Antibiotic Use and Risk of Multiple Sclerosis
Alvaro Alonso1, Susan S. Jick2, Hershel Jick2 and Miguel A. Hernán1
1 Department of Epidemiology, Harvard School of Public Health,
Boston, MA
2 Boston Collaborative Drug Surveillance Program, Boston University,
Lexington, MA
Correspondence to Dr. Alvaro Alonso, Department of Epidemiology,
Harvard School of Public Health, 677 Huntington Avenue, Boston, MA
02115 (e-mail: aalogut{at}alumni.unav.es).
Some reports suggest that bacteria, including Chlamydophila
pneumoniae, could be involved in the etiology of multiple sclerosis.
If that is true, persons who used antibiotics active against these
bacteria, compared with nonusers, might be at lower risk of multiple
sclerosis. Using a 1993–2000 case-control study nested in the United
Kingdom–based General Practice Research Database cohort, the authors
identified 163 multiple sclerosis cases who were followed up for at
least 3 years before their first symptoms (the index date). Up to 10
controls matched to the cases by age, sex, general practice, and
time in the cohort were selected. Exposure to antibiotics was
assessed through computerized medical records. Overall antibiotic
use or use of antibiotics against C. pneumoniae was not associated
with multiple sclerosis risk. However, use of penicillins in the 3
years before the index date decreased the risk of developing a first
attack of multiple sclerosis (odds ratio = 0.5, 95% confidence
interval: 0.3, 0.9 for those who used penicillins for 15 days
compared with no use). In conclusion, use of antibiotics active
against C. pneumoniae was not associated with a decreased risk of
short-term multiple sclerosis. The observed lower risk of multiple
sclerosis for penicillin users needs to be confirmed in other
populations.
anti-bacterial agents; Chlamydophila pneumoniae; multiple sclerosis;
prospective studies
Why doxycycline and roxithromycin (or azithromycin)?
Both are oral, both are active against Chlamydia pneumoniae, both are relatively inexpensive. They are relatively risk-free. They act synergically against test strains of the organism; giving both together would be the equivalent of giving a four-fold increase of each drug were it to be given alone. The drugs work on different steps in the bacterial protein synthesis pathway. Combination therapy reduces the chance of the emergence of resistance. Both drugs pass into the brain. Both reach good levels inside cells. This is very important. Both are well tolerated. Azithromycin is an alternative to roxithromycin. They deplete the organisms slowly: this is very important, as the release of bacterial endotoxins should not be sudden.
Rifampicin may also be considered. It, too, is synergic with doxycycline, penetrates the brain and is active intracellularly. It is not suitable for intermittent use. It is highly active, and, in patients with a large bacterial load, it may give rise to intense reactions.
Why are later short courses of metronidazole to be taken together with these antimicrobials?
Chlamydiae are complex organisms. Long ago their ancestors must once have been free-living bacteria which possessed their own energy-generating pathways. The transformation from EB to RB is an active change, and an active change implies the retention of at least some of these pathways. The ones with the most utility for this purpose would be anaerobic, and thus susceptible to metronidazole.
Doxycycline and roxithromycin block the replicating phase by inhibiting protein synthesis and may be expected to force the organism to maintain itself by using its own primitive anaerobic respiratory mechanisms. In this suspended state it would be susceptible to anti-anaerobic agents such as metronidazole.
This is borne out by clinical evidence. The administration of metronidazole after doxycycline in a patient with likely high-load Chl pneumoniae infection causes a bacteriolytic reaction more severe than that following the original administration of doxycycline.
However, there is a difference: in this leg of treatment there is no risk of the emergence of resistance, for the organism is unable to replicate. Metronidazole need thus be given in courses only as long as can be tolerated.
Five-day courses of metronidazole at three-week intervals, during continuous treatment with doxycycline and roxithromycin, would seem reasonable; at first, metronidazole may be limited to one or two doses on one or two days to judge the severity of reaction.
The eventual aim would be to give all three agents intermittently. This, the final leg of treatment, would entail a 14 day course of doxycycline and roxithromycin, with metronidazole given from day five for five days. (The reason for continuing doxycycline and roxithromycin for a few days after the metronidazole has been stopped is because these drugs both possess anti-inflammatory activity which may prevent a reaction to the organisms killed by metronidazole.) This course would be given once a month. After several months the intervals between the antibiotics would be cautiously extended.
Why this complex antibiotic regime?
The literature is filled with instances of treatment-failure in serologically-proven chronicChl pneumoniae infections of non-CNS systems, whether macrolides, tetracyclines or rifampicin have been used. When the drug is stopped, even after months of treatment, serology rises, and the patient relapses.
The intensive cyclical regime of combined antimicrobials outlined here corrals the pathogen, initially halting replication, then eliminating stalled intracellular forms. Extracellular forms may be depleted by giving N-acetyl cysteine (see below.)
No single antimicrobial agent can be expected to achieve this effectiveness against every phase of the organism's life.