CORE Metadata, citation and similar papers at core.ac.uk

Provided by Elsevier - Publisher Connector

THE JOURNAL OF INVESTIOATIVE DERMATOLOOY vol. 49, No. 5 Copyright 1567 by The Williams & Wilkins Co. Printed in U.S.A.

COLLOID MILIUM: A HISTOCHEMICAL STUDY* JAMES H. GRAHAM, M.D. AND ANTONIO S. MARQUES, M.D. Wagner (1), in 1866, first reported colloidreaction, with and without diastase digestion; milium in a 54 year old woman who showedcolloidal reaction, with and without bovine testicular hyaluronidase digestion for 1 hour at lesions on the forehead, cheeks and nose. In37 C; Movat's pentachrome I stain (2); alcian patients with colloid milium, the involvedblue pH 2.5 and 0.4 (3, 4); aldehyde-fuchsin pH skin is usually hyperpigmented, thickened,1.7 and 0.4 (4), with and without elastase digestion furrowed, nnd covered with multiple 0.5—5(5); Snook's reticulum stain; phosphotungstic hematoxylin stain (PTAH); re- mm dome-shaped, discrete papules. The shiny,action for iron; Fontana-Masson stain for ar- pink or orange to yellowish white translucentgentaffin granules; thiofiavine T fluorescent stain lesions have been likened to vesieles, but are(6, 7); ; alkaline Congo red method firm and only after considerable pressure can(8); stain; a clear to yellow mueoid substance be ex-stain for amyloid (9, 5); toluidine blue (4); and . The crystal violet and methyl vio- pressed from the papules. The lesions involvelet stained sections were mounted in Highman's sun exposed sites including the dorsum of theApathy gum syrup (5) which tends to prevent hands, web between the thumb and indexbleeding and gives a more permanent preparation. finger, knuckles, forearms, malar regions, nose,Formalin-fixed tissue from 1 patient with colloid ears, sides and back of the neck, and temporalmilium was used for preparing frozen sections of unstained sections, utilizing the above methods, areas. In 100 years extending from 1866—1966,and for oil red 0 and Sudan black B fat stains. approximately 100 cases of colloid miliumThe majority of the wet and paraffin-blocked tis- have been reported in the literature, but assues were sectioned at 5 e, but some were cut yet there is no uniform agreement with regard10—40 e thick and stained by the listed methods. to the nature of eolloid. All mounted tissue sections were examined polariscopically (5, 10) utilizing polarizer and The purpose of this paper is to record ouranalyzer lenses. The thiofiavine T and Congo red histopathologie, histochemieal, fluorescent mi-stained sections were examined microscopically for croscopic, and polariseopie observations re-fluorescence with a Leitz Ortholux microscope equipped with an Osram HBO 200 W light source, garding the morphologic features and compo-a UGI excitor, and an ultraviolet absorbing bar- sition of colloid. rier filter. The Congo red stained sections were also examined with the same fluorescent micro- MATERIAL ANDMETHODS scope system, but utilizing a darkfield condenser This report is based on observations of one or(11). With the exceptions given, the procedures more biopsies from S patients with typical lesionswere carried out as outlined in the "Manual of of colloid milium. The specimens were obtainedHistologic and Special Technics" (12). with a 4 or 6 mm cutaneous punch and fixed inFor comparative purposes, skin sections prepared 10% neutral buffered formalin. With the exceptionby the various techniques were studied from 4 of 1 specimen (courtesy of P. J. Alfaro, M.D.,adults with lichen and 1 adult with Bayamon, Puerto Rico), all tissues were processedprimary systemic amyloidosis cutis. for paraffin-blocked sections. Multiple sections stained with hematoxylin and eosin (H & E) CLINICAL DATA were examined in all cases, and additional biopsy sections from 3 of the patients were prepared by The biopsy specimens came from 5 Cauca- the following methods: periodic acid-Sehiff (PAS)sian men and all were clinically diagnosed as having eolloid milium. Their ages were 32, This investigation was supported in part by Research Training Grant No. 2A-5289 (C5), from33, 36, 49 and 65 years respectively. The 2 the National Institute of Arthritis and Metabolicpatients who were 32 and 36 years old came Diseases, U. S. Public Health Service, Bethesda, Maryland 20014. from Philadelphia, Pennsylvania; and the Presented at the Twenty-Eighth Annual Meet-remaining 3 had spent most of their lives in ing of the Society for Investigative Dermatology,the region of Columbus, Georgia; Bayamon, June 18—20, 1967, Atlantic City, New Jersey. * From the Departments of Dermatology andPuerto Rico; and Rio de Janiero, Brazil. For Pathology, Temple University School of Medicineoccupational reasons, 4 of the 5 patients had a and The Skin and Cancer Hospital of Philadelphia, 3322 North Broad Street, Philadelphia, Pennsyl-greater than average amount of exposure to vania 19140. sunlight. All patients showed multiple lesions 497 498 THE JOURNAL OF INVESTIGATIVE DERMATOLOGY A

