A 45-Year-Old Man With Seizures and
Meningitis
Juan Carlos Sarria, MD, Louisiana State University School
of Medicine, New Orleans, La., David M. Mushatt, MD, Tulane
University School of Medicine, New Orleans, La., David H. Martin, MD,
Louisiana State University School of Medicine, New Orleans, La.
[Infect
Med 16(2):84-88, 1999. c 1999 SCP Communications, Inc.]
Our patient is a 45-year-old white male with an unremarkable past medical
history who was brought to the hospital by ambulance with persistent
generalized tonic-clonic seizures. On arrival at the hospital emergency room,
he was given intravenous lorazepam and a loading dose of phenytoin, but his
seizures were still not controlled. He then required intubation for mechanical
ventilatory support. He was subsequently transferred to the medical intensive
care unit for continued management.
There was no history of prior seizures or evidence of head injury. The only
additional information on admission was that he drank alcohol excessively. The
family emphasized, however, that he had never used illicit drugs and had never
smoked. He lives in New Orleans and has had no recent history of travel. The
family history was negative for seizure disorders.
Initial physical examination revealed a temperature of 98.6¸F (37¸C), blood
pressure 137/90mm Hg, pulse 97 beats/minute, respiration 26 breaths/minute, and
oxygen saturation 98% on 40% oxygen by mask. He had no rashes, but there were
minor scalp lacerations and facial abrasions. His pupils were 3mm bilaterally
and sluggishly reactive to light. His neck was supple. His lungs were clear,
and the cardiovascular examination was normal. Abdominal exam revealed no
masses and no organomegaly.
Neurologic assessment indicated a Glasgow Coma Scale of 8. He was described
as lethargic, responding only to painful stimuli. He moved all extremities, and
there was no facial asymmetry. He had positive corneal and gag reflexes. He
demonstrated clonus, symmetric in the lower extremities. There were no
meningeal signs.
Initial laboratory examination yielded a normal hemoglobin, hematocrit, and
platelet count. His WBC count was 12,900/mm3
with 82% segmented neutrophils, 10% lymphocytes, and 8% monocytes.
Electrolytes, glucose, and liver enzymes were normal. Blood alcohol and a blood
drug screen were both negative. A lumbar puncture yielded an opening pressure
of 29cm of water. The CSF contained 21 WBC/mm3
(normal, 0-5) with 99% lymphocytes and 1% granulocytes. There were no RBC. CSF
glucose was 89mg/dL (normal, 40-70), serum glucose was 106mg/dL (normal,
70-110), and protein was 99mg/dL (normal, 5-40). An electroencephalogram (EEG)
showed epileptogenic bilateral cerebral dysfunction, and a CT scan of the brain
was normal
Dr. Mushatt: Seizures in alcoholics are
usually a consequence of acute withdrawal or hypoglycemia related to heavy
alcohol consumption. However, we don't know whether this gentleman is really
what we consider an alcoholic, since excessive intake is quite a subjective
concept. What may be excessive to one observer might represent social drinking
to another. In any case, early laboratory data indicated that our patient had
not consumed alcohol and was not hypoglycemic.
Our differential diagnosis must therefore focus on the abnormal findings in
the CSF, which demonstrated a mild pleocytosis and protein elevation. The white
cells could be attributed to the seizures themselves, although such changes
usually occur in seizures following strokes. There is a continuing debate as to
whether seizures not associated with cerebrovascular accidents produce such
changes. I have personally seen mild pleocytosis in patients with idiopathic
seizure disorders and similar chronic abnormalities in the CSF of alcoholics.
The EEG obtained in this patient suggests diffuse cerebral dysfunction, which
might be due to metabolic or toxic factors. The normal CT scan also rules out
large mass lesions. At 45 years of age this patient's presentation would be
late for idiopathic seizure syndromes, although it is conceivable that a small
scar or brain abscess serving as an epileptogenic focus might be seen on MRI.
We should next consider infectious causes of the CSF changes. If the patient
is truly an alcoholic with cirrhosis, he would have an increased propensity to
gram-negative bacterial infections, although these usually present as
spontaneous bacterial peritonitis caused by Escherichia coli or other
coliforms. The increased incidence of such infections is in part a consequence
of complement abnormalities resulting from impaired liver function.
