CURRENT COMMENT Ongoing Assessment of Therapy in Septic Arthritis By FRANKR. S ~ M I DM.D., , AND RICHARD H. P-, I M.D. of patients with septic arthritis, questions about treatment continue to be points of controversy. Whether antibiotics must be instilled intraarticularly and what type of joint drainage is required are two prime issues, still unresolved. Some physicians favor routine injections of antibiotics into the joint, as well as systemically, and always perform open drainage of the joint with or without continuous Yet closed-space infections in other regions of the body, such as the pleural or subarachnoid space, usually are not handled in this empirical fashion. Antibiotics in these situations are customarily administered solely by the systemic route, and decisions about the required type of drainage are based upon assessment of the condition of the individual patient. Underlying this approach is abundant evidence that antibiotics penetrate enclosed and infected body spaces and achieve bactericidal level^.^ Why is the approach to septic arthritis in dispute? Two reasons may be advanced. First, there is a dearth of data about the transport of antibiotics into joints in contrast to other body spaces.6 Consequently, some clinicians are reluctant to abandon constant intraarticular instillation of antibiotics and to rely on systemic administration alone. Second, the rapidity of joint destruction by products of the inflammatory exudate creates an understandable urgency in the treatment regimen. The consequences of a damaged joint are a constant reminder to the patient and the physician in a way that an adhesive pleura or thickened arachnoid membrane is not. However, it is our contention that the basic principles that govern treatment of closed-space infections in other regions of the body should apply also to the joint. From the Arthritis-Connective Tissue Diseases and Infectious Diseases-Hypersensitivity Sections, Department of Medicine, Northwestern University Medicd Center, Chicago, Illinois. Supported in part b y USPHS Training Grant AM-05069 and Reseurch Grant A&l-11513 from the N a t i d Institute of Arthritis and Metabolic Diseases, National Institutes of Health, Bethesda, Maryland, rmd by grants-in-aid from Wyeth Laboratorfes, Radnor, Pennqllvadu, and Eli Lilly and Co., Indianapolis, Indiana. FRANKR. SCHMID,M.D.: Associate Professor of Medicine, and Chief, Section of Arthritis-Connective Tissue Diseases, Department of Medicine, Northwestern University Medical School, 303 East Chicago Avenue, Chicago, Illinois 60611 . RICHARDH. P-, M.D.: Assistant Professor of Medicine, and Chief, Infectious Disease Section, Veterans Administration Research Hospital, Chicago, Illinois. Reprint requests should be addressed to Dr. Schmid. N THE MANAGEMENT ARTmuTIs AND Antibiotic Transport into loints Early reports suggested that concentrations of drugs achieved intraarticularly were suboptimal during systemic therapy.7-18However, the doses of antibiotics used in early studies were small by present standards, and in some instances the studies utilized nonirdected or nonidamed joints in which the transport of antibiotics might be reduced. More recent evidence indicates that bactericidal concentrations of many antibiotics can be attained in the joint after RHEUMATLS~,VOL. 12,-No. 5 (OCTOBER1969) 529 SCHMID AND PARKER 530 / Studies in our laboratories30 designed to provide information along these lines were 2o01 performed in 15 patients. Antibiotic concentrations were determined on 34 paired lo 5o 0t 1: and simultaneously collected specimens of serum and synovial fluid over variable periSERUM ods of treatment. The infections were caused by staphylococci, gonococci, pneumococci, streptococci, and presumed gonococcal infection. Antibiotics used included penicillin, nafcillin, cloxacillin, cephaloridine, tetracycline, erythromycin, and lincomycin. In Y , <5 5 10 25 50 100 200 all cases, levels of antibacterial activity SYNOVIAL FLUID achieved intraarticularly were either equal Fig. 1.