An animal model of trachoma. II. The importance of repeated reinfection

Transcription

1 An animal model of trachoma II. The importance of repeated reinfection Hugh R. Taylor, Shirley L. Johnson, Robert A. Prendergast, Julius Chandler R. Dawson,** and Arthur M. Silverstein Schachter,* An animal model of chronic cicatrizing trachoma has been produced by repeated ocular inoculation with Chlamydia trachomatis serotype E, a genitally transmitted strain. We have now produced a chronic follicular conjunctivitis in cynomolgus monkeys by repeated inoculation with C. trachomatis serotype A, which had been isolated from an area of endemic trachoma. This disease was similar in all respects to that which followed infection with the serotype E strain. Cynomolgus monkeys inoculated with a single dose of serotype E of C. trachomatis strain developed an acute, self-limited follicular conjunctivitis, which was intense for 4 weeks and then slowly subsided. The organism could be reisolated only during the first 4 weeks after inoculation. On reinoculation at 15 and 30 weeks after the initial infection, these animals demonstrated only a mild and transitory clinical response, and the agent could be recovered for only up to 14 days after inoculation. In contrast, repeated weekly reinoculation with either serotype led to a chronic progressive clinical response in these animals, although after the first 6 weeks the agent was isolated only occasionally. This chronic disease was shown not to be due to hypersensitivity to the egg yolk components in which the organism was grown. These data suggest that the serotype of the chlamydial organism may not be as important in determining the clinical course of disease as is the frequency or persistence of exposure to the chlamydial agent. Although a single inoculum produces an acute follicular conjunctivitis, repeated inoculation is needed to produce the chronic disease characteristic of trachoma in this animal model. (INVEST OPHTHALMOL VIS SCI 23: , 1982.) Key words: Chlamydia, trachoma, secretory antibody, serum antibody, animal model The various stages of trachoma were well described in antiquity, and yet many aspects of the disease remain unexplained today. Al- From the Wilmer Institute, The Johns Hopkins Medical Institutions, Baltimore, Md., the * Department of Laboratory Medicine, University of California, San Francisco, and the **Francis I. Proctor Foundation for Research in Ophthalmology, University of California, San Francisco, Calif. Supported by NIH-NEI grant E Y to Dr. Taylor, by grantey03521todr. Prendergast, bygrantey00427to Dr. Dawson, and by grant EY to Dr. Schachter. Submitted for publication Sept. 28, Reprint requests: Hugh R. Taylor, M.D., Wilmer Institute, Johns Hopkins Hospital, Baltimore, Md though trachoma is first acquired in early childhood, blindness usually does not occur until middle life. Infection in childhood produces a conjunctival follicular response, but follicles are rare in adults. Blindness results from the development of conjunctival scarring with subsequent trichiasis, but the mechanisms by which the early follicular response progresses to conjunctival scarring are not known. In endemic areas, the disease progresses relentlessly in adults in the absence of overt signs of infection, so that the prevalence of potentially blinding lid changes increases for up to 20 years after the peak of inflammatory disease. 1 ' 2 The clinical disease /82/ $00.90/ Assoc. for Res. in Vis. and Ophthal., Inc. 507

2 508 Taylor et al. Invest. Ophthalmol. Vis. Sci. October 1982 usually does not progress in people with moderate conjunctival scarring who leave an endemic area although without intervention, severe scarring will irreversibly lead to blindness, whether or not the person continues to live in an endemic area. Although high levels of specific antibodies against chlamydiae in both serum and tears are common, immunity to persisting infection or reinfection appears to be rather weak. 3 ' 4 The individual may be successfully treated with antibiotics but in an endemic area may quickly redevelop trachoma. There seems to be a clear relation between the level of personal and community hygiene and the prevalence and intensity of trachoma. 1 ' 5 With improvement in hygiene, trachoma disappears. 1 2> 6> 7 ' One unifying theory has emerged to explain many of the perplexing features of endemic trachoma, and this is the importance of repeated episodes of reinfection. 