Histopathology 1979, 3, 421-432

Myospherulosis: further ultrastructura1 observations A.G.ROSE, A.H.TIMME, S.K.PRlCE, M.EMMS & J.DALE Department of Patholog-v, Groote Schuur Hospital and Unhersity of Cape Toion, Cape Town, South Africa Accepted for publication 16 March 1979 ROSE A.G., T I M M E A.H., 3,421-432

PRICE

S.K., EMMSM. & DALE J. (1979)His/opo/ho/ogy

Myospherulosis: further ultrastructural observations The changes of myospherulosis were observed in masses from the buttocks of three patients. Ultrastructural examination of parent bodies and spherules from two patients revealed some spherules to contain dense bodies and filaments. Spherules had a double electron dense wall: the outer layer had a thickness corresponding to that of a cell membrane. Incubation of packed red blood cells with tetracycline ointment produced a similar electron dense deposit on the inner aspect of the cell membrane of the erythrocytes. No filaments or parent bodies were seen. Our findings support the hypothesis that myospherulosis represents altered red blood cells. The filaments seen in our patients may represent polymers of haemoglobin. It is suggested that parent bodies may be derived from histiocytes. Keywords: myospherulosis, red blood cells, haemoglobin, histiocytes

Introduction Myospherulosis is an infrequently reported condition which until very recently had only been described in 12 patients, all East Africans (McClatchie, Warambo & Bremner 1969a, McClatchie & Bremner 1969b, Hutt, Fernandes & Templeton 1971). Kyriakos (1977) described 16 cases of the disease in America involving paranasal sinuses, nose and middle ear. The nature of the condition has been a mystery, and a wide variety of organisms including pathogenic algae have been suggested as possible causes. Several authors noted the similarity in appearance of spherules to red blood cells. Rosai (1978) has presented strong evidence that the spherules are in fact erythrocytes altered by the action of tetracycline ointment. Address for correspondence: Dr A.G.Rose, Department of Pathology, Medical School Observatory, Cape, South Africa.

0309-0167/79/0900-042 I S02.00 0 1979Blackwell Scientific Publications

42 1

422

A.G.Rose et al.

We have encountered three patients with myospherulosis present at injection sites in the buttocks. This material afforded us the opportunity to study spherules and parent bodies by both light and electron microscopy. We were also stimulated to attempt to reproduce the in vitro experiment of Rosai with tetracycline ointment treated red blood cells and to compare such altered erythrocytes by light and electron microscopy with our clinical material. Since electron microscopy has been reported on only one human case of myospherulosis (De Schryver-Kecskemeti & Kyriakos 1977) and Rosai's in ritro study was light microscopic only, the emphasis of this report will be on the ultrastructural findings in the patients and in the in uitro red cell experiment.

Materials and methods PATIENlS

Biopsies showing myospherulosis from patients I and 3 were diagnosed in the past year. A review of our surgical pathology material for the years 1959-1976 covering diagnoses of paraffinoma, injection site response, foreign body reaction, chronic sinusitis and calcifying collagenolysis yielded a third case of myospherulosis. This biopsy had been examined in 1960, 9 years prior to the first description of myospherulosis, and the spherules had been missed. Sections for light microscopy were stained with haemotoxylin and eosin, periodicacid Schiff, Gomori's methenamine silver, Verhoeff-van Gieson, Masson's trichmme, mucicarmine, lipofuscin (Schmorl's stain), Okajima's stain for haemoglobin (Okajima, 19 I 6) and peroxidase (diaminobenzidene) stain. Electron microscopy was performed upon formalin-fixed tissue of patient I and wax-embedded tissue of patient 3 using a method previously described (Rose & Becker 1972). Ultrathin sections were stained with uranyl acetate followed by lead citrate. I N c L' H A r E I> E R Y T H K O c Y T 1:s - T E T R A

cY cLI N E

01 N T M E N T M I X T u R E

I n order to compare the spherules found in our patients with the appearance of erythrocytes altered by treatment with antibiotic ointment we repeated the experiment performed by Rosai (1978) : the lower portions of several test tubes were coated with a thick layer of 3 :(, tetracycline hydrochloride ointment (Achromycin). About 0.5 ml of a preparation of packed human erythrocytes was added to each tube. The preparations were incubated at 37°C in a water bath with occasional shaking for I , 2 and 5 days. Following incubation, a portion of the mixture was smeared on a slide, fixed in absolute ethanol, and stained with haematoxylin and eosin. Another portion was mixed with thrombin solution (100u/ml), dehydrated, embedded in parafin, sectioned and stained with the same stains used for the biopsies. A third portion was taken on days I and 5, fixed as a pellet in glutaraldehyde, and processed for electron microscopy in the same way as the biopsies.