-Antimicrobial activity of antibiotic to or only slightly less than those noted in as determined by tube dilution method in the serum (Fig. 1).Further observations synovial fluid contrasted with level in serum have been made in additional patients and obtained at the same time. Each point repre- confirm these initial results.31 Based upon sents the result obtained from a pair of specimens of synovial fluid and serum. Values are this information, administration of the drug expressed as the reciprocal of that dilution of was started and usually continued by the the specimen at which a bactericidal effect intravenous or intramuscular route. In a few was observed. cases following improvement in synovitis, oral administration was substituted later to systemic administration, provided sufficient complete therapy in the h m knowledge dosage is Nevertheless, the case for that intraarticular inoculation still was not intraarticular inoculation of antibiotics con- required. Although resolution of joint effutinues to be made, even by those who have sions usually occurred within 1week, pershown that antimicrobials cross the synovial sistence of a sterile effusion was observed membrane, because of the fear that only in in several patients. Such an effusion does this way will large concentrations of anti- not necessarily indicate continuing active biotic be present where they are needed.24*25 infection. In our patients, the persistence To overcome this fear it is necessary to was considered due to underlying disease show that systemic administration raises the (rheumatoid arthritis, tophaceous gout, and concentration of the drug in the synovial one patient on immunosuppressive drugs fluid above the minimal level which is bac- for renal homotransplantation). As long as tericidal for the infecting microorganism. such effusions were present, they were asFurthermore, it must be possible to main- pirated and examined. The antibacterial tain such bactericidal concentrations until activity of these fluids continued to be in the microorganism is eradicated from the the same range as the serum activity. area. With one exception, in which animals If our experience is taken together with were infected experimentall~,2~ no sequen- the single determinations of intraarticular tial study of transport of antibiotics into in- drug levels reported previously by others, fected joint fluid has been carried out to direct inoculation of antibiotics into the show that bactericidal intraarticular con- joint space is not a prerequisite for achievcentrations can be maintained over the pro- ing adequate levels in the joint. Indeed in longed time often required for control of some cases such local instillation results in infection. a chemical synovitis added to an already ( 3 4 pairs/lJ patients) 25t I THERAPY IN SEPTIC ARTHRITIS existing inflammatory p r o c e ~ sAn . ~ addi~~~ tional reason for stressing systemic antibiotic therapy-a reason accepted by all clinicians-is that control of infection at points remote from the joint is often required. These extraarticular infections may be either primary infections that led to seeding of the joint initially, or they may represent metastatic infection as a result of the bacteremia that frequently accompanies septic arthritisF3 Regimen for Systemic Antibiotic Therapy The following regimen is proposed for the routine antibiotic management of patients with infectious arthritis. It involves close collaboration between the clinician and the laboratory. However, its technical aspects are not outside the level of skill available in the bacteriological laboratories of most hospitals. 1. Examination of joint fluid should include a gram-stained smear of synovial fluid for bacteria. 2. Culture of joint fluid and other body fluids involved in the infectious process (including two blood cultures) should be completed prior to starting antibiotics. 3. The antibiotic selected should be administered either intramuscularly or intravenously. Oral administration at the onset cannot be relied upon to provide prompt and sustained effective blood levels of the drug. 4. Joint fluid should be aspirated as often ds it accumulates, to allow drainage and to compare antibacterial activity of the specimen with blood specimens obtained simultaneously. This procedure naturally will be discontinued when detectable joint fluid no longer remains. 