A Histological Study of the Nail Region in Normal Human Subjects and in Those Showing Splinter Haemorrhages of the Nail by B

[ 323 ]

A HISTOLOGICAL STUDY OF THE NAIL REGION IN NORMAL HUMAN SUBJECTS AND IN THOSE SHOWING SPLINTER HAEMORRHAGES OF THE NAIL BY B. F. MARTIN AND M. M. PLATTS Department of Anatomy and the Department of Medicine, The University, Sheffield

INTRODUCTION Prior to undertaking a histological investigation to determine the microscopic appearance, location and possible source of the so-called 'splinter haemorrhages' of the nail, a detailed study was made of the soft tissues adjacent to the nail plate (perionychium) in normal subjects, and from this investigation a number of observations concerning the normal histological features of the region were made which warrant description and comment. In a recent clinical study of splinter haemorrhages (Platts & Greaves, 1958), it was found that though they occur most commonly in subacute bacterial endocarditis and in mitral stenosis (conditions in which they are well known to be a feature), they do in fact occur in a wide variety ofconditions and are not uncommon in otherwise normal individuals. In the latter, they usually occur following minor trauma to, or jarring of the fingers, even without direct trauma to the nails: they are sometimes, but not always, accompanied by discomfort and the individual may be unaware of their presence. Macroscopically, 'splinters' appear as tiny linear structures, usually not above 2-3 mm. long, and are arranged in the long axis of the nail. They were found to originate within the distal one-third of the nail, in relation to a pink (vascular) line, situated about 4 mm. proximal to the tip of the finger (PI. 1, fig. 1), and they may be multiple. When first formed they are plum-coloured, but in a day or two, darken to brown or black. The splinters move distally with the growth of the nail and eventually can be scraped from its under surface: splinters so removed were found to give positive colour tests for altered blood, for example, with the benzidine, cyanol and leuco-malachite-green tests. The nature of splinter haemorrhages has been a matter of conjecture, but it has been assumed that they are vascular phenomena, either small haemorrhages or minute emboli. Although they are often found in conditions in which vascular emboli are a feature, suggesting that they too may be embolic in nature (Platts & Greaves, 1958), the fact noted above, that they may occur in healthy subjects following minor trauma to the fingers is more suggestive of a haemorrhagic origin. It is very doubtful whether they are of the nature of petechiae however, since, as shown by Hess's test, the capillary fragility is not increased in cases of splinter haemorrhage (Platts & Greaves, 1958). 324 B. F. Martin and M. M. Platts

