THE WESTERN JOURNAL OF MEDICINE * OCTOBER 1992 *
157 * 4
theoretic advantage of instituting treatment at the time the organism first invades the blood stream and before highgrade bacteremia can develop. Overwhelming postsplenectomy infection is fortunately a rare event, but its very rarity has contributed to the many gaps in understanding how to prevent infection in the first place and how to prevent progression to multiorgan system failure when infection occurs. Immunization is among the least controversial issues, along with patient education-insofar as the facts to transmit are known-and early assessment of febrile illnesses, partly because these are interventions with little intrinsic morbidity and clear analogies to a broad range of other medical situations. The risks and benefits of antibiotic prophylaxis beyond early childhood, and the delineation of basic criteria for the early institution of therapeutic antibiotics, are more difficult to determine because there are more adverse aspects of overtreatment-including toxicity, the selection of resistant organisms, and sometimes the disruption of the quality of life-to be weighed against an unknown success rate in defusing an explosive infection that is uncommon in itself but devastating when it occurs. It has been suggested that multicenter trials are needed to define the optimal use of splenic autotransplantation,2' and it could be argued that a similar approach is needed to define the optimal use of antibiotics to prevent cases like the one Brigden describes. Until and unless trials on such a heroic scale become available, these latter decisions must depend largely on evaluation tailored to each patient's unique situation. BARBARA STYRT, MD Associate Professor of Medicine and Microbiology and Public Health Michigan State University East Lansing
REFERENCES 1. Brigden ML: Overwhelming postsplenectomy sepsis-Still a problem. West J Med 1992 Oct; 157:440 443 2. Cullingford GL, Watkins DN, Watts ADJ, Mallon DF: Severe late postsplenectomy infection. BrJ Surg 1991; 78:716-721 3. Styrt B: Infection associated with asplenia: Risks, mechanisms, and prevention. Am J Med 1990; 88:33N-42N 4. Murdoch IA, Dos Anjos R, Mitchell A: Fatal pneumococcal septicaemia associated with asplenia and isomerism of the right atrial appendages. Br Heart J 1991; 65:102-103 5. Gaston MH, Verter JI, Woods G, et al: Prophylaxis with oral penicillin in children with sickle cell anemia. N Engl J Med 1986; 314:1593-1599 6. Piliero P, Furie R: Functional asplenia in systemic lupus erythematosus. Semin Arthritis Rheum 1990; 20:185-189 7. Eichner ER: Splenic function: Normal, too much and too little. Am J Med 1979; 66:311-320 8. Foster KJ, Devitt N, Gallagher PJ, Abbott RM: Overwhelming pneumococcal septicaemia in a patient with ulcerative colitis and splenic atrophy. Gut 1982; 23:630632 9. Bramley PN, Shah P, Williams DJ, Losowsky MS: Pneumococcal WaterhouseFriderichsen syndrome despite a normal spleen. Postgrad Med J 1989; 65:687-688 10. McElroy PJ, Henderson FI, Brown DL: Immune status and response to immunization with polysaccharide vaccines of a healthy, congenitally asplenic woman. Clin Exp Immunol 1989; 78:402-405 1 1. Corazza GR, Ginaldi L, Zoli G, et al: Howell-Jolly body counting as a measure of splenic function-A reassessment. Clin Lab Haematol 1990; 12:269-275 12. Zwas ST, Samra D, Samra Y, Sibber GR: Scintigraphic assessment of ectopic splenic tissue localization and function following splenectomy for trauma. Eur J Nucl Med 1986; 12:125-129 13. Pabst R, Westermann J, Rothkotter HJ: Immunoarchitecture of regenerated splenic and lymph node transplants. Int Rev Cytol 1991; 128:215-260 14. Eshel Y, Sarova-Pinhas I, Lampl Y, Jedwab J: Autosplenectomy complicating pneumococcal meningitis in an adult. Arch Intern Med 1991; 151:998-999 15. Molin MR, Shackford SR: The management of splenic trauma in a trauma system. Arch Surg 1990; 125:840-843 16. Holdsworth RJ: Regeneration of the spleen and splenic autotransplantation. Br J Surg 1991; 78:270-278 17. Centers for Disease Control: Update on adult immunization-Recommendations of the Immunization Practices Advisory Committee (ACIP). MMWR 1991;
40(No. RR-12):1-94 18. Peter G, Lepow ML, McCracken GH, Phillips CF (Eds): Report of the Com-
463 mittee on Infectious Diseases, 1991, 22nd edition. Elk Grove Village, Ill, American Academy of Pediatrics, 1991 19. Buchanan GR, Siegel JD, Smith SJ, DePasse BM: Oral penicillin prophylaxis in children with impaired splenic function: A study of compliance. Pediatrics 1982; 70:926-930 20. Konradsen HB, Henrichsen J: Pneumococcal infections in splenectomized children are preventable. Acta Paediatr Scand 1991; 80:423427 21. Splenic autotransplantation (Editorial). Lancet 1992; 339:781-782
