Short reports
Ceroid granulomas in the female genital system
K Ooi, C Riley, V Billson, A G Ostor
AbstractThree cases of ceroid granulomas of thefemale genital
system are presented, in-volving the cervix in two and lesions
inthe ovaries and bowel serosa in the other.Ceroid granulomas are
unusual and in-teresting lesions formed when suitablesubstrates
accumulate within macro-phages to such an extent that a
relativelack of biological antioxidants results andauto-oxidation
and conversion to ceroid isfavoured. This may occur in the setting
ofhaemorrhage and necrosis, whether fromtumour necrosis or
associated with en-dometriosis. Other sources of lipids
andlipoproteins include bile, meconium andvernix caseosa.(J Clin
Pathol 1995;48:1057-1059)
Keywords: Lipofuscinosis, ceroid granuloma,
cervix,endometriosis.
Ceroid granulomas are rare lesions which havebeen reported in a
number of sites includingthe female genital system. Although in the
casespresented here the granulomas were incidentalfindings, one of
these (case 2) presented adiagnostic problem.
Figure 1 Effacement of normal cervical architecture in case 3 by
xanthogranuloma(haematoxylin and eosin, original magnification x
20).
Case reportsCASE ONEA 48 year old woman underwent an
electivehysterectomy for persistent menorrhagia des-pite treatment
with norethisterone. Macro-scopically, the specimen consisted of a
uterus80 mm in length, measuring 40 mm betweenthe cornua and 40 mm
in the anterio-posteriorplane. The 25 mm cervix contained a 7
mmcyst-like lesion filled with clotted blood. Histo-logical
sections of cervix revealed a focus ofendometriosis consisting of
endometrial gland-ular epithelium and stroma which merged witha
well circumscribed sheet of polygonal macro-phages laden with a
granular pigment whichwas predominantly golden brown but
whichshaded to grey in some areas. Other findingsincluded inactive
progestogen affected en-dometrium and superficial adenomyosis.
CASE TWOA 50 year old woman with a long history ofendometriosis
underwent a hysterectomy andbilateral oophorectomy with multiple
biopsiesofbowel serosa. Macroscopically, the specimenconsisted of
an intact uterus, measuring 80 mmbetween the comua, 50 mm
anterio-posteriorlyand 100 mm in length. The right ovary con-tained
an 80 mm "chocolate cyst", the leftovary contained a 35 mm
"chocolate cyst" anda smooth haemorrhagic serosal plaque meas-uring
15 x 20 x 5 mm. Microscopic exam-ination revealed extensive
adenomyosis andthe endometrium was in the proliferative phase.The
right ovarian cyst displayed residual en-dometrial type stroma and
glandular epi-thelium with sheets of pigment-ladenmacrophages in
the surrounding tissue. The leftovarian cyst was similar but lacked
endometrialtissue. The pigment was granular and pre-dominantly of
grey brown hue but some of thecells also contained a golden brown
pigmentresembling haemosiderin. Sections from theserosal plaque
showed sheets of similar cells,again containing abundant granular
pigment.These cells were originally described as "de-cidual type"
despite the proliferative nature ofthe endometrium.
CASE THREEAn 84 year old woman with a past history ofdementia
and stroke with residual dysphagia
Department ofAnatomical Pathology,The Royal Women'sHospital,132
Grattan Street,Carlton, Victoria 3053,AustraliaK OoiC RileyV
BillsonA G Ostor
Correspondence to:Dr A G Ostor.Accepted for publication24 April
1995
I Clin Pathol 1995;48:1057-1059 1057
on June 8, 2021 by guest. Protected by copyright.
http://jcp.bmj.com
/J C
lin Pathol: first published as 10.1136/jcp.48.11.1057 on 1 N
ovember 1995. D
ownloaded from
http://jcp.bmj.com/
Ooi, Riley, Billson, Ostor
Pigment profile of ceroid granulomas
Case number
Stain 1 2 3
Haematoxylin and eosin Granular blue/grey Granular blue/grey
Granular brown/greySudan black Black Black BlackPeriodic
acid-Schiff Red Red RedSchmorl's Dark blue Dark blue
Blue-greenZiehl-Neelsen Acid fast (red) Acid fast Acid fastPerls' *
* *S100 Negative Negative NegativeAuto fluorescence Yellow-green
Yellow-green Yellow-green
* Each case was focally positive for haemosiderin but the
pigment was predominantly negative.
