THE JOUR:'IIAL OF I:\'I-'ECTIOUS DISEASES. VOL. 1:\5. :\0. 2 • FEBRL1:\RY 1977 the University of Chicago. All rights reserved.
© 1977 by
Experimental Hematogenous Endophthalmitis Due to Candida: Species Variation in Ocular Pathogenicity From the Department of Medicine, Harbor General Hospital, Torrance; the Research and Medical Service, Veterans Administration Wadsurorth Hospital Center, Los Angeles; and the Department of Medicine, UCLA School of Medicine, Los Angeles, California
John E. Edwards, Jr., John Z. Montgomerie,* Kenji Ishida, Joan O. Morrison, and Lucien B. Guze
Animal studies and case reviews both have demonstrated the propensity for Candida to disseminate to the eye, where it may cause extensive destruction [1-5]. Because our retrospective study of autopsied patients who had hematogenous endophthalmitis due to Candida showed that 78% (28 of 36) had dissemination to other deep organs [2], we used the clinical appearance of ocular lesions, particularly in the setting of increased susceptibility to candida infection, as
an indicator of widespread disseminated disease. However, the sensitivity of the ophthalmologic findings as detectors of dissemination is unknown at present; it will be determined only by a prospective study, including detailed clinical and postmortem examinations of eyes from patients with disseminated disease. Although Candida albicans was once considered the only pathogenic species of Candida, numerous species of Candida are now known to cause extensive infection in humans [6]. Since the diagnosis of hematogenous candida endophthalmitis has been made almost exclusively by histology, and in our review of 76 cases was substantiated only twice by culture (the species was not given in one case; it was C. albicans in the other), the role of the various species of Candida in the etiology of hematogenous endophthalmitis is unknown. These experiments were conducted to determine: (1) whether species of Candida other than albicans are capable of infecting ocular tissues; (2) if so, whether the eye infection with non-albicans species correlates with other deep
Received for publication January 29, 1976, and in revised form July 12, 1976. This work was supported by grant no. EY 0196692 from the U.S. Public Health Service and by grant no. 3330-04 from the Veterans Administration. Dr. Edwards is the recipient of U.S. Public Health Service Academic Investigator Award no. EY0003001. The authors thank Dr. M. David Goodman for his assistance with the renal pathology. Please address requests for reprints to Dr. John E. Edwards, Jr., Division of Infectious Diseases, Department of Medicine, Harbor General Hospital, 1000 West Carson Street, Torrance, California 90509. • Present address: Division of Infectious Diseases, Rancho Los Amigos Hospital, Downey, California 90242.
294
Downloaded from http://jid.oxfordjournals.org/ at University of Iowa Libraries/Serials Acquisitions on June 25, 2015
The ocular pathogenicity of species of Candida other than albicans was studied in the rabbit model of hematogenous candida endophthalmitis by injection of 10L108 colony-forming units of Candida krusei, Candida parapsilosis, Candida guilliermorulii, Candida tropicalis, Candida stellatoidea, and Candida albicans (control). At one and three weeks after infection, rabbits were examined by indirect ophthalmoscopy and were sacrificed for culture of eyes and kidneys. With an inoculum of 108 organisms, C. tropicalis and C. stellato idea infected the kidneys but only occasionally infected the chorioretina and never infected the vitreous. Organisms were cultured only rarely from the kidneys of rabbits infected with C. krusei, C. guilliermondii, and C. parapsilosis; these species were never isolated from eyes. The C. albicans control consistently infected the kidney, chorioretina, and vitreous; approximately equal numbers of C. albicans were isolated from these organs. These data suggest a relative resistance of ocular tissues to hematogenous candida infection with species other than C. albicans.
Candida Endophthalmitis
organ involvement; and (3) the clinical appearance of lesions caused by non-albicans species of Candida as compared with the lesions of hematogenous endophthalmitis due to C. albicans.
295
that had clinical lesions were saved for histology (total, four eyes).