6 " 8 The occurrence of repeated or persistent infection would explain the gradual development of blinding trachoma during a lifelong exposure, and also the reappearance of conjunctival inflammation after treatment or on return to an endemic area. Reinfection may also explain the difference between the clinical picture of trachoma and of inclusion conjunctivitis. In areas of endemic trachoma, repeated episodes of ocular infection are likely to occur, whereas in areas where the genital tract is the major reservoir for C. trachomatis, incidental single episodes of eye infection leading to inclusion conjunctivitis are more likely. Several features of the pathophysiology of trachoma are reasonably clear, and these should be used as the standard with which any experimental model system devised for investigation of this disease must be compared. Many previous investigators have attempted to use experimental animal models to study trachoma, although most have had only limited success. One notable exception is the work of Wang and Grayston, 9 ' 10 who studied chlamydial infection in Taiwan monkeys (Macaca cyclopsis). After a number of inoculations with Chlamydia trachomatis, some of these monkeys developed chronic persistent disease with the eventual appearance of conjunctival scarring and corneal pannus over a period of 10 years. More recently, Monnickendam et al. 11 have reported the development of chronic conjunctivitis with scarring in guinea pigs that were repeatedly reinfected with the guinea-pig inclusion conjunctivitis agent. This model does not use C. trachomatis, however, but the related organism, C. psittaci. We have recently reported a successful animal model system using cynomolgus monkeys infected with the BOUR strain of C. trachomatis. 12 These monkeys demonstrated clinical, pathologic, and serologic evidence of trachoma strikingly similar to the human infection. This article investigates the requirement of repeated exposure to the infectious agent in this model system and examines the specificity of the serotype of C. trachomatis used. Methods Animals and examination techniques. Colonyraised cynomolgus monkeys, ages 18 to 24 months, were used in this study and housed in isolation cages. The detailed examination and specimen collection protocol as previously described was followed. 12 In brief, with animals under general anesthesia with ketamine hydrochloride, a clinical examination was performed with a hand-held slit lamp. Tears and serum were collected for serologic examination by microimmunofluorescence, smears were obtained from the superior tarsal plate for cytologic study, and swabs were taken for reisolation cultures and were inoculated within 2 hr into a McCoy cell tissue culture system. Although both eyes were examined, all specimens were taken from the left eye to eliminate the possibility of enhancing scar development in the right eye. It is important to note, however, that there were no significant or consistent differences between the clinical responses of right and left eyes. Where indicated by the experimental protocol, both eyes were inoculated with 20 /x.1 of inoculum by means of a sterile-tipped pipette at the completion of the examination and specimen collection. Chlamydia trachomatis strains were grown in yolksac of embryonated hens' eggs and diluted to 50% in phosphate-buffered saline. The C. trachomatis strains used were BOUR, E serotype, at a titer of either egg lethal dose (ELD 50 ) or 10 55

3 Volume 23 Number 4 An animal model of trachoma Fig. 1. Clinical response to weekly reinfection with C. trachomatis HAR-1 strain (A serotype), with mean follicular and inflammatory indices for five cynomolgus monkeys and frequency of positive chlamydial reisolation cultures. Reinfection was discontinued after 17 weeks. ELD 50 per ml; and HAR-1 strain, A serotype, at a titer of 10' 4 ELD 50 per ml. The clinical response of each eye was graded for the following individual signs: lymphoid follicles on the conjunctiva of the superior tarsal, superior fornix, superior bulbar areas, and at the limbus; superior tarsal papillae; ocular discharge; and bulbar conjunctival injection. Each sign was graded from 0 (when absent) to a maximum of 3. The numerical scores of these individual signs were combined as previously described 12 to give simplified indices that quantified the clinical response: the "follicular index," which