TUBERCULOSIS SYPHILISTUBERCULOSIS AND SYPHILIS from an extraneous source, cannot be denied. I have observed it in the form of a verrucose tuberculosis of the fingers in old standing

II

TUBERCULOSIS AND SYPHILIS

By H. C. G. SEMON

THAT a man who has syphilis cannot contract a lueticprimary chancre is self-evident. An identical stipulationapplies to tuberculosis, and as nearly everybody is mildlyinfected in infancy or early childhood, explains theextreme rarity of the primary tuberculous chancre. Indirect contrast, therefore, to what pertains in syphilis, itis exceptional for the T.B. chancre to occur in the adult.Certainly I have never seen such a case myself, and whenthe possibility is suspected it should always be subjectedto the most stringent criticism.An example that fulfilled all criteria was published

some years ago in the Dermatologische Wochenschrift. Ayoung woman, a domestic servant, who was " walkingout " with a sailor developed a sore on the chin close tothe labial commissure, about three weeks after herfiance had returned to his ship. Tests for lues havingproved negative, both in the exuding serum and in thehomolateral adenitis which was associated with theprimary lesion, Koch's bacillus was envisaged as a possiblefactor and recovered in scrapings. Subsequently theglands softened and broke down and tubercle bacilli werefound in the secretion. To complete the clinical picture,a patch of lupus vulgaris later made its appearance in theneighbourhood of the chancre, and the source of theinfection seemed proven, when the young man reportedsick on his next leave and was found to be suffering fromlaryngeal tuberculosis. It is interesting to note that theincubation period of three weeks in this case coincidesvery well with that observed in cases of tuberculosis ofthe frenulni and penis in infants, following ritual circum-cision, in which the operator-a phthisical rabbi-wasknown to have staunched, or attempted to staunch,haemorrhage by oral suction.Another recorded site for the primary lesion is the

vulva in the young child when propulsory sliding effortsare being made on a floor previously contaminated by

I59

on April 24, 2020 by guest. P

rotected by copyright.http://sti.bm

j.com/

Br J V

ener Dis: first published as 10.1136/sti.15.3.159 on 1 July 1939. D

ownloaded from

http://sti.bmj.com/

BRITISH JOURNAL OF VENEREAL DISEASES

infected sputum. The propensity of infants and youngchildren to put toys in their mouths accounts for theoccasional infections reported in this locality, and thepossibility of the so-called " tuberculous primary com-plex " should always be borne in mind when a sore andglandular enlargement of some weeks' duration is pre-sented for an opinion. The " chancre " differs from thatof lues in several particulars. The infiltration is softerand the edges are flat and somewhat overhanging, whilethe base has a rather granular appearance and tends tobleed more easily. The infected glands may ba hard atfirst, but rarely of that rubbery consistency so typical ofthe venereal adenitis. They may be painless and discreettoo at an early stage, but they soon become adherent toeach other, soften, and break down with the characteristicinvolvement of the skin. Typical apple-jelly nodules oflupus may develop both in the neighbourhocd of the soreor close to a sinus discharging pus from. caseous glands.It is a sine qua non that no case could be regarded asproved without the microscopic demonstration of thetubercle bacillus, just as no modem syphilologist wouldignore the proof that the dark-ground demonstration ofthe treponema can afford.The difference in the pathogenesis of the two diseases

would seem, then, if traced back to ultimate causes, todepend on the fact that tuberculosis in the large majorityof cases is an air-borne infection and can invade the lungseither directly or through the bronchial lymphatics andglands, and in the food through the alimentary tract-two portals which, because it is so much more vulnerable,and never air-borne, are permanently closed to the lueticinfection. That Koch's bacillus most usually enters byone or other is now generally accepted. It must be sirri-larly conceded that the ensuing infection of the viscerais at first silent or latent in so far as clinical symptoms areconcerned, but demonstrable and commonplace in view ofthe frequency with which skin tests prove positive inapparently healthy individuals. From one aspect thissilent infection must be regarded as salutary, for withoutit, considering how universally distributed the fomitesare, primary cutaneous tuberculosis with its more danger-ous consequences would be more common than it is. Thatnevertheless a tuberculous subject can contract a tuber-culous skin affection, either from his own infection or