5. Antibacterial activity is determined by a simple tube dilution te~hnic.3~ The validity of this method compares favorably with more precise but more difficult bioassay methods utilizing agar diffusion technics. If possible, bacteria isolated from the in- 531 ~fected joint should serve as the best microorganism against which bactericidal levels of the antibiotic can be determined. If the infecting bacteria are not available, then an appropriately sensitive laboratory strain can be used for assay purposes. 6. Continuation of antibiotic, possibly by the oral route, for at least 7-10 days after all signs of joint inflammation have disappeared. I n the series of patients noted above, this protocol resulted in control of the infection in all cases. Sterility of the joint fluid and blood stream was always achieved. Dramatic clinical improvement occurred within a week in those patients without serious underlying disease in whom the diagnosis was made within a few days of onset. This was particularly true in the patients with gonococcal arthritis. Drainage of Joints In addition to achieving bactericidal antibiotic levels in synovial fluid and tissues, the removal of purulent material is of fundamental importance. The presence of retained pus retards the action of many antibiotics by inhibiting the rate of growth of infecting bacteria. Slowly metabolizing bacteria can persist in pus even in the presence of concentrations of the drug well above the minimal bactericidal ~oncentration.3~"~ I n addition, increased intraarticular pressure3* and the enzymatic products of idammation are able to erode cartilage and b 0 n e . ~ ~ - ~ 4 As often and as soon as fluid accumulates in the joint, it should be removed. This can be accomplished almost always at the outset by needle aspiration. Such fluids should be subjected to the usual methods of analysis46 as well as to determination of bactericidal activity. In deeper structures, such as the hip, needle aspiration can be repeated daily until fluid accumulation ceases. The effectiveness of this approach is judged over the course of the first EL7 days of treatment 532 SCHMID AND PARgER by noting whether the volume of drainage is decreasing and whether the character of the fluid, as reflected in cell count, glucose level, and other parameters, is returning toward normal. Adherence to these guidelines will decide whether closed drainage by needle aspiration is adequate. Persistence of effusion beyond this time, however, may require more aggressive attempts at drai11age.3~*~6-52 Incision of the joint space then might be necessary to remove necrotic debris and enter loculated areas of fluid. Incision and drainage at the onset of treatment rather than later in the course may be advisable in infants with septic arthritis of deeper joints, such as the shoulder or hip, where diEculty might be anticipated in securing proper drainage and where the clinical signs of inflammation are ~ b s c u r e d Still . ~ ~later ~ ~in~ the ~ ~course ~ of the disease, it may be necessary to perform a synovectomy and/or reconstructive procedure for a joint that has developed mechanical impairment. This review has supported the view that principles for optimal treatment of closedspace infections can be applied equally well to joints as to other body areas. Analogy to patients with lung abcess is relevant. Control of infection in and around the abcess is achieved by systemic antibiotic therapy. Surgical drainage is deferred unless intrabronchial drainage fails, spontaneous reabsorption does not occur, and/or fistula or other complications develop. Later, segmental lung resection is considered if the residual activity persists. So also with infected joints. Here the physician, however, is in the unique position of being able to monitor his course of action because the site of the inflammation is usually readily accessible. With the aid of a few simple studies, he can make the proper decisions in each case of septic arthritis by utilizing sound principles of antibiotic therapy and needle aspiration and carefully considering what additional benefit would be achieved by open drainage. ACKNOWLEDGMENTS We wish to express our gratitude to Dr. Philip Y. Paterson for his critical review of this manuscript and to Miss Joan Davis for her help in the literature search. REFERENCES 1. Clawson, D. K., and Dunn, A. W.: Management of common bacterial infections of bones and joints. J. Bone Joint Surg. 49A:164, 1967. 