A C T A O P H T H A L M O L O G I C A VOL. 5 7 1 9 7 9
Depai tment of Ophthalmology (Head N Bhkrc),
if rhuc Kommunehoybpztal, Unzwrtzty of Aarhuc, Dmmnrk
THE SPECULAR MICROSCOPIC APPEARANCE
OF CORNEAL GRAFT ENDOTHELIUM DURING AN ACUTE REJECTION EPISODE A Case Report BY
THOMAS OLSEN
T h e specular microscopic appearance of a graft endothelium during a successfully treated acute endothelial rejection episode is reported. Coinciding with the increased thickness of the graft great disturbance in cellular morphology was observed. Individual cells were seen to appear swollen or to show bright intracellular areas. Also abnormalities with respect to cellular intersections were seen with the occurrence of junctional complexes at which more than three cells met. The significance of these abnormalities is discussed. The morphological changes subsided gradually as the graft cleared and the thickness decreased. This study examplifies the value of the non-contact specular microscopic technique which enables a thorough study of the diseased corneal endothelium without traumatizing the cornea and interfering with subsequent specular microscopic examinations.
K q umrd~:specular microscopy - cornea - endothelium -rejection
- graft
T h e appearance of corneal endothelium in response to trauma has been observed by means of specular microscopy following various forms of insult such as air in the anterior chamber (Leibowitz et al. 1974) osmotic dehydration of the cornea (Sherrard 1976, 1978) and ultrasound (Olson et al. 1978). Most of these studies have been carried out on animals, little information is available on the human endothelium which may apply to the clinical situation. The fact that in vi-cm specular microscopy of corneal endothelium requires a translucent cornea hinders to a certain extent a detailed study of human endoReceived February 7 , 1979.
882
thelium following trauma. The case presented here is a n exception to that rule in t h a t it was possible t o study the specular microscopic appearance of a corneal graft e n d o t h e l i u m shortly after onset of a n acute rejection episode.
Methods Central corneal thickness was measured with a modified Haag-Streit pachometer (Sperling & Ehlers 1977). Specular microscopy was performed using a non-contact technique described earlier (Olsen 1979). Case presentation This 50-year-old man had penetrating keratoplasty done for the first time in 1974 due to a vascularized corneal scar after herpetic keratitis. The grafting was successful until one year later where the patient experienced an acute rejection. Two weeks passed before he attended for control. In spite of topical and systemic steroid treatment the graft remained oedematous and became vascularized. Retransplantation was undertaken January 1977. Postoperatively the graft was resutured in the lower left quadrant hut otherwise the postoperative course was uneventful. The case was included as No. 1 in the study of Ehlers & Bramsen (1978). On November 7th 1978 the patient awaked in the morning with blurred vision. On examination the same day the graft was oedematous with epithelial bullae and scattered precipitates on the posterior aspect of graft. N o rejection line was present on this first day after onset, but the cornea was thicker in the lower part than in the upper part. Systemic treatment with prednison 30 mg daily, topical ultracortenol and subconjunctival decadrone phosphate in.jections were started immediately. On this treatment graft thickness decreased
CCT
9
endcIthelial photos
I
0.5 -
1
50 mg
0
0 l
~
,
0 2 L 6 810
~
~
20
~
30
~
LO
~
~
,
90 days after onset
Fig. 1. Diagram showing changes in central corneal thickness (CCT) (mm) of graft during an acute rejection episode. Black horizontal staircase line indicates systemic prednisone treatment.
~
Thomas Olsen
gradually (Fig. 1). O n the 3rd day after onset of symptoms a typical rejection line was present across the lower part of graft. This line was later broken u p into smaller precipitates. Below the line the cornea was constantly thicker than above the line. T h e same tendencdy was found in the ensuing weeks in which the graft gradually cleared. T e n days after onset no precipitates were found with the slit lamp.
