TREATMENT O F PSEUDOMONAS E N D O P H T H A L M I T I S ASSOCIATED W I T H P R O S T H E T I C INTRAOCULAR LENS IMPLANTATION DALE

N.

G E R D I N G , M.D., BROOKS J. P O L E Y , M.D., W E N D E L L D O N A L D P. L E W I N , M.D., AND M A L C O L M D. C L A R K , Minneapolis,

Bacterial endophthalmitis has a poor prognosis for useful vision, and a high risk of loss of the eye. 1 ' 2 When intraocu­ lar infection follows placement of a for­ eign body prosthesis within the eye, the prognosis may be worse. An outbreak of Pseudomonas endophthalmitis associated with the implantation of contaminated intraocular lenses 3 provided a unique op­ portunity to study the outcome of treat­ ment of this infection. We report herein the therapeutic management and longterm follow-up of eight patients treated for Pseudomonas endophthalmitis after implantation of contaminated lenses, five of whom retained useful vision and the artificial lens. S U B J E C T S AND M E T H O D S

On Nov. 10 and 11, 1976, eight pa­ tients underwent cataract extraction and implantation of a Copeland-lens (Iris Plane-Intralens Lot 76-285) in a Minneso­ ta hospital. By Nov. 12, 1976, signs of possible intraocular infection were pre­ sent in all of the patients. The pupil appeared brown in color, the red pupil­ lary reflex was reduced or absent, the corneal reflex was absent, excessive tear­ ing was present, and visual acuity was decreased to counting fingers or light perFrom the Infectious Disease Section, Department of Medicine, Veterans Administration Medical Cen­ ter and University of Minnesota Medical School (Drs. Gerding and Hall); Ophthalmology Profes­ sional Associates (Drs. Poley and LeWin); and Min­ neapolis Internists P. A. (Dr. Clark), Minneapolis, Minnesota. Reprint requests to Dale N. Gerding, M.D., Infec­ tious Disease Section, Veterans Administration Medical Center, 54th St. and 48th Ave. S., Minne­ apolis, MN 55417. 902

H. H A L L , M.D.

M.D.,

Minnesota

ception in all eight patients. In two pa­ tients the cornea was cloudy and a hypopyon was present. All eyes showed a +2 or greater flare on slit-lamp examination, and cells within a coagulum were seen anterior and posterior to the intraocular lens. Tenderness, pain, and conjunctival chemoses were minimal, and only one patient had a purulent conjunctival exudate. Immediate vitreous aspiration through a scleral incision was done under sterile conditions in the operating room on the two most seriously affected patients. Four unused lenses from the same lot were aseptically removed from their containers and cultured. Pseudomonas aeruginosa with the same antibiogram was grown from the eyes of both patients and all four unused lenses. An organism characterized as P. acidovorans was additionally recov­ ered from three unused lenses (same lot) similarly cultured. This organism dif­ fered markedly in antibiotic susceptibili­ ty from the P. aeruginosa strain and was not recovered from vitreous aspirates. All organisms were tested for antibiotic susceptibility by a disk-plate method. 4 Pseudomonas aeruginosa was tested for quantitative susceptibility to gentamicin, tobramycin, amikacin sulfate, and carbenicillin using Mueller-Hinton agar, s and susceptibility to gentamicin, tobra­ mycin, and amikacin was tested in liquid media using Mueller-Hinton broth. 6 Gentamicin content of serum and vitre­ ous fluid was determined by a disk-plate bioassay technique using a carbenicillinresistant Klebsiella pneumoniae, with 24-hour incubation to increase sensitivi­ ty. 7 Carbenicillin assays were also per-

AMERICAN JOURNAL O F OPHTHALMOLOGY 88:902-908, 1979

PSEUDOMONAS

VOL. 88, NO. 5

formed by a 24-hour disk-plate bioassay technique using P. aeruginosa as the test organism, with the p H of the media re­ duced to 5.5 to inactivate any gentamicin in the specimens. Standard curves were prepared in serum for both vitreous fluid and serum specimens. 8 RESULTS