characterized the follicular response, and the "inflammatory index," which summarized the "nonspecific" signs of inflammation. Results Repeated weekly inoculation with serotype A. Both eyes offivemonkeys were inoculated once a week for 17 weeks with an inoculum of HAR-1 (serotype A), which had a titer of ELD 50 per ml. These animals showed a strong clinical response, developing a mucopurulent conjunctivitis within 3 days, with follicles first appearing at 1 week and becoming more notable over the next 7 weeks (Fig. 1). On a given day the grade for a particular sign may have varied by 1 or, rarely, 2 points for some animals. This small variation in clinical response did not alter the overall pattern Fig. 2. Clinical response seen to weekly reinoculation with C. trachomatis BOUR strain (E serotype), with mean follicular and inflammatory indices for six cynomolgus monkeys and frequency of positive chlamydial reisolation cultures. Reinfection was continued for more than 40 weeks. of response for an individual monkey or for the group. This was true for this group and for the subsequent groups. After 4 weeks the inflammatory response became less intense, although it persisted while the eye continued to be inoculated. After the cessation of inoculation at 17 weeks, both the follicular and the nonspecific inflammatory response decreased progressively during the next 11 weeks. The response to repeated reinfection with HAR-1 (serotype A) was similar to that observed in six monkeys inoculated at weekly intervals with BOUR strain (serotype E) at ELD 50 per ml (Fig. 2), although the reinoculation with BOUR strain was not stopped at 17 weeks. The frequency of positive chlamydial reisolation cultures was similar in both of these groups (Figs. 1 and 2), with frequent positive cultures being obtained in the first 2 months. After 2 months, positive cultures were obtained only infrequently, despite the continuation of repeated inoculation and the clinical presence of persistent disease. Routinely, the cultures were taken 7 days after each inoculation, although cultures taken 12, 36, 60, and 80 hr after inoculation were all negative in six monkeys that had received weekly inoculations with BOUR strain for 10 weeks. The serotype-specific antichlamydial re-

4 510 Taylor et al. Invest. Ophthalmol. Vis. Sci. October % 50% -I 4 FOLLICULAR INDEX O O INFLAMMATORY INDEX Fig. 3. Mean serotype-specific antibody levels in tears and serum of five cynomolgus monkeys receiving weekly reinoculation of C. trachomatis HAR-1 strain (A serotype) for 17 weeks NORMAL YOLK SAC WEEKS 24 Fig. 5. Clinical response seen in four cynomolgus monkeys after stopping weekly reinoculation with C. trachomatis BOUR strain cultured in egg yolk. Fourteen weeks after the cessation of reinoculation, the monkeys were challenged with an inoculation of normal yolksac. Mean follicular and inflammatory indices and the frequency of positive chlamydial reisolation cultures are given. Fig. 4. Mean serotype-specific antibody levels in tears and serum of six cynomolgus monkeys receiving weekly reinoculation of C. trachomatis BOUR strain (E serotype) for more than 1 year. sponses for monkeys in each group were generally similar (Figs. 3 and 4). The animals that were reinoculated with BOUR strain showed a brisk response with specific serum IgG and IgM. The serum IgG levels remained high, but although the IgM levels fell somewhat, detectable serum levels were present throughout the period of reinoculation. In contrast, specific IgM was detected in the tears only during the first 7 weeks, although tear IgG levels remained high. Serotypespecific IgA levels in tears developed more slowly but persisted. The monkeys receiving HAR-1 inoculations had a much slower antibody response, both in tears and serum, which may be due to the lower titer of the inocula used. Inoculation with normal yolksac. In another group of four monkeys, weekly inoculation with BOUR strainc. trachomatis (10 32 ELD 50 per ml) was discontinued after 41 to 78 weeks. With the cessation of reinoculation, there was a gradual decrease in the intensity of the clinical disease in these animals, and after 10 weeks their eyes showed minimal hyperemia and only a few residual follicles in the fornices (Fig. 5). This gradual reduction in clinical response after the cessation of reinoculation is similar to that seen in other cynomolgus monkeys when the repeated reinfection with the HAR-1 strain was stopped (Fig. 1). Fourteen weeks after the cessation of weekly reinoculation, the four monkeys were challenged with 20 fx\ of normal egg yolk that was diluted 50% in phosphate-buffered saline and instilled into each eye. No significant