Fm. 1. Colloid Milium. Top. Low power view showing elevations of the epidermis with hyperkeratosis, atrophy and effacement of the rete ridges overlying hyaline masses in the papillary corium. H & E, )< 9. Bottom. A higher magnification illustrates the tendency for peripheral collarette formation. grenz zone, and acidophilic, amorphous, smooth, uniform colloid which is fissured with cleft-like retraction spaces. Changes of solar elastosis are present beneath the colloid deposits. H & E, X 5G. typical of colloid milium and these were lim- HiSTOPATHOLOGIC OBSERVATIONS ited to the dorsum of the hands in 2; the malar Microscopically, H & E stained biopsy sec- regions in 1; 1 had involvement of the malartions from the 5 patients showed similar fea- regions and dorsum of the handa; and the re-tures. The epidermis was elevated at irregular maining patient had discrete papuies on theintervals with hyperkeratosis, intact granular dorsum of the hands and posterior neck. Biop- layer, atrophy, hyperpigmentation of the ba- sies from 4 of the patients came from thesal layer, and effacement of the rete ridges hands, and the patient with lesions limited tooverlying hyaline masses in the papillary the face had a biopsy taken from the left zy-corium (Fig. 1—top). Some sections showcd gomatic region. The stated duration of theperipheral eollarette formation which appeared asymptomatic lesions prior to consulting ato localize the colloid to broadened dermal physician varied from 1 to 6 years. All pa-papillae (Fig. 1—bottom). Generally, a grenz tients were in good general health, but 4 ofzone of relatively uninvolved connective tissue the 5 showed aetinic skin changes. separated the epidermis from the eolloid de- COLLOID MILITJM 499

4

Apt

.4 1) Fio. 2. High power magnification showing spindle andstellate-shaped fibroblasts throughout the colloid. Ciemsa, >< 655. posits, but focally, the hyalinc material ex-rounding and separating aggregates of colloid, tended to the dermoepidermal junction pro-but the morphologic appearance and staining ducing changes of liquefaction degeneration.characteristics of the 2 substances were dis- The colloid appeared as an acidophilic, amor-tinctly different (Fig. 6—left). In a few areas, phous, smooth, uniform substance replacingsmall clumps of elastotic tissue the connective tissue (Fig. 1). The homo-appeared entrapped in the hyaline substance, geneous masses were frequently fissured withbut this may be misleading from the plane of cleft-like retraction spaces (Fig. 1). Spindlesingle sections because deeper in the block and stellate shaped fibroblasts are scatteredcontinuity was usually established with larger throughout the colloid and these cells tend tomasses of solar clastosis at the base or periph- arrange their long axis along the lines of fis-ery of the colloid (Fig. 6—left). suring (Fig. 2). Papillary capillaries course through and adjacent to the hyaline masses, HJ5TOCHEMICAL onsxRvATroNs and their walls sometime show infiltration by The significant results of histochemical the acidophilic substance (Fig. 3—bottom).studies were as follows: The Fontana-Masson Vascular ectasia is usually seen in the grenzstain for melanin showed an abundance of zone and beneath the colloid, and some of thepigment in the basal layer of the epidermis, dilated blood vessels extend into the cleftsand lesser amounts free and within mclano- between the hyaline material (Fig. 3). Oc-phages in the region of the colloid deposits. casionally, a few lymphocytes, histiocytes,No hemosidcrin was demonstrated with the and rarely plasma cells were present laterallyPrussian blue reaction for iron. The colloid and at the base adjacent to the collections ofwas PAS-positive anddiastascresistant colloid. Sections from 4 out of the 5 patients(Fig. 3), showed affinity for Congo red, stained showed basophilic changes of solar elastosis atmctachromatically a purplish to red color with the periphery and beneath the colloid substancemethyl violet and crystal violet (Fig. 4), ap- (Figs. 1—bottom and 5). Some sections showedpeared pink with toluidine blue, and exhibited the abnormal elastotic changes completely sur-a variable amount of colloidal iron and alcian 500 THEJOURNAL OF INVESTIGATIVE DERMATOLOGY H p I

Fic. 3. Top. The colloid is PAS positive. PAS, X 55. Bottom. The PAS-positive and diastase resistant colloid shows a striking contrast with the epidermis, grenz zone, and underlying corium. At the left, the colloid shows infiltration of the wall of a small blood vessel. PAS with Djastase, >< 66.