i6o



j.com/

Br J V


ownloaded from

http://sti.bmj.com/


from an extraneous source, cannot be denied. I haveobserved it in the form of a verrucose tuberculosis of thefingers in old standing pulmonary infections, probablyfrom a contaminated handkerchief, and it is a sign of evilaugury when a lung case develops a peri-anal " super-infection" (probably from swallowed sputum), as suchexamples are rightly called.Jadassohn regards lupus vulgaris as an invariable

super-infection, i.e., that it can only occur in a personalready tuberculous, and it is demonstrable that the dis-tinction is applicable to other tuberculous manifestationson the skin. In syphilis the term appears to be used in amore exact sense in relation to the coincidence of two ormore primary chancres. The local spread or extensionof a tuberculous process is mainly by way of the lym-phatics, and in this respect would seem to differ from anotherwise analogous syphilitic lesion. An interestingexample of this discrepancy is afforded by what usuallyhappens when the alae nasi are affected. In lupus theinfection is at first localised to the nasal mucosa, butsecondary nodules may and often do appear on the skinof the cheeks in the form of lupus patches at any timesubsequently. In the analogous tertiary syphilitic lesionthe treponemata, if they produce a gummatous infiltrationand necrosis, spready by contiguity and are not localisedto the lymphatic path. Hence we never see isolatedgummatous patches on the cheeks as a secondary exten-sion from a nasal focus of the disease. But this is not tosay that cutaneous tuberculosis is not blood-borne.Miliary tuberculosis comes under this heading, andalthough it rarely attacks the skin in the adult, can cer-tainly produce embolism of the skin in infants and youngchildren. Post-exanthematic lupus in young children isanother example of undoubted circulatory transport. Itnot infrequently occurs in the form of numerous widelydisseminated patches on the trunk after recovery frommeasles, or some acute specific fever, and is considered byVolk, Mibelli and v. Pirquet to be the result of a tem-porary failure of antibody production during the febrileperiod. The latter observer found that the previouslypositive Mantou test became negative during that time,but was again obtainable when the patient had recovered.The observation is of valpe and goes to support the con-tention by these authors that during the measles infection

i6i



j.com/

Br J V


ownloaded from

http://sti.bmj.com/


the child is temporarily deprived of the T.B. antibodyprotection stored up in previous months or years. It isinteresting to compare it with what is supposed to happenwhen we infect an early G.P.I. with malaria. Here wespeak of a mobilisation of antibody production to explainthe therapeutic effect and the conversion of a positiveW.R. to a negative. By and large it would appear thatwhile intercurrent febrile reactions are harmful in chronictuberculosis, they are undoubtedly, and in given circum-stances, useful and even indicated in the treatment ofotherwise intractable syphilitic infection. The effect ofintercurrent disease on these two infections is to someextent paralleled by their reactions to pregnancy. It isnotorious that a woman with pulmonary tuberculosis runsconsiderable risks during the periods of pregnancy andlactation, although the chances are that she will give birthto a healthy child. The issues are reversed in the case ofsyphilis, for in this disease active symptoms eitherinvolute or are entirely suppressed, a positive W.R. fre-quently becomes temporarily negative, while the child maybe born with or may shortly develop manifest symptomsof congenital lues. These are facts which no one disputes,but their true explanation is likely to remain obscure andthe subject of theory and discussion for years to come.The clinical differentiation of the cutaneous manifesta-

tions of the two infections, which is my thesis in thispaper, should of course be based on the underlying histo-pathology. But nowadays it is commonly omitted, eitherbecause few of us are competent in the finer details oftheir comparative reactions, i.e., their histology, orbecause the W.R. is a much more rapid, and at least asreliable, discriminatory test.

