Article

Vol. 11. No. 11

985

Eur. J. Clin. Microbiol. Infect. Dis., November 1992, p. 985-989 0934-9723/92/11 0985-05 $ 3.00/0

Association between Staphylococcus aureus Nasopharyngeal Colonization and Septicemia in Patients Infected with the Human Immunodeficiency Virus

T. W e i n k e l , 2 . , R. Schiller 3, E J . F e h r e n b a c h 3, H . D . P o h l e 1

In an attempt to identify risk factors for Staphylococcus aureus septicemia, 136 consecutive HIV-infected patients were investigated for the presence of nasopharyngeal colonization with Staphylococcus aureus and subsequent Staphylococcus aureus infection. Sixty of 136 (44.1%) HIV-infected patients had staphylococci which were detected in the nasopharynx on initial culture compared to 12 of 39 (30.8 %) patients with chronic diseases and U o f 47 (23.4 %) healthy hospital staff. Another 12 HI'V-infected subjects proved to be Staphylococcus aureus carriers on follow-up cultures. Patients with full-bl0wn AIDS had a higher carriage rate compared to subjects who were only HIV-positive (p < 0.05), indicating that Staphylococcus aureus colonized patients were more severely ill. Eight patients with Staphylococcus aureus septicemia were observed, all of whom were carriers; no septicemia occurred in the non-colonized patients (p < 0.01). Colonized patients with neutropenia (< 1000//al) were significantly more likely to develop septicemia (p < 0.01). Nasopharyngeal colonization with Staphylococcus aureus and the presence of an indwelling catheter were established to be factors that help identify patients at risk of acquiring subsequent Staphylococcus aureus infection.

An increased incidence of Staphylococcus aureus septicemia has recently been described in patients with AIDS (1-3), probably as a result of impaired B-cell immune function or neutropenia. Most of these reports associated Staphylococcus aureus septicemia with risk factors for community-acquired Staphylococcus aureus infections, such as intravenous drug abuse (2, 3). However, it was also stated that I/IV-infected patients Without i.v. drug abuse had a higher risk for Staphylococcus aureus septicemia (1), although the reasons for this remained unclear. It has been demonstrated in patients on continuous ambulatory peritoneal dialysis (CAPD) or hemodialysis that nasal carriage of Staphylococcus aureus was significantly associated with subsequent Staphylococcus aureus infection compared to these noncarriers (4-6). A brief report also indicated that HIV-infected individuals had a 1IL Medizinische Kiinik, Universi~tsklinikum Rudolf Vir2chow/Wedding, Freie Universit~it,Berlin, Germany. Medizinische Ktinik, Klinikum Steglitz, Freie Universit~it Berlin, Hindenburgdamm 30, 1000 Berlin 45, Germany. Abteilung fur Medizinisehe Mikrobiologie, Robert-KochInstitut, Berlin, Germany.

high rate of staphylococcal carriage (7). This prompted us to study the frequency of nasopharyngeal Staphylococcus aureus carriage in H I V patients, and to investigate whether the car. rier state correlated with an increased incidence of Staphylococcus aureus septicemia. Furthermore, we compared the carrier rate of the H I V patients with that of two different control groups.

Patients and Methods The study population consisted of consecutively reporting HIV-infected patients who were either hospitalized or treated as out-patients, irrespective of their risk behavior (male homosexuals, i.v. drug addicts, others) or clinical condition. Swabs for demonstration of Staphylo. coccus aureus carriage were taken 1-2 cm inside the anterior nares and from the throat near the tonsillar region, and cultured on appropriate solid and liquid media. Staphylococcus aureus strains were identified by the eoagutase reaction. Swabs from hospitalized patients were taken within 24 hours after admission and once weekly thereafter; swabs from ambulatory patients were taken when the patients reported to our out-patient department. According to the frequency of Staphylococcus aureus positive cultures patients were classified as

986

Eur. J. Clin. Microbiol. Infect. Dis.

persistent carriers (Staphylococcus aureus present in more than 80 % of the cultures), intermittent carriers (20-80 %) or occasional carriers (< 20 %). All patients were investigated for the presence of clinical signs of septicemia, and blood cultures were taken whenever the patient's temperature exceeded 38 *C. The study period lasted eight months; the next four months constituted a follow-up period to detect further eases of Staphylococcus aureus septicemia in patients already ineluded in the study. The antibiotic susceptibility of the organisms isolated was determined using a disk diffusion test. Two groups of individuals served as controls for the nasopharyngeal Staphylococcus aureus carrier state: healthy hospital staff working in an infectious disease unit, and a group of patients with chronic diseases requiring longterm hospitalization (malignancy, hepatic cirrhosis, insulin-dependent diabetes). Follow-up cultures were taken in the same way as described for the HIV-infected subjects. Sigr/ificanee was assumed at the 5 % level (p < 0.05) using the Chi-square test.