MATERIALS AND METHODS The terminal segment of a normal finger was removed from each of ten cadavers at post-mortem, and one from each of ten cadavers showing splinter haemorrhages of the nail; the latter specimens were taken from cases which had suffered a variety of terminal illnesses, though none were available from cases of subacute bacterial endocarditis. In the past, it has been usual to decalcify the terminal bony phalanx prior to sectioning a finger end, but it is preferable to remove the phalanx if this can be achieved without damage to the nail bed. In a recent study, Lewis (1954) removed the phalanx by drilling it out from the volar aspect of the finger by means of a special burr driven by a dental drill. The method used in the present study was to soak the finger end for several days in Heidenhain's 'susa' fixative, and then dissect out the terminal phalanx by first making a complete mid-line incision through the finger end, cutting through the nail and soft tissues down to the phalanx, and finally dissecting the two halves of the finger end completely away from the phalanx. After several more days' fixation in 'susa', the two halves of the finger end were embedded in paraffin wax and one, or occasionally both, were sectioned at 12 ,t on a sledge microtome. As a routine, every tenth section was mounted and stained with haematoxylin and eosin, intervening sections being examined when indicated. With one exception, all specimens were sectioned longitudinally; the single specimen sectioned transversely contained a splinter haemorrhage and showed some interesting features. Although splinter haemorrhages of the toe nails do occur, they were not studied histologically. RESULTS Before dealing with the features shown in cases of splinter haemorrhage, it is necessary to indicate the terminology used in the present work with reference to the soft tissues adjacent to the nail plate and also to deal with the relevant histological features of the region. (a) The normal histology of the nail region The appearance of a normal finger end in longitudinal section is shown in PI. 1, fig. 2. The nail plate (N) tapers posteriorly to the nail root, which is overlapped by the proximal nail fold (P). The deeper layer of the epithelium of this fold is narrower than the superficial layer and differs from it in not possessing obvious papillae. Where the stratum corneum of this fold comes into contact with the nail it is called the eponychium (E), and this, together with the stratum granulosum, passes back along the dorsal surface of the nail to within 1 mm. of the nail root; at this point both terminate and the dorsal matrix begins (DM in PI. 1, fig. 3). The distal free portion of the nail lies on a part of the stratum corneum of the finger tip, called the hyponychium (Hy in PI. 1, fig. 2): together with the stratum granulosum, this extends for only a short distance under the nail; in the specimens examined, it terminated on an average 4 mm. proximal to the finger tip. Immediately distal to the hyponychium is a surface depression (F in PI. 1, fig. 2) called the distal limiting furrow (Lewis, 1954). Histological study of the nail region 325 The main body of the nail rests upon the epithelium of the nail bed, the proximal part of which is differentiated as the volar matrix (VM in P1. 1, fig. 8) and this roughly corresponds in extent with the lunule of the nail. Apart from the matrix area, the nail-bed epithelium is of the same structure as the stratum germinativum (Malpighian layer) of the epidermis, but instead of dermal papillae below it there are longitudinal rows of dermal ridges; these, together with the corresponding epithelial ridges, were well seen in longitudinal sections taken from near the side margin of the finger end, because, due to the curvature of the nail and nail bed, they are here cut tangentially (P1. 1, fig. 7). As the epithelial ridges are traced from the finger tip towards the lunule, they decrease in height and increase in number by splitting (Horstmann, 1957): this fact can be appreciated by comparing figs. 5 and 6 of PI. 1. The epithelium of the volar matrix differs from that of the remainder of the nail bed. In the majority of sections studied, the volar matrix had split during sectioning into two distinct zones of about equal thickness (PI. 1, fig. 3). These two zones are structurally different; the cells of the deeper zone are polygonal, have weakly eosinophilic cytoplasm and rounded vesicular nuclei, whilst those of the superficial zone are flattened, have deeply eosinophilic cytoplasm and dark, shrunken, irregular (pyknotic) nuclei (P1. 1, fig. 4). The cells of the upper, deeply eosinophilic zone have a matted or faintly fibrillar appearance, due to the presence of tonofibrils, and from the superficial layer, cells can be seen entering the nail plate. Intercellular bridges or 'prickles' could be distinguished in this zone, but they were shorter and less distinct than those of the general stratum germinativum of the nail bed. Regarding the connective tissue component of the nail bed, Horstmann's (1957) observation, that the collagen fibre bundles are arranged mainly in a criss-cross lattice arrangement and have scarcely any elastic fibres mixed with them, was confirmed. In the present study elastic fibres were found only among the collagen bundles of the volar aspect and tip of the finger end. Sweat glands were not seen in the connective tissue of the nail bed but