Treatment Strategies in the Prevention of Tuberculosis THE RESURGENCE OF tuberculosis, which began to be reported in 1986, has created new concerns about how to control this persistent and ever-changing disease. Though considered "down, but not out," to paraphrase John Murray, MD, at the annual Amberson Lecture of the American Thoracic Society,t tuberculosis is clearly returning with a vengeance. 2I3 Reasons for its resurgence are the growing number of patients with immunosuppression due to human immunodeficiency virus (HIV) infection4,5 and our aging population6 with the immunodeficiency of age, malnutrition, and the use of drugs that may promote tuberculosis reactivation.5 These major factors plus the presence of immigrants previously infected with Mycobacterium tuberculosis, but without clinically detectable disease and often harboring drug-resistant organisms, create new challenges in our quest to control and eradicate tuberculosis from this country by the year 2010.' This goal will never be achieved using current tuberculosis prevention strategies. Today national policy dictates that isoniazid chemoprophylaxis be given all persons at high risk of contracting active tuberculosis. These are primarily patients with known contacts with infected persons, patients with a well-documented skin test conversion, and patients with chest x-ray film abnormalities as a minimum. Because most of these "high-risk persons" actually have non-culture-proven but active disease, they should be treated, but would it not make more sense to treat with two drugs for greater effectiveness? Socalled preventive therapy with isoniazid alone is widely held to be 90% effective. Does this mean a 10% failure rate? The downside of adding a companion drug such as ethambutol would, of course, be increased cost and some slight increase in toxicity. At this time we have no controlled clinical trials to determine whether isoniazid alone is as effective as the use of two drugs in these patients. Today we face a new tuberculosis epidemic in this country and throughout the world. Accordingly we must develop new strategies to prevent and treat this disease, particularly in high-risk groups, lest the epidemic become uncontrolled. One of the most advocated approaches to tuberculosis prevention and control is the "preventive therapy" of an expanded spectrum of risk groups, including patients infected with HIV or who have the acquired immunodeficiency syndrome (AIDS). Shelley Salpeter, MD, elsewhere in this issue offers specific advice about preventive therapy strategies for the 1990s, including advice about therapy for HIV-infected persons.8 This advice, based on contemporary knowledge of tuberculosis epidemiology and drug susceptibility patterns in this country, represents an extrapolation from inadequate data. None of it is based on solid evidence from controlled clinical trials, particularly in HIV-infected persons. It is known that HIV-infected persons, even those with clinical AIDS, respond well to the major antituberculosis chemotherapy agents given in various combinations, but more anemia
464
EDITORIALS EDITORIALS~~~~~~~~~~~~~~~~~~~~~~~~~~
ABBREVIATIONS USED IN TEXT AIDS = acquired immunodeficiency syndrome BCG = bacille Calmette-Guerin [vaccine] HIV = human immunodeficiency virus
zidovudine was
when
designate
to
used
is,
increasing
and
persons
infected
threshold level of a
suggested
Thus it is
should be
therapy
been
lowest risk has also
considered
reaction in HIV-
significant;
those at
induration before
Possibly
considered.
is
suggested,
substantial tuberculin
that a 5-mm
15 mm of
should have
lowest risk
the
"standard" of 10
units-has been
5 tuberculin
considered to be at
those
reaction in
advised."
tive
the
should
person. ° Lowering
intermediate-strength purified-protein
induration for
of
mm
derivative-that
over what criteria
infected
an
and zidovudine
alone.9
previously accepted
the
for
"threshold"
given
controversy
some
There is also be
chemotherapy
tuberculosis
with
occurred
than
even 2 mm of
preven-
induration
according to current readily distinreaction from the guish pinprick itself. Perhaps all HIV-positive persons should be given isoniazid chemoprophylaxis "preventive therapy," regardless of the degree of might
be taken as But
advice.
