Pathology of 'non-healing (midline) granulomain which a clinical diagnosis of'midline granuloma' or...

J. clin. Path., 1977, 30, 317-327

Pathology of 'non-healing (midline) granuloma"L. MICHAELS AND MONICA M. GREGORY

From the Department ofPathology, Institute ofLaryngology and Otology, University ofLondon

SUMMARY In a histological study of biopsy and postmortem material from 30 cases of nasal diseasein which a clinical diagnosis of 'midline granuloma' or Wegener's granuloma had been given, we

selected 10 cases on the basis of the presence of widespread coagulative necrosis and atypical cells.Evidence is provided that such changes represent a malignant neoplasm of histiocytic lymphomatype. Local invasion and spread to cervical and more distant lymph nodes, spleen, liver, and kidneywere seen in some of the cases. Erythrophagocytic activity was marked in the spleen in three cases

and histiocytic infiltration of the bone marrow in two cases, indicating a more generalised activityof histiocytic cells. Terms such as 'malignant granuloma' should be abandoned. In obstructive andulcerating conditions of the nose efforts should be concentrated on making an accurate histologicaldiagnosis.

A condition variously referred to as non-healinggranuloma, midlinegranuloma, malignant granuloma,granuloma gangrenescens or Stewart's granuloma ischaracterised by a relentlessly progressive ulcerationof the nose and adjacent structures.

Stewart (1933) described 10 such cases in detail,and his paper has been regarded as a classic sourceon the subject. While credit must be given to Stewartfor drawing attention to a clinical condition ofrelentless ulceration of the nose, his acceptance of itas a granulomatous one does not seem to have beenbased on adequate pathological criteria. In at leasttwo of the cases the histology was thought to bearsome resemblance to the 'very atypical spheroidalcarcinomas seen in this region'. In one of the lattercases it is stated that multiple metastases to thelungs were probably present.Friedmann (1955, 1963) was impressed by the

similarity of 'non-healing granuloma' to Wegener'sgranuloma. He suggested that 'non-healing granu-loma' had two forms: (1) Stewart's type, withoutvasculitis, and (2) Wegener's type showing vasculitis.Intermediate forms existed between the two types.This concept and indeed the whole existence of'midline granuloma' as a specific entity has not beenuniversally accepted. Burston (1959) showed thatmore specific entities could often be diagnosed bypersistent investigation in cases initially given the'Paper presented to the Pathological Society of GreatBritain and Ireland, Dublin, July 1975, and as a demon-stration to the same Society, London, January 1976

Received for publication 28 September 1976

designation of midline granuloma. Eichel andMabery (1968) felt that this condition did notactually exist as a specific disease process. Harrison(1974) found that by consideration only of theclinical aspects of the case he was able to decide ona mode of therapy which produced satisfactoryresults. He treated patients with progressive localulceration of the nose as neoplastic by irradiation,while those showing less local but more systemicmanifestations were more likely to be in the Wegener'sgroup and were managed with corticosteroids andazathioprine.The histological findings were not considered by

Harrison to be helpful. The failure of histopatho-logical examination to identify a lesion which in itsbehaviour is neoplastic is rare in modern medicalpractice. In fact, because of the close correlationbetween histology and clinical behaviour, identifi-cation by histological examination has become offundamental importance in the diagnosis andtreatment of most patients with suspected neoplasms.Because of this discrepancy between the cancerousbehaviour of some of the lesions and the apparentlyinflammatory histological appearance of biopsiesfrom them, we investigated histological materialfrom 30 cases, most of them used in the clinical studyof Harrison (1974), with the object of seeking ahistological basis for the lesions which had beenidentified clinically as locally invasive.

Material and methods

Histological sections were examined from 30 cases317

copyright. on S

eptember 2, 2021 by guest. P

rotected byhttp://jcp.bm

j.com/

J Clin P

athol: first published as 10.1136/jcp.30.4.317 on 1 April 1977. D

ownloaded from

http://jcp.bmj.com/

318

with destructive lesions of the nasal passages, in 19of whom a clinical diagnosis of 'malignant (non-healing) granuloma' had been made. In seven cases

tissue from some internal organs was also availablefor histological study; in six cases these were obtainedat necropsy and in one case (case 10, Table 1) spleen,liver, and abdominal lymph nodes had been removedsurgically and were available for study.