were very numerous in the volar aspect and in the tip of the finger end as far as the hyponychium region. Lamellated corpuscles (Vater-Pacini) were also numerous in the same regions. The blood vessels of the nail bed are of particular interest. The veins are thin- walled and are connected with simple capillary loops which project towards the nail-bed epithelium (PI. 1, fig. 10). The arteries were found to be of a special type, in that they have an inner longitudinal and an outer circular coat of smooth muscle, and have no internal elastic lamina (PI. 1, fig. 8). Arterio-venous anastomoses, of the type known as glomi, are well known to be present in the connective tissue of the nail bed. In the specimens studied, they were found not only in the volar aspect and tip of the finger end, but throughout the connective tissue of the nail bed, even below the volar matrix, but none was seen in the proximal nail fold. Reference to fig. 9 of Plate 1 shows the manner in which a glomus is formed from an arterial branch: the branch becomes much coiled and surrounded by concentric layers of connective tissue, whilst the smooth muscle cells of its wall undergo a peculiar epithelioid change. The lumen of the structure often appears to be practically obliterated in histological sections. 326 B. F. Martin and M. M. Platts In the region just below the hyponychium, vessels of a very particular kind are found. They have been described and illustrated by Horstmann (1957). In this region, the subungual dermal ridges give way to large papillae, two to three of which are seen in each longitudinal section (L in PI. 1, fig. 2); they are referred to by Horstmann (1957) as the 'plump' papillae. These papillae contain long, thin-walled looped vessels of large calibre, which are spirally wound. Horstmann (1957) describes them as capillaries and likens them to heating spirals. It is difficult to say whether they should be referred to as veins or capillaries, but for convenience they will be referred to in the present work as capillaries. Examination of the finger ends from cases with splinter haemorrhages revealed that it is these vessels which rupture to give rise to the haemorrhages. (b) The histological features in cases of splinter haemorrhage In section, splinter haemorrhages showed as irregular amorphous masses, varying in outline and size. They were yellowish in unstained sections, and remained yellow after the sections were stained with haematoxylin and eosin, but they gave the typical bluish-green colour reaction for altered blood when treated by the cyanol method. In all cases, the 'splinters' were located in the hyponychium, lying in the angle between the nail plate and the termination of the stratum granulosum (S in PI. 2, figs. 15, 16, 17). A few splinters were found nearer to the tip of the nail and were presumably growing out with it, but in no instance was a splinter found proximal to the termination of the stratum granulosum and hyponychium. Measurements showed that the stratum granulosum and the hyponychium terminate on an average 4 mm. proximal to the finger tip. These histological observations are in keeping with the clinical observation reported above, that splinter haemorrhages first appear about 4 mm. proximal to the finger tip. In most specimens studied, the blood vessels of the nail bed and finger tip were distended and packed with blood cells (B in PI. 2, figs. 15, 16). Due to this factor, the looped, spirally wound vessels in the large papillae below the hyponychium were clearly demonstrated. In a few sections, the looped vessel was seen to terminate below the epithelium in a closely wound knot (V in P1. 2, fig. 11), but far more commonly, the vessel, after a spiral wind, terminated in a simple loop (e.g. V in PI. 2, fig. 13). These vascular loops, in addition to being distended with blood cells, were frequently seen to project towards the hyponychium, with very few layers of epithelial cells overlying the loop (P1. 2, figs. 12 and 13) and, in a few instances, a vascular loop actually projected directly into the hyponychium (P1. 2, fig. 14). Microscopic evidence that a splinter, which appears histologically as an amorphous mass, is developed from a haemorrhage was seen in one instance, and that was in the finger end which was sectioned transversely. The distal part of the splinter in this specimen had the usual appearance of a yellowish mass (S in PI. 2, fig. 18), but to one side of it a small collection of red blood cells (R) lay in a space in the hyponychium. In more proximal sections, a small vessel (V in PI. 2, fig. 19) was seen in close proximity to the hyponychium and was presumably the source of the small haemorrhage. The proximal portion of the splinter itself was of mixed com- position, part of it being composed of the usual yellowish material (S in PI. 2, fig. 19) Histological study of the nail region 327 and the remainder composed of a mass of red blood cells (R): in some sections the two components were intermingled (PI. 2, fig. 20). Further proximally still, the mass was composed solely of blood cells, which could be traced into the epithelium below the hyponychium, and here a vessel was present, lying in the centre of the mass of blood cells (PI. 2, fig. 21). Silver impregnation of adjacent sections showed that the red cell mass lay in a space in the epithelium, with no argyrophil fibres enclosing it: the small central vessel possessed the usual fibrous adventitia.