infection,
evidence of
many experts
that
doubt
I
could
2-mm
a
as
tuberculin
Again,
reactivity,
with this advice is
adequate prospective
bad,
indifferent in
or
why
AIDS,
not do
key questions?
plished
among
highest
those at
compared
chemoprophylaxis
controlled clinical trials
in
is
good,
to answer
have
these
recently successfully accom-
patients
with
risk for the
silicosis,
known to be
development of active and
tuberculosis. '3 In this
with the use of
based on
trials. Because we
infected with HIV or who
Such trials were
virulent clinical
was
patients
Hong Kong
in
that it is not
controlled clinical
whether isoniazid
know
not
do
suggested.'2
been
as has also
the trouble
trial, isoniazid therapy
rifampin and placebo
and with
rifampin versus placebo. In this study both and rifampin proved to be highly effective in preventing the emergence of clinical tuberculosis compared with isoniazid and
both
isoniazid
placebo.
was no more
The combination
effective than either
interesting that thus far, no rifampin resistance developed, but the later emergence drug resistance remains a possibility. similar studies in well-defined Why groups of HIV Answers about the efficacy and safety of isoniazid single drug
this
in
setting.
It was
or
isoniazid
of
not do
infection?
forms
other
or
infected
How
the
neuritis,
Native
such
it
as
is
the
a
isoniazid-related clear
that
drug
is
used
further
Another
alone,
in 19 of 20
Disease Control
to this the risk
from
used suicide
arthritis,
this
of
about 30
recently Moulding and coisoniazid hepatitis in one state
prevention
reactions
toxic
possible emergence
and
fatal
Centers for
frequently
has been known for
Americans,'6"7miscellaneous
becomes
the
a
More
was used for
cases.'5 Add
177
isoniazid
20 cases of
drug
patients,'4 and eral
hepatitis
isoniazid
scarcely reported.
workers found
ported
could be known in
in
but
where
preventive therapy
time,
short
of fatal
fact
years
of
considering the poor prognosis of HIVpatients general and AIDS patients in particular. could isoniazid preventive therapy cause harm? The
relatively
of these
recently
of isoniazid
re-
periph-
drug overdose where agent, particularly in less serious
reactions
drug interactions, and perfectly safe drug. The
and
is not a
isoniazid-resistant
though rare,
organisms,
must also be
where
considered
studied.
problem
with
so-called
preventive therapy
is
conceptual. We use the tuberculin reaction as evidence for infection, but the tuberculin reaction can be negative in the face of florid tuberculosis such as in older patients, in patients with disseminated disease, and in HIV-infected persons with tuberculosis. Conversely, the tuberculin skin test may be strongly positive, that is, more than 15 mm of induration, in many patients with less than a 5% chance of ever having clinical disease. Thus the tuberculin skin test is of only modest value in case finding. In addition, infected persons may have more actively multiplying organisms than is clinically evident, particularly in extrapulmonary forms of disease. Is it prudent to treat with only one drug? Would not a two- or even three-drug combination be more likely to be effective, particularly in cases where the chance of isoniazid resistance is present, such as in immigrants? For any kind of preventive intervention to be recommended for any population requires the establishment of the intervention's validity.'8 The randomized controlled clinical trial is the most conclusive, and expert opinion the least powerful, indicator of validity."8 For all of the above considerations, until guided by controlled clinical trials, the use of "chemoprophylaxis" or preventive therapy must be guided by clinical judgment. Decision analysis focusing on risk:benefit ratios can also be useful.'9'20 The use of decision analysis to guide isoniazid chemoprophylaxis is not new and was first seriously suggested more than a decade ago.2' An accompanying editorial by a major contributor to the knowledge about tuberculosis epidemiology and prospects for its prevention asked for more data at that time.22 Alas, the design and conduct of studies to help guide the choice of isoniazid in various populations did not occur, and we are thus left with therapeutic choices based on model rather than hard data. Now, with HIV infection, controlled clinical trials are desperately needed to guide the choice of preventive therapeutic strategies. A more recent use of decision analysis suggests that ethnicity, age, and sex must be factored into the choice