Fig. 2. Endothelial cells of graft visible in the specular microscope 5 days after onset of rejection. Top: photo from upper part of graft. Centre and buttom: photos from lower, thicker and more heavily attacked part of graft. T h e endothelial morphology is severely disturbed with swelling of cells (examples shown by arrows) intrarellular bright areas (asterixes) and abnormal cellular junctional complexes (circles). Horizontal bar = 100 pm. 884
Non-contuct Sperulnr Microztopv
Specular microscopy For the first 4 days after onset of rejection episode it was not possible to observe and photograph the endothelium due to opacification of the graft.
5th duy after onset (Fig. 2) T h e endothelial cells became discernible. At places no planar reflex from the endothelium was obtainable probably due to great distortion of endothelium. In those places in which the cells were visible, some cells showed dark round intracellular areas which made the cells appear swollen as interpreted from the absent reflex from the center of the cells (arrows in Fig. 2). Such cells also appeared more round than cells without this phenomenon. Other cells showed intracellular variation with respect to intensity of reflected light giving the impression of bright intracellular areas (asterixes in Fig. 2). Cells were larger and more irregular in the lower than in the upper part of cornea. Abnormalities were also found with respect to cellular junctions. Normally cell boundaries meet each other at points where three cells meet. Cell meetings involving four or more cells are practically never seen in normal endothelia (Fig. 6). In some places cells were found to meet four at a time (circles in Fig. 2), thereby greatly disturbing the regular hexagonal pattern.
14th day ufier onset (Fig. 3) Graft almost clear. The endothelial cells now reflected the light in a more even fashion. Intracellular round dark areas were less pronounced but still intracellular bright areas were present. In Fig. 3 abnormal cellular intersections are especially evident showing cell meetings involving more than three cells (circles in Fig. 3).
Fig. 3.
Endothelium of graft 14 days after onset of rejection. Photograph from lower part of graft. Squares indicate abnormal cellular junctions with dark spots occurring in the centre. Other symbols explained in Fig. 2. Bar = 100 pm.
885
Thomas Olsen
Fig. 4. T h e specular microscopic appearance of graft 21 days after onset of rejection. Photomicrograph from lower part of graft. Size and shape of cells show considerable variation with small cells with convex vorders (arrows) appearing between larger cells. Bar = 100 pm.
Fig. 5 . 93 days after onset of rejection the endothelium has reorganized into a more regular pattern. Top: upper part of graft. Buttom: lower part of graft, Note absence of previous marked difference in cell size from top to buttom of graft. Overall cell density is about 460 cells/mm2 which is clearly much lower than on previous photomicrographs. Bar = 100 p.
886
Non-contact Specular Microscopy
Mostly these abnormal junctional complexes consisted of joints of four cells, but in Fig. 3 a junctional complex of five cells is also found (circle in right part of Fig. 3). The cell pattern is far from the normal hexagonal regular pattern, the morphological changes being more pronounced in the lower than in the upper half of cornea.
21st ahy after onset (Fig. 4) T h e graft now appeared clear. Cell size and shape showed still considerable variation with small cells with convex borders (arrows in Fig. 4) appearing between larger irregular cells. Cell meetings where more than three cells met were only occasionally encountered (circle in Fig. 4).
93rd day after
onset (Fig. 5 ) Graft thickness had now stabilized at a value somewhat above the thickness before the onset of rejection. Size and shape of cells were now more uniform but not completely regular. Whereas in the foregoing specular microscopic examinations
Fig. 6. T h e specular microscopic appearance of normal corneal graft endothelium. Top: male, 62, endothelial density 1090 cells/mm*. Buttom: female, 48,endothelial density 1910 cells/mmz. Bar indicates 100 pm.
887
Thomas Oken
there was a considerable difference in size of cells from photographs at the upper and lower part of graft, this difference now was less apparent. The overall number of cells per unit area at this time was clearly much lower than that of normal graft endothelium (Fig. 6) and that of the first specular microscopic examinations, but it was difficult to give figures for this difference due to the large variation in cell density observed within the graft. No clearcut abnormalities within cells or with respect to cellular intersections were found.