Certainly the most common cause of a mild lymphocytic pleocytosis with an
increase in protein but no hypoglycorrhachia is aseptic meningitis, caused by a
number of viruses -- particularly enteroviruses. These infections are generally
mild and not associated with seizures. It would be important to know whether
enteroviral infections were being seen in the community. In our region, most of
these infections occur during the summer months. His presentation is more like
that of meningoencephalitis, and next, I would consider herpes simplex virus
since it is a treatable etiology. I would want to obtain not only viral
cultures for this organism, but also a PCR for more rapid diagnosis. I would
strongly consider beginning acyclovir therapy empirically until the etiology is
resolved.
Arboviruses may also cause encephalitis, and once again epidemiologic
information is quite important. Are we seeing other cases of St. Louis
encephalitis or eastern equine encephalitis in our community? These 2 viral
pathogens are occasionally encountered in southern Louisiana. There is of
course no treatment, but I would obtain serum for antibody determination. I
would also like to save some CSF for viral culture, particularly for
enteroviruses. A PCR for enteroviruses is also available and would be useful if
other tests are negative.
The 2 great masqueraders that are possible etiologies in this patient are Mycobacterium
tuberculosis and Treponema pallidum. The CT scan has failed to
show evidence of tuberculomas and the chest film is clear, but tuberculosis can
present with almost any spectrum of manifestations. Strong evidence against
this etiology is a normal CSF glucose. I would still place a PPD skin test to
help rule out tuberculous meningitis, but often the skin test is negative in
such patients. Therefore, I would save some of the CSF for mycobacterial
culture and perhaps PCR. A mild CSF pleocytosis with a predominance of
lymphocytes is compatible with neurosyphilis, and testing for this organism is
readily accomplished by obtaining a CSF-Venereal Disease Research Laboratory
(VDRL) and appropriate blood tests. This should be done early in the course of
his management.
I think I would begin antimicrobial coverage for the more common bacterial
organisms causing meningitis, such as pneumococcus, meningococcus, and, in this
case, Listeria monocytogenes. As mentioned, I would also consider
beginning acyclovir.
The patient was initially begun on ampicillin and
ceftriaxone at dosages prescribed for the treatment of meningitis. The day
after lumbar puncture was performed, the laboratory reported a positive
CSF-VDRL with a titer of 1:256. The patient was then begun on aqueous
crystalline penicillin G, 24 million units a day, administered as 4 million
units IV every 4 hours; this regimen was continued for 14 days. A diagnosis of
syphilis was subsequently confirmed, with a microhemagglutination assay for
antibody to T pallidum (MHA-TP) in serum. The HIV serum antibody test
was negative. By the second hospital day he was awake and alert, but remained
confused and appeared to be delusional. Repeat neurologic examination revealed
hyperreflexia and bilateral ankle clonus. His hospital course was one of
gradual improvement to the point that he was able to ambulate on his own, but
according to his family, his mental status never returned to normal. He was
discharged to the care of a sister.
Final Diagnosis
Parenchymatous neurosyphilis, general paresis variety Discussion
Syphilis is a systemic disease caused by T pallidum,
with manifestations that vary according to duration of illness. CNS
involvement may occur during any stage of the disease, with such findings as
ophthalmic or auditory dysfunction, cranial nerve palsies, and signs of
meningeal irritation. These syndromes of neurosyphilis were first described
during the sixteenth century. Once penicillin became available in the early
1940s, the incidence of syphilis rapidly declined, as did the incidence of
CNS disease caused by T pallidum. More recently, the AIDS epidemic
has been associated with an increase in the incidence of syphilis and its
varied manifestations. However, overall incidence continues to decline to
where it now appears possible to achieve total eradication. Clinical Features Most of our understanding of neurosyphilis is derived from observations
reported decades ago. In a large collaborative study published in 1932, 3244
patients with syphilis were carefully evaluated.[1] This report well supported the axiom that syphilis
can mimic any disease, and this description certainly applies to
neurosyphilis. Neurosyphilis syndromes were rarely seen as pure entities and
usually overlapped with other disease manifestations. CSF abnormalities were
found in 25% of patients with primary syphilis, 35% of those with secondary
syphilis, and 56% of those with late secondary infection. The clinical presentation of neurosyphilis has changed somewhat since the
prepenicillin era in that asymptomatic or "atypical" disease is
currently more common than the originally described severe CNS
manifestations.[2]
Asymptomatic disease demonstrates a peak incidence 12 to 18 months after
primary infection, and constitutes one third of all neurosyphilis cases.