Initially all patients were presumed to have either an infectious or toxic inflam­ mation of the eye. As a result, treatment was directed at likely bacterial organisms, as well as the use of anti-inflammatory corticosteroids to reduce inflammation. All patients received 40 mg of gentami­ cin via subtenon's capsule per day, 1 g of chloramphenicol every six hours intrave­ nously, and 2 million units of penicillin G every six hours intravenously. Cephalothin (1.5 g) every four hours intrave­ nously was substituted for penicillin G later on the first treatment day, before culture results were available. Each pa­ tient was also given 80 mg of prednisone orally and prednisolone, chlorampheni­ col, and polymyxin B-neomycin-gramicidin ophthalmic drops every two hours. All prosthetic lenses were left in place for the first 24 hours of treatment. On the second treatment day gramnegative bacilli were reported growing from both vitreous aspirates and all of the unused lenses. The clinical appearance of the eyes was unchanged except for Pa­ tient 8, who exhibited an increasing hypopyon. The patient was taken to sur­ gery and the intraocular lens was re­ moved. Antibiotics were continued. On the third treatment day the infecting organism was identified as P. aeruginosa, which was susceptible to gentamicin, carbenicillin, tobramycin, and colistin. At this time systemic antibiotic therapy was changed to 60 to 100 mg of gentamicin every eight hours and 5 g of carbenicillin every four hours intravenously. Chloram­ phenicol and cephalothin were discontin­

903

ENDOPHTHALMITIS

ued. Daily subtenon gentamicin was con­ tinued and administration of gentamicin eyedrops was started. By the fourth treatment day the operat­ ed eye of four patients had cleared. Three patients retained a pupillary coagulum. The eye of one of the three slowly cleared on antibiotic therapy. A total of five pa­ tients were successfully treated without removal of the intraocular lens. Visual acuity after treatment was 6/6 (20/20) in two, 6/9 (20/30) in two, and 6/15 (20/50) in one. Average duration of systemic anti­ biotic therapy was nine days with a mean of 6.9 days of gentamicin and five days of carbenicillin. The outcome of all patients is summa­ rized in Table 1. Patient 6 had a deteriora­ tion in renal function, necessitating a discontinuation of antibiotic therapy. A vitreous coagulum then returned and vitrectomy was performed. Glaucoma fol­ lowed and the eye was removed one month after intraocular lens implantation. Pseudomonas aeruginosa was cultured from the eye at the time of enucleation. Inflammation resolved in the eye of Pa­ tient 8 after removal of the intraocular lens, but the eye became soft and was enucleated eight months later. Patient 7 had a vitrectomy performed on day six, but a pupillary membrane persisted, and one year after intraocular lens implanta­ tion the eye became irritable and phthisi­ cal and was enucleated. The remaining five patients have shown no signs of reTABLE 1 CLINICAL OUTCOME O F PSEUDOMONAS E N D O P H T H A L M I T I S ASSOCIATED W I T H C O N T A M I N A T E D INTRAOCULAR L E N S P R O S T H E S E S

Patient No. 1-5 6,7 8

Vitreous Culture Not done P. aeruginosa P. aeruginosa

Intraocular Lens Removed

Eye Removed

No No Yes

No Yes Yes

CO O

R

3 un­ used lenses

P.

*S designates sensitive; R, resistant.

acidovorans

S

8 7 3 un­ used lenses

Source

Polymyxin B

P. aeruginosa

Organism

TABLE 2

R

S

Gentamicin

R

S

Tobramycin

R

S

Amikacin

R

S

Carbenicillin

S

R

Chloramphenicol

A N T I B I O T I C DISK S U S C E P T I B I L I T Y O F P. AERUGINOSA A N D P. ACIDOVORANS FROM INFECTED PATIENTS AND UNUSED INTRAOCULAR LENSES

VOL. 88, NO. 5

PSEUDOMONAS

current infection one to two years after intraocular lens implantation and have retained visual acuity. Table 2 shows the antibiotic suscep­ tibility pattern of the two Pseudomonas strains isolated in this epidemic. Only P. aeruginosa was recovered from the eyes of infected patients, but both P. aerugino­ sa and P. acidovorans were recovered from the unused intraocular lenses. T h e antibiotic susceptibility patterns of these two organisms were markedly different; P. aeruginosa was sensitive only to genta­ micin, tobramycin, amikacin sulfate, car­ benicillin, and polymyxin B; but P. aci­ dovorans was uniformly resistant to all of these drugs and susceptible only to chloramphenicol, tetracycline, trimethoprimsulfamethoxazole, and sulfisoxazole. Both strains were resistant to cephalothin and ampicillin. The minimal inhibitory concentrations of four antibiotics for P. aeruginosa in broth and agar media are shown in Table 3. Tobramycin was the most active aminoglycoside tested, but the organism was also highly sensitive to gentamicin. Because gentamicin was the treatment drug, the minimum bactericidal con­ centration was determined in MuellerHinton broth and was 2.0 M-g/rnl. Patients 6 and 7 had antibiotic concen­ trations measured in serum a n d vitreous fluid at the time of vitrectomy. Concen­ trations of gentamicin in vitreous fluid were low; however, carbeniciljin con-