Discussion In interpreting the endothelial changes observed in this study it is probably sound to regard them as partly due to the nature of the actual disease and partly due to the secondary healing mechanisms of the endothelium. Graft rejection is generally believed to be a cell-mediated immunological manifestation. With the specular microscope it was not possible to observe any lymphocytes on the endothelium although precipitates were seen with the slit-lamp. It was remarkable, however that the cellular abnormalities in the endothelium were so widely dispersed and occurred at places without precipitates. Furthermore, the rejection line seen with the slit-lamp did not sharply divide the endothelium into an abnormal and a normal area. These non-localized changes may be the result of soluble factors involved in the cellular hypersensitivity reaction, for which a number of mediators are now known to be involved (see David 1973) among them macrophage inhibition factors and factors cytotoxic for target cells. Some of these factors may activate lysosome enzymes. Interestingly, this process has been shown to result in swelling and rounding of target cells and their organelles followed by lysis of the cell (Able et al. 1970). This may be the explanation for the apparent swelling of endothelial cells observed in the present study (Fig. 2). The observed bright intracellular areas observed in the present study bear a certain resemblance to those *eventsndescribed in rabbits (Sherrard 1976, 1978; Leibowitz 1974) after air or osmotic dehydration of cornea. Sherrard showed that some of these events may progress to cell death after which a characteristic rosette is formed by adjacent cells spreading to cover the defect (Sherrard 1976). Such rosettes formed after acute trauma have been described by others in uivo (Olson et al. 1978) and in uitro (Sperling 1978). A characteristic feature for these rosettes is that the centre forms a point at which more than three cells meet depending on the number of cells surrounding the cell which has been destroyed. The junctional complex of five cells described in Fig. 3 is in fact a rosette. The abnormal junctional complexes of four cells, which was of far more frequent occurrence are, however, not identical to such rosettes. 888
Non-contact Sprcular Mzcro \copy
According to Sherrard (1 976) and Olson (1978) these rosettes became reorganized within a few hours leaving a cell pattern almost indistinguishable from a normal endothelium. Based on in vitro studies Sperling (1978) noted that joint meetings of four cells may stay longer than the rosettes and suggested that joints of four cells actually may be formed from the rosettes. The abnormal cellular intersections found in this study, i. e. junctional complexes in which more than three cells meet, may thus represent recent cell death. It was not possible to make valid estimation of endothelial cell loss due to large variation in cell size from one part of the graft to another and the inability to photograph exactly the same area after periods of time. It may be suggested that a great cell loss occurred during the rejection episode (compare Fig. 2 with Fig. 5 ) , and that the greatest cell loss occurred in the lower part of the graft as judged from the greater morphological changes and the greater thickness of graft in this region. Three months after onset (Fig. 5 ) , however, no marked difference in cell density was observed between upper and lower part of graft. This indicates that a certain sliding of cells towards regions of lower cell density has occurred, a phenomenon which has also been described after cataract extraction (Rao et al. 1978). T h e functional impairment of the endothelium accompanying the observed morphological changes was manifested through the changes in graft thickness (Fig.1). T h e return of a clear, thin graft 2-3 weeks after onset of rejection was accompanied by the specular microscope appearance of a planar endothelium with no major intracellular or intercellular abnormalities, although a certain degree of cellular pleomorphism still existed (Fig. 4). This pleomorphism had reduced on the last examination (compare Figs. 4 and 5 ) . This indicates the process of active cell death and the immediate healing response is more deleterious for the barrier function of the endothelium than the process of reorganizing the cellular pattern from a state of pleomorphism to a more regular pattern. In this study non-contact specular microscopy has proved very useful in studying morphologic alterations of endothelial cells following an insult. With the noncontact technique the whole endothelium can be scanned without traumatizing the epithelium and interfering with subsequent specular microscopic examinations.
Acknowledgments T h e technical assistance of Mrs. Anette Poulsen is gratefully acknowledged. This work was supported by Fonden ti1 Lzgevidenskabens Fremme.