Spontaneous reversion occurs in 70% of cases. Acute syphilitic meningitis mimicking viral meningitis is relatively rare,
accounting for only 6% of neurosyphilis cases in Merritt's series.[3] These cases usually occur within a
few years of the initial infection and may be seen as part of the secondary
syphilis syndrome. About 40% of cases of acute syphilitic meningitis present
with cranial nerve abnormalities involving the third, sixth, seventh, or eighth
nerves. Meningovascular manifestations have a peak incidence 4 to 7 years after
infection and are characterized by mononuclear infiltration of the meninges
and focal or diffuse vasculitis, which results in thrombosis and infarction.
The clinical presentation is one of aseptic meningitis, strokes, impaired
sensorium, seizures, and cranial nerve involvement; the seventh and eighth
nerves are most frequently involved. Spontaneous resolution can occur in less
severe cases, and response to therapy is usually complete. Parenchymatous neurosyphilis, a relatively rare form of disease, is
characterized by neuronal loss, demyelination, and gliosis. This entity may
present as either general paresis or tabes dorsalis. General paresis reflects
severe cortical dysfunction with a peak incidence 10 to 20 years after
primary infection. Patients manifest, to varying degrees, psychiatric
changes, poor cognitive function, dementia, delusions of grandeur, pupil
abnormalities (Argyll Robertson), tremors, expressionless facies, and
hyperreflexia. CSF pleocytosis is always present. Depending on the duration
of symptoms, many of the manifestations of general paresis may not be
reversible, and those patients who do improve may be left with significant
residual disabilities. Tabes dorsalis, the most common form of neurosyphilis during the
prepenicillin era, is a result of destruction of dorsal nerve roots in the
spinal cord. This occurs 15 to 25 years after initial infection. Clinical
manifestations include lancinating pain, paresthesias, decreased deep tendon
reflexes, ataxia, incontinence, sensory loss, and cranial nerve
abnormalities. Argyll Robertson pupils may be seen in these patients as well.
Improvement of the neurologic disease associated with this syndrome does not
follow treatment, as the damage to the dorsal spinal cord root is
irreversible. Gummatous neurosyphilis is the most uncommon form, resulting from severe
inflammation and local tissue destruction that produces a mass lesion. Any
area of the CNS can be involved. In a review of 241 neurosyphilitic patients published in 1972,[4] 52% had a prior history of
syphilis, but only 56% of these had been treated. Symptoms included seizures
(24%), ocular changes (12%), impaired consciousness/stroke (11%), and
psychiatric changes (10%). Forty-three percent were asymptomatic. An abnormal
neurologic exam was found in 76% of these cases. Abnormal ophthalmologic
findings were seen fairly frequently and included Argyll Robertson pupils,
ptosis, uveitis, retinitis, and optic atrophy. Although the CSF fluorescent
treponemal antibody-absorption test (FTA-ABS) was positive in all patients,
only 48% had reactive serum VDRL determinations, and 57% had a positive
CSF-VDRL. A CSF pleocytosis >5 WBC/mm3
was seen in 82%, and a protein concentration >45mg/dL in 39%. Seizures are relatively common in neurosyphilis, reported in 24% of 282
patients studied in the mid-1970s.[5]
Focal seizures were seen in 13% of these patients, generalized seizures in
35%, and combined focal and generalized seizures in 52%. In this study,
seizures were more common in patients with meningovascular syphilis. EEG
findings consisted of focal nonspecific slowing in most cases, but in 20% the
EEG was normal. After therapy, 47% of patients remained seizure free, 35% had
less frequent seizure activity, and 17% had persistent seizures. Seizures may
be the only manifestation of neurosyphilis. Testing for syphilis should
therefore be included in the diagnostic workup of any patient with adult-onset
seizures. It is likely that our patient had general paresis based on his mental
status changes, including a delusional thought pattern and hyperreflexia. A
cortical pattern of brain injury was further suggested by the presence of
bilateral ankle clonus. He only very gradually regained the ability to walk,
and he left the hospital unable to care for himself. Though not clearly
associated with this form of neurosyphilis, it was of interest that this
patient had an extraordinarily elevated CSF-VDRL titer. There are no large comparative studies that clearly indicate that
neurosyphilis is more severe in HIV-infected patients, but some evidence
suggests that in these patients, syphilis has a more protracted and malignant
course. Predisposition to CNS involvement is greater, and relapses and
failures after therapy are more common.[6,7]
HIV patients may have an impaired immunologic response to T pallidum.
