J Clin Pathol 1985;38:1039-1044

Disseminated zygomycosis associated witherythroleukaemia: confirmation by lectin stainsEW BENBOW, IW DELAMORE,* RW STODDART, H REID

From the Departments ofPathology and *Haematology, University of Manchester, Manchester

SUMMARY Zygomycosis is not often diagnosed in the United Kingdom, and so the possibleimportance of the findings in a patient with disseminated zygomycosis who had been treated withchemotherapy for erythroleukaemia was not appreciated until histological examination of speci-mens obtained at necropsy provided a presumptive diagnosis. No attempt had therefore beenmade to identify the organism by culture, and lectin binding methods were used to try to compen-sate for this. The characteristics of the hyphae on staining with lectins were similar to thosepreviously shown in Rhizopus oryzae and were unlike those of a wide range of other hyphal fungi.Although definite speciation of the fungus was not achieved, these findings confirm that this was acase of zygomycosis and would seem to represent the first such reported confirmation in theabsence of culture.

Six principal patterns of infection of an organ withzygomycosis are recognised: rhinocerebral, pulmo-nary, disseminated, gastrointestinal, cutaneous, andmiscellaneous.'-3 Each pattern of infection isassociated with a characteristic range of underlyingconditions, with a particularly prominent and welldocumented association between diabetes mellitusand rhinocerebral zygomycosis.245 There is also astrong association between haematological malig-nancy and the development of the disseminatedform of zygomycosis.67 Various other predisposingfactors have been well reviewed by McNulty,8 andseveral cases have occurred in previously apparentlywell subjects.

Opportunistic infections caused by organisms ofthe order Mucorales are confusingly named, forthe terms mucormycosis, phycomycosis, andzygomycosis are often used as if they were inter-changeable. Terminology of disease has clearlyfailed to keep up with changes in taxonomy.Phyomycosis is often used, though taxonomicchanges subsequent to its introduction seem to ren-der it inappropriate.9 Some prefer the older andmore familiar term mucormycosis for infectionsdiagnosed by biopsy without identification of theexact species by culture,'0 although the term hasbeen abandoned by the nomenclature committee ofthe International Society for Human and AnimalMycology." Zygomycosis has the advantage of

Accepted for publication 16 May 1985

reflecting the current name of the class.'2 Most deepinfections ascribable to Zygomycetes are caused bymembers of the genera Rhizopus, Mucor, andAbsidia.3We saw a case of disseminated mucormycosis in a

woman with leukaemia. The presence of opportunis-tic fungal infection was not suspected during life andwas not detected until necropsy. Conventionalspecial stains were supplemented with lectin stainingbecause of the ability of lectins to stain identifiablecarbohydrate residues. The high saccharide contentof fungal walls suggests that lectin staining may be ofpossible value in detecting fungi in histological sec-tions.'3

Case report

A woman aged 66 was transferred to ManchesterRoyal Infirmary in September 1981. She com-plained of lethargy, malaise, oral ulceration, anddysphagia. Peripheral blood counts were as follows:Haemoglobin concentration 7*3 g/dl; white cellcount 120-0 x 109/1(98% blasts, 2% lymphocytes);and platelets 36 x 109/l. Bone marrow smearsshowed complete replacement of the normal bonemarrow cells by myeloblasts. Special stainsconfirmed the diagnosis of acute myeloblasticleukaemia.She was treated with a combination of daunorubi-

cin, cytosine arabinoside, and thioguanine, accord-ing to the protocol laid down by the MedicalResearch Councirs 8th acute myeloblastic leu-

1039

copyright. on M

ay 31, 2020 by guest. Protected by

http://jcp.bmj.com

/J C

lin Pathol: first published as 10.1136/jcp.38.9.1039 on 1 S

eptember 1985. D

ownloaded from

1040

kaemia trial. Remission was confirmed by repeatexamination of bone marrow after two courses ofchemotherapy. She was given two consolidationcourses with the same combination of drugs, andafter she had been in remission for one year lateintensification treatment was started. This consistedof four courses of treatment separated by intervalsof three weeks; each course included cyclophos-phamide 900 mg and vincristine 200 mg intraven-ously at once, cytosine arabinoside subcutaneouslytwice daily for five days, and prednisolone 100 mgorally each day for five days, with chlorpromazine25 mg orally thrice daily for control of nausea ifrequired.