889
Thomas Olsen
References Able M. E., I.ee ,J. & Rosenau W. (1970) 1.vmphocyte-target cell interaction in 71itl-o.Amur. 1. Pnth. 60, 42 1-427. David .J. R. (1973) Lvmphocyte mediators and cellular hypersensitivity. ,\‘mi Eng. ,]. M u d . 288, 143-149. Ehlers N . & Sperling S. (1977) A technical improvement of the Haag-Streit pachometer. Actn ophthnl. (KM.) 56, 53-66. Ehlers N. & Rramsen T. (1978) Management of late corneal graft prohlerns. Actrr op/ithd (Khh.) 56, 984-997. I.eibowit7 H. M., Laing R. A. & Sandstrom M. (1974) T h e effect of air in the anterior chamber . A rrh . Ophthn 1. iC:hicci,goj 92, 22 7-2 3 0 , Olsen T. (1979) Optical principles for estimation of endothelial cell density with the non-contact specular microscope. .4cf(r ophthol. (Khh.) 57, 860-867. Olson 1.. E., Marshall ,J., Rice N. S. C . 8c Andrews K. (1978) T h e effects of ultrasound on the corneal endothelium. I . T h e acute lesion. Brit.J. Ophthal. 62, 134-144. Olson L. E., Marshall J., Rice N. S. C. & Andrews R. (1978) Effects of ultrasound on the corneal endothelium. 11. T h e endothelial repair process. Brit.,]. Ophthal. 62, 145-154. Rao G. N., Shaw L. E. & Aquavella J. V. (1978) Morphological appearance of the healing corneal endothelium. Arch. Ophthal. (Chicago) 96, 2027-2030. Sherrard E. S. (1976) T h e corneal endothelium in uivo: its response to mild trauma. Exp. Eye Res. 22, 347-357. Sherrard E. S. (1978) Characterization of changes observed in the corneal endothelium with the specular microscope. Iniiest. Ophthnl. Vis. Scz. 17, 322-326. Sperling S. (1978) Early morphological changes in organ cultured human corneal endothelium. Actn ophthal. (Kbh.) 56, 785-792.
Author’r addrm: Thomas Olsen, Department of Ophthalmology, k h u s Kommunehospital, DK-8000 Arhus C, Denmark
890
Depai tment of Ophthalmology (Head N Bhkrc),
if rhuc Kommunehoybpztal, Unzwrtzty of Aarhuc, Dmmnrk
THE SPECULAR MICROSCOPIC APPEARANCE
OF CORNEAL GRAFT ENDOTHELIUM DURING AN ACUTE REJECTION EPISODE A Case Report BY
THOMAS OLSEN
T h e specular microscopic appearance of a graft endothelium during a successfully treated acute endothelial rejection episode is reported. Coinciding with the increased thickness of the graft great disturbance in cellular morphology was observed. Individual cells were seen to appear swollen or to show bright intracellular areas. Also abnormalities with respect to cellular intersections were seen with the occurrence of junctional complexes at which more than three cells met. The significance of these abnormalities is discussed. The morphological changes subsided gradually as the graft cleared and the thickness decreased. This study examplifies the value of the non-contact specular microscopic technique which enables a thorough study of the diseased corneal endothelium without traumatizing the cornea and interfering with subsequent specular microscopic examinations.
K q umrd~:specular microscopy - cornea - endothelium -rejection
- graft
T h e appearance of corneal endothelium in response to trauma has been observed by means of specular microscopy following various forms of insult such as air in the anterior chamber (Leibowitz et al. 1974) osmotic dehydration of the cornea (Sherrard 1976, 1978) and ultrasound (Olson et al. 1978). Most of these studies have been carried out on animals, little information is available on the human endothelium which may apply to the clinical situation. The fact that in vi-cm specular microscopy of corneal endothelium requires a translucent cornea hinders to a certain extent a detailed study of human endoReceived February 7 , 1979.
882
thelium following trauma. The case presented here is a n exception to that rule in t h a t it was possible t o study the specular microscopic appearance of a corneal graft e n d o t h e l i u m shortly after onset of a n acute rejection episode.