Xw, 'i 44ie * 4 -R *bf*
Figure 2 Staining characteristics of ceroid showing positivity
for (A) periodic acid-Schiffreagent, (B) Perls' stain, (C)
Ziehl-Neelsen, and (D) Schmorl's reagent (originalmagnification x
400).
and hemiparesis was referred from her nursinghome with a history
of postmenopausal bleed-ing. An examination was performed under
an-aesthetic and a diagnosis ofpyometra was madeand a growth of
Escherichia coli subsequentlyobtained. A total abdominal
hysterectomy wasperformed. The hysterectomy specimen com-prised the
body of the uterus and part of thecervix and measured 18 mm in
length. Onsectioning, a fungating, papillary tumour wasfound,
filling the fundus and most of the bodyofthe uterus. Below this,
the lining ofthe cavityhad a fine nodular cobblestone-like
ap-pearance. Microscopic examination confirmedthe presence of a
poorly differentiated en-dometrioid adenocarcinoma with minimal
in-vasion of the myometrium. In addition, thenormal cervicovaginal
architecture had beenreplaced by a xanthogranulomatous
reactionextending through the entire wall composed ofplump
histiocytes laden with granular yellow-brown pigment (fig 1). In
some areas there wasassociated pseudoepitheliomatous hyperplasiaof
the overlying epithelium alternating withfoci of superficial
ulceration and an underlyingmixed inflammatory infiltrate.
Histochemical and immunohistochemicalstains to define the
granular pigment furtherare summarised in the table and illustrated
infig 2.
DiscussionIn each of the cases the pigment was identifiedas
ceroid, first described as an acid fast brownpigment occurring in
rats with experimentallyinduced cirrhosis of the liver.' Lillie
coinedthe term ceroid from the Greek "keros" becauseof its wax-like
qualities. It was thought ori-ginally to be different from the
lipofuscinswhich are breakdown products formed fromlipids and
lipoproteins by a cascade of reactionsincluding auto-oxidation,
peroxide formationat double bonds and polymerisation.
However,ceroid is now generally considered to be anearly form of
lipofuscin, derived from lipids orlipoproteins which are only
partly oxidised.2Some authors distinguish between the two
pig-ments, referring to lipofuscin as a naturallyoccurring, age
related, wear and tear pigmentand reserve the term ceroid for
pigment foundunder pathological circumstances.3 Likely pre-cursor
substances include unsaturated fattyacids, cholesterol,
phospholipid, and glyco-proteins. As the former are found in
cellmembranes, it has been suggested that cyto-destructive
processes such as haemorrhageor necrosis may result in their
release. Othersources of substrate suggested include me-conium and
vernix.' Bile also contains fattyacids and has been implicated in
the formationof ceroid.3 Vitamin E is an important lipidsoluble
antioxidant and ceroid has been shownto accumulate in vitamin E
deficiency states.4If sufficient quantities are present,
lipofuscinsmay accumulate in macrophages formingceroid granulomas.5
Haemosiderin, pre-sumably derived from degraded erythrocytes,is
frequently found in association with ceroidwithin macrophages and
was present in eachof our three cases.
Similar granulomas have been reported inthe female genital
system within the wall ofendometriotic cysts, in tubo-ovarian
masses, inthe placenta, the ovary, the endometrium, andthe
cervix.5-7
In both cases 1 and 2 the ceroid granulomaswere directly related
to endometriotic deposits.Such an association has been described
pre-viously, and the cyclical changes of cell growth,haemorrhage
and necrosis associated with en-dometriosis seem likely to provide
generousamounts of suitable substrates for the formationof ceroid.
This, however, does not explain whyceroid granulomas are not more
commonlyrecognised. One possible contributing factor isthat
intra-cytoplasmic pigment associated withendometriosis is usually
dismissed as haemo-siderin by the observer without performing
theadditional stains necessary to identify ceroid.'Thus, less
extensive collections of ceroid rep-resenting developing or
intermediate formsmay be overlooked. In our cases both pigmentswere
present but the abundance of ceroid andthe sheets of macrophages
prompted in-vestigation.
Al-Nafussi et al5 described a case of ceroidgranuloma of the
cervix similar to that of ourcase 1 and postulated that tampon
usage res-ulted in a chronic ulcer with embedded fibrefragments
which lead to the accumulation ofceroid. As the patient had had her
last men-
1058
on June 8, 2021 by guest. Protected by copyright.
http://jcp.bmj.com
/J C
lin Pathol: first published as 10.1136/jcp.48.11.1057 on 1 N
ovember 1995. D
ownloaded from
http://jcp.bmj.com/
Ceroid granulomas in the female genital system
strual period eight years before presentation,this seems
somewhat tenuous and the authorsthemselves suggest endometriosis as
an al-ternative underlying factor.A ceroid granuloma ofthe female
peritoneum
was described by White and Chan,9 who sug-gested that it
represented an involutional phe-nomenon in a case of ectopic
decidual changeinvolving the pelvic peritoneum in a pregnantwoman.
They surmised that decidual cells maybe engulfed by macrophages and
provide thesubstrate for ceroid formation. Although nohistory of
endometriosis was given, it is in-teresting to note that
pre-existing peritonealendometriosis may undergo decidual
changeduring pregnancy. Ishizaki4 also postulated thatdegeneration
and necrosis in trophoblastic anddecidual cells in placentas may
result in ceroidproduction. In our second case the lesionsoccurred
in the setting of established en-dometriosis, and in the ovary the
sheets ofpigment-laden macrophages merged with anobvious deposit of
endometriosis.