Myospherulosis

423

O S M I U M TREATED TETRACYCLINE OINTMENT

A sample of ointment was placed in processed for electron microscopy.

2%

osmium tetroxide for 3 days and then

Results PATIENTS

The main clinicopathological features of our three patients with myospherulosis are listed in Table I . All three patients were adults who presented with discomfort due to masses in the buttocks. Two patients had bilateral masses. Since the appearance Table

I.

Clinocopathological features of three patients with myospherulosis Pa tients : ( I ) M.K.

Age (Yd Race Sex Known duration of lumps No. of lesions Greatest diameter of lesions (cm) Site Parent bodies (pm) Spherules diameter (pm) Giant cells Refractile foreign material Areas like calcifying collagenolysis

36 col F 2 days 2

3

buttocks 36-180 5

+ +

(2)

V.J.

(3) A.N.

57

32

white M 2 4 months 3

co I F 6 months 3 3 buttocks

12

buttocks 36-55 5

+

57

5

+ + +

of the spherules, parent bodies and host response were similar in all three patients, a composite description of the pathology will be given. The lesions were situated in the subcutaneous fibro-adipose tissue in patients I and 3, and in relation to striated muscle in case 2. On sectioning, the lesions werc characterized by solid yellow areas and cystic spaces filled with gelatinous yellowbrown material encompassed by prominent swathes of fibrous tissue. Some cyst walls had a golden-brown colour. Microscopically, the cystic areas had a lining composed of histiocytes and giant cells with surrounding dense fibrosis. Lying free within the cystic spaces and amongst histiocytes in the wall were numerous parent bodies (Figures I & 2) which contained variable numbers of spherules. Free-lying degenerating spherules were also seen within acellular fibrous tissue. Other spherules were evoking a giant cell response (Figure 3). Some of the spherules stained positively with Okajima’s stain, peroxidase and

424

A.G.Rose et 01.

Figure I . Parent body containing numerous spherules. Several spherules contain a dense spherical opacity. Histiocytcs surround the parent body. Patient M.K. x 150. H & E.

Figure

2.

Smaller parent body containing fewer spherules lies close to inner lining of cystic space.

A disintegrating parent body lies just above the intact one. Patient A.N. x 250. H & E.

Figure 3. Disruption of a parent body has released free spherules into the tissues where they arc evoking a giant cell and histiocytic response. Patient A.N. x 60. H & E.

Figure 4. Electron micrograph shows portion of electron dense wall of parent body (top) and a few spherules. Pores (gaps) are seen in the wall of one spherule (top leff). Patient M.K. x 7200.

. Figure 5. Spherule contains numerous filaments plus several rounded electron densities, some of which are separated by a halo from a thin, partially encircling membrane. Arrows indicate the 90-100 8, thick outer layer of the spherule’s double electron dense wall. Patient M.K. x 1 8 000.

426

A.G.Rose et al.

Schmorl’s stain. Negative results were obtained when the spherules were stained with PAS and Gomori’s methenamine silver. Masson’s trichrome stained the spherules a reddish-brown colour. Whilst the parent bodies showed a wide variation in diameter (36-180 pm), the spherules uniformly measured about 5 pm in diameter (Table I). In patient 3, two out of the three lumps showed advanced fibrosis and calcification, giving an appearance like that seen in calcifying collagenolysis. One of the three masses contained refractile foreign material suggesting a relationship to injected material. Electron microscopy on formalin-fixed tissue from patients I and 3 showed the parent bodies to have a single electron dense structureless wall which enclosed the contained spherules i n a continuous fashion (Figure 4). The spherules often showed focal defects in their walls as well as focal internal thickenings. Most of the spherules were empty or contained only scanty granular material, but a few contained numerous filaments (Figures 5 & 6). The filaments were 180 A wide and were either aggregated into bundles of two to four filaments or arranged singly. The filaments varied in length from 4.4 to 1 8 p n . Occasional spherules contained scanty small rounded electron densities (Figure 5 ) . The spherules appeared to have a double electron dense wall; the outer layer measured 90-100 8, thick which corresponds to the

Figure 6. Elcctron micrograph showing 180 8, thick filaments in two closely adjaccnt spherules attaching tolthe inner aspect of the electron dense wall of each. Patient A.N. x 45 000.