In early October 1983 a falling peripheral bloodcount heralded a relapse of leukaemia. The mor-phological pattern of the bone marrow smear wasnow that of erythroleukaemia, and furtherchemotherapy with daunorubicin, cytosinearabinoside, and thioguanine was given. She becamepancytopenic, and during this period she developedinflammatory changes in the left orbit. She com-plained of severe pain and throbbing in the eye andwas found to be slightly feverish. Flucloxacillin andbenzylpenicillin were given intravenously, togetherwith chloramphenicol eye drops.When a swab from the eye was reported to have

grown coagulase negative staphylococci theintravenous antibiotics were changed to pipericillinand tobramycin. A subsequent report that Strep-tococcus milleri had been grown from blood culturesled to the substitution of benylpenicillin for pipericil-lin. By this stage, however, severe proptosis of theleft eye had occurred, and the conjunctiva wasreported as looking chemotic and necrotic (Fig. 1).

Benbow, Delamore, Stoddart, Reid

Cerebral angiography was hampered by patientmovement and a poor cardiac output, but the inter-nal carotid arteries appeared to be normally patent;the venous phase was not visualised. Because of herpersistent fever and continuing deteriorationgranulocyte transfusions were given, but she died inMarch 1984.

FINDINGS AT NECROPSYExternal examination showed moderate jaundice,with extensive ecchymoses and mild bilateral ankleoedema. The left periorbital tissues were greatlyswollen and showed a deep red discolouration simi-lar to that after fairly recent bruising.

Examination of the external surface of the brainshowed pronounced softening and discolouration ofthe frontal lobes; this was more extensive on the left,where it affected a conical zone of about 4 cm indepth at the anterior pole of the brain (Fig. 2). Thecut surface in this area was friable and was a deeppink colour. The intracranial portion of the leftinternal carotid artery contained firm thrombus,which extended into the left anterior cerebral artery.The heart weighed 480 g, with left ventricular

hypertrophy. The myocardium appeared pale andflabby but contained no focal lesion. Coronaryatheroma was generally of only minor degree. Thelungs were focally firm, especially in the upper lobes.There was an annular ulcer at the lower end of the

oesophagus, which measured 3-5 cm in the long axisof the viscus. The margins of this lesion were a littlethickened, but did not appear to extend through thewall. The liver contained many nodules, ranging indiameter from 0-3 to 1*2 cm, each with a homogen-ous yellowish cut surface. The intervening hepatic

Fig. 1 Left eye, one week before death.

copyright. on M

ay 31, 2020 by guest. Protected by

http://jcp.bmj.com

/J C

lin Pathol: first published as 10.1136/jcp.38.9.1039 on 1 S

eptember 1985. D

ownloaded from

Lectin stains for zygomycosis

Fig. 2 Part ofhorizontal slice through fixed brain. Lesions are present in both

frontal lobes.

parenchyma was rather pale. The spleen bore anirregular pale lesion measuring 2-0 x 1-5 x 1P5 cmon its diaphragmatic aspect, together with many tinynodules within the parenchyma. The kidneysshowed particularly prominent vascular markingsbut no other macroscopic abnormality. The otherorgans appeared unremarkable.

CONVENTIONAL HISTOLOGYThe frontal lobes were infarcted, and many blood

vessels in the abnormal areas and in the adjacentmeninges contained fungal hyphae. The hyphaecould be seen penetrating vessel walls (Fig. 3) andwere visible within infarcted brain tissue. Hyphaldiameter was variable, branching was irregular (Fig.4), and septa were sparse. Indeed, many of theapparent cross walls were pseudosepta formed bykinking of the hyphae. The density of hyphaleosinophilia seemed to vary from area to area.The hepatic nodules comprised liver plates, show-

:1,- t:

%*, 0

to

A

Fig. 3 Hyphae within thrombus in lumen ofanintracerebral vessel. Some are also present in the wall.(Grocott stain.) Original magnification x 40.

Fig. 4 Fungal hyphae ofvariable diameter, with irregularbranching and pseudoseptae. Left internal carotid artery.(Grocott stain.) Original magnifiation x 170.

1041

copyright. on M

ay 31, 2020 by guest. Protected by

http://jcp.bmj.com

/J C

lin Pathol: first published as 10.1136/jcp.38.9.1039 on 1 S

eptember 1985. D

ownloaded from

1042

ing coagulative necrosis with infiltration by fungalhyphae. The splenic lesions were similarly formed offungal infiltrate, with a rather more regulararrangement of hyphae than that found in the brain,within a background of necrotic tissue. Tiny fungalcolonies were present in the kidneys. Sections of thelower oesophagus showed shallow ulceration withan overlying pseudomembrane of necrotic mucosaand fibrinous exudate; fungal hyphae were presentonly in the superficial layers of this membrane. Noevidence of fungal infiltration was found in sectionsof the heart or lungs.

