THE DIAGNOSIS AND TREATMENT OFTUBERCULOUS CERVICAL ADENITIS

By BASIL ARMSTRONG, M.B., B.S.Medical Superintendent of The Royal Sea Bathing Hospital, Margate

Tuberculous glands of the neck may bedivided clinically and pathologically into threemain groups, the first two of which arecommon, the third rather infrequent.

(a) The first is that in which the glandularenlargement is consecutive to a tuberculous in-fection of the tonsils or nasopharynx, morerarely to an infected tooth socket, to a localizedtuberculous infection of the skin of the face(lupus) or to a tuberculous lesion of the eye orear.

In these cases the appropriate gland, drain-ing the structure primarily involved, firstenlarges, and from it the infection spreads toother glands in the cervical chain, but ingeneral in a downward direction, though lateralspread may occur even to the extent of in-volving glands on the opposite side of the neck.Tonsillar and nasopharyngeal infections are byfar the commonest, and in their case the glandfirst involved is situated high up in the cervicalchain just behind the angle of the jaw. Insevere cases, the spread may be such as toinvolve the whole length of the deep cervicalchain as well as submaxillary, submental andeven occipital glands, but the important featureis that it takes place mainly in a downwarddirection.

It is frequently stated that the tuberclebacillus can pass through an intact mucousmembrane without producing a lesion at itspoint of entry and can reach and involve agland draining the region concerned, and thereset up a primary tuberculous infection. Thisstatement is most often made in regard to theinfection of cervical and mesenteric glands.

I do not believe it to be true, but believe that asufficiently careful search would reveal theprimary infection to be at the point of entry,be it tonsil or intestinal mucosa. The in-fection is therefore exactly on all fours withthe primary complex of lung and hilarinfection.

(b) In the second important group, thelymphatic spread is from below upwards, thefirst glands of the neck observed to be involvedbeing usually in the supraclavicular region.The involvement is consecutive to an extensivemediastinal gland tuberculosis, in its turnresulting from a lung infection of variableintensity, and may ultimately attack a largenumber of glands in the neck, but the spreadtakes place mainly from below upwards, andthe glands which are visible and palpable are,it must be remembered-and this is an im-portant point in deciding upon a line oftreatment-probably only a small proportionof the whole number affected. They are, as itwere, the apples on the uppermost branches ofthe tree, the lower branches of which are con-cealed from sight, but are doubtless also heavilyladen with fruit. (Fig. i.)

(c) A third type of tuberculous cervicalglands probably results from a haematogenousspread. These cases, which are somewhat un-common are characterized by the rapid and al-most simultaneous involvement of a largenumber of glands throughout both sides of theneck and of the axillae likewise. The glandstend to be large, fleshy, less prone to becomeadherent to their fellows and less liable tocaseate and soften than those of the two pre-

;87by copyright.

on 11 October 2018 by guest. P

rotectedhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.23.260.287 on 1 June 1947. Dow

nloaded from

POST-GRADUATE MEDICAL JOURNAL

ceding groups-in fact, they approximate tothe type of glands seen in Hodgkin's diseaseand may be difficult to distinguish from themwithout a biopsy.

Tuberculous glands occur mainly in childrenof all ages and in adolescents, although no ageis immune. The greatest frequency is probablyin children between the ages of five and ten,and falls steadily as age increases, becomingrelatively uncommon after the age of 20. Thelarger proportion of cases falling into the firstgroup-i.e. those due in the main to tonsillarand nasopharyngeal infections-occurs inchildren. The second and third groups, whichare necessarily associated with a mediastinalgland tuberculosis of some chronicity, aremore likely to be seen in adolescents and youngadults. Males and females are about equallyaffected.The infection may be either by the bovine

or human strain of the bacillus. In the firstgroup, bovine infection probably pre-dominates, though to a -less degree than isusually stated. Of the two other groups, theinfection is much the most frequently by thehuman strain of the bacillus.