Figure 7. Electron micrograph of electron dense spherular wall showing gaps (pores). The wall has a double electron dense structure, the outer layer (arrows) having a thickness similar to that of a cell membrane. Patient M.K. x 90 000.

Figure 8. Electron micrograph of two adjacent spherules. Pores are present in the walls of both structures. The lower spherule contains filaments and rounded electron densities, while the upper fractured spherule contains a large membrane bound spherical remnant of uncertain nature.

Figure 9. Appearance of erythrocyte after I day's incubation with tetracycline ointment. Irregular deposits o f electron dense material are seen on the inner aspect of the cell membrane. x 1 5 000. Figure 10. Aftcr 5 days' incubation with tetracycline ointment the erythrocytes show moreabundant electron dense deposits. The similarity to the ultrastructural appearances of myospherulosis (sec Figure 4) is obvious. i( 20 coo.

Myospherulosis

429

thickness of a cell membrane (Figures 5 & 7). One spherule contained a spherical remnant of unknown nature (Figure 8). INCUBATED ERYTHROCYTES-TETRACYCLINE

OINTMENT MIXTURE

Light microscopic examination of smears and paraffin embedded material showed some loss of haemoglobin from erythrocytes, but we were unable to reproduce the formation of parent bodies with contained altered erythrocytes produced by Rosai (1 978). However, at the ultrastructural level the samples taken on days I and 5 showed the erythrocytes to have developed an electron dense deposit on the inner aspect of the cell membrane resulting in an appearance strikingly similar to that seen in the spherules of our 2 patients (Figures 9 & 10). Like the spherules, the electron dense layer showed focal gaps and internal thickenings. No filaments or parent bodies were seen. OSMIUM-TREATED TETRACYCLINE OINTMENT

A small sample of ointment immersed in 2% osmic acid stained black in colour within I hour. Ultrastructural examination confirmed the osmiophilia.

Discussion The light and electron microscopic findings in our patients and the ultrastructural findings in the erythrocytes incubated with tetracycline ointment support the interpretation of Rosai (1978) that myospherulosis represents collections of erythrocytes altered by a foreign substance. The giant cell response observed in the tissues is probably being evoked by the foreign material deposited within the erythrocytes. The fact that one is dealing with altered red blood cells explains why the spherular structures have failed to correspond to the known morphology of any algal, fungal or protozoal organism. The siting of the lesions in the buttocks of our three patients is suggestive of a relationship to injections. Only one patient could remember having been injected in the buttocks. One other also had refractile foreign material within macrophages at the site of the myospherulosis supporting the injection hypothesis. In the East African patients the buttocks and limb girdles were involved and the possibility of injections by witchdoctors was mooted (Hutt et al. 1971). The 16 cases of Kyriakos (1977) involving the ear, nose and nasal sinuses were related to the use of petrolatum-based tetracycline ointment. De Schryver-Kecskemeti & Kyriakos (1977) were able to produce myospherulosis in experimental animals by the use of petrolatum-based antibiotic ointments. It is likely that in all of these situations erythrocytes are brought into contact with petrolatum-based preparations. As far as we are aware, electron microscopy has been performed upon only one other human case of myospherulosis (De Schryver-Kecskemeti & Kyriakos 1977). Like them we observed electron dense walls in the parent bodies and spherules; scanty small

430

A.G.Rose

et

al.