Speaking broadly, the histological differentiation oftubercular from tertiary syphilitic lesions is based on thefact that tubercles due to Koch's bacillus are- more fre-quently multiple, while the gumma is usually a solitaryinfiltration with a strong tendency to necrose in its centre.Giant cells may occur in, both, but in tuberculosis theyhave a more typical appearance and arrangement, mostcharacteristically seen in lupus vulgaris, which, oncememorised, could hardly be mistaken for any othergranuloma. The lupus nodule is a collection of tubercles,composed for the most part of so called epithelioid cellsaggregated in multiple groups, in which not infrequently

I62



j.com/

Br J V


ownloaded from

http://sti.bmj.com/


actual tubercle bacilli, the cause of their proliferation, canbe stained and demonstrated. The giant cells of suchtubercles contain nuclei arranged like a horseshoe in theperiphery, in contra-distinction to those seen in syphilisand other granulomatous infiltrations in which no suchuniformity is demonstrable. The coalescence of severalsmall groups of tubercles to form one or two large onesaffords an explanation of the relatively common tuber-culous lesion with a polycyclic border with multiple pointsof caseation, as against the gumma which has a round oroval form and a tendency to central necrosis. The verylarge number of tuberculides-which include T.B. verru-cosis, scrofuloderma, miliary T.B. cutis, lichen scrofulo-sorum, papulonecrotic tuberculides, Darier's and Boeck'ssarcoids, angiolupoid and many other less determinedforms-preclude a detailed account of their variousdeviations from the typical tubercle. On the whole,however, it may be asserted that the resemblances to thetrained eye are distinct and rarely liable to confusion withthe histology of clinically similar syphilides.These considerations embolden me to tackle my task

in a rather unorthodox manner, and I propose to discussthe subject according to the regional areas which may beinvolved by the manifestations of the two diseases in lateor early stages of their incidence.

(i) THE SCALP

If we except lupus erythematosus, which as you mayknow is thought to be of tuberculous origin by someauthorities, tuberculosis manifests itself but rarely in thissituation, so rarely that a case of lupus vulgaris capitis isa curiosity. Gumma of the scalp, on the other hand,would be a commonplace but for salvarsan. Its ravagesbefore the war not infrequently involved the outer tableof the skull, and patients sometimes succumbed to aseptic meningitis. We are rarely on debatable groundmoreover when we are consulted for loss of hair. Themoth-eaten appearances of the luetic secondary stage,with or without roseolar rash, are never seen in diffusefall of hair due to tuberculosis. I remember being con-culted for this symptom by a former house physician whodied some months later of acute tuberculosis involvingboth her lungs, infection from which had doubtless been

I63



j.com/

Br J V


ownloaded from

http://sti.bmj.com/


the cause, entirely unsuspected at the time, of heralopecia. The cicatricial localised types of follicularinfection are rarely attributable to either disease, andalthough it is always advisable to examine the W.R. inchronic cases, I have yet to see a case in which theoetiology was syphilitic.

It is hardly likely that lupus erythematosus of thescalp could be mistaken for any luetic condition. As onthe face, it is characterised by atrophy, scaling and hyper-keratosis at the mouths of the follicles affected. The slowand inevitable destruction of the follicles leads eventuallyto permanent baldness of a strictly localised type. In nocase does L.E. affect the bone, and I have never seenfrank ulceration or the crusting which a tertiary syphiliswould cause.