Diagnosis of Tuberculous Adenitis of the neck.The diagnosis is made largely on the history ofa slowly occurring, usually painless, enlarge-ment of the glands of the neck. To differ-entiate between the cases falling into the twocommon groups, it is important, where thedisease is already widespread throughout theneck, to discover whether enlarged glandswere first noticed in the upper cervical groupor in the supraclavicular region; in earlier andless extensive cases, the question presen'ts nodifficulty.

In the descending type of infection, thedisease is usually apyrexial, or almost so, andthe general health little affected; in theascending type, constitutional symptoms aremore likely to be present, and there may beevening pyrexia of I00° to 1020 with pallor,loss of weight and anaemia.

In the class, in which an extensive haemato-genous involvement occurs, the general dis-turbance is again likely to be severe.The glands will. present a different appear-

ance according to the stage at which they arefirst seen.

In the early stages they are usually some-

what fleshy and elastic, and not adherent to oneanother or to surrounding structures. Later,however, periadenitis occurs, the earlier in-fected glands tend to coalesce into a con-siderable mass, many of them becoming firmand tending to caseate, while the more recentlyinfected glands, lying beyond the periphery ofthe mass, are still discrete, soft and onlymoderately enlarged.

Later the caseous glands may break downand abscess formation result. The abscessmay be superficial ab initio, as where a super-ficial gland is concerned, or a deeply seatedabscess may send a prolongation to the surface,which burrows its way through the muscle andfascia and eventually becomes subcutaneous-the so-called ' collar stud' abscess. As theabscess approaches the surface, the skinbecomes thin, distended and bluish-red incolour, and frequently ruptures, with thedevelopment of a sinus, sometimes compliCatedby the occurrence of a surrounding scro-fuloderma.

In the case of cervical adenitis secondary toa mediastinal infection, a radiograph will oftenshow "enlarged mediastinal glands, with orwithout a recognizable pulmonary lesion.

Differential Diagnosis(i) Simple adenitis due to oral or skip sepsis,

as in the presence of a focus of sepsis in thetonsils or teeth, of impetigo or of pediculosisof the scalp is usually easily recognized. Itruns a more acute course, with fever, pain andsome general malaise. The glands usuallyrecede when the primary focus of infection istreated, but may soften, point and dischargepus.

(2) Streptococcal adenitis, due to a strep-tococcal tonsillitis, has an acute onset with sorethroat and high temperature and considerableconstitutional disturbance. The glands arehard and tender at first. They usually resolvewith the disappearance of the causal infection,but may suppurate or may remain swollen fora long time, in which case, differentiation froma tuberculous adenitis may be difficult. How-ever, the history of an acute onset and atuberculin test, if negative, will help inarriving at a correct conclusion.

(3) Hodgkin's Disease (Lymphadeonoma) hasa gradual onset, and glands in other parts of the

288 7utze I1947by copyright.

on 11 October 2018 by guest. P

rotectedhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.23.260.287 on 1 June 1947. Dow

nloaded from

ARMSTRONG: Tuberculous Cervical Adenitis

Fig. i. Tuberculosis of Mediastinal Glands.

C1

)eune I 947 :Z9by copyright.

on 11 October 2018 by guest. P

rotectedhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.23.260.287 on 1 June 1947. Dow

nloaded from

290 POST-GRADUATE MEDICAL JOURNAL June I947

F..

Fig. 2. TIpe of neck splint described in text.

by copyright. on 11 O

ctober 2018 by guest. Protected

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.23.260.287 on 1 June 1947. D

ownloaded from

ARMSTRONG: Tuberculous Cervical Adenitis

Fig. 5. A bad scar, the presence of which was advertised by stitch marks due totoo late removal of sutures, lying in front above the level of the- natural crease(marked by ink line). As a result of this the infra-mandibular branch? of thecervical division of the facial nerve was damaged, producing a permanent weakness

of the muscles of the corner of the mouth as shown in the photograph.

_7une- 4 291by copyright.

on 11 October 2018 by guest. P

rotectedhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.23.260.287 on 1 June 1947. Dow

nloaded from

POST-GRADUATE MIEDICAL JOURNAL

Fig. 6. Permanent weakness of the muscles of the left corner of the mouth resultingfrom damage at the time of operation to the infra-mandibular branch of thecervical division of the Facial Nerve. The placing of the anterior extremitv of theincision is at too high a level-in this case at about 3- in. above the upper natural

crease.