Selected sections were stained with periodic acidSchiff, Gram, and Grocott's methenamine silverstains. Each of these standard special stains outlinedthe fungus with greater clarity than the haematoxy-lin and eosin stain, especially in sites such as the liverand within the thrombus, where the degree ofeosinophilia of the hyphae resembled that of theadjacent tissue.The value of the Gram stain varied between tis-

sues. The hyphae stained well with the magentacounterstain and were particularly clearly defined inthe spleen. Unfortunately, the Gram stain appearedto enhance the fibrin strand artefact seen in a bloodclot and was therefore of little value in defining theextent of fungus within blood vessels. Periodic acidSchiff and Grocottfs stains were of more value andclearly defined the features and extent of fungalinfection in all the affected tissues. Indeed, thoughthe presence of fungus had been detected in sectionsstained with haematoxylin and eosin, its extent hadbeen considerably underestimated until sectionstreated with special stains became available.

LECTIN STAINSA panel of lectin stains (Table) were applied usingthe avidin-biotin technique described by Hsu andRaine,'4 modified by the exclusion of incubationwith mouse liver powder and by using different con-centrations of reagents to achieve a ratio of 1:4 forthe components of the peroxidase complex labelledwith avidin and biotin.'5 In addition, a selection oflectins labelled with fluorescein were used.'3

Table Identity and affinities oflectin stains applied

Lectin Affinity

Wheatgerm agglutinin + + +Concanavalin A*Garden pea agglutinin +Lentil agglutinin +/-Erythrophytohaema glutinin + +Soy bean agutininPeanut agglutininWinged pea agglutinin +

Concanavalin A and soy bean agglutinin staining is difficult toquantify because of staining of surrounding material.

Benbow, Delamore, Stoddart, Reid

Fig. 5 Hyphae outlined by lectin shown by avidin-biotinmethod. (Wheat germ agglutinin.) Original magnification x170.

The various lectin stains showed considerable dif-ferences in affinity (Table). The hyphae wereclearly, though faintly, stained using the avidin-biotin method (Fig. 5), and a semiquantitativeassessment of density of staining could be made eas-ily. Lectins labelled with fluorescein gave denserstaining of hyphae, but the non-specific backgroundstaining tended to be distracting (Fig. 6). Similarbackground staining was seen on avidin-biotin stain-ing with concanavalin A and, to a lesser degree, withsoy bean agglutinin. In both cases hyphae were out-lined by negative staining.The pattern of lectin staining was compared with

that obtained with several fungal hyphae from theNational Type Collection at Kew (BM Herbertsonand RW Stoddart, unpublished data). This showedan affinity pattern similar to that of R oryzae, andthere were appreciable differences from the stainingpatterns of Mortierella wolfl, R arrhizus, Absidiacorymbifera, Aspergillus flavus, A jumigatus, Aniger, and all of the dermatophytes investigated todate.

Discussion

The Zygomycetes are a class of fungi usually foundin soil, humus, and dung'6; they cause decay in manykinds of fruit and vegetables'7 and are occasionallypathogenic in domestic animals.'8 Spores of variousmembers of the class have been widely detected in

copyright. on M

ay 31, 2020 by guest. Protected by

http://jcp.bmj.com

/J C

lin Pathol: first published as 10.1136/jcp.38.9.1039 on 1 S

eptember 1985. D

ownloaded from

Lectin stains for zygomycosis

...- A

Fig. 6 Fluorescein labelled lectin staining ofhyphae.Surrounding fibrinous clot is distractingly stained. (Wheatgerm agglutinin, fluorescein isothiocyanate.) Originalmagnification x 170.

air, including air in a ward in a London teachinghospital.'9 Skin tests suggest sensitisation in a sub-stantial minority of patients with allergic diseasesand in controls,20 and sputum from apparently heal-thy men may be contaminated.2'Our failure to diagnose zygomycosis in this case

would appear to have been due to a low index ofsuspicion engendered by the apparent rarity of thedisease in this country. Few well documented caseshave been reported from the British Isles,22-24though Symmers was able briefly to discuss thefindings in "many" cases of British origin.25Authors from several continents have described

increases in the number of reported cases ofmucormycosis,257-2 a change that is partly due to anincrease in the incidence of certain of the predispos-ing factors, partly to greater international mobility,and partly to greater awareness of the condition.25Increasingly aggressive chemotherapy and morewidely used antibacterial treatment may also beimvortant.7Zygomycosis arising in association with ery-

throleukaemia is rare, presumably because of therarity of erythroleukaemia, though the association isby no means unknown.'28 It is not possible to sepa-rate the relative importance of disease and treat-ment in inducing susceptibility to this opportunisticinfection, though several cases after renal, 2930 car-diac,29 and bone marrow transplantation3' would

suggest that cytotoxic treatment has an importantrole. Zygomycosis after renal failure without trans-plantation is less common, and in such cases thepatient may also have diabetes mellitus, or at leastglucose intolerance induced by steroids.2230