905

ENDOPHTHALMITIS

centrations were surprisingly high at 96 and 140 M-g/ml, and exceeded simul­ taneous serum values by two- to fourfold (Table 4), and minimal inhibitory con­ centration values by three- to fourfold (Table 3). DISCUSSION

Infections of implanted prosthetic de­ vices such as heart valves, cerebral ven­ tricular shunts, and artificial joints fre­ quently require removal of the prosthesis for successful treatment. 9 - 1 1 In the only previously reported outbreak of infection of intraocular lens implants, a fungal or­ ganism, Paecilomyces lilacinus, was the cause and lens removal was required in most cases, 12 but treatment was hampered by lack of an effective chemotherapeutic drug. 1 3 On the basis of this previous expe­ rience we initially predicted a poor prog­ nosis for intraocular lens preservation in our outbreak. Surprisingly, the intraocu­ lar lens was retained in five of our eight patients and recovery of vision was excel­ lent in all five. Since only three of the eight patients had vitreous cultures done, there is the possibility that not all of the patients were infected. We think this is unlikely, based on the clinical findings of loss of red reflex, loss of visual acuity, and coagulum surrounding the intraocular lens in all eight patients. In contrast, three patients underwent intraocular lens implantation with another type of lens during the out-

TABLE 3 Q U A N T I T A T I V E A N T I B I O T I C S U S C E P T I B I L I T Y O F P. AERUGINOSA I S O L A T E D F R O M CASES O F ENDOPHTHALMITIS ASSOCIATED WITH INTRAOCULAR LENS IMPLANTATION*

Antibiotic

Mueller-Hinton Broth M I C (n-g/ml)

Mueller-Hinton Agar MIC (p.g/ml)

Gentamicin Tobramycin Amikacin Carbenicillin

0.25 < 0.125 1.0 -

0.5 < 0.25 8.0 32.0

*MIC designates minimum inhibitory concentration.

906

AMERICAN JOURNAL OF OPHTHALMOLOGY

NOVEMBER, 1979

TABLE 4 S I M U L T A N E O U S A N T I B I O T I C C O N C E N T R A T I O N S IN SERUM AND V I T R E O U S

Patient No. 7 6

Antibiotic Gentamicin Carbenicillin Gentamicin Carbenicillin

Dose 80 5 80 5

mg g mg g

Time Since Last Dose 9 2 4 1

hr hr hr hr*

Antibiotic Concentration (M-g/ml) Serum Vitreous 2.1 22.0 3.0 83.0

Undetectable 96.0 1.7 140.0

* Exact time not recorded; time estimated from dosing schedule.

break period, and none exhibited any of the clinical signs of infection shown by these eight patients. Additionally, all four unused lenses from Lot 76-285 grew Pseudomonas species when cultured, suggesting a high level of contamination. Early signs of severe inflammation (hypopyon, eyelid edema) in two of the pa­ tients alerted the attending physician to the more subtle signs (loss of red reflex, diminished visual acuity) in the remain­ der of the patients. This enabled the phy­ sician to treat the patients with antibiotics at an early stage of infection. Of the initial drugs chosen for empiric treatment (peni­ cillin, cephalothin, chloramphenicol, and gentamicin), only gentamicin was effec­ tive in vitro against the P. aeruginosa isolated from vitreous fluid (Table 2). Considering the low level of gentamicin found in one of the vitreous aspirates (Table 4), the effectiveness of this drug can be questioned. In contrast, carbeni­ cillin achieved high vitreous concentra­ tions, but was not begun until the third day of treatment. Thus, the exact role of each drug in the successful therapy of these infections remains unclear. Barza and associates 8 studied the intra­ ocular penetration of carbenicillin in rabbits and found poor penetration of vitreous with both systemic and subconjunctival administration. Extremely high doses of carbenicillin (100 mg/kg of body weight, intravenously) resulted in occa­