rounded electron densities were seen within some spherules. We interpret the thin partially encircling membranes seen in proximity to some of these densities as being part of the erythrocyte cell membrane. However, unlike the latter study, we encountered several spherules which contained numerous I 80 8, thick filaments arranged singly or in small bundles. A further new observation was the finding that the spherules had a double electron dense wall, the outer layer having a thickness similar to that of a cell membrane. One fractured spherule was observed to contain a spherical structure limited by a membrane; the nature of this structure is unknown. The double electron dense wall of the spherules was similar in appearance to that seen in the tetracycline ointment treated erythrocytes. In both, the outer unit membrane appears to represent the erythrocyte cell membrane and the thicker electron dense inner layer probably represents a deposit of foreign material with staining properties consistent with lipid derived from the ointment or injected oily suspension. The deposit had a similar appearance ultrastructurally to that of the osmium tetroxide stained tetracycline ointment. In the ointment treated erythrocytes this layer increased from day I (Figure 9) to day 5 (Figure 10). The filaments (which were encountered i n the biopsies only) are less easy to explain. Our initial reaction was that the filaments might represent some organized internal structure and be evidence that the spherules are organismal in nature. However, it has been shown that haemoglobin S of sickle cell disease may form polymers of about 170-190 A in diameter with mostly a uniformly dense substructure (White & Heagan 1973). The size of such haemoglobin rods corresponds to the size of our filaments. I n gels of normal haemoglobin one gets the formation of tubular polymers (microtubules). I t is possible that the filaments seen in some of our spherules represent polymers of haemoglobin. Unfortunately our patients have been lost to follow-up so we are unable to test whether the two patients whose spherules contained filaments have the sickling trait. We do not know why we failed experimentally to reproduce the light microscopic spherules and parent bodies produced so convincingly by Rosai (1978). However, the ultrastructural coating of the inner aspect of the erythrocyte cell membrane by an electron dense layer, presumably of foreign material, in the tetracycline ointment treated erythrocytes produced an appearance virtually identical to that of the spherules seen in the patients. An unsolved problem is what gives rise to the parent body structure. I t would appear unlikely that erythrocytic membranes would open and unite to form a continuous sphere but, as Rosai (1978) points out, the positive reaction to this envelope with the stains for red blood cells suggests that this has indeed occurred. In our ultrastructural study we failed to demonstrate an outer unit membranelike structure in the parent bodies. However, the inner electron dense deposit seen in the spherules was identical in appearance to that of the parent body. An alternative hypothesis based on an anecdotal observation (Figure I I ) is that the parent body may be derived from the cell membranes of histiocytes, which have engulfed the erythrocytes, the foreign material depositing in relation to this membrane. Figure I I shows parent body-like structures formed following the ingestion by histiocytes of extravasatcd red blood cells in an area of extra-retinal haemorrhage in a patient suffering from a malignant melanoma of the choroid. In some of the formations one sees only the encompassing cell membrane, the histiocyte’s nucleus being

Myospkerirlosis

43 1

Figure 11. Anecdotal observation of parent body-like structures formed by histiocytic ingestion of erythrocytes in an area of extra-retinal haemorrhage. (The pigment present is melanin derived from a nearby necrotic choroidal malignant melanoma.) x 600.

displaced to one side or not seen in the plane of section. Possible support for a histiocytic origin for the parent body is the observation of Rosai (1978) that one of the original Ugandan cases (which he reviewed) had parent bodies containing degcnerative neutrophils. One concludes that the formations seen in myospherulosis represent altered erythrocytes. The nature of the parent body awaits further elucidation. Another curious feature worthy of study is why the foreign material appears to deposit only on the inner aspect of the erythrocyte membrane.

References DE SCHRYVER-KECSKEMETI K . & KYRIAKOS M. (1977) Myospherulosis. An electron-microscopic study of a human case. American Journal of Clinical Pathology 67, 555-561 H u n M.S.R., FERNANDES B.J.J. & TEMPLETON A.C. (1971) Myospherulosis (subcutaneous spherulocytic disease). Transactions of the Royal Society of Tropical Medicine arid Hygicne 65,182-188 KYRIAKOS M. (1977) Myospherulosis of the paranasal sinuses, nose and middle ear. A possible iatrogenic disease. American Journal of Clinical Pathology 67, I I 8-130 MCCLATCHIE S . , WAHAMBO M.W. & BREMNER A.D. (1969a) Myospherulosis. A previously unreported disease? American Journal of Clinical Pathology 51,699-704 MCCLATCHIE S. & BREMNER A.D. (1969b) Unusual subcutaneous swellings in African patients. East Africun Medical Journal 46, 625-633 K. (1916) On the selective staining of the erythrocyte. Afiatomical Record 11, 295 296 OKAJIMA ROSA] J . (1978) The nature of myospherulosis of the upper respiratory tract. American J o u r d of Clinical Pathology 69,475-481

ROSEA.G. & BECKERW.B. (1972) Disseminated Herpes-virus hominis (Herpes simplex) infectionretrospective diagnosis by light and electron microscopy of paraffin wax-embedded tissues. Journal of Clinical Pathology 25, 78-87 WHITEJ.G. & HEACANB. (1973) Fine structure of hemoglobin polymerization. In Sickle Cell Disease Diagnosis, Management, Education and Research, eds H.Abramson, J.F.Bertles & D.L.Wethers, pp. 104-129. C.V. Mosby, St Louis

Myospherulosis: further ultrastructural observations.

Histopathology 1979, 3, 421-432 Myospherulosis: further ultrastructura1 observations A.G.ROSE, A.H.TIMME, S.K.PRlCE, M.EMMS & J.DALE Department of Pa...
4MB Sizes 0 Downloads 0 Views