5 Volume 23 Number 4 An animal model of trachoma 511 Fig. 6. Clinical response seen to a single inoculation of C. trachomatis BOUR strain (low titer) in six cynomolgus monkeys. Fifteen weeks after the initial inoculation, the animals were rechallenged with a second low-titer inoculum. After a further 15 weeks they received a second rechallenge, this time with a high-titer inoculum. The mean follicular and inflammatory indices and the frequency of positive chlamydial reisolation cultures are given. reaction was noticed after this challenge, although one animal developed minimal ocular discharge 3 days after inoculation. There was no exacerbation of the follicular response in any animal. Reisolation cultures remained negative throughout, and there was no change in antibody titers. Single inoculations with serotype E. Six monkeys were given a single inoculation in both eyes of BOUR strain (serotype E), ELD 50 per ml. Within 3 days an acute, mucopurulent conjunctivitis had developed, although follicles appeared more slowly during the first 3 weeks (Fig. 6). The response, however, was self-limiting, with spontaneous resolution commencing at 4 weeks. Apart from residual follicles, mainly in the superior fornix, the eyes were "quiet" by 10 weeks. C. trachomatis was reisolated during the first weeks on days 3, 7, 10, 14, 21, and 28 after inoculation, but thereafter cultures were repeatedly negative (Fig. 6). Fifteen weeks after the initial inoculation, the animals were rechallenged with a second single inoculum of BOUR strain at the same titer ( ELD 50 per ml). This was followed by an abrupt exacerbation of the clinical disease, with a transient conjunctivitis and in- Fig. 7. Mean serotype-specific antibody levels in tears and serum of six cynomolgus monkeys receiving three single inoculations of C. trachomatis BOUR strain 15 weeks apart. crease in the number of follicles noted 3 days after inoculation. This abbreviated infection had essentially resolved by 3 weeks, and the residual clinical response continued to diminish over the next 12 weeks. C. trachomatis was reisolated only on days 10 and 14 after the second inoculation. A third inoculum with BOUR strain of a higher titer (10 55 ELD 50 per ml) was given 15 weeks after the second. Again, an acute, self-limited infection developed by the third day after inoculation, but it also had essentially resolved within 2 weeks. Positive reisolation cultures were obtained on days 3, 7, and 10 after inoculation. The clinical response of these animals is contrasted with the response seen in animals repeatedly reinfected at weekly intervals with BOUR strain (Fig. 2). Specific antichlamydial antibodies were not detected in either the serum or tears before the first inoculation (Fig. 7). After the first inoculation, serotype-specific IgM and IgG appeared in serum. Although the IgM levels fell after 2 weeks, serum IgG levels remained high. Serum antibody levels did not change significantly after the second and third inocula, although there was a slight and transient increase in serum IgM after each exposure. The serotype-specific IgM response in tears was similar to that in serum,

6 512 Taylor et at. Invest. Ophthalmol. Vis. Sci. October 1982 although IgM antibodies could not be detected in tears after 6 weeks. Specific IgG levels in tears appeared after the first inoculation, but unlike the serum levels, tear IgG levels fell after 3 weeks and also showed a brief spike with each reinoculation. Specific IgA levels in tears rose more slowly than IgG or IgM and showed an equivocal rise with each reinoculation (Fig. 7). Discussion The results of this study confirm the importance of repeated reinfection in the development of this animal model of trachoma with each of the two serotypes ofc. trachomatis used. The disease produced experimentally in monkeys closely resembles the human disease in that it is a chronic follicular conjunctivitis that involves the superior tarsal conjunctiva and progresses to conjunctival scarring. 