7tine 1 947292by copyright.

on 11 October 2018 by guest. P

rotectedhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.23.260.287 on 1 June 1947. Dow

nloaded from

ARMSTRONG: Tuberculous Cervical Adenitis

body may be enlarged also. In its later stages,splenic enlargement may be present, and thereis associated anaemia with increased cosinophilcount and often some degree of leucocytosis.The glands are large, discrete, elastic and non-tender. They do not tend to fibrose or caseote.The condition of tuberculous glandular in-fection described above as being of probablyhaematogenous origin, may closely resembleHodgkin's disease. The tuberculin reaction,if negative, may help in diagnosis, but theremoval of a gland for biopsy will often benecessary.

(4) Lymphatic Leukaemia shows a picture ofsevere illness with widespread lymphatic en-largement, high fever, splenic enlargement anda characteristic blood picture.

(i ) Lymphosarcoma-a condition very rarein children-may suggest a tuberculous in-fection, but its rapidly progressive course willusually suggest the advisability of the per-formance of a biopsy.

TreatmentMany cases of tuberculous cervical adenitis,

particularly those coming under treatment intheir early stages, will make a spontaneousrecovery. The enlarged glands, which havenot gone on to caesation, diminish in size andmay cease to be palpable, while some largecaseous glands may, in the course of time,become encapsulated and fibrose and shrink tosuch an extent as no longer to constitute anyserious danger. Such glands may eventuallycalcify completely or in part. But the processwhere caseous glands are concerned is slowand uncertain, and for a long time a dangerwill persist of their softening and formingabscesses with a possibility of rupture and aformation of ugly scars.

In general, all cases of whatever type should,at first, be given a course of from three to sixmonths of constitutional treatment.

In the cases falling into the first group-i.e.those in which the glandular involvement issecondary to an infection of the tonsils ornasopharynx, of the eye (e.g. tuberculous con-junctivitis), of the ear (e.g. tuberculous otitismedia or externa), or of the skin, the appro-priate treatment of the primary focus is ofparamount importance.

I have no hesitation in saying that, in the

case of tonsillar or nasopharyngeal infection,the tonsils and adenoids should be removed assoon as the patient's general condition permits,the only exception being where a glandularabscess is present or is threatening, in whichcase it may be well to postpone tonsillectomyuntil the abscess has been successfully dealtwith.

Constitutional treatment should be carriedout, whenever possible, at the seaside or in thecountry. If possible the patient should beabsolutely at rest and confined to bed. Helio-therapy and open-air treatment is of greatvalue; the diet should be generous; extramilk should be given, and in many cases theadministration of halibut or cod liver oil willbe beneficial.The use of a splint to restrict movement of

the head and neck is helpful in all cases,particularly in disease of the upper cervicalglands.A simple splint, easily made by anyone

possession a small amount of mechanical skill,is illustrated. It consists, in the main, of astrip of duralumin sheet, running up verticallybehind the upper thorax and neck, some 5 in.or more wide behind the thorax, and 2 in. ormore wide behind the neck, and extendingfrom waist to occiput This is bent to fit thecurves of the upper thorax, the back of theneck and head, and is reinforced by means ofa strip of duralumin bar riveted to it. At itsupper end it carries a 5 in. strip of duralumin,extending forwards on either side of the h-adin the form of a horseshoe, encircling theposterior two-thirds of the head. The anteriorone-third is closed by means of a leather straprunning across the forehead, and another strapruns from side to side below the chin. Thesplint is anchored to the thorax by means of,padded axillary straps and a wide webbingband around the waist. A thin layer of whitefelt is glued to the surface of the splint incontact with the body and head, and the wholeis completed by a covering of unbleachedcalico. (Fig. 2.)Thus treated, many cases of cervical

adenitis will improve to such an extent as tomake the consideration of an operative removalof the diseased glands unnecessary.A proportion will remain, however, in which

the size of the glands and their chronicity will

yutne I 1947 293by copyright.