sional vitreous concentrations as high as 50 (xg/ml, but they did not reach the levels observed in the two patients we studied. However, some aqueous carbenicillin concentrations in inflamed rabbit eyes were similar to those achieved in our patients. A possible explanation for our high vitreous levels may be contamina­ tion by aqueous at the time of vitrectomy. Because simultaneous serum levels were much lower than the vitreous levels (Table 4), we cannot attribute the high vitreous levels to contamination by blood. Treatment of bacterial endophthalmitis remains a challenging problem. Early in­ stitution of antibiotic, proper choice of antibacterial spectrum, and adequate de­ livery of antibiotic to the site of infection have been shown to influence out­ come. 1 4 " 1 6 A wide variety of organisms cause postoperative ocular infections, 15 necessitating initial empiric therapy with a broad spectrum of antibiotics, because delay in initiation of effective treatment worsens the prognosis. The aminoglycoside antibiotics (gentamicin, tobramycin, and amikacin) cannot be relied upon alone to be effective against all of the wide array of potential gram-negative pathogens (as evidenced by the intrao­ cular lens contamination in this outbreak with aminoglycoside-resistant as well as sensitive strains of Pseudomonas). Injec­ tion of antibiotics directly into the eye (subconjunctival, subtenon's, intracam-

PSEUDOMONAS

VOL. 88, NO. 5

eral, intravitreous) enhances the concen­ tration of drug in eye tissues, 1 7 - 2 0 but eye toxicity may occur, 21 particularly with intravitreal instillation. 1 8 - 2 0 Clearly more experimental work is required with newer antibiotics to establish safe and effective modes of therapy. Our experience in this outbreak suggests that endophthalmitis associated with intraocular lens implanta­ tion can be successfully treated without removal of the intraocular lens. Addition­ ally, vitreous concentrations of carbeni­ cillin were found to be surprisingly high, suggesting an effective therapeutic role for this drug when susceptible infecting organisms are present. SUMMARY

Eight patients were treated for Pseudomonas endophthalmitis associated with the implantation of contaminated in­ traocular lenses. All patients showed clinical signs of infection (loss of red reflex, diminished visual acuity, and in­ traocular lens coagulum) and P. aeruginosa was isolated from vitreous aspirates and unused lenses of the same lot. Antibi­ otic treatment was initiated with systemic penicillin G, cephalothin, and chloramphenicol as well as subtenon-injected gentamicin. After identification of the or­ ganism, treatment was continued with systemic carbenicillin and gentamicin and subtenon-injected gentamicin. The intraocular lens was left in place for the first 48 hours of treatment in all eight patients. Five patients were successfully treated without removal of the intraocular lens and attained visual acuity of 6/6 (20/20) to 6/15 (20/50). Three patients (the two most seriously infected and one in whom antibiotics were discontinued) eventually lost their infected eye. Vitre­ ous concentrations of gentamicin were good in one patient (1.7 |xg/ml) and undetectable in another. Carbenicillin concen­ trations in vitreous (96 and 140 M-g/nd) were high in two patients sampled. Endo­

ENDOPHTHALMITIS

907

phthalmitis in the presence of a prosthetic intraocular lens can be successfully treat­ ed in some patients without removal of the prosthesis. ACKNOWLEDGMENTS

Thomas W. Purcell, M.D., John S. Andrews, M.D., William J. Paule, M.D., William B. Torp, M.D., and John Washburn B.A., assisted in patient care and outbreak investigation.