Sections from all blocks of tissue were stained withhaematoxylin and eosin (H and E). In addition,sections of all nasal tissues, lymph nodes, andselected other organs were stained by the periodicacid Schiff method, Gordon and Sweet's stain forreticulin, and the Unna-Pappenheim (methyl green-

pyronin) methods. A preliminary survey ofH and Estained sections from all biopsies of the nasalpassages in these cases was carried out withoutrecourse to the clinical history.

Nasal biopsy material from four cases (cases 1, 4,5, and 6) was examined electron microscopically.Difficulty was experienced in each case in obtainingareas in which cells were sufficiently well preservedfor adequate ultrastructural observation to be made,but a few suitable areas were obtained. Nasal biopsymaterial was fixed in 3% glutaraldehyde for electronmicroscopy, washed in Sorensen's buffer, and em-bedded in Araldite. The sections were examinedwith a Siemens Elmiskop II microscope.

Results

PRELIMINARY OBSERVATIONSIn the preliminary examination of H and E stainedsections from biopsies of the nasal passages of all30 cases, 17 cases were found to show inflammatorychanges only. These changes included tuberculoidgranulomatous inflammation as well as non-specificchronic inflammation. Vasculitis was present insome. All 11 cases of clinical Wegener's granulomawere found in this group. Since a study of thehistological aspects of Wegener's granuloma was

not an aim of this work, these cases will not befurther referred to.

In three nasal biopsies the histological appearanceswere in the category of lymphoma. In these casesan initial clinical diagnosis of 'midline granuloma'had been made, and earlier biopsies had shown no

neoplasm, but subsequent biopsies had revealeddefinite lymphoma classified as histiocyticlymphoma,well-differentiated lymphocytic lymphosarcoma, andplasmacytoma respectively. These three cases willnot be further referred to. The remaining 10 casesall showed in one or more of the nasal biopsies acombination of widespread necrosis and groups ofatypical cells, the cytological appearances of whichsuggested a malignant origin (Figs. 1 and 2). We

L. Michaels and Monica M. Gregory

believe that this combination of histological changes,which we will refer to as NACE (necrosis withatypical cellular exudate) represents the micro-scopical counterpart of the 'lethal' form of theclinical condition described as 'malignant granuloma',and the rest of this report will deal with the featuresof the 10 cases showing this change.

PATHOLOGICAL FEATURES OF CASES WITHNACEThe atypical cells in cases showing NACE were alittle larger than mononuclear inflammatory cells(approximately 15-20 microns) with nuclei thatvaried from round to irregularly elongated and bent,often with irregular protuberances and an irregularlystaining nuclear membrane. The chromatin materialwas rather scanty and irregularly distributed.Nucleoli were often prominent. Binucleate formswere present and occasional mitotic figures were seenin each case. Cytoplasm varied from scanty toabundant and the cytoplasmic membrane wasindefinite. Phagocytised basophilic debris wasfrequently present in the cytoplasm of the cells (Figs.2 and 3). A fine reticular framework honeycombedthe cells individually (Figs. 4 and 5). Zones composedofsuch cells varied from very small areastoprominentportions of the available biopsy material. A variableamount of chronic inflammatory cellular infiltration,mainly lymphocytes and plasma cells, was alsopresent. Histiocytes of normal appearance were seenin most cases and in several cases formed unbrokengroups which occupied several low-power fields. Theatypical cells described above formed homogeneousgroups in about half the cases but in the other caseswere interspersed as varying numbers of isolatedcells among plasma cells, lymphocytes, and histio-cytes.

Necrosis was always prominent and was usuallyof the coagulative type. It was seen as a spotty changebetween the tumour cells and also forming large,eosinophilic, acellular areas which often compriseda prominent feature of the biopsy material. In thelatter areas a fine honeycomb of reticulin was presentwhich was identical in appearance with that of theatypical cellular areas (Figs. 4 and 5).NACE was seen in some biopsy material to be

occupying a mucosal position beneath epitheliumand in the vicinity of seromucinous glands, but moreoften it was without such anatomical localisation,there being no epithelial tissue in the biopsy.