DISCUSSION The terminology used in this work with reference to the soft tissues adjacent to the nail plate follows common usage. It may be noted, however, that Horstmann (1957) applies the term 'hyponychium' to the whole of the epithelium underlying the nail plate, whereas this region is usually referred to as the epithelial component of the nail bed, and he applied the term 'Sohlenhorn' to the tissue usually referred to as the hyponychium. Yet, he applies the term 'eponychium' in the usual way to the thin layer of cornified tissue which overlies the base of the nail. It is doubtful whether Horstmann's application of the terminology is as logical as that usually employed. No special study was made of the nail plate in this investigation, since this has received detailed attention by Lewis (1954) and Horstmann (1957). Certain observations on the normal histology of the adjacent soft tissues were made, however, which warrant comment. It was confirmed that the nail matrix extends for a short distance on to the dorsal aspect of the nail root (Lewis, 1954; Horstmann, 1957), and that this dorsal matrix terminates where the stratum granulosum of the proximal nail fold begins: Horstmann found it to be 0.8 mm. long in a child's nail, and in the present study it was found to be 1 0 mm. long in the adult. The length of the matrix on the deep aspect of the nail root (volar matrix) measured 5 mm. on an average in the specimens studied. Horstmann (1957) states that there is not a clear demarcation between the volar matrix and the remainder of the epithelium underlying the nail plate, but this was not confirmed; on the contrary it was found that the demarcation was readily seen, due to the fact that the epithelium of the volar matrix differs from that of the remainder of the epithelium deep to the nail plate in possessing two distinct zones of cells, which usually separate during sectioning. The structure of the blood vessels of the finger end is of interest. It was found that the arteries possess an inner longitudinal and an outer circular coat of smooth muscle and have no internal elastic lamina. Arterio-venous anastomoses (glomi) are known to be common in the finger tip and connective tissues of the nail bed, and in the present study they were found to extend proximally deep to the volar nail matrix, but were not seen in the proximal nail fold. They are generally believed to act as temperature regulators by controlling blood flow. The veins of the region are thin walled; those below the nail-bed epithelium are connected to simple capillary loops but those below the hyponychium region are connected to loops which are long, of wide bore and spirally wound. Horstmann (1967) suggests that the latter, which he likens to heating spirals, may be concerned with temperature regulation at the extremities of the fingers, and that the blood flow through them is controlled by the arterio-venous anastomoses of the nail bed. 328 B. F. Martin and M. M. Platts From the present study, it is clear that so-called splinter haemorrhages of the nails are indeed haemorrhagic in origin, and that the source of the haemorrhages is from the special, spirally wound capillary loops in the large papillae which lie below the hyponychium. These vessels undoubtedly produce the pink line which is normally seen macroscopically under the nail, about 4 mm. proximal to the tip of the finger, and in relation to which splinter haemorrhages were observed to occur clinically. In the majority of specimens examined from cases of splinter haemorrhage, there was considerable engorgement of the vessels of the finger end with blood cells, and in many instances the special capillary loops were not only distended with blood cells but were seen to be separated from the hyponychium by only a few layers of epithelial cells, and on occasions to be in direct contact with it. It is not clear whether the capillary loops are brought to this close proximity with the hyponychium by vascular congestion or whether it is an inherent feature which predisposes the individual to suffer from splinter haemorrhages, particularly if the vessels become dilated. It is evident, however, that these long, thin-walled vessels would be particularly liable to rupture if distended, even from relatively minor trauma. That they rupture into the hyponychium is not surprising, in view of their close proximity to it. Although no emboli were seen in the vessels of the specimens examined, the possibility has not been excluded that in some instances a small embolus may lodge in a vessel prior to its rupture. After the blood has entered the hyponychium it undergoes a change, and becomes transformed to an amorphous mass, yellow in colour when viewed in section, but giving the usual colour reactions for altered blood. It is probable that the blood soon undergoes transformation after entering the hyponychium. This is suggested by the clinical observation that the haemorrhages, which are plum-coloured when first formed, become brownish or black within a day or two, and would account for the fact that in only one specimen of the series studied was an actual haemorrhage found, and even in this case part of the haemorrhage had already undergone the change to amorphous yellowish material. A clinical parallel to this example was a case of a recently formed splinter haemorrhage, observed by one of us (M. M. P.), in which the proximal part of the splinter was fluid in nature and faded on pressure, whereas the distal part was unaltered by pressure. Regarding the nature of the material composing the splinters, it was thought that it might be related to the substance known as ceroid. This substance is described as an acid-fast brown pigment, which appears as yellow globules in sections (Lillie, 1954), and one variety of it is produced when free red blood corpuscles come into contact with fatty material; in fact, Hartroft (1951) has produced it experimentally from these sources, both in vitro and in vivo, and he has also shown (Hartroft, 1952) that it is a constituent of atheromatous plaques (haemoceroid). The stratum corneum contains fatty material (Cowdry, 1932), so that, theoretically, contact between free red blood corpuscles and this tissue might produce ceroid. Some of the tests which are used for the demonstration of ceroid (see Lillie, 1954) were therefore carried out on a few sections of splinters. It was found that they are not sudanophilic, whereas sudanophilia is a characteristic of ceroid. Nevertheless, they show some of the characteristics of ceroid, namely, they are acid-fast (to carbol fuchsin), they give Histological study of the nail region 329 no reaction with Perl's test for iron, and with the periodic acid-Schiff (PAS) test small areas of the splinters react. Thus, the results of the tests applied, though somewhat equivocal, suggest that the material may be related to ceroid. Although in the present study, splinter haemorrhages were seen only within the distal one-third of the nail, haematomata have been described as occurring proximal to this region, and such extravasations become incorporated into the nail substance (Alkiewicz, 1933). Recently, Loewenthal (1958) described four cases of psoriasis of the nails in which splinter haemorrhages occurred and, although they were distal to the lunule, it is apparent from his illustration that at least some of them were well proximal to the hyponychium area. He found that they became incorporated into the nail substance, which suggests that the general nail-bed epithelium as well as the matrix contributes to nail formation. On removing a splinter haemorrhage from the nail substance he found that it gave a positive benzidine test for blood, and he considered the haemorrhage to be caused by rupture of dilated capillaries of the nail bed. It is concluded therefore that, whereas splinter haemorrhages normally occur deep to the distal one-third of the nail, after direct trauma or in disease of the nail itself they can occur between the nail and nail bed, and then they become incorporated into the nail substance. In local disease conditions, however, it is likely that the mechanism of formation is different from that described in the present work.