blue (pH 2.5) positive material separating,stained similar to fibrin with PTAH, and coating, and surrounding the hyaline de-this suggests the presence of fibrin or fibrin posits. This latter substance is bovine testic-precursors as a component of colloid. The PAS- ular hyaluronidase labile and negative withpositive and diastase resistant material may alcian blue (pH 0.4), identifying it as hyalu-in part be due to the presence of fibrin, serum ronic acid. Some of the colloid masses showedproteins, and complex mucosaccharides bound uniform staining with aldcbyde-fuehsin pHto proteins through covalent linkages. The 1.7 and 0.4, probably indicating the presencepentachrome stain colored the colloid a pale of sulfated acid mucopolysaccharides. Elasticgrey to light yellow which was tinctorially tissue in areas of solar elastosis, and a mod-distinct from collagen and elastic tissue. Oc- erate increase in mast cells were also best dem-casionally, narrow septae occupied by blood onstrated with adchydc-fuchsin(Fig. 5).vessels, collagen fibrils, elastic fibers, and solar Elastase digestion removed the aldehyde-fuch-elastotie tissue completely surrounded and sin reactive elastic tissue and abnormal clas-divided small accumulations of colloid. The totic fibers of solar clastosis, but had no effectareas of solar elastosis surrounding some of on the colloid or mast cells. Some of the colloidthe colloid masses was striking (colored black 502 THEJOURNAL OF INVESTIGATIVE DERMATOLOGY

- --- 4 —

Fio. 6. Collojd milium. Left. Solar elastotie tissue separating nodules of colloid. Multiple sections show that the abnormal elastic tissue is in direct continuity with septae at the periphery and beneath the colloid substance. Aldehyde-fuehsin, >< 445. Right. View show- ing granularity of the colloid and a delicate network of argyrophihe fibers surrounding individual hyaline nodules. Snook's Reticulum, X 445.