7 ' 12 Although limbal follicles occur, they have not led to the development of Herbert's pits or pannus. However, corneal pannus itself rarely causes blindness in humans and most trachomatous blindness occurs, either directly or indirectly, as a result of conjunctival scarring. The monkeys followed up by Wang and Grayston 9 ' 10 over a 10 year period did eventually develop corneal pannus (superficial vascularization) as part of their disease. It has been proposed that the serotype of the C. trachomatis organism determines the clinical picture of the resulting infection 13 and that this is an inherent characteristic of the different serotypes. The chlamydiae species C. trachomatis includes 15 serotypes. 8 Serotypes A, Ba, B, and C have been isolated almost exclusively from typical trachoma cases in areas where the disease is endemic. Serotypes D to K are some of the most common sexually transmitted pathogens, and they cause disease of the genital tract, eye, and respiratory tract. In some cases, however, the genitally transmitted strains may cause eye disease with the features of endemic trachoma, such as conjunctival and corneal scarring and vascularization. 7 ' 14 Furthermore, trachoma serotypes may be isolated from people who have the clinical diagnosis of inclusion conjunctivitis. 0 It is of considerable interest, therefore, that in our studies in monkeys the clinical diseases produced by weekly reinoculation of C. trachomatis were similar, regardless of whether the serotype A organism (HAR-1) or the E serotype (BOUR) was used. This E- serotype organism was initially isolated from an adult with acute follicular conjunctivitis in an urban area of California, whereas the A serotype was isolated from a child with classic endemic trachoma in Saudi Arabia. Our observations in this animal model suggest that frequent reinfection or prolonged exposure to the chlamydial agent is the key determinant of the subsequent clinical disease. Our studies indicated that repeated reinfection with either the A or the E serotype will produce the chronic follicular disease. This persistent inflammatory response seems to be a direct result of the repeated inoculation of the chlamydial agent, and the response persists as long as reinoculation continues. When the continued exposure of the eye to the chlamydial agent is stopped, the inflammation resolves completely during 2 to 3 months. The patterns of positive reisolation cultures from animals receiving repeated reinfections of both A and E serotypes were similar. It is especially interesting that although chlamydiae were readily reisolated during the first 8 weeks, they were recovered only infrequently after this time, despite weekly reinfection and the persistence of clinical disease. It has been pointed out by Dawson et al. 15 that the presence of chlamydial agent correlates closely with the intensity of the inflammatory response, especially the papillary response. The decline in positive cultures in our animals tended to follow the decline in the clinical nonspecific inflammatory index. The monkeys in this study have had an essentially pure chlamydial infection. Although not performed on a regular basis, bacteriologic cultures have not grown other ocular pathogens. Bacteria were only rarely seen on cytologic smears, and there has been no clinical evidence of secondary bacterial superinfection. This study has not examined the potentially confounding or synergistic ef-

7 Volume 23 Number 4 An animal model of trachoma 513 feet of bacterial infection, which in man could also increase chlamydial transmission by increasing the amount of discharge and thus increasing the frequency of reinfection. Alternatively, bacterial superinfection may lead to a simulated picture of repeated reinfection by blocking the clearing of chlamydiae from the conjunctiva. The animals given a single inoculation of C. trachomatis developed an acute, selflimited follicular conjunctivitis similar to inclusion conjunctivitis of man. Although unlike the naturally acquired or experimental human cases, keratitis has not been a noteworthy feature of this monkey model. For 4 weeks after the first inoculation, the disease seen in animals receiving a single inoculation was similar to that seen in animals receiving repeated inoculations. After 4 weeks, however, there was a spontaneous resolution of infection in the animals that had received only a single inoculation. An acute follicular conjunctivitis after a single inoculation of C. trachomatis, including A serotype, has been reported previously. 