SUMMARY 1. Sections from normal adult finger ends as well as from those showing splinter haemorrhages of the nail have been studied, and the normal histology of the region is described. 2. Splinter haemorrhages were found to lie in the hyponychium, in the angle between the nail plate and the termination of the stratum granulosum. They are amorphous and yellowish in section, but give colour reactions for altered blood. One splinter was composed partly of yellowish material and partly of blood cells. It is suggested that the splinters may be composed of a material related to ceroid. 3. The vessels from which the haemorrhages originate are the large spirally wound capillaries which lie in the large dermal papillae deep to the hyponychium. We should like to record our thanks to Prof. F. Davies for his helpful criticism in the preparation of the manuscript, to Mr J. H. Kugler and Miss C. J. Crockford for their technical assistance, and to Dr J. L. Edwards for the supply of post- mortem material.

REFERENCES ALKizwIcz, J. (1933). Zur Histopathologie der Hamatome des menschlichen Nagels. Arch. f. Dermat. 168, 411-419. COWDRY, E. V. (1982). Special Cytology. 2nd ed., vol. 1. The skin, p. 19. New York: Paul B. Hoeber Inc. HARTROFT, W. S. (1951). In vitro and in vivo production of a ceroid-like substance from erythro- cytes and certain lipids. Science, 113, 678-674. HARTROFT, W. S. (1952). Ceroid-like pigments, hemoceroid and hyaloceroid, in atheromatous lesions in human subjects. Amer. J. Path. 28, 526-527. 330 B. F. Martin and M. M. Pkatts HoRsimANN, E. (1957). Der Nagel. In von Mdllendorff's Handbuch der mikroskopischen Anatomie des Menschen, III/3. Berlin: Springer. LEWIS, B. L. (1954). Microscopic studies of fetal and mature nail and surrounding soft tissues. Arch. Dermat. 70, 732-747. LILLIE, R. D. (1954). Histopathologic Technic and Practical Histochemistry. New York: The Blakiston Co. Inc. LOEWENTHAL, L. J. A. (1958). Streifige Punktblutungen bei Nagelpsoriasis. Der Hautartz. 9,92. PLATTs, M. M. & GREAVES, M. S. (1958). Splinter haemorrhages. Brit. med. J. 2, pt. 1, 143-144