to red with resorcin-fuchsin and purple toSometimes, there was enhancement of staining blue-black with aldehyde-fuchsin) and empha-with Congo red in 10—15 z paraffin sections, size the general absence of elastic fibers andand when longer periods of staining were used collagen in the colloid substance (Fig. 5). (sometimes overnight or even 24—72 hours). Snook's silver stain demonstrated a granularity of the colloid and a delicate network of argy- FLUORESCENT OBSERVATIONS rophilic fibers surrounding individual collec- tions of colloid (Fig. 6—right). This feature Frozen and paraffin sections stained with may account in part for the characteristicthioflavine T showed a uniform yellow to fissuring seen in routine H & E sections (Fig.green fluorescence under ultraviolet light and 1). It is likely that the cleft-like spaces arecontrasted with the dull gray of the adjacent artefacts from tissue processing, and occurepidermis and connective tissue. Paraffin sec- along normal lines of cleavage which in situtions cut at 15 p. gave optimal results, and contain blood vessels, fibroblasts, reticularthis is apparently a quantitive factor of en- fibers, and ground substance material. Sectionshanced fluorescence in the presence of a greater prepared with the Giemsa stain were partic-volume of organized structure. Congo red ularly good for demonstrating the spindle-stained sections of eolloid appeared pink under shaped or fusiform fibroblasts (Fig. 2). Thebrightfield fluorescence microscopy. When a oil red 0 and Sudan black B stains for lipidsdarkfield condenser was used, colloid stained showed a moderately positive staining reactionwith Congo red showed a moderate red fluores- which uniformly colored the colloid. With thecence and stood out well when contrasted with exception of the fat stains, the results observedthe gray connective tissue fibers. Paraffin sec- in paraffin sections wore usually preferable fortions 10—15 p., Stained with Congo red gave most of the histochemical studies performed.the most satisfactory results. COLLOID MILIUM 503

POLARISCOPIC OBSERVATIONS The PAS-stained sections of amyloidosis cutis Congo red stained sections of colloid miliumusually show a less intense color reaction than examined under polarized light show areas ofseen for colloid. The PTAH stain is essentially hirefringence, dichroism and even some degreenegative for amyloid, but demonstrates some of trichroism. Dichroism refers to a change inreactive material in the colloid deposits. Oil color that varies with the plane of polarization.red 0 stained frozen sections of lichen amy- The birefringcnt color of colloid can be changedloidosis were essentially negative and the from yellow with shades of orange to bluish-Sudan black B reaction was only weakly green by rotating the polarizer or analyzerpositive, as compared to a moderately positive lenses while keeping the deposits under directstaining of colloid by these methods. observation. Paraffin or frozen sections 10—15 Thioflavine T stained sections of amyloi- p. enhanced the prominence of birefringencedosis cutis tend to fluoresce a yellowish-green of the colloid deposits and could even be dem-similar to colloid, but this was a variable onstrated in unstained frozen sections of thosefeature. In general, the thioflavine T stained showing only minimal affinity for Congo red.sections of primary systemic amyloidosis cutis Paraffin sections cut at 15 p. and stainedshowed the most prominent fluorescence, but with thioflavine T showed prominent brightsome 15 p. paraffin sections of colloid exhibited silvery birefringence of the colloid, but di-a brillant yellow to green color. The thicker chroism was minimal. Thick sections stainedsections of 10—15 p. were usually best for dem- by techniques other than Congo red and thio-onstrating fluorescence in the thioflavine T flavine T also exhibit areas of birefringence,stained sections of lichen amyloidosis. but dichroism is essentially absent other than Birefringence and dichroism of amyloid the color of the usually seen in the tissuedeposits were minimal to absent in Congo red sections. Toluidine blue stained paraffin sec-stained 5 p. paraffin sections of lichen amy- tions 15 p. thick showed a striking silveryloidosis, whereas, the phenomenon was striking birefringence with a bluish sheen of the colloidin primary systemic amyloidosis cutis. Sec- deposits. The