Results Materials and Methods
Downloaded from http://jid.oxfordjournals.org/ at University of Iowa Libraries/Serials Acquisitions on June 25, 2015
Experimental animals. Female New Zealand white rabbits, aged eight to 10 weeks, were used. Organisms and infection. C. albicans, Candida tropicalis, Candida stellatoidea, Candida parapsilosis, Candida guilliermondii, and Candida krusei were obtained from patients and from the American Type Culture Collection (Rockville, Md.). Organisms were identified and speciated by gram stain, germ-tube forma tion, fermentation patterns, and carbohydrate assimilation tests. Suspensions (1 ml) of each of the test species grown for 20 hr in brain-heart infusion (BHI) broth (Difco, Detroit, Mich.) were adjusted by centrifugation and washing to the desired concentrations, as measured by pour plates and counts of du. The suspensions of yeast-phase organisms were injected into the marginal ear vein. Study of infected eyes. Just before the rabbits were killed, their eyes were dilated with 1% Cyclogylw (cyclopentolate hydrochloride; Alcon Laboratories, Fort Worth, Tex.) and were examined with the indirect ophthalmoscope (Xonics, Waltham, Mass.). The number of ocular lesions was determined, and the size of lesions was graded on a scale of 1-4. Rabbits were killed by injection of Diabutolv (sodium pentobarbital; Diamond Laboratories, Des Moines, Iowa) into the marginal ear vein one and three weeks after injection. Animals were sacrificed on days 7 or 8 and on days 21 and 22. Intraocular tissues were cultured quantitatively by homogenization of tissue and determination of the number of dujg of tissue [4]. The candida organisms recovered were confirmed for species. Two eyes per five rabbits were selected randomly for histologic studies. In the first two experiments, coronal sections (2 mm) from the centers of all kidneys were examined histologically; renal tissues were not saved for histology in the last two experiments. All eyes that were from rabbits inoculated with non-albicans species of Candida and
In an initial experiment the size of the inoculum of all species of Candida was 105 du in 1 ml. Each species was injected into 12 rabbits, except for the C. albicans control, which was injected into three. Rabbits that received C. tropicalis had a low-grade infection in the kidney (average, 1.9 log dujg of tissue) but not in the eye. Only a few rabbits that received C. stellatoidea had renal infections (average, log 0.22 cfujg of tissue). No lesions were seen by ophthalmoscopy. Control rabbits that received C. albicans had significant infections in the chorioretina, vitreous, and kidney (averages, 2.4, 1.4, and 1.3 log cfujg of the respective tissues). No other species produced infection. In a second experiment in which larger inocula were used (10 7 cfu in 1 ml for non-albicans species and 105 cfu for C. albicans), C. krusei and C. parapsilosis did not produce infection in any tissue, and C. guilliermondii infected only three kidneys from the group of 10 rabbits. (Each species, including the C. albicans control, was injected into 10 rabbits.j A total of six small intraocular lesions were seen clinically in three eyes of the 10 rabbits infected with C. stellatoidea, and a questionable small lesion was seen in one eye of the 10 rabbits that received C. tropicalis. Only one of these lesions (C. stellatoidea) was dissected successfully for histologic examination; organisms were seen within that lesion. All lesions were seen at one week after infection. Although the lesions were small and few in number, they appeared to be identical to the early lesions caused by C. albicans. When assessed by the criterion of the number of cfujg of tissue, the kidneys of rabbits infected with C. albicans, C. stellatoidea, and C. guilliermondii showed statistically significant improvement by the third week. In a third experiment, three strains of each of the test species (inoculum size, 106-107 cfu) were injected into four rabbits each (12 rabbits per species). No organisms were recovered from the eyes of any animals, and no lesions were seen
Edwards et al.
296
Table 1. Results of iv inoculation of rabbits with large doses of each of four species of Candida.
Species of
Candida (no. of efu)
albicans (lOS) tropicalis (10 8 ) stellatoidea (10 8 ) parapsilosis (10 8 )
No. of rabbits tested
10
4t 10 10
Candida isolated (average no. [log] of efu/g of tissue}" Chorioretina Vitreous
3.1 0.4 0.3
o
Kidney
2.5
2.9
o o o
6.4 3.0 0.6
"Combined results of right and left organs and of animals sacrificed at first and third weeks after infection. t All animals were sacrificed or died before the third week.