EXPLANATION OF PLATES PLATE 1 Fig. 1. Finger nail showing two splinter haemorrhages (S) in relation to a pink (vascular) line (V). x3. Fig. 2. L.S. adult finger end, showing nail plate (N), proximal nail fold (P), eponychium (E), hyponychium (Hy), large papillae deep to the hyponychium (L) and the distal limiting furrow (F). H.& E. x4 5. Fig. 3. L.S. through nail root. Note the termination of the stratum granulosum (Gr) of the proximal nail fold, the dorsal matrix (DM) and the volar matrix (VM); the latter has split during preparation into two zones. H. & E. x 54. Fig. 4. Volar matrix, to show the two zones of cells; those of the upper zone are deeply eosinophilic, of matted appearance, and have pyknotic nuclei. Note cells entering the nail plate. H.&E. x185. Fig. 5. T.S. distal part of nail bed to show the epithelial ridges. H. & E. x 48. Fig. 6. T.S. proximal part of nail bed. The epithelial ridges are shorter here and have increased in number by splitting. H. & E. x 48. Fig. 7. L.S. from side margin of finger end, to show the subungual epithelial ridges, which have been cut tangentially. H. & E. x 20. Fig. 8. T.S. artery of nail bed. Note inner longitudinal (IL), and outer circular coat (OC) of smooth muscle and absence of internal elastic lamina. H. & E. x 150. Fig. 9. Arterio-venous anastomosis (glomus). H. & E. x 82. Fig. 10. Simple capillary loop below epithelium of nail bed. H. & E. x 100.

PLATE 2 Fig. 11. Complex capillary knot (V) in large papilla below the hyponychium (Hy). H. & E. x 68. Fig. 12. Capillary loop of wide bore (V), separated from hyponychium (Hy) by only a few layers of epithelial cells. H. & E. x 45. Fig. 13. Large spirally wound capillary (V) ending in a loop, and almost in contact with the hyponychium (Hy). H. & E. x 68. Fig. 14. Capillary loop (V) projecting into the hyponychium (Hy). H. & E. x 68. Fig. 15. Splinter haemorrhage (S) in the hyponychium, lying in the angle between the nail plate (N) and the stratum granulosum (Gr). Vessels (B) are filled with blood cells. H. & E. x 23. Fig. 16. Splinter haemorrhage (S) in hyponychium. Vessels (B) are distended with blood cells. H. &E. x23. Fig. 17. Splinter haemorrhage (S) in hyponychium. H. & E. x 23. Fig. 18. T.S. finger end. Splinter haemorrhage (S) in hyponychium. A small collection of blood cells (R) lies to the right of the splinter. H. & E. x 125. Fig. 19. Section proximal to that in fig. 18. The splinter haemorrhage is here composed partly of yellowish material (S) and partly of blood cells (R). The small vessel (V) to the right was probably the source of the blood cell collection seen in fig. 18. H. & E. x 125. Fig. 20. Section proximal to that in fig. 19. The splinter haemorrhage is composed partly of yellowish material (S) and partly of blood cells (R). H. & E. x 125. Fig. 21. Section proximal to that in fig. 20. The splinter haemorrhage is now composed entirely of blood cells (R), lying in a space in the nail-bed epithelium. A vessel (V) occupies the centre of the haemorrhage. H. & E. x 125. Journal of Anatomy, Vol. 93, Part 3 Plate 1

N Hy

I. r ( F

'..y4 .@

2 .4 I

- - tA - I:.fU-t.. - - -rjw.-.. -, a-t - IL

?' I.^ _ _ t t-*

-- - .4S. -'-----..- - -- -

MARTIN AND PLATTS-HISTOLOGICAL STUDY OF THE NAIL REGION (Facing p. 330) Journal of Anatomy, Vol. 93, Part 3 Plate 2

I ",* , I * -,. I

L .. ..

z S- oF ^wt~s - _ 2 A!~~~~I

- v -~- w ,

V r1N

X!A , w ,.),1E{_.,99 1= S~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~h1;¢vw,-.m.^/k

; ~ Y ~ ~ Gr / Li * >-A~~~~~~~sji5L ~~~~~ CY;Wm;wf ;\,a. ,;>_,/jt11111*-"'''fl.~~~~~~~~~~~~~~~

R I

16 S R s R t' s.V

, , -

MARTIN AND PLATTS-HISTOLOGICAL STUDY OF THE NAIL REGION