phenomenon of dichroism andtions 10—15 p. stained with Congo red and trichroism is best illustrated with Congo redthioflavine T enhanced birefringence in lichen staining of 30 p. sections, and the promi-amyloidosis and this was a similar feature nence of color changes seem directly related todescribed for colloid milium. The characteris- the affinity of the dye for the colloid deposits.tic of trichroism under polarized light is a Thick sections of colloid probably enhancedifference seen in colloid milium as compared to the yellow to green dichroism of amy- birefringence quantitatively. Morphologically, the birefringent components of colloid appearloidosis cutis. granular and fibrillar, and probably represent an abnormal scleroprotein different from col- DISCUSSION lagen and elastin. Current opinions in the literature are varied regarding the exact composition of the hyaline COMPARATIVE OBSERVATIONS WITH material demonstrated in routine II & E AMYLOIDO5IS CUTIS stained sections of colloid milium. Some inter- The significant results of histopathologic,pret colloid as abnormal elastic tissue (13, 14), histochemical,fluorescent and polarizationothers consider it a form of collagen degen- studies were as follows: Cutaneous amyloideration (15—17), while there are those that feel and colloid deposits show similar characteris-it is an alteration involving both collagen tics with some minor differences. A commonand elastin (18—21). Percival and Duthie (22) feature is that amyloid, like colloid showsfavored that colloid was not a connective tis- variability of staining reactions, particularlysue change, but a protein infiltration similar with Congo red and toluidine blue. Methylto amyloid. Subsequently, Zoon, Jansen and violet and crystal violet stained sections ofHovenkamp (23), and Becker and Wilson amyloidosis cutis show the amyloid deposits(24), by paper chromatography showed that have a tendency to appear pink to red, whereascolloid has an amino acid composition iden- colloid is usually purple with elements of red.tical with serum proteins. They (23, 24) dem- 504 THEJOURNAL OF INVESTIGATIVE DERMATOLOGY onstrated that hydroxyproline was absent inwere a few scattered fibrillar elements showing colloid and serum, but present in collagen anda soft residue of birefringence which faded clastin, and concluded that colloid originatesout toward the center of the colloid nodule. from serum proteins. Woolridge and FrerichsElectron microscopic studies by Piredda (16) (25), in 1960, reported a 12 year old boy withshowed the amyloid in lichen amyboidosis a clinical type of amyloidosis involving thesubstance mixed with fibrillar elements which cheeks, nose, and sides of the neck. They (25)appeared as intensively altered collagen. Pi- interpreted the deposited material as proteinredda (16) concluded that colloid represented in character, and that the patient showeda progressive degeneration of collagen and is features of amyloidosis cutis and colloid mil-not a foreign substance deposited in the skin ium. Woolridge and Frerichs (25) speculatedwith secondary morphologic changes. that future investigations may prove colloid Our observations of birefringence and color and amyloid are closely related substances, butchanges (dichroism and trichroism) in sec- finally classified their patient as cutaneoustions of colloid milium indicated to us several amyloidosis. Sullivan and Ellis (21) reportedyears ago that colloid was probably structur- an unusual case resembling that of Woolridgeally similar to amyloid, and that additional and Frerichs (25), and mentioned that thereultrastructure studies would demonstrate this are many clinical and histologic similaritiesto be true. Personal communication from of colloid and amyloid. Hashimoto and Bereston (35) regarding pre- Polariscopic studies have demonstrated thatliminary electron microscopic studies of col- amyloid stained by Congo red exhibits hire-bid milium are as follows: colloid appears to fringence and dichroism (7, 10, 26—31). Thesebe amorphous with amyloid-like filaments. A characteristics indicated an underlying or-striking feature is that every small or large ganized structure might be present, and sug-island of colloid is surrounded by fibroblasts gested that a study of the ultrastructure mightwhich show intracytoplasmic precursor ma- be fruitful. Tn 1959, Cohen and Calkins (32),terial. Higher