Granulomatous hepatitis in association with fever and a palmar rash

Secondary syphilis may present without the classical featu res of the disease and should be considered in patients acutely ill with abnormal liver function tests and a nonspecific pattern on liver biopsy. Secondary syphilis may present in association with granulomatous hepatitis and serological testing should be performed if the etiology of the hepatic granulomas is unclear. While the finding of a disproportionate elevation of alkaline phosphatase relative to serum bilirubin may be a useful pointer towards the diagnosis this pattern is not specific and not consistently observed. The definitive answer to the question of whether syphilitic hepatitis exists awaits the development of comprehensive rests to exclude viral pathogens. Can J Gastroenterol 1987;1( 1):37-40

red to hospital with headache and fever.She had been well until two weeks prior to admission when she experi• e nced rhe onset of bifrontal headache, mild photophobia and intermittent nausea.The patient had consulted her family doctor one week before admission and metronidazole was prescribed for symptoms of vaginal discharge thought to be due to trichomonas infection.Three days before admission, fever, rigors and generalized joint stiffness were noted .There was no history of sore throat, skin .' Figure 1) Left Mono1111clear in/iltrace in 1he porr.alrrncc.sR ight Noncascating granuloma rash or arthritis.She denied the use of intravenous drugs and had never received a blood transfusion.The patient lived alone having been recently separated from her husband.S he drank alcohol rarely and in minimal amounts.The patient was caking thyroxine because of hypothyroidism resulting from a previous partial thyroid resection for hyperthyroidism l5 years previously.There was a history of pneumonia in the distant past and a hysterectomy and bila eral salpingo-oophoreccomy had been performed eight years previously for symptoms of pelvic pain; no specific diagnosis was made.Functional inquiry was unremarkable except for a long history of occasional left lower quadrant abdominal pain.
Physical examination revealed temperature of 39°C.irregular pulse of 80 beats/min and blood pressu re l 30/90 mm Hg.The patient was orientated and ale rt but appeared ill.There was no evidence of skin rash.Examination of the mouth , head and neck was unremarkable.The re was no evidence of lymphadenopathy.No meni ngeal signs were elicited and fundoscopy revealed normal optic d iscs.The chest was clear to auscu ltation and heart sounds were normal with no audible murmurs.The liver and spleen were not enlarged.The muscles were not tender and there was no evidence of arth ritis.Vaginal examination was not performed.
A chest x-ray showed borderli ne ca rdiomegaly but the lung fie lds were clear.An electrocardiogram showed sin us rhyth m with ventricular bigeminy and trigeminy.An echocardiographic examination of the heart showed no evidence of valvular lesions.
An ultrasonographic examination of the liver showed no dilatation of the intra-or extrahepatic bile ducts and no focal abnormalities.An inrravenouscholangiogram was attempted but visualization of the bile ducts was not obtained.On the 29th d ay of the illness the characte ristic rash of secon dary sy philis appeared (Figure 2) and su bseq uently the FT A-ABS a nd VDRL were reported to be positive.Although the patient had denied a history of casu al sexual contact she now admitted to a brief e ncounter with a male who had a penile wart o r ulce r.High dose intravenous penici llin was started and con tinu ed fo r IO days.
The fever, rash and jo int pain hecame worse on th e fi rst day of treatment (J a risch -Herx heimcr reactio n) but su bseque nt resolution o f the clinical and laboratory abnormalities occurred.Figure 3 shows the resolution o f the liver function abnormalities.The VDRL titre.which was initially positive at 1/256, fell to :, negative titre of 1/64.