on 11 October 2018 by guest. P

rotectedhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.23.260.287 on 1 June 1947. Dow

nloaded from

POST-GRADUATE MNEDICAL JOURNAL

make it unlikely that they will heal and absorbspontaneously within a reasonable time, and ofthese cases those falling into the first groupshould be treated by the radical excision of allthe enlarged glands. By this means, a com-plete cure of the condition may be obtainedwith the reasonable prospect of insignificantdisfigurement of the neck, if any at all.At this point, it is worth while to discuss the

desirability or otherwise of operation upon thecases falling into groups (b) and (c), viz., thosewhich have either an ascending lymphatic in-fection from a group of intrathoracictuberculous glands, and those in which theglandular involvement is due to a haemicspread. Both groups are bad subjects torextensive surgery, particularly the latter.

In the former group, it is likely that glands,here and there, will break down and formabscesses. In general, these should be treatedas conservatively as possible. The widest ex-cision practicable will only remove a smallproportion of the diseased glands, the operationis likely to be severe and will perhaps besufficient to upset the resistance of the patientwhich we have been building up laboriously bymonths of constitutional treatment. Treatthese cases, then, by aspiration, where possible,of their abscesses, and failing this by a smalltransverse incision with the evacuation of theircontents and light curetting.

In parenthesis, it may be said that theaspiration of tuberculous abscesses of the ne-kis less successful than in the case of coldabscesses in most other situations. The pus isalmost always very thick and a very large-bored needle must be used and successiveaspirations tend to become more and moredifficult, the lumen of the needle being almostinvariably blocked by tough fibro-caseousdebris. However, the manoeuvre is alwaysworth attempting in the case of large deep-seated abscesses, in the hope that one or twoaspirations will cause them to dry up; wherethe abscess has become superficial and the skinis stretched and discoloured, incision andevacuation is the best line of treatment, for inthese circumstances even if the skin does notbreak down spontaneously a lumpy brownishsemi-keloid scar results which is more con-spicuous than the scar of the small incision.

If the skin over a superficial abscess breaksdown and sloughs, as is not infrequently thecase, this is not a major disaster. The removalof the contents of the deeper part of the abscesswith a curette will usually be followed by theproduction of a flat ulcer with a base coveredwith short healthy granulations to which asmall Thiersch graft may readily be caused toadhere and an excellent cosmetic resultobtained.

In the case of the large fleshy glands result-ing from a haemic spread, the temptation tooperate must be sternly resisted-a tempta-tion which is all the stronger because, in suchcases, there is little tendency for the glands tocaseate or to become adherent and their re-moval, therefore, is a matter of great simplicity.In the few cases, however, where, in myexperience, operation has been performed, ithas been poorly tolerated and an early andextensive recurrence has resulted. Moreover,in this class of case, given adequate andsufficiently prolonged constitutional treatment,the tendency for spontaneous absorption andresolution to occur is very marked.

Reverting, then, to the first group of cases,those in which the infection may be termed adescending one, it has been seen that, after areasonable period of conservative treatment hasbeen given, there will still be a proportion ofcases in which large caseous glands remain andtequire operative removal. One advantage ofa preliminary period of rest and constitutionaltreatment is that, in addition to the buildingup of the patient's general resistance, the peri-adenitis which causes the glands to becomematted together and adherent to surroundingstructures will have largely disappeared, andthe operation is thereby rendered easier, andwhat is perhaps more important, any smallseed-like glands which may be missed atoperation are much less likely to enlargesubsequently, and to cause recurrences, thanif surgical intervention had been undertaken,prematurely. The only exception to the ruleof waiting for from three to six months is inthe case of glands which are breaking down toform abscesses. The difficulty, already men-tioned, of treating . these abscesses con-servatively makes it worth while to operateearly to avoid a disfiguring scar, even at aslightly greater risk of a recurrence.

Yune I1947by copyright.

on 11 October 2018 by guest. P

rotectedhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.23.260.287 on 1 June 1947. Dow

nloaded from

ARMSTRONG: Tuberculous Cervical Adenitis

OperationThe object of operation in these cases is to

achieve the fullest possible removal of diseasedglands, with the smallest amount of disfigure-ment of the patient.The success of the operation will be judged

-and quite rightly-by the patient and his orher relatives, but the inconspicuousness orotherwise of the scar, and it is worth whilegoing to some extra trouble to obtain an almostinvisible scar.