There was evidence of invasion of adjacentstructures by the NACE tissue. In four cases it hadinvaded nerve sheaths (Fig. 6). In two cases invasionof skeletal muscle was present, and in one of these(case 5) the masseter muscle was infiltrated by palegrey tissue with the appearance of NACE histo-

copyright. on S



j.com/

J Clin P


ownloaded from

http://jcp.bmj.com/

Pathology of'non-healing (midline) granuloma'

Fig. 1 Nasal biopsyfrom a case of'midlinegranuloma'. On the leftis a cellular area.Towards the right thisgives way to a zonecomposed largely ofcoagulation necrosis.Haematoxylin andeosin (HandE) x 110

Fig. 2 High-powerview ofsquare markedin Fig. 1. Note irregu-larity ofmany ofthecell nuclei ('atypicalcells'). Some typicallymphocytes andplasmacells are also present.Towards the right thereis coagulation necrosis.HandE x 355

;......>

V4..

319

copyright. on S



j.com/

J Clin P


ownloaded from

http://jcp.bmj.com/

L. Michaels and Monica M. Gregory'W-114w Mr.

"ik

la .e.

bal-.4oa 11 .,;,.-: -56-1 .,

AW.ir,L

Fig. 3 'Atypical cells'in nasal biopsyfromanother case of'midlinegranuloma'. HandEx 355

Fig. 4 Nasal biopsyfrom case of'midlinegranuloma'. Area ofNACE. Notepatchynecrosis at bottom rightandlamina ofbone atbottom left. HandEx 110

4.f .#;0

4t"c,\

* ..~ ,, - ----4-*,rF17:,._*-2p*s .a..* a_. eS z,,J . ........w *.* %.^ *;I-S* oX *u-' ~'s V

320

copyright. on S



j.com/

J Clin P


ownloaded from

http://jcp.bmj.com/

Pathology of 'non-healing (midline) granuloma'

Fig. 5 Sectionfromparaffin block adjacentto that ofFig. 4. Thereis afine reticulinpatternincluding the zone ofnecrosis. Gordon andSweet's reticulin stainx 185

Fig. 6 Nerve bundle inbiopsy ofnasal region incase of'midline granu-loma'. There is intenseinfiltration with 'atypicalcells'. HandE x 110

'hA."~~~~~~A

321

copyright. on S



j.com/

J Clin P


ownloaded from

http://jcp.bmj.com/


logically (Fig. 7). In a further two cases histologicalevidence of invasion of nasal bone by NACE waspresent (Fig. 8).

Vascular changes were evident in some of thecases. Two of these changes-fibrous thickening andfoam-cell infiltration of the intima of small arteries-were found only in previously irradiated cases. Intwo cases there was involvement of the walls ofsmall vessels by ACE (Fig. 9), and in five cases, twoof them not irradiated, fibrin thrombi were found insmall vessels within areas of NACE. In no case withNACE was there seen the necrotising vasculitischaracteristic of polyarteritis nodosa or Wegener'sgranuloma.

ELECTRON MICROSCOPICAL OBSERVATIONSElectron microscopical observation of the 'atypical

cells' in the cases with NACE showed peripherallyarranged masses of chromatin and prominentnucleoli in nuclei. Cytoplasm contained roughendoplasmic reticulum, occasional mitochondria,and frequently prominent phagosomal bodies(Fig. 10).

CLINICAL FEATURES OF CASES WITH NACEOn review of the clinical features of the 10 patientsselected by the presence of NACE without priorknowledge of the clinical history, it was found thatnine of the 10 were males and the ages of presentationranged from 23 to 72 years with a median age of 52.In all patients there had been a history of unilateralnasal obstruction, often for many years; in fivepatients submucous resection of the nasal septumhad been carried out on one or two occasions with

Fig. 7 Slices of left massetermuscle at necropsyfrom caseof 'midline granuloma' showingextensivefacial ulceration (case6). The muscle is severelyinfiltrated by pale grey tissue,which was histologicallycharacteristic ofNA CE.