Results

During the eight-month study period 136 consecutive HIV-infected patients were investigated, 118 (86.8 %) of w h o m w e r e m a l e s a n d 18 (13.2 %) females. The median age was 35.7 years (range 23-56 years). Sixty-eight (50 %) patients were classified as having full-blown AIDS, 36 (26.5%) as having AIDS-related complex (ARC), and 32 (23.5 %) as being HIV-positive. Risk factors for HIV infection were male homosexuality in 92 (67.6 %), i.v. drug abuse in 34 (25 %), homosexuality and i.v. drug abuse in 2 (1.5 %), and unknown in 8 (5.9 %). A total of 1124 swabs were taken from these 136 individuals (average of 8.3 swabs per person). S t a p h y l o c o c c u s a u r e u s was isolated more frequently from the nose (172 positive swabs) than

the throat (121 positive). Sixty of the 136 (44.1%) HIV-infected patients were culture positive on admission to hospital or on the first examination as an out-patient (Table 1). A m o n g the controls, 12 of 39 (30.8 %) patients with chronic diseases were carriers (median age 57 years; range 30--83 years; 20 males, 19 females), as well as 11 of 47 (23.4 %) healthy hospital staff (median age 33.3 years; range 21-59 years; 21 males, 26 females). Thus, HIV-infected patients had a significantly higher (p < 0.05) carrier rate compared to healthy hospital staff, but not compared to patients with other chronic diseases. The patients with underlying chronic diseases consisted of subgroups whose initial staphylococcal colonization rate was as follows: diabetics on insulin 33.3 % (3 of 9), liver cirrhosis patients 33.3 % (4 of 12), patients with malignancies 27.8 % (5 of 18). Twelve HIV-infected subjects proved to be S t a p h y l o c o c c u s a u r e u s carriers on follow-up cultures, so that overall 72 of the 136 (52.9 %) patients were S t a p h y l o c o c c u s a u r e u s positive. These 12 individuals were merely colonized and only two of them developed staphylococcal septicemia later on. Fifteen (20.8 %) were persistent carriers, 48 (66.7 %) intermittent carriers, and 9 (12.5 %) occasional carriers. Persistent carriers had a higher rate of resistant S t a p h y l o c o c c u s aureus strains compared to intermittent and occasional carriers; resistance was most marked to penicillin, tetracycline and erythromycin. No methicillin-resistant Staphylococcus aureus strains (MRSA) were detected. The 12-hospitalacquired staphylococcal strains showed a higher frequency of antibiotic resistance compared to those isolated from patients colonized on admission. The treatment of opportunistic infections with cotrimoxazole, clindamycin or rifampicin resulted in a change of category of three subjects from initially colonized to non-colonized on fol-

Table 1: Rates of Staphylococcus aureus carriage in the different study groups based on the results of the

initial cultures. Study group

No. of carriers/total no.

Percentage of carriers

HIV-infected patients Male homosexuals i.v. drug addicts Homosexuals + i.v.drug addicts Patients with unknown risk factors

60/136 37/ 92 18/ 34 1/ 2 4/ 8

44.1 40.2 52.9 50 50

Hospitalized patients with other chronic diseases

12/ 39

30.8

Healthy hospital staff

11/ 47

23.4

Vo1.11,1992

987

Table 2: Patient characteristics and Iaboratory findings in eight patients with Staphylococcus aureus septicemia detected in the study period.

Patient Sex no.

HIV status

Risk factor

CD4 count Granulocyte (per/al) count (per lal)

Nasal S. aureus

initially

Portal of entry

carrier status

S. aureus

homosexuality

51

900

positive

i.v. catheter

persistent

?