It should not be necessary to point out thatthe incision should never run in the long axisof the neck parallel to the border of thesternomastoid muscle, as the resulting scarwill always stretch widely and become veryobvious, but it is insufficient to say that itshould run transversely and parallel to thenatural creases of the.neck. It must actuallycoincide most meticulously with one of thesenatural creases of which there are usually twoin the upper neck in the positions shown inthe diagram. (Fig. 3.)

It will usually be found that one or other ofthese creases is conveniently placed in respectto the main mass of glands to be removed. Ifthe glands are relatively few and situated highup under the anterior margin of the sterno-mastoid muscle, the upper crease may bechosen, but if they are numerous and extendfor some distance down the course of theinternal jugular vein, and particularly if theylie partly or largely in the posterior triangle ofthe neck, making it likely that the posteriorborder of the sternomastoid will have to befreed from its sheath to facilitate their removal,then the line of the lower of these creasesaffords the best access and offers two additionaladvantages, the first that the inframandibularbranch of the cervical division of the facialnerve supplying the muscles of the corner ofthe mouth is less liable to be injured, the secondthat it is possible through this incision to clearaway glands for a considerable distance downthe reck; this may obviate the necessity ofmaking a second incision except in the case ofglands lying low down in the supraclavicularregion. The disadvantage of using the lowercrease is that it makes the freeing of the upperextremity of the gland mass, which may extendup almost to the base of the skull, more difficultand increases the risk of a troublesome

Fig. 3. The natural creases of the neck.

haemorrhage through the tearing of a smallbranch of the internal jugular in this veryinaccessible situation.

If it is necessary to use a second incisionlow down in the neck it must, of course, followthe rule already laid down and must not onlyrun transversely but must coincide closelywith a natural crease.As it is usually very difficult, particularly

when the patient is anaesthetized, with theneck placed somewhat on the stretch, and inthe presence of a large mass of glands, todefine accurately the line of the natural creasesof the neck, my own practice is always tospend a few minutes on the day proceeding theoperation, in marking out, first with ink andthen by needle scratch the exact line to befollowed in incising the skin. At the sametime three or four scratch marks are made atintervals, running at right angles to the line ofthe proposed incision. If these are made 24

Yune I1947 295by copyright.

on 11 October 2018 by guest. P

rotectedhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.23.260.287 on 1 June 1947. Dow

nloaded from

POST-GRADUATE MEDICAL JOURNAL

'.i

Fig. 4. The author's pattern of Self-Retaining Retractor. It consists of a metal ring, I4 ins. in diameter, held inposition over the wound by a flexible metal arm to which it is attached by a universal joint. The flexible arm in itsturn is fastened by a clamp to the edge of the table. The ring is made in three sections for ease of sterilization.The retractors, which are of various sizes, are attached through short coiled springs to lengths of ' ladder ' chain,

which can be hooked in appropriate positions over small blunt studs placed on the periphery of the metal ring.They can, of course, be used to retract muscles as well as skin edges.

hours before operation, they are much moreeasily recognized when it comes to sewing upthe skin, than if made as the operation starts,and accurate appositior. to the wound edges isgreatly facilitated.The anaesthetic should preferably be ad-

ministered intratracheally. The patient lieson his back with his head turned to the soundside, and a small pillow is placed behind theshoulders and lower neck.The incision is made in the predetermined

line and must be of such a length as to allowgood access to the region to be exposed, andno longer. It should not extend backwardsbehind the posterior border of the sterno-mastoid muscle further than is absolutelynecessary, for in this region, scars easily tendto assume a slightly keloid character, andremain somewhat raised and reddened.

The incision extends through the skin andunderlying fat. If the platysma is welldeveloped, it should not be divided in thisincision, but should be preserved and latersplit longitudinally in the course of its fibres,close to the anterior border of the sterno-mastoid. It can then be sutured with a fewinterrupted catgut stitches at the end of theoperation, a procedure which diminishes therisk of a sunken scar after the removal of alarge gland tumour. This depressed scar,when it occurs, is very conspicuous and maymar an otherwise excellent cosmetic result.