322

copyright. on S



j.com/

J Clin P


ownloaded from

http://jcp.bmj.com/


Fig. 8 Nasal biopsyfrom case of'midlinegranuloma'. There is in-filtration ofbone byNACE tissue. Noteatypical cells in centralmarrow cavity andcoagulative necrosisoutside bone to right.HandE x 110

Fig. 9 Nasal biopsyfrom case of'midlinegranuloma'. There isinfiltration ofthe wall ofa small artery with'atypical cells'.HandE x 335

W g J .:'i .............. i....... ...SEX, ,---. 323

copyright. on S



j.com/

J Clin P


ownloaded from

http://jcp.bmj.com/


Fig. 10 Electron micrograph of'atypical cell'from nasal biopsy in a case of'midline granuloma' showing NACE.Note low chromatin content of nucleus andphagosomes in cytoplasm. Uranyl acetate and lead citrate x 10 000

little or no relief of the obstruction. In most of thepatients there was an accompanying unilateral orbilateral maxillary sinusitis for which, in two patients,a Caldwell-Luc operation had been performed. Ineach case the nasal obstruction had eventually givenway to ulceration of parts of the nasal cavity withvarying degrees of involvement of the nasal septum,turbinate bones, ethmoid sinuses, hard palate, andlateral wall of the nose. In case 1 there was extensionof the disease process to the orbit, with proptosis,and to the base of the skull. In another patient theulcerative process massively involved the anteriorwalls of the maxillary antra, the whole of the nose,the palate, and front of the mouth. Each patient wastreated with radiotherapy, but it is not possibleadequately to assess the effects of this treatment onthe course of the disease as the treatment was admin-istered in different centres in different dosages andat different stages of the disease. Four patients diedfive years, four years, three years, and six monthsrespectively after the onset of the disease. Death wasrelated to infection in three cases and massive

haemorrhage from iatrogenic duodenal ulcerationin one case (Table 1).

SYSTEMIC INVOLVEMENT IN CASES WITH NACEHistological evidence of involvement by a similar orrelated pathological process outside the nasal areawas seen in some of the cases. Such changes took thefollowing forms: (a) lymphomatous change in spleenand/or lymph nodes; (b) tumour-like deposits; (c)infiltration of the liver; (d) erythrophagocytosis;(e) histiocytic infiltration of bone marrow.

(a) Cervical lymph nodes at necropsy in cases 1, 2,and 3 and abdominal lymph nodes and spleen atlaparotomy in case 10 were enlarged on grossinspection. Histologically there was loss of normalarchitecture and replacement by atypical lymphoidcells similar in appearance to the cells describedabove in NACE. Multinucleate cells were common.Areas of necrosis were frequent. The changes werethose of lymphomatous involvement of lymph nodesand spleen.

(b) In one kidney in case 2, a 1-5 cm deposit vwas

324

copyright. on S



j.com/

J Clin P


ownloaded from

http://jcp.bmj.com/


Table 1 Features of10 cases with involvement ofnasalpassages by NACE

Case Age Sex Course ofdisease Length ofhistory Involvement

Nasalpassage Cervical Systemiclvmph node

1 23 M Erosion ofskull base. 6 mth Died Ethmoid, sphenoid + + Necropsy: SinusoidsMerningitis of liver

2 51 M Progressive destruction of nasal 5 yr Died Septum, nasopharynx, + Necropsy: Kidneytissues. Terminal Strep. pyogenes maxillary sinus depositsinusitis

3 55 F Progressive destruction ofnasal 5 yr Died Septum, maxillary sinus, + Necropsytissues. Radical removal R. nasopharynxmaxilla one month before death.Septicaemia

4 48 M No further progression after 3 yr A & W Septum, turbinatesirradiation

5 58 M No further progression after 3 yr A &W Maxillary sinus, - -irradiation ethmoids

6 53 M Massive progressive involvement 3 yr Died Whole nose and palate - Necropsyof nose and front skull. Deathfrom massive haemorrhage fromduodenal ulceration due tosteroid therapy

7 65 M No further progression after 2 yr A &W Septum, maxillary sinus -radiotherapy