2. Compere, E. L., Metzger, W. I., and Mitra, R. N.: The treatment of pyogenic bone and joint infections by closed inrigation (circulation) with a nontoxic detergent and one or more antibiotics. J. Bone Joint Surg. 49A:614, 1967. 3. Ortiz, A. C., and Miller, W. E.: Treaiment of a septic joint. Southern Med. J. 54:594, 1961. 4. Steindler, A.: Pyogenic arthritis. Bull. N.Y. Acad. Med. 27:101, 1951. 5. Hoeprich, P. D., and Ward, J. R.: The Fluids of parenteral Body Cavities. N~~ yo&, Grune & Stratton, 1959, p. 10. 6. c,,.,jss,p. H., Ira:Changes produced in the synovial membrane and synovial fluid by disease (instructional course lecture). J. Bone Joint Surg. 48A:873, 1964. 7. Rammelkamp, C. H., and Keefer, C. S.: The absorption, excretion and distribution of penicillin. J. Clin.Invest. 22:4!25, 1943. 8. Herreil, W. E., Nichols, D. R., and Heilman, D. H.: Penicillin. Its usefulness, limitations, diffusion and detection, with analysis of 150 cases in which it was employed. J.A.M.A. 125:1003, 1944. 9. Bdboni, v. G.,Shapiro, 1. M., and Kydd, D. M.: fie penetration of penicillin into joint fluid following intramuscular administration. h e r . J. Med. sci, 210:588, 1945. 10. Florey, M. E., and Heatley, N. G.: Systemic of penicillin by absorption from body cavities. Lancet 1:748, 1945. 11. M c A d a , 1. w. J . 2 DuSid, 3. P v C h a b o r , S. W., and McCall, A.: Penicillin treatment of serous cavity infections. Lancet 2:843, 1945. 12. ow, E. Ma, Meads, M-9 Brown, B.9 wilmx, c.,and Finland, Me: Penicillin levels in Serum and in some body fluids during systemic and local therapy. J. Lab. clin. Med. 30:809, 1945. 13. Hirsh, H. L., Feffer, H. L., and O’Neil, C. B.: A study of the diffusion of penicillin across THERAPY IN SEPTIC ARTHRITIS serous membranes of joint cavities. J. Lab. Clin. Med. 31:535, 1946. 14. Jones, G. B.: The behavior of penicillin in synovial cavities. J. Bone Joint Surg. 30B:107, 1948. 15. Linsell, W. D., and Fletcher, A. P.: Laboratory and clinical experience with terramycin hydrochloride. Brit. Med. J. 2:1190, 1950. 16. Werner, C. A., Knight, V., and McDermott, W.: Absorption and excretion of terramycin in humans: comparison with adeomycin and chloramphenicol. Proc. SOC. Exp. Med. Biol. 74:261, 1950. 17. Jocson, C. T.: Diffusion of antibiotics through the synovial membrane. J. Bone Joint Surg. 37A:l07, 1955. 18. Bogacki, B., Szulc, H., Sarnowski, M., Majewski, C., Kubiak, E., and Leja, Z.: The effect of intramuscular and intra-articular administration of penicillin and streptomycin injections in experimental bacterial infection of the knee joint. Narzad. Ruchr. i Ortop. Polska. 25.573, 1960. In Polish. 19. Geraci, J. E., Heilman, F. R., Nichols, D. R., Wellman, W. E., and Ross, G. T.: Some laboratory and clinical experiences with a new antibiotic, vancomycin. Proc. Staff Meet, Mayo Clin. 31:564, 1956. 20. Pulaski, E. J., and Tubbs, R. S.: Inhibitory effects of kanamycin and diffusion into various body fluids. Antibiot. Med. Clin. Therap. 6:589, 1959. 21. Willikens, R. F., Healey, L. A., and Decker, J. L.: Acute infectious arthritis in the aged and chronically ill. Arch. Intern. Med. 106:354, 1960. 22. Viek, P.: Concentration of sodium nafcillin in pathological synovial fluid. In: Antimicrobial Agents and Chemotherapy. Ann Arbor, American Society for Microbiology, 1962, p. 379. 23. Viek, P., and Santangelo, S. C.: Pyarthrosis: a surgical emergency. J. Int. Coll. Surg. 37:88, 1962. 