.E. and the latex test depends heavily on the quality of the antisera employed. Some commercial antisera, produced for use in agglutination tests, do not produce a precipitation reaction and are thus unsuitable for use in c.I.E. Particular difficulty has been experienced with antisera to the group-B meningococcal polysaccharide antigen, but antisera have now been prepared which give a good precipitation reaction with this antigen and with the very similar antigen of the Kl strain of E. coli, which is responsible for many cases of neonatal meningitis. 17 Gram-negative bacteria multiplying within the C.S.F. can produce enough endotoxin to cause gelation of limulus lysate, so the limulus test can be used to diagnose this form of meningitis. 20 21 A diagnostic success-rate of nearly 100% has been obtained in meningococcal and haemophilus meningitis, whilst tests were negative in meningitis due to gram-positive organisms and in other neurological conditions. The limulus test has been adapted for use as a side-room investigation21 but great care is needed to protect the reagents and containers from endotoxin contamination since this will cause false-positive reactions. The limulus test is a very sensitive indicator of the presence of gram-negative meningitis but it cannot distinguish between different gram-negative organisms. It is thus less helpful clinically than tests which provide an exact bac-
philus,l ’2
teriological diagnosis. The prognosis of pyogenic meningitis varies,
1. Feigin, R.
4. McCracken, G. H. ibid. p. 706. 5. Rake, G. J. exp. Med. 1933, 58, 375. 6. Maegraith, B. G. Lancet, 1935, 1, 545. 7. Alexander, H. E. J. clin. Invest. 1937, 16, 207. 8. Edwards, E. A. J. Immun. 1971, 106, 314. 9. Greenwood, B. M., Whittle, H. C., Dominic-Rajkovic, O. Lancet, 1971, ii, 519. 10. Tobin, B. M., Jones, D. M. J. clin. Path. 1972, 25, 583. 11. Higashi, G. I., Sippel, J. E., Girgis, N. I., Hassan, H. Scand. J. infect. Dis. 1974, 6, 233. 12. Coonrod, J. D., Rytel, M. W. Lancet, 1972, i, 1154.
13. 14.
Fossieck, B., Craig, R., Paterson, P. Y. J. infect. Dis. 1973, 127, 106. Whittle, H. C., Tugwell, P., Egler, L. J., Greenwood, B. M. Lancet, 1974, ii, 619. 15. Ingram, D. L., Anderson, P., Smith, D. H. J. Pediat. 1972, 81, 1156. 16. O’Reilly, R. J., Anderson, P., Ingram, D. L., Peter, G., Smith, D. H. J. clin. Invest. 1975, 56, 1012. 17. McCracken, G. H., Sarff, L. D., Glode, M. P., Mize, S. G., Schiffer, M. S., Robbins, J. B., Gotschlich, E. C., Orskov, I., Orskov, E. Lancet, 1974, ii, 246. 18. Newman, R. B., Stevens, R. W., Gaafar, H. A. J. Lab. clin. Med. 1970, 76, 107. 19. Severin, W. P. J. J. clin. Path. 1972, 25, 1079. 20. Nachum, R., Lipsey, A., Siegel, S. E. New Engl. J. Med. 1973, 289, 931. 21. Ross, S., Rodriguez, W., Controni, G., Korengold, G., Watson, S., Khan, W. J. Am. med. Ass. 1975, 233, 1366.
1278
patients dying within a few hours of admishospital, others recovering rapidly. Certain are of prognostic value-in pneumococcal signs meningitis, for example, coma on presentation and a low c.s.F. white-cell count indicates a gloomy prognosis-but on clinical grounds alone the outcome often cannot be foretold. Bacterial antigen some
sion
to
and endotoxin levels in c.s.F. have therefore been quantitated to see if they could be of help in prognosis. Antigen concentration in c.s.F. is closely related to the c.s.F. bacterial count2 and a relationship between antigen titre and prognosis might therefore be expected. This has proved to be the case, for in meningococcal, 22 21 pneumococcal,2a haemophilus,l 12 16 and E. coli17 meningitis those patients with severe brain damage and late sequelae had a higher concentration of antigen in c.s.F. than patients who recovered uneventfully. However, in most investigations there was considerable overlap between the two groups. Antigen usually disappears from the c.s.F. 24-48 hours from the start of treatment. Longer persistence of antigen is ominous. Persistence of antigen for more than 48 hours after the start of treatment, especially if accompanied by a poor clinical response, may be an indication for a change of therapy. Antigen in the serum, particularly if present in high titre or if present for more than 2 or 3 days, is another bad sign in haemophilus,t6 E. coli,t7 and meningococcal
meningitis.22 Detection of bacterial products in c.s.F. and serum is proving a useful adjunct to the management of patients with pyogenic meningitis. It allows rapid bacteriological diagnosis and, early in the course of their illness, it singles out the patients with a poor prognosis. It should not be regarded as a replacement for routine bacteriological investigations but as a supplement, for the diagnostic success-rate of antigen detection and culture combined is greater than that of either technique alone. In addition, culture may identify one of the more un-usual causes of bacterial meningitis and can provide valuable information on antibiotic sensitivities.