16 " 20 When the animals that had received a single inoculum were rechallenged with a second inoculation 15 weeks later, a milder infection developed. The clinical disease was much attenuated and it ran an abbreviated course. The eyes had effectively returned to their prechallenge condition within 4 weeks. A similar resistance to reinfection has been described in the owl monkey and in human subjects. 14 ' 20 Despite the use of a much higher titer of inoculum at the third challenge, the clinical response was still much less severe than that seen with the primary inoculation, and it was only slightly more intense than that obtained with the second lowtiter challenge. The positive reisolation cultures of C. trachomatis confirmed that reinfection occurred at each challenge. It is noteworthy, however, that there was a greater lag period before the positive cultures were obtained with the second inoculation. The first positive cultures were obtained 10 days after the second inoculation, whereas the lag period was 3 days for the primary and tertiary (high-titer) challenges. The clinical response seen with rechallenge appears to be a specific reaction to the chlamydial agent, since not only could C. trachomatis be recovered in isolation cultures, but also no response was seen after the inoculation of normal yolksac into the eyes of other animals that had previously received repeated inoculations of Chlamydia- infected yolksac. It is of interest that Wang et al. 21 described a topical ocular hypersensitivity reaction to yolksac in monkeys that had previously received intramuscular injections of an egg-yolk derived vaccine. The serologic findings in our animals that received the single inoculation demonstrate a brisk local (tear) and systemic (serum) response, with persisting IgG levels and transient levels of IgM. Despite detectable levels of specific antichlamydial IgG and IgA in tears, reinfection still occurred with reinoculation. After reinfection, there was a transient local increase in IgG and IgA levels in tears but no systemic response. This is consistent with a localization of the infection to the eye, and localized antibody production. Alternatively, the local inflammation may have caused increased transudation of serum-derived antibodies. In the animals that received repeated reinoculations, it is of great interest that disease occurred and persisted despite the high antibody levels in tears. It should also be noted that in these animals specific antichlamydial serum IgM levels did not fall but remained high as long as the repeated inoculations were continued. This in contrast to the IgM levels in the serum of animals receiving a single inoculation, which fell progressively after an initial peak. The sustained elevation of serum IgM in the repeatedly reinfected animals may be a feature of continued antigenic exposure. The serum IgM response was not seen until 2 to 4 weeks after the single rechallenge inoculations given at 15 and 30 weeks, and this was similar to the response seen to the primary infection. This suggests that specific serum IgM antibody to C. trachomatis may be a useful marker of recent or current infection but that IgG antibody can be interpreted only as evidence of infection at some time in the past.

8 514 Taylor et al. Invest. Ophthalmol. Vis. Sci. October 1982 Thesefindingshave several important clinical implications for the control of endemic trachoma. Because of the importance of persistent or repeated exposure in the establishment of the clinical disease, it may be unnecessary to eliminate the ubiquitous chlamydiae from the environment to prevent blinding trachoma. It appears possible that a reduction in the infectious "pool" to such a level that episodes of reinfection are either relatively infrequent or do not occur may be sufficient to eliminate the chronic blinding disease. The reduction of the frequency of episodes of reinfection probably explains the mechanism by which an improvement in living conditions or hygiene reduces the prevalence and severity of trachoma. x> 2 Further, the failure to alter this frequency of reinfection explains why isolated attempts at individual chemotherapy are usually ineffective and followed by reappearance of the disease if the person remains in, or returns to, an endemic area. The frequency of reinfection might also explain the diflferent clinical presentations of trachoma and inclusion conjunctivitis. Jones 5 has drawn a parallel between the "ocular