24. Balboni, V. G.: Other forms of infectious arthritis. In: Hollander, J. (Ed.): Arthritis and Allied Conditions (ed. 7 ) . Philadelphia, Lea & Febiger, 1966, p. 1037. 25. Bardenheier, J. A,, 111, Morgan, H. C., and Stamp, W. G.: Treatment and sequelae of experimentally produced septic arthritis. Surg. Cynec. Obstet. 122:249, 1966. 26. Eli Lilly and Co.: Profile of an Antibiotic. Indianapolis, Indiana, 1966, p. 23. 27. Rapp, G. F., CrifEth, R. S., and Hebble, W. M.: The permeability of traumatically inflamed synovial membrane to commonly used antibiotics. J. Bone Joint Surg. 48A:1534, 1966. 533 28. Drutz, D. J., Schaffner, W., Hillman, J. W., and Koenig, M. G.: The penetration of penicillin and other antimicrobials into joint fluid. J. Bone Joint Surg. 49A:1415, 1967. 29. Morgan, H. C., Stamp, W. G., and Nickel, J. S.: Quantitative barriers of infected synovium to penicillin. Southern Med. J. 58:710, 1965. 30. Parker, R. H., and Schmid, F. R.: Transport of antibiotics into infected joints. Arthritis Rheum. 11:503, 1968. Abstract. 31. Parker, R. H., and Schmid, F. R.: Unpublished observations. 32. Argen, R. J., Wilson, C. H., Jr., and Wood, P.: Suppurative arthritis. Arch. Intern. Med. 117: 661, 1988. 33. Cluff, L. E., Reynolds, R. C., Page, D. L., and Breckenridge, J. L.: Staphylococcal bacteremia and altered host resistance. Ann. Int. Med. 89:859, 1968. 34. Parker, R. H., Chiu, V. S. W., and Hoeprich, P. D.: Penicillin N therapy of enteric bacillary infections. Arch. Int. Med. 111:799, 1963. 35. Eagle, H.: Experimental approach to the problem of treatment failure with penicillin. Amer. J. Med. 13:389, 1952. 36. Wood, W. B., and Smith, M. R.: An experimental analysis of the curative action of penicillin in acute bacterial infections. 1. The relationship of bacterial growth rates to the antimicrobial effect of penicillin. J. Exp. Med. 103: 487, 1956. 37. Rogers, D. E.: Observations on the nature of staphylococcal infections. Bull. N. Y. Acad. Med. 3525, 1959. 38. Phemister, D. B.: The effect of pressure on articular surfaces in pyogenic and tuberculous arthridites and its bearing upon treatment. Ann. Surg. 80:481, 1924. 39. Gaines, L. M., and Shulman, L. E.: In: Proceedings 23rd Annual Meeting, American Rheumatism Association. Bull. Rheum. Dis. 10: 195, 1959. 40. Lack, C. H.: Chondrolysis in arthritis. J. Bone Joint Surg. 41B:384, 1959. 41. Ziff, M., Gribetz, H. J., and Lospalluto, J.: Effect of leucocyte and synovial membrane extracts on cartilage mucoprotein. J. Clin. Invest. 39:405, 1960. 42. Curtiss, P. H., Jr., and Klein, L.: Destruction of articular cartilage in septic arthritis. I. In vitro studies. J. Bone Joint Surg. 45A:797, 1963. 43. Curtiss, P. H., Jr., and Klein, L.: Destruction of articular cartilage in septic arthritis. 11. In vivo studies. J. Bone Joint Surg. 47A:1595, 1965. 44. Evanson, J. M., and Krane, S. M.: Studies 534 on collagenase from rheumatoid synovium in tissue culture. J. Clin. Invest. 47:2639, 1968. 45. Schmid, F. R., and Ogata, R. I.: Synovial fluid evaluation in joint disease. Med. Clin. N. h e r . 49:165, 1965. 46. Hirsh, H. L., Feffer, H. L., and Dowling, H. F.: The treatment of bacterial arthritis with penicillin. New Eng. J. Med. 234:853, 1946. 47. Altemeier, W. A., and Largen, T.: Antibiotic and chemotherapeutic agents in infections of the skeletal system. J.A.M.A. 150:1462, 1952. 48. Chartier, Y., Martin, W. J., and Kelly, P. J.: Bacterial arthritis: experiences in the treatment of 77 patients. Ann. Int. Med. 50:1462,1959. 49. Ward, J., Cohen, A. J., and Bauer, W.: SCHMID AND PAFUCER The diagnosis and therapy of acute suppurative arthritis. Arthritis Rheum. 3:522, 1960. 50. Baitch, A.: Recent observations of acute suppurative arthritis. Clin. Orthop. 22:157,1962. 51. Cohen, A. S., and Kim, J. C.: Acute suppurative arthritis. Mod. Trends Rheum. 1347, 1966. 52. Cobbs, C. G.,and Kaye, D.: Arthritis in infectious diseases. ch. Orthop. 57:57, 1968. 53. Obletz, B. E.: Suppurative arthritis of the hip joint in infants. Clin. Orthop. 22:27, 1962. 54. Stetson, J. W.,DePonte, R. J., and Southwick, W. 0.: Acute septic arthritis of the hip in children. Clin. Orthop. 56:105,168.