8 50 M No further progression after 2 yr A &W Roofofnose -

radiotherapy9 72 M No further progression after I yr A &W Palate, maxillary sinus - -

radiotherapy10 41 M Pancytopenia and hepato- 9 mth A Palate, lateral wall - Laparotomy: Spleen,

solenomegaly during irradiation abdominal lymphtreatment ofnasal lesion nodes, bone marrow

A &W = alive and well

found in the cortex. Histologically this was composedof small cells similar to those of the NACE areas inthe nasal lesion. Large areas of necrosis were present.At the edges of this deposit there was infiltrationbetween the renal tubules by the abnormal cells.As mentioned above, the left masseter muscle in

case 6 showed grey infiltration which was composedhistologically of tissue with NACE appearance. Thismuscle was at the edge of the advancing zone ofmassive facial ulceration in this patient so that it ispossible that the tissue reached it by direct invasionfrom the nasal area rather than by some form ofsystemic spread.

(c) The sinusoids of the liver showed in case 1

infiltration by 'atypical' lymphoid cells, some ofthem multinucleate. In case 10 there was infiltrationof portal areas and parenchyma by groups of cellssimilar to those causing lymphomatous enlargementof abdominal lymph nodes and spleen.

(d) Case 10 presented with a destructive nasallesion with histological changes of NACE. Duringthe course of radiotherapy to the nose this patientdeveloped pancytopenia, and hepatosplenomegalywas detected. The spleen was removed and biopsiesof abdominal lymph nodes and liver were taken.These tissues all showed erythrophagocytosis as wellas the changes of malignant lymphoma and thecondition was classified as malignant histiocytosis

(histiocytic medullary reticulosis). This patient willbe presented in greater detail in a separate report.A study was made for the presence oferythrophago-cytosis in the other cases in which NACE had beendetected in the nasal biopsy. Liver, spleen, andlymph node sections were available in four othercases, and the degree of erythrophagocytosis seen inthese tissues is listed in Table 2. Two cases (1 and 3)showed what is listed as a 'moderate' degree oferythrophagocytosis, by which is meant to a lesserdegree than was present in the spleen of case 10 butnevertheless sufficiently prominent to be observed inmost high-power fields. Sections of liver and lymphnodes in case 1 showed what is listed in Table 2 asslight 'erythrophagocytosis', by which is meant thathistiocytes containing erythrocytes in their cytoplasmwere seen in occasional high-power fields.

Table 2 Erythrophagocytosis in liver, spleen, and lymphnodes in cases with NACE in nasal biopsy

Case Liver Spleen Lymph node

I Slight Moderate Slight2 Absent Absent Absent3 NA Moderate NA4 Absent Absent NA10 Slight Marked Marked

NA = not available for histological study

325

copyright. on S



j.com/

J Clin P


ownloaded from

http://jcp.bmj.com/


(e) In case 10, normal and abnormal histiocyticcells were seen in excessive numbers in the smears ofbone marrow. Some of these cells contained phago-cytised red cells. Similar changes were seen in thebone marrow of another case of 'midline granuloma'not included in this report, in which the nasal biopsyshowed changes of NACE (Kay, 1976).

Discussion

In a histological study of biopsies from 30 cases inwhich a clinical diagnosis of 'midline granuloma' orWegener's granuloma had been given,we selected 10cases on the basis of a common histological feature,the presence of widespread necrosis and atypicalcells (NACE). The cytological features of these cellswere such as to suggest a malignant nature by*criteria usually used by pathologists. It must beadmitted that such criteria are not always valid inhistological sections, eg, fibroblasts can show4malignant' features in organising granulation tissueswhere there is no neoplastic activity. The atypicalcells seen in these nasal biopsies, however, weresometimes arranged in broad groups with a finereticulin pattern surrounding individual cells. Asimilar reticulin pattern was present in the areas ofnecrosis which were also a feature of the biopsies ineach case. 'Atypical' cells, often accompanied byareas of necrosis, were seen infiltrating nerve, bone,muscle, and arterial walls in some of these cases.There was, therefore, histological evidence of aninvasive and destructive neoplasm composed of'atypical cells', ie, a malignant neoplasm, in these 10cases. The lesion not only appeared to be locallymalignant, but in several cases also showed evidenceof systemic dissemination. Cervical or abdominallymph nodes in four cases and spleen in one caseshowed not only the presence of malignant cells butobliteration of the normal architecture by thismalignant infiltration. In one case a deposit of thesame cells was present in the renal cortex, and in twocases there was infiltration of the liver.We believe that the malignant neoplasm in these