DISCUSSION
The patient presented with a history suggestive o f ,in acu te infectious or inflammatory process.Physical examination and exte nsivc radio logical investigation did not reveal an ob vious sou rce of sepsis and cultures were negative.The most striking labo ratory fi ndings were th e a bnorm;il liver function tests and the d isproportion:i tcly high a lkaline phosphatase relative to scrum bi lirubin.Hepatic ultrasound showed no ev idence of biliary obstruction and a subsequent li ver biopsy showed portal infiltration.a lobu lar hepatitis a n d numerous granulomas.
Hepatic infiltration with granu lomas may occur in a va rie ty of illnesses ( 1,2) but in this setti ng tuberculosis seemed a strong possibi lity.However, the chest was clear a nd the marrow rcve;iled no evid ence of mycohacteria.Sarcoidosis was considered but the absence of hilar lym ph adenopathy and arthritis argued against this.Hepatic grnnu lomas have been reported in associ:ition wirh connective tissue diseases such as lupus crythematosis ( 3) but the an tinu clear antibodies were negative O n day 29 o f the patient's illness a rash appeared with the rather characteristic distribution o f seconda ry syphilis.Syphilis had nor been consid ered on presentati on because of the absence of mucous membrane lesions, lymphad cnopathy or a rash .Furthermore, the patien t initially denied a history of casual sexual contact.
The high titre VDRL and FTA-ABS, the Jarisch-Herxhcimer reaction and a Gronulomolous hepatitis suhscqucnt fall in titre afte r treatment with penici llin confirmed the clinica l diagnm is.Further su pport for the diagnosis of secondary syphilis was given by the patient's revel:itio n of casual scxu:, I contact with a male who had a penile lesion .
Cli nically evident hepatic in vo lvement in second a ry syphi lis is not common.Hahn (4).in:, remarkable study of an estimated I 0,000 cases o f early syphilis, reported that jaundice occurred in only 0.05qo.H owever. b iochemical evidence of hepatic invo lve ment.as in thc present case, has been reported to occur in IO to 91 % of patients with thc disease (5.6).
A disproportionate elevation of alkaline phosphatase rcl:itive to the serum bilirubin is occasionally observed in this condition (7).A similm biochemical pa ttern may be seen in granulomatous hepatitis ( 8) hut in nei ther conditio n is this a consistent finding (5,6).The reasons fo r this biochemica l pattern arc unclea r.
Even if hepatic in volve ment is present the biopsy findings can be rather nonspecific and very variable .T he problem of id entifying a specific h istological patte rn is compounded by the possibil ity of si multaneo us infection with vira l agents in a population at risk for both syph ilis an d viral hepatitis Nor in frequently an ulcerating chancre may be present in the rectum (7).It is possible that at least some of the nonspecific altera tions in hepatic morphology may be related to th is.Spiroche tes. as in th is case.are rarely found in the liver but this may represent diffic ul ties in meth odology.Ir ha s been postulated that the hepatic damngc is mediared by immune complexes (9).
The portal areas arc almost invariably infiltrated hy mononuclear cells and occasionally by polymorphonuclcar leukocytes.Pericholangitis may be observed (7) but cholcstasis is uncommon .The hepatic inflamm;ition may predominate around the central veins occasionally in association with perivcnu lar fibrosis ( 5).Kupffer cell proliferation may be prominen t (6) .Hcparocellu la r necrosis between portal areas and cen tral ve in has been reported ( 10).Granulo matous infiltration , as in this patient, may occur ( 11) b u t this is not com mon.The most consisten t pattern of parenchymal involvement appears to he foca l collections of mononuclear cells wi th little hepatocellular necrosis.
T he wide variation in clinical, biochemical and histological features, and the failure to exclude other pathogens has led to a continuing scepticism as co whether the hepatic damage observed in secondary syphilis is related to the spirochetes (12).In the present case, alcohol intake was minim al, metronidazole has not been reported to cause hepatic granulomas and screening for hepatitis B was negative.We did not test for hepatitis A but a slide test fo r Epstein-Barr virus was negative.We cannot exclude the possibility of non-A, non-B hepatitis but neither the biochemical nor the histologic features were particularly suggestive of a viral etiology.
The resolution of abnormal liver function tests began before treatment was started.This has previously been reported (13) and may be illustrative of the tendency for secondary syphilis to resolve to a latent phase.
The Jarisch-Herxheimer reaction is a featu re of the treatment of syphilis.In the case reported by Young et al ( 13) the Jarisch-Hl\rxheimer reaction was accompanied by both clinical and biochemical exacerbation .A deterioration in liver enzymes was not seen in the present case.

ACKNOWLEDGEMENTS:
The authors acknowledge the assistance of Drs P Fernandez, I Bowmer and S Chittal in the presentation of this manuscript.
L'hepatite syphilitique: Existe-t-il ou non ?L E SECOND DEGRE DE LA SYPHILIS PEUT SE PRESENTER SANS LES PARTICULARITES classiques de la maladie et chez les patients severement malades avec une fonction anormale des rests du foie et un modele non precis de la biopsie du foie, la diagnose de la syphilis devrait etre consideree.Le second degre de la syphilis peut se presenter en association avec l'hepatite gran ulomas et le test serologique devrait etre execute si l'eriologie de l'hepatique granulomas n'est pas claire.Quoique la constatation disproportion nee de l'elevation de l'alcaline phosphatase relatif au serum bilirubine serait un ind ice profitable vers le diagnostique, ce modele n'est pas specifique et pas toujours observe regulierement.La rcponse definitive a la question si l'hepatite syphilirique existe ou non attend le dcveloppement des tests comprehensifs pour exclure tour virus qui sont susceptibles de donner l'hepatite.