by the non-albicans species of Candida, and (2) to assess the correlation between the magnitude of the positive cultures and the extent of histological changes. In general, the number of positive cultures correlated with the degree of infection as determined histologically. However, there was a significant number (19 of 48) of kidneys that yielded positive cultures but were normal on histologic examination. This phenomenon was noted in kidneys with positive cultures in the low range, generally below the average of 3.0 log c£ujg of tissue. When the number of infecting organisms reached 5.0 log dujg, all kidneys were positive on histology. Undoubtedly, one of the reasons' for the lack of correlation between results of cultures and of histologic examination of specimens with smaller numbers of Candida was the sampling error inherent in the taking of a single section through the kidney. On histologic examination of renal tissue, microabscesses, approximately 0.5 mm in size, consisting of a suppurative-granulomatous reaction were seen; this type of lesion was found in both cortex and medulla. The lesions caused by the non-albicans species were essentially identical to those due to C. albicans. Although C. krusei was nonpathogenic, as determined by culture, when the inoculum consisted of '~108 cells, one kidney showed a typical microabscess, which was presumably due to C. krusei. Discussion
Variations in the pathogenicity of Candida species in experimental animals have been known for years [7-11]. All of the species used in this experiment were known to be pathogenic for both animals and humans. Because previous experiments showed C. albicans to be more pathogenic in rabbits than other species of Candida [7, 8], we used large-size inocula in an attempt to establish ocular infection in rabbits. The results of these experiments demonstrate the ability of C. tropicalis and C. stellatoidea to cause extensive renal infection in the rabbit model without causing significant eye infection. However, in two of 42 rabbits receiving C. tropicalis and three of 42 receiving C. stellaioidea, positive chorioretinal cultures were obtained, a fact that demonstrates a low level of ocular pathogenicity for these species. In these animals,
Downloaded from http://jid.oxfordjournals.org/ at University of Iowa Libraries/Serials Acquisitions on June 25, 2015
(except in the C. albicans controls). Low-grade renal infection was established in rabbits infected with C. parapsilosis, C. guilliermondii, C. stellatoidea, and C. tropicalis (average, log 0.98, 0.23, 1.1, and 3.1 cfujg, respectively). None of the strains of C. krusei were pathogenic. A fourth experiment was performed with use of larger inocula. Results are shown in table I. C. guilliermondii infected only two of 20 kidneys and C. krusei only one of 20. These two species caused no eye infection. C. tropicalis and C. stellatoidea both caused extensive kidney infection. Of the rabbits killed at one week after infection, only one (one of 20 eyes) from the group that received C. stellatoidea had infection in the chorioretina (average, log 3.1 c£ujg), and two (two of eight eyes) from the group that received C. tropicalis had chorioretinal infections. All other ocular tissues were negative on culture. No lesions were seen in these eyes. The control rabbit that received C. albicans had both eye and kidney infections. In the first, second, and fourth experiments, the infection in the eyes of the control rabbits inoculated with C. albicans and in the kidneys of the rabbits inoculated with C. stellatoidea and C. tropicalis showed statistically significant improvement between the first and third weeks. On histologic examination a predominately suppurative response was seen in the one eye lesion due to C. stellatoidea that was sectioned successfully. A single organism in the pseudohyphal form was present. In experiments no. I and 2, sections of tissue from each kidney were examined histologically for two purposes: (1) to determine the nature of the renal pathology caused
Candida Endophthalmitis
297
2.
3.
4.
5.
6.
7.
8. 9.
10.
11. 12.
References 1. Fishman, L. S., Griffin,
J. R.,
Sapico, F. L., Hecht, R.
Hematogenous candida cndophthalmitis-a complication of candidemia. N. Engl. J. Med. 286:675-681, 1972. Edwards, J. E., Jr., Foos, R. Y., Montgomerie, J. Z., Guze, L. B. Ocular manifestations of candida septicemia. Review of seventy-six cases of hematogenous candida endophtha1mitis. Medicine (Balt.) 53:4775, 1974. Hoffmann, D. H. Die experimentelle endogene Entzundung des Augeninnern durch Candida albicans. Ophthalmoskopische, histologische, und mikrobiologische Studien zum Ablau der Infektion beim Kaninchem.Ophthalmologica (Basel) 151:1-84, 1966. Edwards, J. E., Jr., Montgomerie, J. Z., Foos, R. Y., Shaw, V. K., Guze, L. B. Experimental hematogenous endophthalmitis caused by Candida albicans. J. Infect. Dis. 131:649-657, 1975. Michelson, P. E., Stark, W., Reser, E., Green, W. R. Endogenous candida endophthalmitis. Report of 13 cases and 16 from the literature. Intern. Ophthalmol. Clin. 11:125-147, 1971. Hurley, R. Pathogenicity of the genus Candida. In H. I. Winner and R. Hurley [ed.], Symposium on candida infections. E. and S. Livingstone, Ltd., London, 1966, p. 13-25. Goldstein, E., Grieco, M. H., Finkel, G., Louria, D. B. Studies on the pathogenesis of experimental Candida parapsilosis and Candida guilliermondii infections in mice. J. Infect. Dis. 115:293-302, 1965. Redaelli, P. Experimental moniliasis. J. Trop. Med. Hyg. 27:211-213,1924. Mankowski, Z. T. The experimental pathogenicity of various species of Candida in Swiss mice. Trans. N. Y. Acad. Sci. 19:548-570, 1957. Hasendever, H. F. Comparative pathogenicity of Candida albicans for mice and rabbits. J. Bacteriol. 78:105-109, 1959. Hurley, R., Winner, H. I. The pathogenicity of Candida tropicalis. J. Pathol. Bacteriol. 84:33-38, 1962. A1-Doory, Y., Baker, C. A. Comparative observations of ultrastructure of five species of Candida. Mycopathol. Mycol. Appl. 44:355-367,.1971.
