Вы находитесь на странице: 1из 7

Ocular Tuberculosis

by C. Michael Samson, M.D.

The term "ocular tuberculosis" encompasses any infection by Mycobacterium tuberculosis, or one
of three related mycobacteria species (sp. bovis, africanum, and microti), in, on, or around the
eye. Historically, authors have used the terms "primary" and "secondary" ocular tuberculosis. The
definition of primary ocular tuberculosis (TB) has varied: some authors have used it to describe
ocular disease in the absence of systemic involvement, while others take it to describe disease in
which the eye is the initial port of entry of the mycobacterium into the body. econdary ocular
tuberculosis is defined as ocular involvement as a result of seedin! by hemato!enous spread
from a distant site, or direct invasion from ad"acent areas, like the sinus or cranial cavity.
There are known cases of ocular involvement with other species of mycobacteria. Thou!h many
share similar clinical presentations as TB, they are considered distinct separate patholo!ic
TB has e#isted since ancient times. $ra!ments of spinal columns from %!yptian mummies show
evidence of tuberculous patholo!y. &n ancient 'reece, the disease was known as consumption,
and was uniformly fatal. Hippocrates even cautioned his collea!ues, warnin! them not to visit
patients with consumption, because the inevitable death of the patient mi!ht dama!e their
&n the mid ()
century, sanatoriums were created* these sites, in con"unction with improved
social and sanitary conditions, helped control TB+s prevalence. &n (,,-, .obert /och discovered
a stainin! techni0ue to demonstrate M. tuberculosis. &n (,)1, 23rays were discovered, !ivin!
physicians a new tool with which to follow the disease. &n ()45, treptomycin was proved to be
safe in humans, and the followin! year, was administered for the first time to a patient critically ill
with TB, and with ama6in! results as this terminally sick patient recovered in the followin!
7aitre38an is credited with the earliest written description of ocular tuberculosis in (9((, when he
described a case of an iris lesion, which eventually caused perforation. &n (,5:, 'uenea de
7ussy was the first to reco!ni6e choroidal tubercles. The year after acid3fast stain was
discovered, mycobacteria or!anisms were found in ocular tissue in a case of panophthalmitis.
The number of TB cases worldwide corresponds with economic conditions: the hi!hest
incidences are seen in the countries of ;frica, ;sia, and <atin ;merica that have the lowest '=>.
The ?H@ estimates that , million people !et TB every year, )1A in developin! countries. =early
5 million people die from TB each year.
$rom ()15 throu!h (),4, the number of TB cases fell, mostly in industriali6ed countries: the rate
in the B dropped appro#imately CA per year durin! this period. However, around (),1, there
was an increase in number of new cases* new cases in the B rose from --,-:( in (),1 to
-1,5(5 in ())5. everal factors have been sited to e#plain why this occurred:
(. H&D
-. Hi!h rate of immi!ration from countries with a hi!h incidence of TB
5. &nade0uate fundin! for TB control and other health efforts
4. %mer!ence of multi3dru! resistance of TB
However, since ())5, the number of new cases has be!un to drop a!ain, both worldwide and in
the B* new cases of TB in the B had dropped to -(,559 in ())C. $actors that have contributed
to this decrease include new and better treatments for H&D, increased physician awareness and
better institution of Eirect @bserved Therapy (E@T), and the use of multi3dru! therapy. The main
lesson is that TB is still present, and may continue to be a si!nificant patho!en !oin! into the new
millenium: physicians must always keep it in mind in the relevant clinical settin!s.
There has been a dramatic chan!e in the epidemiolo!y of ocular tuberculosis since the
century. Eurin! that time, TB was considered a common cause of uveitis. &t was considered
so common, that prominent ophthalmolo!ists of the time were able to classify TB uveitis into
different types. ome investi!ators of the time found TB to account for up to (:A of all uveitis
cases. &n the ()4:s, 'uyton and ?oods placed TB as the cause of ,:A of all !ranulomatous
uveitis. @ver the subse0uent decades, the number of uveitis cases attributed to M.
tuberculosis declined. There are many reasons for this, most citin! the fact that other previously
unknown etiolo!ies, like sarcoid, to#oplasmosis, and Histoplasmosis, were now reco!ni6ed. ;lso,
with new dia!nostic tests, dia!nostic criteria for ocular tuberculosis have become more strict.
Today, ocular tuberculosis is e#tremely rare. $or e#ample, there was a recent publication of an
epidemiolo!ical review of the cases of uveitis seen over a - year period at a uveitis referral clinic
in &ndia, a country in which TB is endemic. @f a total of (-95 cases, they only found 1 cases of
ocular TB, despite fairly broad dia!nostic criteria. &t should be noted that in all 1 cases, the
dia!nosis was definitive, in that they were able to demonstrate acid3fast bacilli in either ocular
fluid or in tissue specimen. ince the (),:+s, reports in the literature cite TB as an etiolo!y of
uveitis from : F 4A.