resolution studies will be nec- demonstrated that amyloid had a finely fibrousessary, but the dimensions of the filaments component. Hashimoto, Gross and Lcver (33),with some beaded pattern in colloid miium showed the amyloid in lichen amyloidosisare compatible with similar features demon- consisted of filaments embedded in an amor-strated for amyloid (33). phous ground substance, and concluded that Histochemically, with only minor differen- the material is an abnormal secretion productces, colboid is similar to amyloidosis cutis. laid down by fibroblasts. Gafni, Merker, Shi-Our study and those of many others on amy- bolet, Sohar and Heller (34), studied a scalploidosis cutis document the variability in the amyloid tumor and demonstrated fibroblastsstaining reactions for amyloid. Variation in showing signs of active secretion in a sea ofresults of staining colloid and amyloid can amyloid filaments. They (34) thought theiroccur from laboratory to laboratory, and even observations supported the concept that amy-from numerous technical problems in the same bid is a variant of one of the normal sclero-laboratory. Technical problems undoubtedly proteins. In a recent review article, Cohenarc responsible for some conflicting results (31) indicated that fine amyloid fibrils arc areported for colboid and amyloid, but a con- consistent finding in all types of amyloidosis,sideration is that these substances arc of var- and this substance is formed locally by retic-iable composition which will alter the staining ubu-cndothelial cells in human and experi-reactions from case to case, and in different mental animal tissues. The amyloid fibrilstissues in the same patient. Congo red has arc protein in nature, probably containingbeen the most variable stain for colloid and small amounts of , and are em-amyloid in our laboratory, but longer staining bedded in ground substance, which itselfand/or thicker tissue sections have usually may be normal or abnormal (31). given positive results. Even with minimal Piredda (16) examined 10—12 sections ofstaining reaction, Congo red enhances hire- colloid milium polariscopically and concludedfringence and is mainly responsible for color that birefringencc was essentially absent. Re-changes exhibited by colloid and amyloid view of the article (16) seems to suggest thereunder polarized light. The phenomenon of COLLOID MILIUM 505 dichroism occurs as a result of altering thematerial positive, and not the color or color plane of light by rotating the polarizer lensesintensity. while examining Congo red stained sections Regarding the origin of colloid, it is our of colloid or amyloid. Puchtler, Sweat andopinion that this substance is closely related Levine (7)concludedthat Congo red stainingto amyloid occurring in the skin, but the of amyloid is by a nonionic binding by hy-etiology of both diseases remains a mystery. drogen bonding between the hydroxyl groups ofWhatever the causative factor or factors are amyloid and the amino groups of the dye.in colloid milium, they undoubtedly stimulate Gueft and Ghidoni (36) suggested that Congothe dcrmal fibroblasts to function in an ab- red staining of amyloid may be analogous tonormal way leading to the synthesis of a cotton fiber staining, and the double fibrilpathologic sclcroprotein, analogous to its nor- configuration may physically sequester themal counterparts, collagen, clastin and kcratin. fiat Congo red molecule, forming an inclusionStudies (38—40) indicate that dcrmal fibro- compound. It is probable that Congo red andblasts are the source of acid mucopolysac- other show affinity for colloid similar tocharidcs in the ground substance and pre- amyloid because of the fibrillar nature of thesecursors of collagen and elastic tissue. In 2 substances. Birefringcncc of thiofiavinc Tpathologic conditions such as colloid milium and stained sections of colloid and primary sys-amyloidosis cutis, fibroblasts are influenced to temic amyloidosis cutis was striking. One canproduce a pathologic scleroprotein and greater speculate that thioflavinc T may show affinityamounts of acid mucopolysaccharidc (par- for the fibrils of colloid and amyloid similarticularly hyaluronic acid) at the expense of to that suggested for Congo red. There is littlethe normal formed elements, and this results doubt that the organized structures demon-in partial to complete replacement of retic- strated