Methods Central corneal thickness was measured with a modified Haag-Streit pachometer (Sperling & Ehlers 1977). Specular microscopy was performed using a non-contact technique described earlier (Olsen 1979). Case presentation This 50-year-old man had penetrating keratoplasty done for the first time in 1974 due to a vascularized corneal scar after herpetic keratitis. The grafting was successful until one year later where the patient experienced an acute rejection. Two weeks passed before he attended for control. In spite of topical and systemic steroid treatment the graft remained oedematous and became vascularized. Retransplantation was undertaken January 1977. Postoperatively the graft was resutured in the lower left quadrant hut otherwise the postoperative course was uneventful. The case was included as No. 1 in the study of Ehlers & Bramsen (1978). On November 7th 1978 the patient awaked in the morning with blurred vision. On examination the same day the graft was oedematous with epithelial bullae and scattered precipitates on the posterior aspect of graft. N o rejection line was present on this first day after onset, but the cornea was thicker in the lower part than in the upper part. Systemic treatment with prednison 30 mg daily, topical ultracortenol and subconjunctival decadrone phosphate in.jections were started immediately. On this treatment graft thickness decreased
CCT
9
endcIthelial photos
I
0.5 -
1
50 mg
0
0 l
~
,
0 2 L 6 810
~
~
20
~
30
~
LO
~
~
,
90 days after onset
Fig. 1. Diagram showing changes in central corneal thickness (CCT) (mm) of graft during an acute rejection episode. Black horizontal staircase line indicates systemic prednisone treatment.
~
Thomas Olsen
gradually (Fig. 1). O n the 3rd day after onset of symptoms a typical rejection line was present across the lower part of graft. This line was later broken u p into smaller precipitates. Below the line the cornea was constantly thicker than above the line. T h e same tendencdy was found in the ensuing weeks in which the graft gradually cleared. T e n days after onset no precipitates were found with the slit lamp.
Fig. 2. Endothelial cells of graft visible in the specular microscope 5 days after onset of rejection. Top: photo from upper part of graft. Centre and buttom: photos from lower, thicker and more heavily attacked part of graft. T h e endothelial morphology is severely disturbed with swelling of cells (examples shown by arrows) intrarellular bright areas (asterixes) and abnormal cellular junctional complexes (circles). Horizontal bar = 100 pm. 884
Non-contuct Sperulnr Microztopv
Specular microscopy For the first 4 days after onset of rejection episode it was not possible to observe and photograph the endothelium due to opacification of the graft.
5th duy after onset (Fig. 2) T h e endothelial cells became discernible. At places no planar reflex from the endothelium was obtainable probably due to great distortion of endothelium. In those places in which the cells were visible, some cells showed dark round intracellular areas which made the cells appear swollen as interpreted from the absent reflex from the center of the cells (arrows in Fig. 2). Such cells also appeared more round than cells without this phenomenon. Other cells showed intracellular variation with respect to intensity of reflected light giving the impression of bright intracellular areas (asterixes in Fig. 2). Cells were larger and more irregular in the lower than in the upper part of cornea. Abnormalities were also found with respect to cellular junctions. Normally cell boundaries meet each other at points where three cells meet. Cell meetings involving four or more cells are practically never seen in normal endothelia (Fig. 6). In some places cells were found to meet four at a time (circles in Fig. 2), thereby greatly disturbing the regular hexagonal pattern.
14th day ufier onset (Fig. 3) Graft almost clear. The endothelial cells now reflected the light in a more even fashion. Intracellular round dark areas were less pronounced but still intracellular bright areas were present. In Fig. 3 abnormal cellular intersections are especially evident showing cell meetings involving more than three cells (circles in Fig. 3).
Fig. 3.
Endothelium of graft 14 days after onset of rejection. Photograph from lower part of graft. Squares indicate abnormal cellular junctions with dark spots occurring in the centre. Other symbols explained in Fig. 2. Bar = 100 pm.
885
Thomas Olsen
Fig. 4. T h e specular microscopic appearance of graft 21 days after onset of rejection. Photomicrograph from lower part of graft. Size and shape of cells show considerable variation with small cells with convex vorders (arrows) appearing between larger cells. Bar = 100 pm.