The skin flaps are undermined widely in alldirections; the external jugular vein isdivided between ligatures. Sinuses should notbe excised but should be divided sub-cutaneously in the undermining of the skinflaps. A self retaining retractor to hold back

- Yu-e 1947296by copyright.

on 11 October 2018 by guest. P

rotectedhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.23.260.287 on 1 June 1947. Dow

nloaded from

ARMSTRONG: Tuberculous Cervical Adenitis

the skin edges is very helpful and liberates bothhands of the assistant for more useful purposes.

The platysma muscle having been splitlongitudinally, the anterior border of thesterno mastoid muscle is exposed over the fullextent of the swelling formed by the glandmass and is retracted backwards. At thisstage, it is wise to find and identify the spinalaccessory nerve as it enters the deep surface ofthe muscle' in order to preserve it fromsubsequent injury.The enlarged glands will now be exposed

and should be seized by suitable forceps andpulled backwards and the fascial tissue ontheir deep surface divided near the lower endof the operation area, where the internal jugularis most easily exposed. Dissection of thegland mass is most easily carried out by meansof curved Mayo's scissors and swab pressure,and it the plare of dissection is kept as close Pspossible to the glands little difficulty is ex-perienced in separating them from the vein, thetributaries of which should be clamped andtied as encountered.

The lower end of the mass being freed,dissection is continued upwards until thespinal accessory nerve is encountered. Inorder to clean and preserve it, division of thegland mass may be necessary. Keeping con-stantly in the areolar tissue plane immediatelyadjacent to the glands themselves and usingscissor and gauze diss ,ction, the uppermostmember of the chain will eventually be reachedand the pedicle above this clamped and dividedand the mass removed.

'In those cases where the glands extend intothe posterior triangle, it will be necessary, ashas already been said, to free the posterioredge of the sternomastoid muscle, and herethe spinal accessory nerve is in greater danger,as it lies more superficially. However, bykeeping close to the glands themselves, and byexercising sufficient care and patience, it shouldbe possible to avoid injury to this and to anyother important structure.

The removal of the glands being completed,the most careful attention must`be given tohaemostasis, all divided vessels being ligaturedwith fine catgut. A twisted silkworm gut orfine rubber drain is inserted into the deadspace under the sternomastoid muscle and isbrought out at the anterior angle of the wound,unless a conveniently placed sinus had existedtowards the front of the lower flap, when itsorifice may advantageously be used for thispurpose. This blood drain may be removed atthe end of 24 hours. Haemostasis secured, theplatysma is united by fine interrupted catgutsutures, and the skin' closed with ophthalmicsilkworm gut.The greatest care must be exercised in

securing perfect apposition of the skin edges.Childe's approximating forceps used by anassistant, thus allowing the surgeon the use ofboth hands in tying the sutures, is an excellentinstrument in securing this end. The suturesmust be numerous and must be placed closetogether-at a distance of not more than iin. apart.

Closed in this way, the skin wound will besufficiently soundly healed to allow the re-moval of the stitches' on the third or fourthda'y. Left longer than this, stitch scars willresult and nothing is more unsatisfactory thanto have the presence of an almost invisiblescar advertised by a double line of ugly whitedots. (Figs. 5 and 6.)The best dressing after operation is a strip

of i in. gauze ribbon soaked in Mastisol orFriar's Balsam and laid lengthwise over thewound. This is covered by layers' of gauze andwool, held in position by a figure-of-eightbandage round the neck and skull. The gauzeribbon dressing, which allows the removal ofthe blood drain without disturbance, should berenewed atter the stitches are taken out.A neck splint may, with advantage, be worn

for three or four weeks atter the operation.Convalescence should be of several months

dutation, preferably spent at the seaside or inthe countrv.

3¢une I1947 297by copyright.

on 11 October 2018 by guest. P

rotectedhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.23.260.287 on 1 June 1947. Dow

nloaded from