Downloaded from http://jid.oxfordjournals.org/ at University of Iowa Libraries/Serials Acquisitions on June 25, 2015
the vitreous was never infected; the infection was confined to the chorioretina. With the inoculum sizes used in this model, the other species of Candida (C. parapsilosis and C. krusei) were not pathogenic for the eye. These data indicate that although C. stellatoidea and C. tropicalis can infect the chorioretina, they are very insensitive indicators of the presence of renal infection. This finding contrasts with the observation that ocul~r infection with C. albicans mirrors infection in the kidney [4]. The reason for this apparent lack of pathogenicity of non-albicans species of Candida for the eye is open for speculation. Detailed comparative electron microscopy of these species has failed to identify any significant difference in their morphology that would elucidate virulence factors [12]. Some possibilities are that (1) ocular tissues contain growth inhibition factors to which C. albicans is resistant; (2) non-albicans species of Candida, because of clearance mechanisms, may not reach the eye in large numbers; and (3) C. albicans produces germ tubes more rapidly than other species, and possibly germ-tube formation is important in the establishment of ocular infection. The results of these experiments, if applicable to the human, indicate a decreased likelihood of spread to the eye in persons with disseminated candidiasis caused by non-albicans species of Candida.
© 1977 by
Experimental Hematogenous Endophthalmitis Due to Candida: Species Variation in Ocular Pathogenicity From the Department of Medicine, Harbor General Hospital, Torrance; the Research and Medical Service, Veterans Administration Wadsurorth Hospital Center, Los Angeles; and the Department of Medicine, UCLA School of Medicine, Los Angeles, California
John E. Edwards, Jr., John Z. Montgomerie,* Kenji Ishida, Joan O. Morrison, and Lucien B. Guze
Animal studies and case reviews both have demonstrated the propensity for Candida to disseminate to the eye, where it may cause extensive destruction [1-5]. Because our retrospective study of autopsied patients who had hematogenous endophthalmitis due to Candida showed that 78% (28 of 36) had dissemination to other deep organs [2], we used the clinical appearance of ocular lesions, particularly in the setting of increased susceptibility to candida infection, as
an indicator of widespread disseminated disease. However, the sensitivity of the ophthalmologic findings as detectors of dissemination is unknown at present; it will be determined only by a prospective study, including detailed clinical and postmortem examinations of eyes from patients with disseminated disease. Although Candida albicans was once considered the only pathogenic species of Candida, numerous species of Candida are now known to cause extensive infection in humans [6]. Since the diagnosis of hematogenous candida endophthalmitis has been made almost exclusively by histology, and in our review of 76 cases was substantiated only twice by culture (the species was not given in one case; it was C. albicans in the other), the role of the various species of Candida in the etiology of hematogenous endophthalmitis is unknown. These experiments were conducted to determine: (1) whether species of Candida other than albicans are capable of infecting ocular tissues; (2) if so, whether the eye infection with non-albicans species correlates with other deep
Received for publication January 29, 1976, and in revised form July 12, 1976. This work was supported by grant no. EY 0196692 from the U.S. Public Health Service and by grant no. 3330-04 from the Veterans Administration. Dr. Edwards is the recipient of U.S. Public Health Service Academic Investigator Award no. EY0003001. The authors thank Dr. M. David Goodman for his assistance with the renal pathology. Please address requests for reprints to Dr. John E. Edwards, Jr., Division of Infectious Diseases, Department of Medicine, Harbor General Hospital, 1000 West Carson Street, Torrance, California 90509. • Present address: Division of Infectious Diseases, Rancho Los Amigos Hospital, Downey, California 90242.