Fig. 5 . 93 days after onset of rejection the endothelium has reorganized into a more regular pattern. Top: upper part of graft. Buttom: lower part of graft, Note absence of previous marked difference in cell size from top to buttom of graft. Overall cell density is about 460 cells/mm2 which is clearly much lower than on previous photomicrographs. Bar = 100 p.
886
Non-contact Specular Microscopy
Mostly these abnormal junctional complexes consisted of joints of four cells, but in Fig. 3 a junctional complex of five cells is also found (circle in right part of Fig. 3). The cell pattern is far from the normal hexagonal regular pattern, the morphological changes being more pronounced in the lower than in the upper half of cornea.
21st ahy after onset (Fig. 4) T h e graft now appeared clear. Cell size and shape showed still considerable variation with small cells with convex borders (arrows in Fig. 4) appearing between larger irregular cells. Cell meetings where more than three cells met were only occasionally encountered (circle in Fig. 4).
93rd day after
onset (Fig. 5 ) Graft thickness had now stabilized at a value somewhat above the thickness before the onset of rejection. Size and shape of cells were now more uniform but not completely regular. Whereas in the foregoing specular microscopic examinations
Fig. 6. T h e specular microscopic appearance of normal corneal graft endothelium. Top: male, 62, endothelial density 1090 cells/mm*. Buttom: female, 48,endothelial density 1910 cells/mmz. Bar indicates 100 pm.
887
Thomas Oken
there was a considerable difference in size of cells from photographs at the upper and lower part of graft, this difference now was less apparent. The overall number of cells per unit area at this time was clearly much lower than that of normal graft endothelium (Fig. 6) and that of the first specular microscopic examinations, but it was difficult to give figures for this difference due to the large variation in cell density observed within the graft. No clearcut abnormalities within cells or with respect to cellular intersections were found.
Discussion In interpreting the endothelial changes observed in this study it is probably sound to regard them as partly due to the nature of the actual disease and partly due to the secondary healing mechanisms of the endothelium. Graft rejection is generally believed to be a cell-mediated immunological manifestation. With the specular microscope it was not possible to observe any lymphocytes on the endothelium although precipitates were seen with the slit-lamp. It was remarkable, however that the cellular abnormalities in the endothelium were so widely dispersed and occurred at places without precipitates. Furthermore, the rejection line seen with the slit-lamp did not sharply divide the endothelium into an abnormal and a normal area. These non-localized changes may be the result of soluble factors involved in the cellular hypersensitivity reaction, for which a number of mediators are now known to be involved (see David 1973) among them macrophage inhibition factors and factors cytotoxic for target cells. Some of these factors may activate lysosome enzymes. Interestingly, this process has been shown to result in swelling and rounding of target cells and their organelles followed by lysis of the cell (Able et al. 1970). This may be the explanation for the apparent swelling of endothelial cells observed in the present study (Fig. 2). The observed bright intracellular areas observed in the present study bear a certain resemblance to those *eventsndescribed in rabbits (Sherrard 1976, 1978; Leibowitz 1974) after air or osmotic dehydration of cornea. Sherrard showed that some of these events may progress to cell death after which a characteristic rosette is formed by adjacent cells spreading to cover the defect (Sherrard 1976). Such rosettes formed after acute trauma have been described by others in uivo (Olson et al. 1978) and in uitro (Sperling 1978). A characteristic feature for these rosettes is that the centre forms a point at which more than three cells meet depending on the number of cells surrounding the cell which has been destroyed. The junctional complex of five cells described in Fig. 3 is in fact a rosette. The abnormal junctional complexes of four cells, which was of far more frequent occurrence are, however, not identical to such rosettes. 888
Non-contact Sprcular Mzcro \copy