of isoniazid chemoprophylaxis in persons who are not HIVpositive.20 In general, the benefit of isoniazid chemoprophylaxis appears to favor all groups except black women older than 50. But many would remain cautious about using this drug for so-called prevention for those older than 35. The guidelines and advice concerning tuberculosis chemoprophylaxis offered by Salpeter, if carefully followed, will probably do more good than harm-at least until there is better guidance through controlled clinical trials. Finally, are there alternatives to be considered in the prevention of tuberculosis? Although the official position of the American Thoracic Society and the Centers for Disease Control is against the use of the bacille Calmette-Guerin vaccine (BCG) in this country, certain experts in our country and elsewhere in the world disagree with this dictum.22-24 Of course, HIV-positive patients should not be given BCG because their impaired immune state might foster the development of clinical BCG infection. But why not use BCG in young, otherwise healthy persons who are tuberculinnegative, such as HIV-negative health care workers? The vaccine has been used worldwide as a control measure since 1921, with the greatest demonstrable success in children.25 Unfortunately, various BCG strains have been used in different countries with different results. Also,there is no identifiable immunologic counterpart of protective immunity.24The argument that vaccination will negate the value of the tuber-
THE WESTERN JOURNAL OF MEDICINE
o
OCTOBER 1992
0
157
0
465
4
culin skin test as a case-finding tool is countered by evidence that the age at vaccination has an important bearing on tuberculin skin testing 10 to 25 years later.26 Infants vaccinated with BCG in Canada had no greater tuberculin reaction than nonvaccinated infants in a follow-up study.26 Should not infants in HIV-infected high-risk areas such as New York City be vaccinated? Nearly everywhere I have- traveled outside the United States, I am asked wThiy we do not use BCG in certain highrisk groups as is commonly done elsewhere in the world including Canada. All I can answer, with a shrug, is, "There is no reason. It's just national policy." My views on the prevention of tuberculosis may stir controversy. That is intended because it is high time we reconsider the dogma and policies of the past decades that have failed to control tuberculosis in this country. THOMAS L. PETTY, MD Director of Academic and Research Affairs Professor of Medicine
University of Colorado
Health Sciences Center Denver REFERENCES 1. Murray JF: The white plague: Down and out, or up and coming? J. Bums Amberson lecture. Am Rev Respir Dis 1989; 140:1788-1795 2. Rieder HL, Cauthen GM, Kelly GD, Block AB, Snider DE Jr: Tuberculosis in the United States. JAMA 1989; 262:385-389 3. Rieder HL, Cauthen GM, Comstock GW, Snider DE Jr: Epidemiology of tuberculosis in the United States. Epidemiol Rev 1989; 11:79-98 4. Handwerger S, Mildvan D, Senie R, McKinley FW: Tuberculosis and the acquired immunodeficiency syndrome at a New York City hospital in 1978-1985. Chest 1987; 91:176-180 5. Barnes PF, Bloch AB, Davidson PT, Snider DE Jr: Tuberculosis in patients with human immunodeficiency virus infection. N Engl J Med 1991; 324:1644-1650 6. Morris CD: Pulmonary tuberculosis in the elderly: A different disease? Thorax 1990; 45:912-913 7. Centers for Disease Control: A strategic plan for the elimination of tuberculosis in the United States. MMWR 1989; 38:269-272 8. Salpeter S: Tuberculosis chemoprophylaxis. West J Med 1992 Oct; 157:421424 9. Antoniskis D, Easley AC, Espina BM, Davidson PT, Barnes PF: Combined toxicity of zidovudine and antituberculosis chemotherapy. Am Rev Respir Dis 1992; 145:430-434 10. Graham NMH, Nelson KE, Solomon L, et al: Prevalence of tuberculin positivity and skin test anergy in HIV- I -seropositive and -seronegative intravenous drug users. JAMA 1992; 267:369-373 11. American Thoracic Society: Diagnostic standards and classification of tuberculosis. Am Rev Respir Dis 1990; 142:725-735 12. Jordan TJ, Lewit EM, Montgomery RL, Reichman LB: Isoniazid as preventive therapy in HIV-infected intravenous drug abusers-A decision analysis. JAMA 1991; 265:2987-2991 13. Hong Kong Chest Service/Tuberculosis Research Centre, Madras/British Research Council: A double-blind placebo-controlled clinical trial of three antituberculosis chemoprophylaxis regimens in patients with silicosis in Hong Kong. Am Rev Respir Dis 1992; 145:36-41 