REFERENCES 1. Forster, R. K.: Endophthalmitis. Diagnostic cultures and visual results. Arch. Ophthalmol. 92:387, 1974. 2. Golden, B., and Meek, E. S.: Intraocular infec­ tions. In Hoeprich, P. D. (ed.): Infectious Diseases, 2nd ed. Hagerstown, Harper and Row, 1977, pp. 1179-1182. 3. Bauer, H., Ozols, L., Poley, B., Gerding, D., Andrews, J. S., Washburn, J., Stambaugh, G., Smith, E. W. P., Chambers, C. F., Lewis, J. N., Chin, J., and Hall, W. H.: Endophthalmitis associated with implantation of intraocular lens prosthesis—United States. Morbidity Mortality Weekly Rep. 25:369, 1976. 4. Matsen, J. M., and Barry, A. L.: Susceptibility testing. Diffusion test procedures. In Lennette, E. H., Spaulding, E. H., and Truant, J. P. (eds.): Manual of Clinical Microbiology, 2nd ed. Washing­ ton, D. C , American Society for Microbiology, 1974, pp. 422-427. 5. Ericsson, H. M., and Sherris, J. C : Antibiotic sensitivity testing. Report of an international collab­ orative study. Acta Pathol. Microbiol. Scand. [B] 217 (Suppl.):79, 1971. 6. Washington, J. A., and Barry, A. L.: Dilution test procedures. In Lennette, E. H., Spaulding, E. H., and Truant, J. P. (eds.): Manual of Clinical Microbiology, 2nd ed. Washington, D. C , American Society for Microbiology, 1974, p p . 414-416. 7. Lund, M. E., Blazevic, D. J., and Matsen, J. M.: Rapid gentamicin bioassay using a multipleantibiotic-resistant strain of Klebsiella pneumoniae. Antimicrob. Agents Chemother. 4:569, 1973. 8. Barza, M., Baum, J., Birkby, B., and Weinstein, L.: Intraocular penetration of carbenicillin in the rabbit. Am. J. Ophthalmol. 75:307, 1973. 9. Wilson, W. R., Jaumin, P. M., Danielson, G. K., Giuliani, E. R., Washington, J. A., II, and Geraci, J. E.: Prosthetic valve endocarditis. Ann. Intern. Med. 82: 751, 1975. 10. Shurtlefr, D. B., Foltz, E. L., Weeks, R. D., and Loeser, J.: Therapy of Staphylococcus epidermidis infections associated with cerebrospinal fluid shunts. Pediatrics 53:55, 1974. 11. Kettlekamp, D. B.: Infected total joint re­ placement. Arch. Surg. 112:552, 1977. 12. Webster, R. G., Martin, W. J., Pettit, T. H., Rhodes, J., Boni, B., Midura, T., and Skinner, M. D.:

908

AMERICAN JOURNAL OF OPHTHALMOLOGY

Eye infections after plastic lens implantation. Mor­ bidity Mortality Weekly Rep. 24:437, 1975. 13. O'Day, D. M.: Fungal endophthalmitis caused by Paecilomyces lilacinus after intraocular lens implantation. Am. J. Ophthalmol. 83:130,1977. 14. May, D. R., Ericson, E. S., Peyman, G. A., and Axelrod, A. J.: Intraocular injection of gentami­ cin. Single injection therapy of experimental bacte­ rial endophthalmitis. Arch. Ophthalmol. 91:487, 1974. 15. Forster, R. K., Zachary, I. G., Cottingham, A. J., Jr., and Norton, E. W. D.: Further observa­ tions on the diagnosis, cause, and treatment of endophthalmitis. Am. J. Ophthalmol. 81:52, 1976, 16. Cottingham, A. J., Jr., and Forster, R. K.: Vitrectomy in endophthalmitis. Arch. Ophthalmol. 94:2078, 1976.

NOVEMBER, 1979

17. Golden, B.: SubTenon injection of gentami­ cin for bacterial infections of the eye. J. Infect. Dis. 124 (Suppl.):S271, 1971. 18. Peyman, G. A., May, D. R., Ericson, E. S., and Apple, D.: Intraocular injection of gentamicin. Toxic effects and clearance. Arch. Ophthalmol. 92:42, 1974. 19. Zachary, I. G., and Forster, R. K.: Experimen­ tal intravitreal gentamicin. Am. J. Ophthalmol. 82:604, 1976. 20. Graham, R. O., Peyman, G. A., and Fishman, G.: Intravitreal injection of cephaloridine in the treatment of endophthalmitis. Arch. Ophthalmol. 93:56, 1975. 2 1 . Kobetz, L. E., Bussanich, M. N., and Rootman, J.: Toxic effects of subconjunctival dicloxacillin. Can. J. Ophthalmol. 13:206, 1978.

O P H T H A L M I C MINIATURE

At age 13 to 14 while hunting, Theodore Roosevelt found "that he could not hit anything." Even more puzzling was the fact that friends using the same gun seemed to be able to bag the invisible. The truth was slow to dawn on him. "One day they read aloud an advertisement in huge letters on a distant billboard, and I then realized that something was the matter, for not only was I unable to read the sign, but I could not even see the letters. I spoke of this to my father, and soon afterwards got my first pair of spectacles . . . while much of my clumsiness and awardness was doubtless due to general characteristics, a good deal of it was due to the fact that I could not see, and yet was wholly ignorant that I was not seeing." E d m u n d Morris, The Rise of Theodore Roosevelt Coward, McCann, and Geoghegan, 1979