Irradiation of the Thyroid Gland THE people of Hiroshima and Nagasaki were exposed to external gamma and neutron radiation from the nuclear bombs "Fat Man" and "Little Boy", but not to any significant fallout. An accident during a thermonuclear explosion at Bikini in 1954 unfortunately subjected a group of Marshall 22.
Whittle, H. C., Greenwood, B. M., Davidson, N. McD., Tomkins, A., Tugwell, P., Warrell, D. A., Zalin, A., Bryceson, A. D. M., Parry, E. H. O., Brueton, M., Duggan, M., Oomen, J. M. V., Rajkovic, A. D. Am. J. Med. 1975, 58, 823. 23. Hoffman, T. A., Edwards, E. A. J. infect. Dis. 1972, 126, 636. 24. Tugwell, P., Greenwood, B. M., Warrell, D. A. Q. Jl Med. (in the press).
Islanders and Japanese fishermen to acute radiation from fallout-the only population so affected. They have been under surveillance ever since; weI have commented on some of the earlier reports,’ and a twenty-year review has now appeared.2 The fallout was deposited on people’s skin and surroundings and on food and water supplies. This resulted in very high epidermal doses, mostly from nonpenetrating beta rays, and whole-body gamma doses estimated at 175 rads on Rongelap and 14-69 rads in neighbouring islands. Because many radioisotopes of iodine were produced and ingested, thyroid doses were higher, calculated at a total amount of about 220-450 rads to adults and 700-1400 rads to children on Rongelap (because of many uncertainties these are rough estimates). Initially, transient nausea and depression of leucocyte-counts were common, and there were severe skin burns, all of which healed. In the succeeding years a few ill-effects were manifest, such as an increase in miscarriages and stillbirths, and chromosome abnormalities in blood-cells. (There has also been an increase in the incidence of non-thyroid cancer in the heavily exposed group, including a case of acute myeloid leukaemia, but the numbers are small; so far no skin cancers have appeared.) Then growth retardation was noted in 5 of the 19 children exposed before ten years of age. Soon evidence ofmyxoedema became apparent in 2 of them, and when satisfactory methods of measuring thyroxine were devised, low levels were found (the protein-bound iodine levels had, confusingly, been normal, apparently because of an iodoprotein in Marshallese unrelated to thyroid hormone). Thyrotrophin levels were normal. Nine years after exposure, thyroid nodules began to appear (but not in the 2 myxoedematous boys). By 1974, 27 of the 86 exposed people on Rongelap had nodules, with smaller numbers in the less heavily exposed people of the other islands. These nodules were all found clinically; scanning might have revealed larger numbers.3 The highest incidence of nodules has been in the heavily exposed group who were less than ten years old at the time of the accident. 1 of the 4 children exposed in utero developed a thyroid adenoma. 24 of these patients have had thyroid operations, revealing benign lesions in 21 and carcinoma in 3. Many glands had hairlike vessels on the surface, reminiscent of thyroids treated with radioiodine. Many of the adenomas were papillary; some patients were given tracer doses of 1311 before operation, and autoradiography showed diminished or absent function in the nodules. Most of the thyroids contained many minute encapsulated lesions 1. Lancet, 1966, ii, 580; ibid. 1968, i, 625. 2. A Twenty-Year Review of Medical Findings in a Marshallese Population Accidentally Exposed to Radioactive Fallout. By R. A. CONARD and others. Brookhaven National Laboratory, Upton, New York, 1975. 3. Favus, M. J., Schneider, A. B., Stachura, M. E., Arnold, J. E., Yun Ryo, U., Pinsky, S. M., Colman, M., Arnold, M. J., Frohman, L. A. New Engl. J. Med. 1976, 1294, 1019.
