Phlyctenular keratoconjunctivitis has been associated in the past with poor, undernourished, sickly children, with tuberculosis strongly implicated. Duke-Elder1 described the pitiable position of these children: “During the day the child hides away in a dark corner, burying his face in his hands; and during the night he curls up under the blankets.”

Phlyctenular keratoconjunctivitis has a worldwide distribution. It is found most commonly during the first and second decades of life in children living in crowded, impoverished quarters. All observers report a higher incidence (60% to 70% of reported cases) in girls than in boys.7, 8 Fritz and coworkers9 studied 346 Alaskan native children with corneal scarring apparent in 143 cases. The involvement was bilateral in 112 cases (78%) and unilateral in 31 (22%).

Philip and coworkers7 found a characteristic distribution of corneal opacities of varying severity according to age. In their study of 4,635 Eskimos, opacities were infrequent among children younger than 4 to 5 years. However, 53% of school-aged children and 45% of adults exhibited scarring. Opacities were more frequent among females than among males. Of 1,197 children who were tested with tuberculin, 93% of those with opacities were reactors to 10 tuberculin units of purified protein derivative (i.e., intermediate-strength tuberculin), whereas only 34% of those without scarring were reactors. Among tuberculin reactors, there was no relation between degree of skin-test hypersensitivity and corneal opacity.

Phlyctenular keratoconjunctivitis is uncommon in areas where the incidence of tuberculosis is low. Likewise, the prevalence of corneal opacity was found to be highest in Alaskan households with the greatest crowding and lowest standard of living. As the incidence of tuberculosis in Alaskan natives has fallen from a death rate of 650 in 100,000 population in 1950 to 10 in 100,000 population in 1966, so has the incidence of phlyctenular keratoconjunctivitis, although exact figures are difficult to determine.10

In the United States, where the rate of positive tuberculin test results is relatively low, phlyctenular keratoconjunctivitis may occur in adults as well as in children, and it is likely to occur in patients free of tuberculosis. Sorsby8 reported that phlyctenular keratoconjunctivitis occurs more often in spring and summer than in late autumn or winter.

There is some debate about the pathophysiology and causes of phlyctenular keratoconjunctivitis. A clear analogy exists between phlyctenules of the conjunctiva and bacterids of the skin.11 Bacterids are nodules in the skin that are comparable histologically, as well as in their allergic origin, to phlyctenules of the conjunctiva and cornea. Although the tuberculid is the most common bacterid, dermatologists describe other types, such as the moniliid and trichophytid, which result from sensitization to proteins from organisms other than the tubercle bacillus. The organism is not present in the skin reaction, which is hypersensitive to organisms present elsewhere in the body. The tubercle bacillus has not been found in the phlyctenular lesion itself.

Although the precise mechanism by which phlyctenules are produced has not been determined, it is assumed from clinical evidence that the patient has been sensitized to the offending antigen in the past. In patients with tuberculosis, this sensitization presumably occurred as part of a bacteremia from an early infection in the lungs or lymph glands. Nontuberculosis phlyctenular keratoconjunctivitis caused, for example, by Staphylococcus probably comes from the pathogenic staphylococci that inhabit the lid margins of all people from time to time. The attack of phlyctenular keratoconjunctivitis may be precipitated by the presentation of the antigen to the sensitized ocular tissue, either by the bloodstream in the event of a recrudescence of a focus of infection elsewhere or by an exogenous inoculation of bacteria into the conjunctival sac from the lid margin.

Thygeson and coworkers12 reported that they were not able to reproduce the earlier successes of Gibson13 and Rosenhauch14 in inducing the formation of phlyctenules by instilling tuberculin or Staphylococcus aureus into the conjunctival sacs of tuberculous rabbits. Consequently, Thygeson and coworkers concluded that no model was available for studying the phlyctenules and cellular infiltrates typical of phlyctenular keratoconjunctivitis.

Mondino and coworkers,15 however, have reported a successful experimental model of phlyctenulosis in which rabbits were immunized by intradermal and intramuscular injection of phenol inactivated S. aureus. All animals challenged topically with viable S. aureus developed conjunctival hyperemia and edema in the first week. In 26 of the 28 rabbits, vascular invasion of the peripheral cornea from the limbus developed in an average of 33 days (range: 14 to 49 days). In 8 of 22 rabbits, elevated, nodular lesions of the cornea developed in an average of 29 days (range: 14 to 42 days). Elevated, whitish nodular infiltrates developed approximately 3 mm from the limbus in an average of 54 days (range: 43 to 60 days). Nodular infiltrates gradually flattened within a month after topical challenge with S. aureus was discontinued, leaving a vascularized scar. The nodular corneal infiltrates were found in a subepithelial location and were composed of vessels, polymorphonuclear leukocytes, and mononuclear cells, including lymphocytes, plasma cells, and macrophages. The peripheral corneal infiltrates that were separated from the limbus by a lucid interval were found in the anterior stroma beneath the corneal epithelium and were composed of polymorphonuclear leukocytes and mononuclear cells. The pathologic findings seen in the above infiltrates were not suggestive of a direct infection since no gram-positive organisms were found. Further studies in this rabbit model with S. aureus cell-wall immunization were associated with development of corneal phlyctenules and catarrhal infiltrates, suggesting that cell-wall antigen was responsible for these lesions.16 Three major components of S. aureus cell wall are known to exist; however, only rabbits immunized to one of these components, ribitol teichoic acid (RTA), which is coupled to sheep red blood cells, developed corneal phlyctenules after topical application of viable S. aureus.17 Skin tests for delayed hypersensitivity in these RTA immunized rabbits were negative, suggesting that antibody was more important in the development of corneal lesions. Further studies measured serum, tear, and corneal antibody titers to RTA in rabbits immunized by various routes (intradermal subconjunctival, topical) with staphylococcal antigens.18 Analysis of the data suggests that corneal antibody levels to RTA may be influenced by exposure to staphylococcal antigens not only in the external eye but also at sites remote from the eye.

Infection with parasites has been associated with the development of phlyctenular keratoconjunctivitis. Al-Hussaini and coworkers19 studied the stools of 471 patients suffering from the phlyctenular disease and found that 63% of them had Hymenolepis nana ova in their stools, as compared to 11% of the controls. All patients had H. nana-immune sera. Age incidence and geographic distribution of H. nana infection and phlyctenular keratoconjunctivitis are parallel. Al-Amry et al.20 reported a case of a 6-year-old boy from Saudi Arabia with recurrent corneal phlyctenulosis who had stool analysis positive for H. nana. The parasite was treated and eradicated with praziquantel, and the child’s signs and symptoms of phlyctenulosis resolved without recurrence.

Mounting evidence exists for an association of chlamydial infection and phlyctenular disease. Thygeson11 reported a case of lymphogranuloma venereum phlyctenulosis in 1954. Subsequently, Bialasiewicz and Holbach21 described two patients with phlyctenular disease and evidence of chlamydial conjunctival involvement. More recently, Culbertson and coworkers22 studied 17 cases of phlyctenular keratoconjunctivitis in children and teenagers ranging in age from 3 to 18 years. The workup in 10 of these patients included chlamydial cultures and/or monoclonal immunofluorescent antibody stain of conjunctival smears for Chlamydia antigen. Five of these patients (50%) had at least one positive test for Chlamydia.