14. Moulding TS, Redeker AG, Kanel GC: Twenty isoniazid-associated deaths in one state. Am Rev Respir Dis 1989; 140:700-705 15. Snider DE, Caras GJ: Isoniazid-associated hepatitis deaths: A review of available information. Am Rev Respir Dis 1992; 145:494-497 16. Brown CV: Acute isoniazid poisoning. Am Rev Respir Dis 1972; 105:206-216 17. Cash JM, Zawada ET: Isoniazid overdose-Successful treatment with pyridoxine and hemodialysis. West J Med 1991; 155:644-646 18. Brown EY, Viscoli CM, Horowitz RI: Preventive health strategies and the policy makers' paradox. Ann Intern Med 1992; 116:593-597 19. Colice GL: Decision analysis, public health policy, and isoniazid chemoprophylaxis for young adult tuberculin skin reactors. Arch Intern Med 1990; 150:25172522 20. Jordan TJ, Lewit EM, Reichman LB: Isoniazid preventive therapy for tuberculosis-Decision analysis considering ethnicity and gender. Am Rev Respir Dis 1991; 144:1357-1360 21. Taylor WC, Aronson MD, DelBanco TL: Should young adults with a positive tuberculin test take isoniazid? Ann Intern Med 1981; 94:808-813 22. Comstock GW: Evaluating isoniazid preventive therapy: The need for more data. Ann Intern Med 1981; 94:817-819 23. Smith DW: Why not vaccinate against tuberculosis? Ann Intern Med 1970; 72:419-422
6
24. HaroAS: Twenty years later-Evaluation of the results of a national mass BCG vaccination in Finland. Scand J Respir Dis [Suppl] 1972; 80:153-169 25. Fine PEM, Rodrigues LC: Mycobacterial vaccines. Lancet 1990; 335:10161020 26. Menzies R, Vissandjer B: Effect of bacille Calmette-Guerin vaccination on tuberculin reactivity. Am Rev Respir Dis 1992; 145:621-625
Nonsteroidal Anti-inflammatory Drugs and Ulcers ALTHOUGH ASPIRIN has been used for medicinal purposes for more than a century, only in the past five years have the gastrointestinal complications of this agent and the other nonsteroidal anti-inflammatory drugs (NSAIDs) become widely recognized as serious problems. As reviewed by Richard Babb, MD, elsewhere in this issue, NSAIDs have deleterious effects on the entire gastrointestinal tract from the esophagus to the colon, although the most obvious clinical effect is on the gastroduodenal mucosa.' My comments focus on a few points regarding the etiology of NSAIDinduced injury and the clinical consequences of this injury. Nonsteroidal anti-inflammatory drugs may cause or may precipitate peptic ulcers, but little is known about the actual pathophysiologic mechanisms involved. A number of controversial points pertain to the delivery to the mucosa, the mode of action, and the nature of the injury caused by NSAIDs.
NSAID Delivery to the Gastroduodenal Mucosa These drugs are delivered to the gastric and duodenal mucosa by three routes: * Aspirin and the other NSAIDs that are soluble at both acidic pH and weak acids can be dramatically and rapidly concentrated in the gastric mucosa by pH-dependent trapping. Weak acids are un-ionized at an acidic pH of gastric secretions and can freely cross the gastric cell membranes. When these weak acids encounter the neutral pH of the cell interior, however, they give up a hydrogen ion and, once negatively charged, are unable to exit the cell. Concentrated in mucosal cells, their potential damaging effects are amplified, contributing to the damage occurring within minutes after intragastric administration. Antisecretory agents reduce the amount of acute mucosal injury from aspirin in part
by diminishing pH-dependent trapping. * Administering NSAIDs parenterally can also increase the risk of ulcer, often without the superficial mucosal damage typically seen with their topical administration. For example, ketorolac tromethamine (Toradol) given daily at an intramuscular dose of 90 mg for five days produced invasive antral ulcers in four of five subjects tested.2 Two points are important: a drug's systemic action may mediate an ulcer risk, and the absence of acute injury does not guarantee the
absence of an ulcer risk. * Certain NSAIDs, such as indomethacin and the oxicams, enter the enterohepatic circulation, thereby producing high intestinal concentrations of active drug. These NSAIDsin that are recirculated cause intestinal damage, resulting or focal bleeding, perforation,3 and strictures from ulcers submucosal injury. The enterohepatic circulation of NSAIDs contributes to duodenal mucosal drug exposure and also to gastric mucosal exposure (to the extent of duodenalgastric reflux). The relative contributions to the NSAID-induced ulcer risk