However, there is a lack of correlation between progression of disease to
involvement of the CNS and CD4-cell concentrations. Prevalence of syphilis in
HIV ranges from 5% to 44%, and 2% to 58% of HIV-infected patients with
syphilis develop neurosyphilis.[5]
The diagnosis of neurosyphilis in HIV-infected patients is difficult because
false-positive and false-negative serologic tests are common. In addition,
CSF pleocytosis is seen with HIV infection alone and may confuse
interpretation of laboratory changes, which should suggest neurosyphilis or
other CNS infections. Diagnosis Despite the increasing importance of syphilis in the HIV era, there have
been few advances in the diagnosis of this disease. Because T pallidum
cannot be cultured, serologic tests are the main means of obtaining a
diagnosis. The CSF-VDRL is an acceptable screening test; it is highly specific,
but the sensitivity is relatively low (30% to 70%).[7,8] A nonreactive result therefore
does not exclude the diagnosis. CSF treponemal tests (microhemagglutination T
pallidum, FTA-ABS) are highly sensitive, but less specific. These tests
are broadly reactive at various stages of syphilis. False-positive results
occur when Ig crosses the blood-brain barrier or when blood contaminates CSF
during lumbar puncture. A nonreactive result excludes the diagnosis, but a
reactive result is not necessarily diagnostic. CSF WBC and protein are considered indicators of CNS disease activity, but
they too are nonspecific, especially in HIV patients. CSF pleocytosis and
high protein are seen in up to 30% of HIV patients without neurosyphilis or
other CNS infection. Other tests or ratios of tests designed to detect intrathecal antibody
production (T pallidum hemagglutination [TPHA] index, IgG TPHA
ratio, quantitative MHA-TP) have been developed in an attempt to improve
specificity. The TPHA index takes into consideration the integrity of the
blood-brain barrier by comparing CSF and serum albumin concentrations.[7] These tests have been studied in
HIV-infected and uninfected patients with a nonreactive CSF-VDRL and appear
to have utility, but their accuracy remains unproven and will continue to be
unproven due to the lack of a "gold standard" test for the
diagnosis of neurosyphilis.[8]
PCR amplification of T pallidum nucleic acids in the CSF has been
attempted; however, the role of this technique is unclear because of the
small number of patients studied thus far.[9]
Treatment There have been no large, prospective studies evaluating newer antibiotic
regimens for neurosyphilis since penicillin was evaluated in the 1950s.
Though never rigorously studied because of its very potent activity against T
pallidum and very rare reports of treatment failures, penicillin remains
the drug of choice. The current recommendation for neurosyphilis is 18 to 24
million units of aqueous crystalline penicillin G, administered as 3 to 4 million
units IV every 4 hours for 10 to 14 days.[10]
Some experts have suggested that 8 to 10 days of therapy are adequate.
Occasional failures have been reported, even with 14 days of therapy.
Treatment effectiveness is difficult to assess because the organism cannot be
cultured or visualized in CSF specimens. Because of concern for persistence
of slowly dividing organisms after therapy, particularly in AIDS patients,
most experts recommend at least 2 weekly IM injections of penicillin G
benzathine at a dosage of 2.4 million units after the completion of IV
therapy. All other regimens, usually considered for penicillin-allergic patients,
are less well studied. The most popular alternative is doxycycline 100mg PO
bid for 28 days. However, in general, penicillin-allergic patients with
neurosyphilis should be desensitized and treated with penicillin. Ceftriaxone
1g IV or IM once daily × 10 to 14 days has been evaluated in
HIV-infected patients; a failure rate of 23% was reported in 1 series.[11] Follow-up A repeat examination of the CSF is recommended every 6 months for 2 years.
CSF pleocytosis should resolve rapidly, while protein concentrations return
to normal more slowly. The CSF-VDRL is the last parameter to become negative.
Persistence of CSF-VDRL at lower titers is not necessarily indicative of
active infection nor of any necessity for additional therapy if the CSF
pleocytosis and protein concentration are improving. Patients with persistent
CSF abnormalities are at higher risk for disease progression.
|
|
|
||||||||||||||||||