294
Downloaded from http://jid.oxfordjournals.org/ at University of Iowa Libraries/Serials Acquisitions on June 25, 2015
The ocular pathogenicity of species of Candida other than albicans was studied in the rabbit model of hematogenous candida endophthalmitis by injection of 10L108 colony-forming units of Candida krusei, Candida parapsilosis, Candida guilliermorulii, Candida tropicalis, Candida stellatoidea, and Candida albicans (control). At one and three weeks after infection, rabbits were examined by indirect ophthalmoscopy and were sacrificed for culture of eyes and kidneys. With an inoculum of 108 organisms, C. tropicalis and C. stellato idea infected the kidneys but only occasionally infected the chorioretina and never infected the vitreous. Organisms were cultured only rarely from the kidneys of rabbits infected with C. krusei, C. guilliermondii, and C. parapsilosis; these species were never isolated from eyes. The C. albicans control consistently infected the kidney, chorioretina, and vitreous; approximately equal numbers of C. albicans were isolated from these organs. These data suggest a relative resistance of ocular tissues to hematogenous candida infection with species other than C. albicans.
Candida Endophthalmitis
organ involvement; and (3) the clinical appearance of lesions caused by non-albicans species of Candida as compared with the lesions of hematogenous endophthalmitis due to C. albicans.
295
that had clinical lesions were saved for histology (total, four eyes).
Results Materials and Methods
Downloaded from http://jid.oxfordjournals.org/ at University of Iowa Libraries/Serials Acquisitions on June 25, 2015
Experimental animals. Female New Zealand white rabbits, aged eight to 10 weeks, were used. Organisms and infection. C. albicans, Candida tropicalis, Candida stellatoidea, Candida parapsilosis, Candida guilliermondii, and Candida krusei were obtained from patients and from the American Type Culture Collection (Rockville, Md.). Organisms were identified and speciated by gram stain, germ-tube forma tion, fermentation patterns, and carbohydrate assimilation tests. Suspensions (1 ml) of each of the test species grown for 20 hr in brain-heart infusion (BHI) broth (Difco, Detroit, Mich.) were adjusted by centrifugation and washing to the desired concentrations, as measured by pour plates and counts of du. The suspensions of yeast-phase organisms were injected into the marginal ear vein. Study of infected eyes. Just before the rabbits were killed, their eyes were dilated with 1% Cyclogylw (cyclopentolate hydrochloride; Alcon Laboratories, Fort Worth, Tex.) and were examined with the indirect ophthalmoscope (Xonics, Waltham, Mass.). The number of ocular lesions was determined, and the size of lesions was graded on a scale of 1-4. Rabbits were killed by injection of Diabutolv (sodium pentobarbital; Diamond Laboratories, Des Moines, Iowa) into the marginal ear vein one and three weeks after injection. Animals were sacrificed on days 7 or 8 and on days 21 and 22. Intraocular tissues were cultured quantitatively by homogenization of tissue and determination of the number of dujg of tissue [4]. The candida organisms recovered were confirmed for species. Two eyes per five rabbits were selected randomly for histologic studies. In the first two experiments, coronal sections (2 mm) from the centers of all kidneys were examined histologically; renal tissues were not saved for histology in the last two experiments. All eyes that were from rabbits inoculated with non-albicans species of Candida and
In an initial experiment the size of the inoculum of all species of Candida was 105 du in 1 ml. Each species was injected into 12 rabbits, except for the C. albicans control, which was injected into three. Rabbits that received C. tropicalis had a low-grade infection in the kidney (average, 1.9 log dujg of tissue) but not in the eye. Only a few rabbits that received C. stellatoidea had renal infections (average, log 0.22 cfujg of tissue). No lesions were seen by ophthalmoscopy. Control rabbits that received C. albicans had significant infections in the chorioretina, vitreous, and kidney (averages, 2.4, 1.4, and 1.3 log cfujg of the respective tissues). No other species produced infection. In a second experiment in which larger inocula were used (10 7 cfu in 1 ml for non-albicans species and 105 cfu for C. albicans), C. krusei and C. parapsilosis did not produce infection in any tissue, and C. guilliermondii infected only three kidneys from the group of 10 rabbits. (Each species, including the C. albicans control, was injected into 10 rabbits.j A total of six small intraocular lesions were seen clinically in three eyes of the 10 rabbits infected with C. stellatoidea, and a questionable small lesion was seen in one eye of the 10 rabbits that received C. tropicalis. Only one of these lesions (C. stellatoidea) was dissected successfully for histologic examination; organisms were seen within that lesion. All lesions were seen at one week after infection. Although the lesions were small and few in number, they appeared to be identical to the early lesions caused by C. albicans. When assessed by the criterion of the number of cfujg of tissue, the kidneys of rabbits infected with C. albicans, C. stellatoidea, and C. guilliermondii showed statistically significant improvement by the third week. In a third experiment, three strains of each of the test species (inoculum size, 106-107 cfu) were injected into four rabbits each (12 rabbits per species). No organisms were recovered from the eyes of any animals, and no lesions were seen
Edwards et al.