Zygomycosis in patients with leukaemia is particu-larly associated with periods of neutropenia6 and,once established, tends to advance rapidly; bloodvessel walls are easily breached, and infarction afterthrombosis induced by fungus facilitates the spreadof disease. Although mucormycosis is often wide-spread at the time of detection when leukaemia isthe underlying condition, it may start its course inthe nose and orbit; this is apparently the secondmost common portal of entry after the lung in thoseimmunosuppressed following chemotherapy.32Thrombosis of an internal carotid artery, which wasa striking feature of this case, is described in at leasta third of patients affected cerebrally.33The facility with which the organism spreads once

infection is established, coupled with the rarity ofthe condition, has led to a considerable amount ofexperimental work aimed at determining what thebody's normal defence against mucormycosis mightbe. It would seem that these defences are multifac-torial, with components in both the serum and thephagocytic cell system.34 It is also clear that mechan-isms exist that inhibit both spore germination andhyphal growth.35

Opportunistic infection is often a disaster becauseit may kill a patient with an otherwise treatable orcurable disease. As chemotherapeutic regimensbecome more potent and rates of remission improveit becomes necessary to guard against a parallelincrease in opportunistic infections. The possibilityof nosocomial zygomycosis3036 makes each case anindicator of potentially similar disease in otherpatients. Many of the cases reported concernpatients whose zygomycosis was discovered only atnecropsy, often at a stage when culture was nolonger possible. This is a problem that is common invarious other opportunistic fungal infections also,and so techniques, such as lectin staining, that offerpossible improvements in our ability to recogniseand characterise such infections are to be welcomed.

We thank Mr GW Rogers and Mrs J Crosby forphotographic help, and Dr CJP Jones and Mr PWard for technical help.

References

'Baker RD. Mucormycosis-a new disease? J Am M Ass1957; 163:805-8.

2 Straatsma BR, Zimmerman LE, Gass JDM. Phycomycosis. Aclinicopathologic study of 51 cases. Lab Invest 1962;11:963-87.

1043

copyright. on M

ay 31, 2020 by guest. Protected by

http://jcp.bmj.com

/J C

lin Pathol: first published as 10.1136/jcp.38.9.1039 on 1 S

eptember 1985. D

ownloaded from

1044

3 Lehrer RI, Howard DH, Sypherd PS, Edwards JE, Segal GP,Winston DJ. Mucormycosis. Ann Intern Med 1980; 93:93-108.

4 Pillsbury HC, Fischer ND. Rhinocerebral mucormycosis. ArchOtolaryngol 1977; 103: 600-4.

5 del Real Mora 0, Quezada JZ, Mendoza CA, Diaz ES, de LeonSP, Ruiz-Palacois G. Mucormicosis. Informe de 14 cases. RevInvest Clin 1983;35:237-40.

6 Baker RD. Leukopenia and therapy in leukemia as factor pre-disposing to fatal mycoses, mucormycosis, aspergillosis andcryptococcosis. Am J Clin Pathol 1962;37:358-73.

Meyer RD, Rosen P, Armstrong D. Phycomycosis complicatingleukemia and lymphoma. Ann Intern Med 1972;77:871-9.

8 McNulty JS. Rhinocerebral mucormycosis: predisposing factors.Laryngoscope 1982;92: 1140-3.

Emmons CW, Binford CH, Utz PJ, Kwon-Chung KJ. Medicalmycology, 3rd ed. Philadelphia: Lea and Febiger, 1977:254.

'° Gordon MA. Nomenclature of fungi pathogenic to man and ani-mals. Sabouraudia 1978; 16:312-3.

" Vanbreuseghem R, Gentles JC, Mackenzie DWR, Male 0,Mantoviani A. Nomenclature of mycoses. Sabouraudia1980; 18:78-84.

12 Webster J. Introduction to fungi. 2nd ed. Cambridge: CambridgeUniversity Press, 1980.

3 Stoddart RW, Herbertson BM. The use of fluorescein-labelledlectins in the detection and identification of fungi pathogenicfor man: a preliminary study. J Med Microbiol 1978; 11: 315-24.