Clinical Presentation
The most common manifestation of ocular involvement is uveitis, usually presentin! as a chronic
anterior uveitis, panuveitis or as a choroiditis. Holland and Helm, at the 8ules %ye &nstitute, wrote
a review published in urvey of @phthalmolo!y in ())5, and found a total of 4: cases of
histolo!ically confirmed intraocular TB in the literature, datin! back to (,C). There were several
conclusions one could draw from the data. $irst, the a!e ran!ed from ( yr to 91, showin! that any
a!e can be affected. There was no !ender predilection. The ma"ority of cases were chronic. The
papers indicated that most presentations were !ranulomatous, which is in concordance with
current thinkin!, but many reports did not specify the type of inflammation. However, there are
histolo!ically3confirmed cases of intraocular TB that present as non3!ranulomatous uveitis, so
non3!ranulomatous uveitis does not e#clude TB. &ris nodules were rare. The ma"ority of cases
were unilateral, and both anterior and posterior se!ments were involved.
@ne other important note from the older literature was that when tuberculosis was more
prevalent, and TB was still considered one of the important causes of uveitis, it was rarely seen in
patients with active pulmonary tuberculosis. @ne study reported only 5 cases of iritis amon! (:95
cases of pts with systemic TB. ;nother series reported no cases of iritis in a series of (::: pts
with pulmonary TB. Eonahue, in the ()C:Gs, did a review of (:,151 patients bein! treated at TB
sanatoriums, and only found -, cases of iritis.
@n the other hand, the presence of choroidal lesions, with or without inflammation, is stron!ly
correlated with systemic disease, and is an indicator of hemato!enous spread of mycobacteria. ;
review in ()4, showed choroidal tubercles were present in -,A of cases of miliary TB. &n ()C4,
another study looked at cases of pulmonary TB without evidence of miliary TB, and showed a
similar prevalence of choroidal tubercles.
Bpon e#amination of the literature, several characteristics of the choroidal tubercles become
evident. The ma"ority are unilateral, and can ran!e in si6e from (34 to several disk si6es in
diameter. 7ost of the lesions are found around the posterior pole, and althou!h there are cases
of 1:3C:H lesions, there are usually less than 1. The lesions appear yellow early on, and become
more pi!mented as time passes. @ccasionally, they can be associated with an overlyin! serous
There are other ocular manifestations of TB. The retina is rarely involved primarily, but there are
case reports of retinal vasculitis or vascular occlusions with associated periphlebitis. Tuberculous
con"unctivitis has been reported several times in the literature* usually unilateral, chronic
con"unctivitis, occasionally associated with con"unctival mass or ulceration. 7ost do not have
systemic manifestations of TB, and may represent primary ocular tuberculosis.
Tuberculosis can also present as orbital mass, or as a eyelid abscesses. There are biopsy3
proven cases of TB scleritis. >hylectenulosis is a Type &D Hypersensitivity reaction that presents
as an inflammatory mass on the cornea, and can be associated withStaphylococcus aureus, as
well as tuberculosis.
There are two other ocular entities that are related to TB. The first is %ales+ disease. This is a rare
disorder that is not associated with M. tuberculosis specifically, but with a positive tuberculin skin
test. &t is characteri6ed by recurrent vitreous hemorrha!es in youn! men. .etinal periphlebitis is
the prominent findin!, alon! with peripheral capillary nonperfusion. &n the few cases where
patholo!y was e#amined, there was no evidence of mycobacterium infection.
The other entity is uveitis associated with tuberculin skin testin!. This is a bilateral !ranulomatous
anterior uveitis that responds to topical and oral steroids. The few reported cases did not find
evidence of systemic TB, and resolved without the need for anti3tuberculous medications.
$i!ures (I-. $undus photo!raphs and correspondin! fluorescein an!io!ram of presumed
choroidal tubercle, nasal to the disk. The patient was stron!ly >>E reactive, and was treated with
anti3tuberculous medications. >hotos courtesy of 8oseph ?alsh, 7.E., Jhairman of
@phthalmolo!y at the =ew Kork %ye I %ar &nfimary.
$i!ures 5I4. $undus photo and correspondin! fluorescein an!io!ram of same patient, after C
months of treatment with anti3tuberculous medications. Jourtesy of 8oseph ?alsh, 7.E.,
Jhairman of @phthalmolo!y at the =ew Kork %ye I %ar &nfirmary.