cases is a form of lymphoma. This view is based onthe appearance of the tumour cells, on the reticulinpattern, and on the mode of infiltration of lymphnodes, spleen, and liver. Classification of non-Hodgkin's lymphoma is at present in a state ofchange (Bennett et al., 1974; Dorfman, 1974;Gerard-Marchant et al., 1974). The appearances ofthe lymphoma cells as seen by light and electronmicroscopy would suggest that the neoplasm shouldbe classified as histiocytic lymphoma by the newerterminology. The feeble chromatin production andevidence of phagocytic activity in the rather abundantcytoplasm indicate an origin from histiocvtes.

The clinical behaviour of these neoplasms is inkeeping with a malignant lymphoma of rather slowactivity. The lesion would seem, in both histologicalfeatures and clinical activity, to be identical withthat described as 'polymorphic reticulosis' byEichel et al. (1966).

Erythrophagocytosis is frequently seen in normallymphoid tissue in small amounts and may be foundin increased amounts in haemolytic conditions(Smith, 1958). The presence of relatively largedegrees of erythrophagocytosis in three of five of thecases with NACE, in which the relevant tissues wereavailable for histological examination, indicates thatthe lymphomatous tumour, which we believe to bethe pathological basis of these cases, may be accom-panied by a generalised enhanced phagocytic activityof histiocytes. In case 10 this was extreme enoughto warrant a diagnosis of malignant histiocytosis. Inthis case normal and abnormal histiocytes wereabundant in the bone marrow smears, and thisfinding has been noted in another case of 'midlinegranuloma' showing NACE (Kay, 1976). Theseobservations would suggest that the neoplasms foundin these cases have a functional similarity and arelationship to normal histiocytes. It is of interestthat the cells in the cases of 'midline granuloma'described by de Faria et al. (1957) were induced tophagocytise saccharated iron oxide in vivo, as aresult of which the authors described the lesion as areticulohistiocytoma.We wish to emphasise that none of the 10 cases

showed classical lymphoma on nasal biopsy but thatthe conclusion that the histological change which wehave designated as NACE represents a malignantlymphoma is based on an analysis of all the findingsin these cases. Large areas of the nasal biopsy tissueshowed inflammatory changes only, and in somecases NACE tissue was found in only one of threeor more biopsies. The sparsity of malignant tissuetogether with the large amount of necrosis in thesecases accounts for the designation 'granuloma' thathas been given to this lesion from the time of itsfirst description, in spite of its behaviour as amalignant neoplasm. A further similarity to aninflammatory lesion is the small size of the tumourcells and their resemblance to inflammatory histio-cytes. It is noteworthy that although some necrosiswas present in the cervical lymph nodes and systemicdeposits of neoplasm, neither this nor inflammationwas ever as extensive as in the nasal biopsy material.Nasal tumours are particularly prone to secondaryinfection. Biopsies which have to be relatively smallin this situation are usually taken from superficialmucosa where secondary changes are likely to bemost marked. In some of our cases in whom thedesignation midline or malignant 'granuloma' had

326

copyright. on S



j.com/

J Clin P


ownloaded from

http://jcp.bmj.com/


been given, small doses of radiotherapy had beenused as treatment to the nose. This producedwidespread necrosis in the tissue and also vascularchanges which probably enhanced the necrosis.