4 Hsu S-M, Raine L. Versatility of biotin-labelled lectins andavidin-biotin-peroxidase complex for localization of carbohy-drate in tissue sections. J Histochem Cytochem 1982;30: 157-61.

X- Hsu S-M, Raine L, Fanger H. The use of avidin antibody andavidin-biotin-peroxidase complex in immunoperoxidase tech-nics. Am J Clin Pathol 1981;75:816-21.

16 Emmons CW. Natural occurrence of opportunistic fungi. LabInvest 1962;11: 1026-32.

71 Harter LL, Weimer JL. Decay of various vegetables and fruits bydifferent species of Rhizopus. Phytopathology 1922; 12:205-12.

18 Monga DP, Mohapatra LN. A compilation of published reportsof mycoses in animals in India. Mycopathologia 1980;72: 3-11.

9 Noble WC, Clayton YM. Fungi in the air of hospital wards. J GenMicrobiol 1963;32: 397-402.

20 Shivpuri DN, Agarwal MK. Studies on the allergenic fungalspores of the Delhi, India, metropolitan area. Clinical aspects.J Allergy 1969;44:204-13.

21 Comstock GW, Palmer CE, Stone RW, Goodman NL. Fungi inthe sputum of normal men. Mycopathologia 1974;54:55-62.

22 Kurrein F. Cerebral mucormycosis. J Clin Pathol 1954;7: 141-4.23 Symmers WStC. Silicone mastitis in "topless" waitresses and

Benbow, Delamore, Stoddart, Reidsome other varieties of foreign-oody mastitis. Br Med J1968;3: 19-22.

24 Helenglass G, Elliott JA, Lucie NP. An unusual presentation ofopportunistic mucormycosis. Br Med J 1981;282: 108-9.

25 Symmers WStC. Deep seated fungal infections currently seen inthe histopathologic service of a medical school laboratory inBritain. Am J Clin Pathol 1966;46:414-37.

26 Hotchi M, Okada M, Nasu T. Present state of fungal infections inautopsy cases in Japan. A statistical survey of all autopsy casesduring the ten-year period from 1966 to 1975. Am J ClinPathol 1980;74:410-6.

27 Marchevsky AM, Bottone EJ, Geller SA, Giger DK. The chang-ing pattern of disease, etiology and diagnosis of mucormycosis.Human Pathol 1980;11:457-64.

28 Rosenberg SW, Lepley JB. Mucromycosis in leukemia. Oral Surg1982;54: 26-32.

29 Bottone EJ, Weitzman I, Hanna BA. Rhizopus rhizopodiformis:emerging etiological agent of mucormycosis. J Clin Microbiol1979;9:530-7.

30 England AC, Weinstein M, Ellner JJ, Ajello L. Two cases ofrhinocerebral zygomycosis (mucormycosis) with commonepidemiologic and environmental features. Am Rev Respir Dis1981; 124:497-8.

3' Myskowski PL, Brown AE, Dinsmore R, et al. Mucormycosisfollowing bone marrow transplantation. J Am Acad Dermatol1983;9: 111-5.

32 Krick JA, Remington JS. Opportunistic invasive fungal infec-tions in patients with leukaemia and lymphoma. Clin Haematol1976;5:249-310.

3 Landau JW, Newcomer VD. Acute cerebral phycomycosis(mucormycosis). Report of a pediatric patient successfullytreated with amphotericin B and cycloheximide and review ofthe pertinent literature. J Pediatr 1962;61:363-85.

34 Chinn RYW, Diamond RD. Generation of chemotactic factorsby Rhizopus oryzae in the presence and absence of serum:relationship to hyphal damage mediated by human neutrophilsand effects of hyperglycaemia and ketoacidosis. Infect Immun1982;38: 1123-9.

Waldorf AR, Peter L, Polak A. Mucormycotic infection in micefollowing prolonged incubation of spores in vivo and the roleof spore agglutinating antibodies on spore germination.Sabouraudia 1984;22:101-8.

36 Gartenberg G, Bottone EJ, Kensch GT, Weitzman I. Hospital-acquired mucormycosis (Rhizopus rhizopodiformis) of skinand subcutaneous tissue. Epidemiology, mycology and treat-ment. N Engl J Med 1978;299:1115-8.

Requests for reprints to: Dr EW Benbow, Department ofPathology, Stopford Building, University of Manchester,Oxford Road, Manchester M13 9PT, England.

copyright. on M

ay 31, 2020 by guest. Protected by

http://jcp.bmj.com

/J C

lin Pathol: first published as 10.1136/jcp.38.9.1039 on 1 S

eptember 1985. D

ownloaded from