(2) THE FACE, INCLUDING THE EARS AND NOSE

It is in this area that we are most often confronted byinteresting problems of diagnosis. Speaking generallyas regards tuberculous manifestations, young personsusually of the female sex are more prone to them thanmales. They are insidious in their onset and very chronicin their course. Lupus vulgaris is of course by far thecommonest example of tuberculous infection in this area.It usually begins in childhood, and females are more fre-quently affected than males. An insidious onset andextreme chronicity, as compared with tertiary syphilis,are the rule. The latter will do more damage in a monththan lupus in a year and may involve bony structures,which L.V. never does, although on the nose and earscartilaginous structures eventually succumb. The clas-sical " apple-jelly " nodule, when clearly demonstrableon vitropressure, is a diagnostic feature which practicallysettles the nature of the lesion, but it is not alwaysapparent. Cases of lupus" tumidus" and other aberrantvarieties including the "sarcoids occur in whichdiascopy fails us, and in these we are thrown back on thehistology and W.R. findings. I recently favoured adiagnosis of " sarcoid " (Schaumann), in a middle agedman, with subcutaneous nodular infiltration over the leftside of the temporal and malar regions. But the W.R.was strongly positive, and the lesions cleared up rapidlyon the usual antiluetic therapy.

164



j.com/

Br J V


ownloaded from

http://sti.bmj.com/


The common tendencv of the lupus nodule or tubercleto recur in scars is a feature of great clinical value in thedifferential diagnosis from tertiary syphilis, and onewhich, in my opinion, overrides any other objectivesymptom when doubt arises. Recurrences in general aremuch more frequent in cutaneous tuberculosis than theyare in its luetic imitators.That cutaneous gummata could be mistaken for scro-

fuloderma is not outside my experience. The illustrationI am circulating is an instance in point. The patient wasa woman in the early forties who some weeks after theremoval of alleged tuberculous glands in the neck-thequadrilateral incision is well depicted-developed infiltra-tions of a dusky colour mainly at the margins, and at onespot, just external to the thyroid cartilage, a smallcircular ulcer. The W.R. was found to be strongly posi-tive and would have saved the patient an unnecessaryoperation if it had been performed when she was firstseen.

Scrofuloderma is more apt to occur on the neck thanon the face, for the simple reason that it is nearly alwayssecondary to some underlying hypodermic tuberculousinfection, most commonly glandular in origin. Cervicaltuberculous adenitis is still unfortunately an all toocommon occurrence in childhood, and its co-existence orrecent history afford the diagnostic clue in nearly everycase of so-called scrofuloderma. When something re-sembling it occurs for the first time in the adult a pre-ceeding W.R. should always anti-date any surgical pro-cedure. In rare instances scrofuloderma, sinus formationand an ascending tuberculous lymphangitis may occur onthe legs. As in the cervical type, children are morecommonly the victims than adults.A somewhat rare manifestation of lupoid type is the

so-called lupus miliaris (or follicularis) facei. It usuallyoccurs in women in the third decade and may be confusedwith rosacea in which, in addition to erythema, brownishpapules persist on vitro-pressure. That confusion with apapular or miliary-papular syphilide might arise is con-ceivable, but outside my experience. In the latterdisease the papules are not likely to be confined to theface. On the contrary, the chest and back are the sitesof election, and the papules, distributed in clusters orgroups, would not display the symmetrical arrangement

I65



j.com/

Br J V


ownloaded from

http://sti.bmj.com/


of the tubercular lesions. Scarring in the form of smallpits would tend to occur in both, but more marked insyphilis, in which, too, pigment disturbances are more aptto be associated. Laboratory assistance other than theW.R. should rarely be required.

Ulcerations of the lips, gums and buccal mucosa are anoccasional source of difficulty, but this is rare nowadaysowing to .the Wassermann test and the rapid healinginduced by one or two salvarsan injections in the venerealcases. Clinically they differ considerably too, for intuberculosis the ulcer is usually painful and has a moreeroded and frayed appearance than the clean cut orpunched out characters of syphilitic destruction.

T.B. TONGUE ULCERATION AND GLOSSITISVery rarely tuberculous ulcers can also be punched

out. I have under my care at the present time an elderlyspinster who has attended the Royal Northern Hospitalfor the past twelve years for a circular indolent ulceration(rivalling that of the rodent), involving the lower lip andangle of the mouth symmetrically. If I could show herto you I doubt if anyone here would hesitate to label thedeep sunken cicatrices syphilitic, yet the W.R. has alwaysbeen negative, and in the course of the disease shesuddenly developed a tuberculous iritis which respondedexceedingly well to treatment by tuberculin injections,which also benefited the cutaneous manifestations.