by fluorescent microscopy of thiofiavinculum, collagen and elastin. Like amyloid, T stained sections of colloid and amyloidcolloid can probably be classified as a rare in account for the brillant bircfringencc seenvivo solid pathologic protein which causes when the same sections arc examined underhuman disease. The presence of an amino acid polarized light. Toluidine blue, crystal violetspectrum in colloid milium which is similar and methyl violet stain colloid and amyloidto serum proteins is easily explained on the mctachromatically principally because of thebasis of periodic deposition of blood elements presence of hyaluronic acid, and probably to ain areas of byaline formation. Hyalinc in- lesser degree because of sulfated acid muco-filtration of small blood vessels in the region polysaccha rides. The reason for toluidinc blueof colloid deposits undoubtedly weakens these enhancing birefringence of colloid under po-structures resulting in fragility and escape of larized light even after fading of the mcta-serum proteins. This feature is analogous to chromasia is difficult to explain, but one ofthat occurring in lesions of primary systemic the reasons for the affinity that all the dyesamyloidosis cutis in which the blood vessel show for colloid may be physical rather thanwalls arc infiltrated by amyloid resulting in chemical because of fibrillar structures. vascular fragility, with gross evidence of Staining of colloid with thioflavinc T haspctcchiae and ccchymoscs in the region of the generally been reported as giving negativepapnlcs and nodules. results (6, 21). Heyl (37), in 1966, reported 4 The influence of sunlight as the cause of cases of clinically and histologically typicalcolloid milium is frequently mentioned in the colloid milium which stained positive withliterature and textbooks of dermatology. The thiofiavinc T. Heyl (37) mentioned the nega-absence of solar clastosis in one of our pa- tive results reported by Kurban (6) andtients, others reported in the literature, and thought the contrasting results could be ex-the rare occurrence of cutaneous colloid de- plained by usc of different optical systems orgeneration on covered parts of the body (15) variation in the pre-staining treatment oftends to minimize sunlight as a primary cause tissue blocks. Our results with thiofiavinc Tof colloid milium. Because of the striking showed a yellow to green fluorescence of colloiddistribution of colloid milium to sun-exposed deposits and we agree with Heyl (37) thatareas of the body and its more common oc- fluorescence is the criterion for judging thecurrence in Southern regions, it is possible 506 THE JOURNAL OF INVESTIGATIVE DERMATOLOGY that aetinie rays may act as a promoting fac- stain for fluorescence microscopy of amyloid. J. Histochem. Cytochem., 13: 693, 1965. tor after unknown initiating agents have12. Armed Forces Institute of Pathology. Manual triggered the disease. of Histologic and Special Staining Technies, ed. 2. New York, The Blakiston Division, McGraw-Hill Book Company, Inc., 1960. SUMMARY 13. Ferreira-Marques, J. and van Uden, N.: Die elastosis colloidalis conglomerata. Arch. Our histopathologie, histoehemical, fluores- •Derm. Syph., 192: 2, 1950. cent microscopic, and polaroseopie studies14. Findlay, G. H.: On elastase and the elastic dystrophies of the skin. Brit. J. Derm., 66: strongly suggest that eolloid is similar to amy- 16, 1954. bid. Our observations of fluorescent staining15. Reuter, M 3. and Becker, S. W.: Colloid de- of colboid with thioflavine T and birefringenee generation (collagen degeneration) of the skin. Arch. Derm. Syph. (Chicago), 46: 695, of the hyaline substance under polarized light 1942. indicate an organized fibrillar structure. It is16. Piredda, A.: Le alterazioni ultrastrutturali nello pseudo milio colloide. Giorn. It. Derm. our opinion that colloid represents an ab- Sif., 99: 496, 1958. normal seleroprotein synthetized by dermal 17. Agius, 3. R. G.: Colloid pseudomilium. Brit. 3. Derm., 74: 55, 1963. fibroblasts, and this is a solid pathologic pro-18. Arnold, H. L., Jr.: Colloid pseudomilium. Arch. tein different from collagen and elastin. The Derm. Syph. (Chicago), 48:262,1943. fibrillar component of colloid is embedded in19. Prakken, J. R.: Colloid and senile degeneration of the skin. Aeta Derm.