296
Table 1. Results of iv inoculation of rabbits with large doses of each of four species of Candida.
Species of
Candida (no. of efu)
albicans (lOS) tropicalis (10 8 ) stellatoidea (10 8 ) parapsilosis (10 8 )
No. of rabbits tested
10
4t 10 10
Candida isolated (average no. [log] of efu/g of tissue}" Chorioretina Vitreous
3.1 0.4 0.3
o
Kidney
2.5
2.9
o o o
6.4 3.0 0.6
"Combined results of right and left organs and of animals sacrificed at first and third weeks after infection. t All animals were sacrificed or died before the third week.
by the non-albicans species of Candida, and (2) to assess the correlation between the magnitude of the positive cultures and the extent of histological changes. In general, the number of positive cultures correlated with the degree of infection as determined histologically. However, there was a significant number (19 of 48) of kidneys that yielded positive cultures but were normal on histologic examination. This phenomenon was noted in kidneys with positive cultures in the low range, generally below the average of 3.0 log c£ujg of tissue. When the number of infecting organisms reached 5.0 log dujg, all kidneys were positive on histology. Undoubtedly, one of the reasons' for the lack of correlation between results of cultures and of histologic examination of specimens with smaller numbers of Candida was the sampling error inherent in the taking of a single section through the kidney. On histologic examination of renal tissue, microabscesses, approximately 0.5 mm in size, consisting of a suppurative-granulomatous reaction were seen; this type of lesion was found in both cortex and medulla. The lesions caused by the non-albicans species were essentially identical to those due to C. albicans. Although C. krusei was nonpathogenic, as determined by culture, when the inoculum consisted of '~108 cells, one kidney showed a typical microabscess, which was presumably due to C. krusei. Discussion
Variations in the pathogenicity of Candida species in experimental animals have been known for years [7-11]. All of the species used in this experiment were known to be pathogenic for both animals and humans. Because previous experiments showed C. albicans to be more pathogenic in rabbits than other species of Candida [7, 8], we used large-size inocula in an attempt to establish ocular infection in rabbits. The results of these experiments demonstrate the ability of C. tropicalis and C. stellatoidea to cause extensive renal infection in the rabbit model without causing significant eye infection. However, in two of 42 rabbits receiving C. tropicalis and three of 42 receiving C. stellaioidea, positive chorioretinal cultures were obtained, a fact that demonstrates a low level of ocular pathogenicity for these species. In these animals,
Downloaded from http://jid.oxfordjournals.org/ at University of Iowa Libraries/Serials Acquisitions on June 25, 2015
(except in the C. albicans controls). Low-grade renal infection was established in rabbits infected with C. parapsilosis, C. guilliermondii, C. stellatoidea, and C. tropicalis (average, log 0.98, 0.23, 1.1, and 3.1 cfujg, respectively). None of the strains of C. krusei were pathogenic. A fourth experiment was performed with use of larger inocula. Results are shown in table I. C. guilliermondii infected only two of 20 kidneys and C. krusei only one of 20. These two species caused no eye infection. C. tropicalis and C. stellatoidea both caused extensive kidney infection. Of the rabbits killed at one week after infection, only one (one of 20 eyes) from the group that received C. stellatoidea had infection in the chorioretina (average, log 3.1 c£ujg), and two (two of eight eyes) from the group that received C. tropicalis had chorioretinal infections. All other ocular tissues were negative on culture. No lesions were seen in these eyes. The control rabbit that received C. albicans had both eye and kidney infections. In the first, second, and fourth experiments, the infection in the eyes of the control rabbits inoculated with C. albicans and in the kidneys of the rabbits inoculated with C. stellatoidea and C. tropicalis showed statistically significant improvement between the first and third weeks. On histologic examination a predominately suppurative response was seen in the one eye lesion due to C. stellatoidea that was sectioned successfully. A single organism in the pseudohyphal form was present. In experiments no. I and 2, sections of tissue from each kidney were examined histologically for two purposes: (1) to determine the nature of the renal pathology caused
Candida Endophthalmitis
297
2.