FOREARMS AND HANDSThe commonest tuberculous manifestation here is the

verrucose tuberculide, and both in its evolution andappearance it is so characteristic that I have yet to see acase in which confusion with syphilis might arise. But itis otherwise with palmar lesions. Relatively common inthe latter disease, the manifestations are scaly andcircinate in type without much tendency to ulcerate,unless fissuring can be described under that heading.Tuberculides are exceedingly rare in this situation,although the verrucose variety affects the dorsum of thehands and knuckles, and is characterised by hyperkera-tosis more often than by the development of the apple-jelly tubercle. When the warty character prevails on the

i66



j.com/

Br J V


ownloaded from

http://sti.bmj.com/


palms or soles it is much more likely to be syphilitic thantuberculous.

So far as I am aware the nail substance is not affectedby T.B. directly, while luetic ungual dystrophies andeven congenital absence of the nails were by no meansuncommon before the days of pre-natal anti-luetic treat-ment.

THE TRUNKDifficulty may sometimes arise in diagnosing between

tuberculides such as lichen scrofulosorum or the papulo-necrotic tuberculide and the large and small papularsyphilides, and it must be remembered that the formermay conceivably occur in children with a positiveW.R. The syphilides are usually universal and maybe associated with lesions of the mucous membranes.They are split-pea in size, smooth and rotund, of abrownish-red colour and more often pigmented thantheir tuberculous analogues. The small papular, acumin-ate or miliary papular syphilide is follicular in localisationwith a mark-ed tendency to grouping-the corymboseconfiguration, which is highly characteristic of theiraetiology. There are three main varieties of the tuber-culides to be differentiated: the miliary acute cutaneous,the lichenoid, and the papulo-necrotic. The first namedmay be part of a generalised tuberculosis, with a hightemperature and a fatal issue. Measles or some otherexanthem may have preceded it, and the cutaneous lesionsare pinhead to small pea-sized nodules, sometimes groupedin clusters, which, if the child survives, usually breakdown and ulcerate. Lichen scrofulosorum is also a diseaseof childhood. It is relatively benign and symmetricallydistributed over the trunk and consists of small shinypapules, frequently aggregated in patches, which may bescaly and associated commonly with keratosis of thefollicles. This feature not infrequently gives rise todifficulty in the diagnosis from lichen planus. The latter,however, is always intensely pruritic and never ends inpitting or other evidence of scarring. The papulo-necrotic tuberculide is perhaps as frequently seen on theextensor surfaces of the limbs as on the trunk. Hereagain adults are rarely affected. The lesions are solitary,like the gumma, and not symmetrical as a rule. They aresmall and subcutaneous in origin and take a week or

V.D. I67 N



j.com/

Br J V


ownloaded from

http://sti.bmj.com/


longer to rise to the epidermic level where they becomeinflamed and undergo necrosis and ulceration, which isinevitably followed by deep pitting as a more or lesspermanent relic in all cases. Exacerbations and relapsesover some years is a common history, and these criteria,taken as a whole, are so uniform that the chances of con-fusion with papular syphildes or small gummata areunlikely. Errors of diagnosis between these two sub-cutaneous lesions do occasionally occur, however, andmust have been more common before the W.R. test wasavailable.

It is highly probable that N.A.B. injections, which aresometimes given with good results in the demonstrablytuberculous nodules, were first introduced under the mis-apprehension that they were subcutaneous gummata.