promiscuity" associated with endemic trachoma and the sexual promiscuity associated with inclusion conjunctivitis. It would appear that in an area with relatively good hygiene, episodes of reinfection occur only infrequently, and inclusion conjunctivitis is then the ocular infection that is recognized clinically. In areas of endemic trachoma where hygiene is poor, repeated episodes of reinfection occur commonly and therefore the clinical picture of trachoma develops. Grayston 22 recognized the importance of reinfection in trachoma when he proposed that inclusion conjunctivitis may represent the first infection, whereas trachoma was the "reinfection syndrome." Other factors, as yet unrecognized, may also enhance the chronicity of the chlamydial infection in some children in areas of endemic trachoma. Our studies indicate that in monkeys a single inoculation with E serotype produces an acute, self-limited follicular conjunctivitis similar to inclusion conjunctivitis in man. Similar findings have been reported with the A serotype > ' 20 Repeated reinfection, however, appears to be necessary for the development of the chronic follicular disease characteristic of trachoma and can be produced with either E or A serotype. REFERENCES 1. Royal Australian College of Ophthalmologists: Report of the National Trachoma and Eye Health Program of the Royal Australian College of Ophthalmologists, Sydney, Dawson CR, Daghfous T, Messadi M, Hoshiwara I, and Schachter J: Severe endemic trachoma in Tunisia. Br J Ophthalmol 60:245, Nichols RL, Oertley RE, Fraser CEO, MacDonald AB, and McComb DE: Immunity to chlamydial infections of the eye. VI. Homologous neutralization of trachoma infectivity for the owl monkey conjunctivae by eye secretions from humans with trachoma. J Infect Dis 127:429, Collier LH, Sowa J, and Sowa S: The serum and conjunctival antibody response to trachoma in Gambian children. J Hyg (Camb) 70:727, Jones BR: The prevention of blindness from trachoma. Trans Ophthalmol Soc UK 95:16, Grayston JT, Yet LJ, Wang S-P, Kuo C-C, Beasley RP, and Gale JL: Pathogenesis of ocular Chlamydia trachomatis infections in humans. In Nongonococcal Urethritis and Related Infections, Hobson D and Holmes KK, editors. Washington, D. C, 1977, American Society for Microbiology, pp Jones B: Changing concepts of trachoma and its control. Trans Ophthalmol Soc UK 100:25, Schachter J and Caldwell HD: Chlamydiae. Annu Rev Microbiol 34:285, Wang S-P and Grayston JT: Pannus with experimental trachoma and inclusion conjunctivitis agent infection of Taiwan monkeys. Am J Ophthalmol 63: 1133, Wang S-P and Grayston JT: Trachoma in the Taiwan monkey. Ann NY Acad Sci 98:117, Monnickendam MA, Darougar S, Treharne JD, and Tilbury AM: Development of chronic conjunctivitis with scarring and pannus, resembling trachoma in guinea pigs. Br J Ophthalmol 64:284, Taylor HR, Prendergast RA, Dawson CR, Schachter J, and Silverstein AM: An animal model for cicatrizing trachoma. INVEST OPHTHALMOL VIS SCI 21:422, Thygeson P: Historical review of oculogenital disease. Am J Ophthalmol 71:975, Dawson C, Wood TR, Rose L, Hanna L, and Thygeson P: Experimental inclusion conjunctivitis in man. III. Keratitis and other complications. Arch Ophthalmol 78:340, Dawson CR, Jones BR, and Darougar S: Blinding and non-blinding trachoma: assessment of intensity

9 Volume 23 Number 4 An animal model of trachoma 515 of upper tarsal inflammatory disease and disabling lesions. Bull WHO 52:279, Thygeson P, Dawson CR, Hanna L, Jawetz E, and Okumoto M: Observations on experimental trachoma in monkeys produced by strains of viruses propagated in yolksac. Am J Ophthalmol 50:907, Mordhorst CH: Experimental infections and immunogenicity of TRIC agents in monkeys. Am J Ophthalmol 63:1603, Collier LH: Experimental infection of baboons with inclusion blennorrhea and trachoma. Ann NY Acad Sci 98:188, Provost PJ and Vickers JH: Attempted immunization against trachoma infection in baboons. Am J Vet Res 33:599, Murray ES, Fraser CEO, Peters JH, McComb DE, and Nichols RL: The owl monkey as an experimental primate model for conjunctival trachoma infection. In Trachoma and Related Disorders Caused by Chlamydial Agents, Nichols RL, editor. Amsterdam, 1971, Excerpta Medica, p Wang S-P, Grayston JT, and Alexander ER: Trachoma vaccine studies in monkeys. Am J Ophthalmol 63:1615, Grayston JT: Symposium on trachoma. Biology of the virus. INVEST OPHTHALMOL 2:460, 1963.