Case 3 did not appear to be related to en-dometriosis but
clearly the presence of anecrotic, poorly differentiated,
endometrial ad-enocarcinoma would provide ample substratefor ceroid
production in the form of cellulardebris. It is interesting that
the granulomasdeveloped in the cervix even though the tumourwas
located in the uterine fundus. The reasonfor this is unclear.The
lesions we have described bear many
similarities to the necrotic pseudoxantho-matous nodules of
ovary and peritoneum de-scribed by Clement et al,'0 which they
ascribedto an unusual end stage manifestation of en-dometriosis.
The cases presented here, how-ever, lacked the well circumscribed
nodularity,central necrosis and hyalinised collagen de-scribed by
Clement et al. Moreover, the granu-lomas in our series were
intimately related to,or merged with, endometriotic foci while
thenecrotic nodules predominantly occurred else-where in the
peritoneum close to endometrioticfoci. The presence of both ceroid
and haemo-siderin within sheets of macrophages, the agedistribution
(older women between 40 and 72years of age in their series) and the
association
with endometriosis are similar, and suggest thatthe two lesions
represent different stages in theevolution of endometriosis.
Ceroid granulomas may mimic other lesions,both macroscopically
and histologically, de-pending on their location. In the cervix,
en-dometrium and ovary they can be distinguishedfrom malakoplakia
by the absence ofMichaelis-Guttman bodies, from decidual change by
theclinical setting and by morphological and stain-ing
characteristics, and from endometriosis bythe lack of glands
surrounded by endometrialstroma. Endometrial stromal foam cells
arepresent in endometrial hyperplasia or car-cinoma and lack the
staining profile of ceroid.Infections (including mycobacteria)
should beconsidered and further stains including Ziehl-Neelsen
performed if clinically warranted. Inthe peritoneum the
differential diagnoses in-clude malakoplakia, ectopic decidual
reaction,infection, necrotic pseudoxanthomatous nod-ules and
peritoneal melanosis, whether benignor associated with malignant
melanoma.
We are grateful to Dr Kevin Bendall for referring case 2.
1 Lillie RD, Fullmer HM. In: Histopathologic, technic
andpractical histochemistry. 4th edn. New York: McGraw
Hill,1976:519-21.
2 Bancroft JD, Stevens A (eds). In: Theory and practice
ofhistological techniques. 2nd edn. Edinburgh: Churchill
Liv-ingstone, 1983:250-2.
3 Amazon K, Rywlin AM. Ceroid granulomas of the gallbladder. Am
J Clin Pathol 1980;73:123-7.
4 Ishizaki Y Ceroid in the placenta. With special reference
tomeconium staining thereof Am J Obstet Gynecol 1960;80:245-51.
5 Al-Nafussi AI, Hughes D, Rebello G. Ceroid granuloma ofthe
uterine cervix. Histopathology 1992;21:282-4.
6 Shintaku M, Sasaki M, Baba Y Ceroid-containing his-tiocytic
granuloma of the endometrium. Histopathology1991;18:169-72.
7 Reagan JW. Ceroid pigment in the human ovary. Am JObstet
Gynecol 1950;59:433-6.
8 Clement PB. Pathology ofendometriosis. PatholAnnu
1990;25:245-83.
9 White J, Chan Y-F. Lipofuscinosis peritonei associated
withpregnancy-related ectopic decidua. Histopathology
1994;25:83-5.
10 Clement PB, Young RH, Scully RE. Necrotic pseudo-xanthomatous
nodules of ovary and peritoneum in en-dometriosis. Am J Surg Pathol
1988;12:390-7.
11 Thomas W Jr, Sadegheih B, Fresco R, Rubenstone AI,Stepto RC,
Carasso B. Malacoplakia of the endometrium,a probable cause of
postmenopausal bleeding. Am J ClinPathol 1978;69:5637-41.
12 Fechner RE, Bossart MI, Spjut HJ. Ultrastructure of
en-dometrial stromal foam cells. Am 7 Clin Pathol
1979;72:628-33.
J Clin Pathol 1995;48:1059-1061
Pseudoangiosarcomatous carcinoma of thegenitourinary tract
Department ofHistopathology,Bolton GeneralHospital,Bolton
Correspondence to:Dr M A Pitt,Department ofHistopathology,Royal
Preston Hospital,Sharoe Green Lane North,Fulwood, Preston PR2
4HT.Accepted for publication30 May 1995
M A Pitt, G Morphopoulos, S Wells, D L Bisset
Abstract year old woman, the vulvectomy specimenTwo cases of
pseudoangiosarcomatous contained an irregular ulcerated
tumour,carcinoma ofthe genitourinary tract, aris- infiltrating the
left labia and extending intoing in the vulva in one and the
bladder in the clitoris. In case 2, a 59 year old woman,the other,
are presented. In case 1, an 84 the excised bladder showed diffuse
thick-
1059
on June 8, 2021 by guest. Protected by copyright.
http://jcp.bmj.com
/J C
lin Pathol: first published as 10.1136/jcp.48.11.1057 on 1 N
ovember 1995. D
ownloaded from
http://jcp.bmj.com/