According to Sherrard (1 976) and Olson (1978) these rosettes became reorganized within a few hours leaving a cell pattern almost indistinguishable from a normal endothelium. Based on in vitro studies Sperling (1978) noted that joint meetings of four cells may stay longer than the rosettes and suggested that joints of four cells actually may be formed from the rosettes. The abnormal cellular intersections found in this study, i. e. junctional complexes in which more than three cells meet, may thus represent recent cell death. It was not possible to make valid estimation of endothelial cell loss due to large variation in cell size from one part of the graft to another and the inability to photograph exactly the same area after periods of time. It may be suggested that a great cell loss occurred during the rejection episode (compare Fig. 2 with Fig. 5 ) , and that the greatest cell loss occurred in the lower part of the graft as judged from the greater morphological changes and the greater thickness of graft in this region. Three months after onset (Fig. 5 ) , however, no marked difference in cell density was observed between upper and lower part of graft. This indicates that a certain sliding of cells towards regions of lower cell density has occurred, a phenomenon which has also been described after cataract extraction (Rao et al. 1978). T h e functional impairment of the endothelium accompanying the observed morphological changes was manifested through the changes in graft thickness (Fig.1). T h e return of a clear, thin graft 2-3 weeks after onset of rejection was accompanied by the specular microscope appearance of a planar endothelium with no major intracellular or intercellular abnormalities, although a certain degree of cellular pleomorphism still existed (Fig. 4). This pleomorphism had reduced on the last examination (compare Figs. 4 and 5 ) . This indicates the process of active cell death and the immediate healing response is more deleterious for the barrier function of the endothelium than the process of reorganizing the cellular pattern from a state of pleomorphism to a more regular pattern. In this study non-contact specular microscopy has proved very useful in studying morphologic alterations of endothelial cells following an insult. With the noncontact technique the whole endothelium can be scanned without traumatizing the epithelium and interfering with subsequent specular microscopic examinations.
Acknowledgments T h e technical assistance of Mrs. Anette Poulsen is gratefully acknowledged. This work was supported by Fonden ti1 Lzgevidenskabens Fremme.
889
Thomas Olsen
References Able M. E., I.ee ,J. & Rosenau W. (1970) 1.vmphocyte-target cell interaction in 71itl-o.Amur. 1. Pnth. 60, 42 1-427. David .J. R. (1973) Lvmphocyte mediators and cellular hypersensitivity. ,\‘mi Eng. ,]. M u d . 288, 143-149. Ehlers N . & Sperling S. (1977) A technical improvement of the Haag-Streit pachometer. Actn ophthnl. (KM.) 56, 53-66. Ehlers N. & Rramsen T. (1978) Management of late corneal graft prohlerns. Actrr op/ithd (Khh.) 56, 984-997. I.eibowit7 H. M., Laing R. A. & Sandstrom M. (1974) T h e effect of air in the anterior chamber . A rrh . Ophthn 1. iC:hicci,goj 92, 22 7-2 3 0 , Olsen T. (1979) Optical principles for estimation of endothelial cell density with the non-contact specular microscope. .4cf(r ophthol. (Khh.) 57, 860-867. Olson 1.. E., Marshall ,J., Rice N. S. C . 8c Andrews K. (1978) T h e effects of ultrasound on the corneal endothelium. I . T h e acute lesion. Brit.J. Ophthal. 62, 134-144. Olson L. E., Marshall J., Rice N. S. C. & Andrews R. (1978) Effects of ultrasound on the corneal endothelium. 11. T h e endothelial repair process. Brit.,]. Ophthal. 62, 145-154. Rao G. N., Shaw L. E. & Aquavella J. V. (1978) Morphological appearance of the healing corneal endothelium. Arch. Ophthal. (Chicago) 96, 2027-2030. Sherrard E. S. (1976) T h e corneal endothelium in uivo: its response to mild trauma. Exp. Eye Res. 22, 347-357. Sherrard E. S. (1978) Characterization of changes observed in the corneal endothelium with the specular microscope. Iniiest. Ophthnl. Vis. Scz. 17, 322-326. Sperling S. (1978) Early morphological changes in organ cultured human corneal endothelium. Actn ophthal. (Kbh.) 56, 785-792.
Author’r addrm: Thomas Olsen, Department of Ophthalmology, k h u s Kommunehospital, DK-8000 Arhus C, Denmark
890