@ur knowled!e of the patholo!y of ocular tuberculosis is limited, not only because of the rarity of
cases, but also because rarely does one have the need or opportunity to remove a piece of the
eye, let alone the entire eyeball. However, such instances do occur, and they occur in one of four
circumstances: after an eye becomes phthisical from chronic, drawn out inflammatory episodes*
from an acute tuberculous panophthalmitis* enucleation when a tumor is suspected (like
retinoblastoma or a ciliary body tumor)* and lastly, if the patient dies of systemic tuberculosis.
Histolo!ic e#am usually shows !ranulomatous inflammation of the choroid, characteri6ed by
lymphocytes, epithelioid cells, and !iant cells, in the presence of caseatin! necrosis. @verlyin!
retina can be involved. 'ranulomatous inflammation can e#tend into the iris and ciliary body, but
frank !ranulomas in these areas are not common. The sclera involvement can ran!e from mild to
frank perforation.
Eefinitive dia!nosis is achieved by identifyin! the M. tuberculosis in ocular tissue or fluid.
Historically, this proved difficult and confirmation usually only followed enucleation of the eye.
However, in modern days, there are several tools that have madeit possible to obtain a definitive
dia!nosis while keepin! the eye intact.
>ars plana vitrectomy allows removal of intraocular specimens. ;cid3fast stains may identify
mycobacteria immediately, and cultures can be used to identify antibiotic sensitivities and
resistance. ; team of retinal specialists at the Bniversity of ?isconsin published a report of
dia!nosin! TB choroiditis by chorioretinal endobiopsy, usin! standard three3port pars plana
vitrectomy techni0ue. >ost3operatively, the patient did well, vision improved from J$ to -:L9:,
with an epiretinal membrane as the only post3operative conse0uence.
7ore recently, >J. has been used to dia!nosis ocular TB. Jase reports show that both a0ueous
and vitreous specimen may be used to make the dia!nosis. However, false3positives are an
e#pected difficulty when usin! a test that is e#tremely sensitive.
However, there are many times when a uveitic patient is not a sur!ical candidate, or one is not
able to demonstrate M. tuberculosis from an ocular specimen. &n these circumstances, it is
important to try find systemic evidence of TB. Jhest 23ray findin!s shows infiltration, sometimes
with evidence of cavitation or pleural effusion. ;cid3fast stains and cultures can be done from
urine, spinal fluid, or sputum.
; common dilemma is the patient who presents with uveitis and demonstrates a positive
tuberculin skin test. ince the treatment of proven active TB infection is different from prophyla#is
in >>EH patients without evidence of active TB, the decision falls on the ophthalmolo!ist to
decide if the uveitis is indeed a si!n of active TB infection. &t is probably not indicated to treat
every uveitis patient with a positive >>E with lon!3term anti3tuberculous medication in the
absence of other evidence of tuberculosis, but each case should be treated individually, lookin!
at the risks and benefits of treatin! or not treatin! with TB medications.
?hen TB is confirmed, treatment is be!un immediately, since treatment will last at least C
months, and will sometimes re0uire treatment up to - years. Treatment for ocular tuberculosis is
the same as for pulmonary tuberculosis. Because of the prevalence of resistant TB, especially to
isonia6id (&=H) and rifampin, multi3dru! therapy is now used routinely. &f primary isonia6id
resistance e#ceeds 4A in the re!ion, it is recommended to use 4 dru! therapy. Bnfortunately,
there are only a few states where resistance is less than 4A: these include ;rkansas, ;ri6ona,
=evada, 7innesota, 7aine, =orth I outh Eakota, and ?yomin!. &=H, rifampin, and
pyra6inamide (>M;) are used for the first - months* a fourth dru!, streptomycin or ethambutol, is
also used until susceptibilities are available. $or the second part of treatment, &=H and rifampin
are used for 4 months, for a total of C months therapy. The $E; has licensed fi#ed3dose
combination dru!s of this re!imen: .ifater (isonia6idLrifampinLpyra6inamide) and .ifamate
(isonia6idLrifampin). There is an alternative ) month re!imen for people that cannot tolerate >M;:
&=HL.if is used, with ethambutol or streptomycin until susceptibilities are confirmed.
T ! H"#
There are some points to keep in mind when considerin! Tuberculosis in the H&D infected patient.
$irst, there are multiple case reports of H&D patients with biopsy or culture proven systemic TB
who do not respond to tuberculin skin testin!: tuberculin skin testin! is not reliable in this !roup.
@cular tuberculosis manifestations are the same as in immunocompetent patients, and
disseminated choroiditis is the most commonly cited manifestation. =ot only does their
immunocompromised status retard recovery, but dru! malabsorption is additionally a common
problem in H&D infected individuals, mandatin! lon!er therapy.