3.
4.
5.
6.
7.
8. 9.
10.
11. 12.
References 1. Fishman, L. S., Griffin,
J. R.,
Sapico, F. L., Hecht, R.
Hematogenous candida cndophthalmitis-a complication of candidemia. N. Engl. J. Med. 286:675-681, 1972. Edwards, J. E., Jr., Foos, R. Y., Montgomerie, J. Z., Guze, L. B. Ocular manifestations of candida septicemia. Review of seventy-six cases of hematogenous candida endophtha1mitis. Medicine (Balt.) 53:4775, 1974. Hoffmann, D. H. Die experimentelle endogene Entzundung des Augeninnern durch Candida albicans. Ophthalmoskopische, histologische, und mikrobiologische Studien zum Ablau der Infektion beim Kaninchem.Ophthalmologica (Basel) 151:1-84, 1966. Edwards, J. E., Jr., Montgomerie, J. Z., Foos, R. Y., Shaw, V. K., Guze, L. B. Experimental hematogenous endophthalmitis caused by Candida albicans. J. Infect. Dis. 131:649-657, 1975. Michelson, P. E., Stark, W., Reser, E., Green, W. R. Endogenous candida endophthalmitis. Report of 13 cases and 16 from the literature. Intern. Ophthalmol. Clin. 11:125-147, 1971. Hurley, R. Pathogenicity of the genus Candida. In H. I. Winner and R. Hurley [ed.], Symposium on candida infections. E. and S. Livingstone, Ltd., London, 1966, p. 13-25. Goldstein, E., Grieco, M. H., Finkel, G., Louria, D. B. Studies on the pathogenesis of experimental Candida parapsilosis and Candida guilliermondii infections in mice. J. Infect. Dis. 115:293-302, 1965. Redaelli, P. Experimental moniliasis. J. Trop. Med. Hyg. 27:211-213,1924. Mankowski, Z. T. The experimental pathogenicity of various species of Candida in Swiss mice. Trans. N. Y. Acad. Sci. 19:548-570, 1957. Hasendever, H. F. Comparative pathogenicity of Candida albicans for mice and rabbits. J. Bacteriol. 78:105-109, 1959. Hurley, R., Winner, H. I. The pathogenicity of Candida tropicalis. J. Pathol. Bacteriol. 84:33-38, 1962. A1-Doory, Y., Baker, C. A. Comparative observations of ultrastructure of five species of Candida. Mycopathol. Mycol. Appl. 44:355-367,.1971.
Downloaded from http://jid.oxfordjournals.org/ at University of Iowa Libraries/Serials Acquisitions on June 25, 2015
the vitreous was never infected; the infection was confined to the chorioretina. With the inoculum sizes used in this model, the other species of Candida (C. parapsilosis and C. krusei) were not pathogenic for the eye. These data indicate that although C. stellatoidea and C. tropicalis can infect the chorioretina, they are very insensitive indicators of the presence of renal infection. This finding contrasts with the observation that ocul~r infection with C. albicans mirrors infection in the kidney [4]. The reason for this apparent lack of pathogenicity of non-albicans species of Candida for the eye is open for speculation. Detailed comparative electron microscopy of these species has failed to identify any significant difference in their morphology that would elucidate virulence factors [12]. Some possibilities are that (1) ocular tissues contain growth inhibition factors to which C. albicans is resistant; (2) non-albicans species of Candida, because of clearance mechanisms, may not reach the eye in large numbers; and (3) C. albicans produces germ tubes more rapidly than other species, and possibly germ-tube formation is important in the establishment of ocular infection. The results of these experiments, if applicable to the human, indicate a decreased likelihood of spread to the eye in persons with disseminated candidiasis caused by non-albicans species of Candida.