-venereol. (Stock- a matrix of acid mueopolysaccharides, com- holm),31: 713, 1951. plex mucosubstanees, and solidified serum20. Torres, E. A., Radice, 3. C., Villafane, N. U. and Vaccaro, S.: Seudomilium Coloide. proteins. Prensa med. Argent., 44: 566, 1957. 21. Sullivan, M and Ellis, F. A.: Facial colloid REFERENCES degeneration in plaques. Arch. Derm. (Chi- eago),84: 816, 1961. 1. Wagner, E.: Colloid-Milium of the skin. Arch.22. Percival, G. H. and Duthie, D. A.: Notes on d. Heilk., 7: 463, 1866. a ease of colloid pseudomilium. Brit. 3. 2. Movat, H. Z.: Demonstration of all connective Derm. Syph., 60: 399,1948. tissue elements in a single section: Penta-23. Zoon, 3. 3., Jansen, L. H. and Hovenkamp, A.: chrome stains. Arch. Path. (Chicago), 60: The nature of colloid milium. Brit. J. Derm., 289, 1955. 67: 212, 1955. 3. Johnson, W. C., Graham, J. H. and Helwig,24. Becker, J. F. and Wilson, H. T. H.: Colloid E. B.: Histoehemistry of the acid mucopoly- Milium. Brit. J. Derm., 68: 345,1956. saccharides in cutaneous calcification. J. In-25. Wooldridge, W. E. and Frerichs, J. B.: Amy- vest. Derm., 42: 215, 1964. loidosis: A new clinical type. Arch. Derm. 4. Johnson, W. C.: Histochemistry of cutaneous (Chicago), 82: 230, 1960. ground substance. In: Application of Modem 26. Missmahl, H. P. and Hartwig, M.: Polarisa- Investigative Techniques to Experimental tionsoptische untersuchungen an der amy- Pathology, p. 33. ed. by Bajusz, E. and loid-substanz. Virehow. Arch. Path. Anat., Jasmin, M., Basel/New York, S. Karger, 1966. 324:489, 1953. 5. Lillie, R. D.: Histopathologic Technic and 27.Puehtler, H., Sweat, F, and Kuhns, 3. G.: On Practical Histoehemistry, pp. 557—558, 519, the binding of direct cotton dyes of amyloid. 101, 11—13. ed. 2. New York, The Blakiston 3. Histochem. Cytochem., 12: 900, 1964. Division, McGraw-Hill Book Company, Inc., 28. Rosenthal, S. I., Puchtler, H. and Sweat, F.: 1965. Paper chromatography of dyes. Method to 6. Vassar, P. S. and Culling, C. F. A.: Fluorescent investigate vagaries in histological staining. stains, with special reference to amyloid and Arch. Path. (Chicago), 80: 190, 1965. connective tissues. Arch. Path. (Chicago),29. Sweat, F. and Puchtlcr, H.: Demonstration of 68: 487, 1959. amyloid with direct cotton dyes. Arch. Path. 7. Kurban, A. K.: Fluorescent stain for amyloid: (Chicago), 80: 613, 1965. An easy histological differentiation between30. Shirahama, T. and Cohen, A. S.: A Congo red cutaneous amyloidosis, colloid milium and staining method for epoxy-embedded amyl- senile elastosis. Bull. Johns Hopkins Hosp., oid. 3. Histochem. Cytochem., 14: 725, 1966. 107:320,1960. 31. Cohen, A. S.: Studies on the amyloid fibril. 8. Puehtler, H., Sweat, F. and Levine, M.: On the In: The Connective Tissue, p. 81. ed. by binding of Congo red by amyloid. J. Histo- Wagner, B. M. and Smith, D. E., Williams chem. Cytoehem., 10: 355,1962. 9. McManus, J. F. A. and Mowry, R. W.: Stain- and Wilkins Co., Baltimore, 1967. ing Methods: Histologic and Histochemieal,32. Cohen, A. S. and Calkins, E.: Electron micro- pp. 150—151. New York, Paul B. Hoeber, scopic observations on a fibrous component Inc., Medical Book Dept., Harper and Row, in amyloid of diverse origins. Nature, 183: 1960. 1202, 1959. 10. Pearse, A. 9. E.: Histochemistry, Theoretical33. Hashimoto, K., Gross, B. G. and Lever, W. F.: and Applied, pp. 281—287. Boston, Little, Lichen amyloidosis:Histochemieal and Brown and Co., 1960. electron microscopic studies. 3. Invest. 11. Puchtler, H. and Sweat, F.: Congo red as a Derm., 45: 204, 1965. COLLOID MILIUM 507

34. Gafni, J., Merker, H. J., Shibolet, S., Sohar, E. in dermatopathology. Transact. St. Johns and Heller, H.: On the origin of amyloid: Hosp. Derm. Soc., 52:84,1966. Study of an amyloid tumor in multiple38. Morris, C. C. and Goodman, G. C.: Production mycloma. Ann. Tnt. Med., 85:1031,1966. of acid mucopolysaccharides by fibroblasts 35. Hashimoto, K. and Bereston, E. S.: Colloid in cell culture. Nature, 188:407,1960. Milium: An electron microscopic study (in39. Zelickson, A. S.: Fibroblast development and preparation). fibrogenesis. Arch. Derm., 8:497,1963. 36. Gueft, B. and Ghidoni, J. J.: The site of for-40. Fullmer, H. M.:Thehistochemistry of the mation and ultrastructure of amyloid. Amer. connective tissues. In: Intern. Rev, of Con- J. Path., 43:837,1963. nective Tissue Res., Vol 3, p. 1. New York, 37. Hcyl, T.: Amyloid staining with thioflavine T Academic Press Inc., 1965.