It is possible that there does exist a disease inwhich there is a relentlessly progressive ulcerationconfined to the nose and surrounding area but inwhich repeated biopsies show inflammatory changesonly. We have not, however, found such a case, incompany with Burston (1959), Eichel and Mabery(1968), and McGuirt and Rose (1976). We would,therefore, suggest that the concept of 'midlinegranuloma' and its synonyms be abandoned. Ifbiopsy of the nose in a patient with a destructivelesion shows non-specific inflammatory changes only,particularly if these changes are accompanied byconsiderable necrosis, further biopsies should becarried out. The pathologist may then be satisfiedthat there is either a specific inflammatory change(eg, tuberculosis or sarcoid) or a histologicalappearance indicating a neoplastic condition, but wedo not feel that it is justified to diagnose a conditionwith malignant potential on the basis of non-specific chronic inflammatory changes only.The use of terms such as 'malignant granuloma'

for cases that are in the malignant neoplasticcategory and their classification together withWegener's granuloma have been, we believe, detri-mental to the treatment of the patient. On themistaken assumption that the process is a near-inflammatory or an autoimmune one, large doses ofsteroids have been given and radiotherapy is usedin doses that are not sufficient for a malignantlymphoma. We suggest that the mainstay of treat-ment for histiocytic lymphoma of the nose shouldbe adequate tumour doses of x rays. The cervicalarea should probably be irradiated also. We wouldfurther suggest that consideration be given to the useof surgical methods to remove as much as possibleof the neoplasm. Surgery is sometimes a veryeffective therapy for lymphoma, for example, in thegastrointestinal tract, and might be useful in thisarea where infiltration of bone by the tumourprotects against some of the destructive effect ofradiotherapy on the tumour cells. Cytotoxic drugs

might also play a useful part, particularly where thereis systemic involvement by the growth.We wish to thank Professor D. F. N. Harrison forhelp and encouragement in carrying out this histo-logical study based mainly on his clinical material.Dr H. E. M. Kay kindly gave help and advice on thebone marrow changes of case 10 and providedinformation on another case at the Royal MarsdenHospital. Miss Eileen Buckle helped with the type-script and Mr D. Connolly with the photography.

References

Bennett, M. H., Farrer-Brown, G., Henry, K., andJelliffe, A. M. (1974). Classification of non-Hodgkin'slymphomas (Letter). Lancet, 2, 405-406.

Burston, H. H. (1959). Lethal midline granuloma: is it apathological entity? Laryngoscope, 69, 1-43.

de Faria, J. L., Cutin, M., Morgante, A. P., and Ferri,R. G. (1957). Malignant granuloma of the face: con-tributions to its nosology. Arch. Otolaryng., 62,255-262.

Dorfman, R. F. (1974). Classification of non-Hodgkin'slymphomas. Lancet, 1, 1295-1296.

Eichel, B. S., Harrison, E. G., Jr., Devine, K. D., Scanlon,P. W., and Brown, H. A. (1966). Primary lymphomaof the nose including a relationship to lethal midlinegranuloma. Amer. J. Surg., 112, 597-605.

Eichel, B. S. and Mabery, T. E. (1968). The enigma of thelethal midline granuloma. Laryngoscope, 78, 1367-1386-

Friedmann, I. (1955). The pathology of malignantgranuloma of the nose. J. Laryng., 69, 331-341.

Friedmann, 1. (1963). The'pathologyofmidlinegranuloma.Proc. roy. Soc. Med., 57, 289-297.

Gerard-Marchant, R., Hamlin, I., Lennert, K., Rilke, F.,Stansfeld, A. G., and van Unnik, J. A. M. (1974).Classification of non-Hodgkin's lymphomas (Letter)Lancet, 2, 406 408.

Harrison, D. F. N. (1974). Non-healing granulomata ofthe upper respiratory tract. Brit. med. J., 4, 205-209.

Kay, H. E. (1976). Personal communication.McGuirt, W. F. and Rose, E. F. (1976). Lethal midlinegranuloma: a pathological spectrum. J. Laryng., 90,,459-466.

Smith, F. (1958). Erythrophagocytosis in human lymph-glands. J. Path. Bact., 76, 383-392.

Stewart, J. P. (1933). Progressive lethal granulomatousulceration of the nose. J. Laryng., 48, 657-701.

327

copyright. on S



j.com/

J Clin P


ownloaded from

http://jcp.bmj.com/

Pathology of 'non-healing (midline) granulomain which a clinical diagnosis of'midline granuloma' or...

Documents

Transcript of Pathology of 'non-healing (midline) granulomain which a clinical diagnosis of'midline granuloma' or...