>atients should be followed for side effects of medication. &sonia6id, rifampin, and pyra6inamide
have each been associated with dru!3induced hepatitis* liver en6ymes should be followed, as well
as warnin! the patients of symptoms of hepatito#icity. .ifampin is associated with
thrombocytopenia: complete blood counts should be followed as well. >yra6inamide is associated
with hyperuricemia, but acute !out is not common, and asymptomatic hyperuricemia is not an
indication to stop the medication. &sonia6id is associated with peripheral neuropathy: patients with
conditions associated with neuropathy should also take pyrido#ine, which has been shown to
prevent isonia6id3associated neuropathy.
treptomycin has been associated with hearin! loss. %#perts recommend a baseline audiometry
before be!innin! therapy.
%thambutol is associated with optic neuritisL optic neuropathy* the development of si!ns or
symptoms of this complication may warrant discontinuation of the dru!.
&ntraocular tuberculosis can cause complications seen in other types of uveitis: cataracts,
!laucoma, and in posterior disease, retinal detachment. ubretinal neovasculari6ation has also
been reported, and responded to retinal photocoa!ulation in at least two reports.
The most common reason sited for treatment failure in pulmonary tuberculosis is nonadherance
to the therapeutic re!imen. &n =ew Kork Jity in ())(, a study showed ,)A of (,) patients failed
to complete therapy. ,:A of those brou!ht back failed therapy a second time. &sonia6id
resistance in =ew Kork Jity in ())( was -5A. Then, Eirect3@bserved3Therapy (E@T) became a
routine part of treatment. E@T involves the direct monitorin! of the patient takin! the pills by a
health care worker. Jurrent recommendations are that all patients should be treated under E@T
because there is no way to predict which patients will not comply with the re0uired si# to nine
month therapy. .ecently, &ionia6id resistance in =ew Kork Jity is down to around 1A.
There is little data on pro!nosis of ocular TB, a!ain because of rarity of cases. However, it is
reasonable to assume that TB in the eye re0uires prolon!ed treatment to eradicate the or!anism,
"ust as it does in other or!ans.

Ocular Tuberculosis
Manolette $o%ue, M.D.
True or $alse
(. The most common ocular manifestation of tuberculosis is uveitis. T.B%. &t usually presents as
a chronic anterior uveitis, panuveitis, or as a choroiditis.
-. Bnlike systemic patholo!y, ocular histolo!ical e#amination of tuberculous lesions are usually
non3!ranulomatous. $;<%. Histolo!ic e#amination shows !ranulomatous inflammation of the
choroids, characteri6ed by lymphocytes, epitheliod cells, and !iant cells, in the presence of
caseatin! necrosis. The overlyin! retina can be involved. 'ranulomatous inflammation can
e#tend into the iris and ciliary body, however, frank !ranulomas in these areas are rare. cleral
involvement can ran!e from mild to frank perforation.
5. Eefinitive dia!nosis of ocular tuberculosis is made by the demonstration of iris nodules.
$;<%. Eefinitive dia!nosis is achieved by identifyin! the or!anism Mycobacterium
tuberculosis in ocular tissue or fluid. ;cid3fast stains, cultures, and >J. analysis have been used
to identify M tuberculosis.
4. %thambutol is contraindicated in the treatment of ocular tuberculosis because of its association
with optic neuropathy. $;<%. 7onitorin! for the development of si!ns or symptoms relatin! to
the use of ethambutol is advised. The presence of si!ns or symptoms of optic neuropathy may
re0uire discontinuation of the use of ethambutol.
1. &ntraocular administration of anti3tuberculous dru!s is the route of choice for treatment of
intraocular tuberculosis. $;<%. The treatment of intraocular tuberculosis is the same as that
!iven for pulmonary tuberculosis. 7ultidru! therapy (isonia6id, rifampicin, pyra6inamide,
streptomycin or ethambutol), is !iven.
>%;.< NB%T&@=
(. ?hat is the etiolo!ic a!ent for ocular tuberculosisO ;nswer: Mycobacterium tuberculosis. &t is
an acid3fast bacilli.
-. ?hat is the most sensitive dia!nostic procedure for isolatin! Mycobacterium tuberculosisO
;nswer: >J. analysis.
5. Eescribe the clinical findin!s of a choroidal tubercle. ;nswer: 7a"ority are unilateral and ran!e
in si6e from (P4 to several disc si6es in diameter. They are mostly locali6ed in the posterior pole.
These lesions may number from less than 1 to more than C: lesions. The appearance is yellow
early on, and pi!mentation occurs in time. @ccasionally, choroidal tubercles are associated with
serous retinal detachments.
4. ?hat are the dru!s of choice (E@J) for ocular tuberculosisO ;nswer: &soni6id, rifampicin,
pyra6inamide, streptomycin or ethambutol.
1. ?hat is the most common complication of anti3tuberculosis dru! useO ;nswer: Eru!3related
hepatoto#icity. The patient should be advised of symptoms of hepatoto#icity and liver en6ymes
should be monitored.