Special consideration of the effects of syphilis as com-pared with tuberculosis in the lower extremities is scarcelynecessary, for there is no essential difference in type towhat is seen on the hands. The only exception we mustconsider is erythema induratum (Bazin).Erythema induratum, or Bazin's disease, is worthy of

separate consideration in view of its relative frequency inout-patient clinics. It is a disease of young women andstarts as symmetrical hypodermic nodules in the lowerthird of the legs. An antecedent tuberculous history,familial or personal, is not invariably obtained; but theMantoux reaction in my experience is practically alwayspositive, sometimes strongly so, and seems to me to pre-sent difficulties to recent attempts to explain the aetiologyon circulatory deficiencies. Frequently the nodules tendto ulcerate, and when they do so the question of a lueticorigin may arise. But such ulcers are characteristicallyseen in older women and then usually in the neighbour-hood of the knee joint. They occur in the late or tertiarystages of the disease and are therefore not symmetrical,while the punched out, clean-cut edges of the cutaneousgummatous ulcer are never observed.

Histologically, Bazin's disease is mainly a fat necrosisand it is this feature that has suggested to ProfessorTelford and other workers the possibility of a circulatoryaetiology, spastic in type, and amenable to surgicalmeasures such as sympathectomy. Whatever the aetio-logy, it is undeniable that very striking cures have beenattained by surgery. I am happy to record one such in

i68



j.com/

Br J V


ownloaded from

http://sti.bmj.com/


my own series. The patient, an unmarried woman oftwenty-six, had suffered from Bazin's disease, associatedwith Raynaud's syndrome and chilblains, in the upperextremities for five years. No treatment had had mucheffect and she was almost crippled by her symptoms. Apre-ganglionic sympathectomy, both cervicle and lumbar,was successfully carried out by Mr. Lawrence Abel, andrestoration of a normal circulation with immediate heal-ing of her ulceration and cure of the Raynaud symptomsresulted within a week.

ANO-GENITALIt is in this region that difficulties in diagnosis most

often arise, not so much between syphilis and tuber-culosis, but in the differentiation of luetic from otherulcers and non-ulcerative infiltrations and granulomata,mostly contracted in coitus, either in Europe or thetropics. With these latter we are not concerned in thispaper, and it is fortunate that tuberculosis here is rela-tively rare, at any rate, in the healthy adult. I lhave yetto see a primary tuberculous chancre for the reasons Ihave tried to explain in my introductory remarks. But ifan infant is brought to you with a penile chancre orvulvar ulceration, the aetiology may conceivably betuberculous. I would go further and say that it is morelikely to be tuberculous than syphilitic.Lupus vulgaris is rare on the genitalia but is by no means

uncommon round the anus and on the buttocks, mostly inchildren, although of course- it may have persisted sincechildhood into adult life. Destructive progress is notmarked, far less so than on the face. Its differentiationfrom tertiary syphilis rarely offers any difficulty to thosefamiliar with the clean cut, circinate lesions of cutaneousgummata. In contradistinction, the edges of a tuberculousulcer are sloping or shelving, the base raised or nearlylevel with the rest of the surface and covered usuallyby easily bleeding, unhealthy granulations. In the neigh-bourhood can often be seen numbers of the characteristic" apple-jelly " nodules which are best demonstratedby vitropressure. Peri-anal sinuses are more likely to betuberculous than luetic. But even more characteristic,practically diagnostic, is the tendency to local recurrencesin scar tissue-a feature not observed in syphilis and one

I69 N 2



j.com/

Br J V


ownloaded from

http://sti.bmj.com/


in which the phenomena of local immunisation must beclosely concerned. Cicatrisation in tuberculous cases ismore likely to be followed by deformities and atresia, dueto contraction of scar tissue, than in the parallel syphiliticinvolvement, in which specific therapy invariably leadsto rapid healing with the formation of innocuous supplecicatrices.

This article and also the article Pulmonary Tuberculosisand Syphilis are based on addresses delivered before theMedical Society for the Study of Venereal Diseases,April 28th, I939.

I70



j.com/

Br J V


ownloaded from

http://sti.bmj.com/

TUBERCULOSIS SYPHILISTUBERCULOSIS AND SYPHILIS from an extraneous source, cannot be denied. I have observed it in the form of a verrucose tuberculosis of the fingers in old standing

Documents