77

1265

negative

i.v.catheter

intermittent

drugaddiction

40

3300

positive

i.v. catheter

AIDS

drugaddiction

3

180

positive

drugs + i.v. catheter

intermittent

5

M AIDS

homosexuality

58

803

positive

i.v. catheter

persistent

6

M AIDS

homosexuality

9

560

positive

implanted i.v. catheter

7

M AIDS

homosexuality

9

2952

positive

cutaneous lesions*

intermittent

8

F ARC

drug addiction

288

9877

negative

drugs

occasional

1

M AIDS

2

M AIDS

3

F AIDS

4

F

occasional

intermittent

* Herpes zoster, syphilis.

low-up on cultures. Furthermore, only two of the nine subjects on long-term antimycobacterial therapy including rifampicin proved to be COlonized and none of them developed septicemia. Disease progression from A R C (Staphylococcus aureus carriage 55.6 %) to AIDS (58.8 %) did not influence the prevalence of Staphylococcus aureus carriage, but subjects who were HIV infected only were less frequently colonized (37.5 %), and the difference between AIDS and asymptomatic HIV-infected persons was significant (p < 0.05). The CD4 count was not a factor that influenced colonization within the subgroups: no significant difference was observed between colonized AIDS patients (mean CD4 count 48//al, range 0176) compared to non-colonized AIDS patients (45//al, range 3-146), colonized A R C subjects (220//al, range 111-341) compared to noncolonized A R C subjects (243/lal, range 98-370), or colonized HIV-infected patients (416/pl) compared to non-colonized patients (398/~d). Granulocytopenia (< 1000//al) was seen in 10 of 72 (13.9 %) colonized patients compared to 4 of 64 (6.3 %) non-colonized individuals. Patients with Staphylococcal colonization and neutropenia had a significantly higher risk of developing septicemia (4 of 10 = 40 %) compared to subjects With a neutrophil count > 1000/btl (4 of 62 = 6.5 %; P < 0.01). Furthermore, a difference could be seen

between out-patients (11 of 29 Staphylococcus carriers = 37.9%) and hospitalized patients (61 of 107 carriers = 57 %), although this did not reach a significant level.

aureus

During the eight-month study period six cases of Staphylococcus aureus septicemia were observed

in the study group of 136 HIV infected patients, and another two occurred during the follow-up period. The clinical and laboratory findings in these patients are summarized in Table 2. All eight patients with septicemia had a presumed portal of entry which was either an i.v. catheter or implanted i.v. catheter (n = 6), i.v, drug abuse (n = 2) or cutaneous lesions facilitating dissemination (n = 1). In comparison only 38 of 64 (59.4 %) colonized subjects who did not develop septicemia had such predisposing factors (p < 0.05). Staphylococcus aureus septicemia was seen only in Staphylococcus aureus carriers (8/72 = 11.1%), and not in non-colonized patients. This difference is highly significant (p < 0.01). In each of the eight cases of septicemia, blood culture and nasopharyngeal isolates yielded identical antimicrobial resistance patterns. One of the eight septicemia patients died on the day antistaphylococcal treatment was started. Autopsy revealed concomitant staphylococcal pneumonia and endocarditis. Another subject had a systolic murmur with vegetations suggesting endocarditis. The seven surviving patients received treatment with flucloxacillin.

988

Discussion

The most common opportunistic infections in patients with AIDS are due to microorganisms that take advantage of a T-cell defect, such as Pneumocystis carinii, Toxoplasma gondii, cytomegalovirus and mycobacteria. B-cell dysfunction can also predispose to infection in these patients (8, 9), as.well a s neutropenia which is usually caused by antimicrobial or anti-HIV treatment regimens with ganciclovir, zidovudine or cotrimoxazole. Staphylococcus aureus septicemia has been observed with increasing frequency in HIV-infected subjects (1-3). It ranks directly after Mycobacterium avium-intracetlulare in the order of frequency of bacterial pathogens associated with septicemia in AIDS patients. While mycobacterial infection is linked pathogenetically to a Tcell defect (10), Staphylococcus aureus infections are more likely to occur as a result of neutropenia or possibly B-cell dysfunction. This study reports on the relationship between nasopharyngeal carriage of Staphylococcus aureus in HIV-infected patients and their subsequent risk of infection, Staphylococcus aureus infection in HIV-infected patients with their inherent immunosuppression is always a lifethreatening event, leading to a high rate of morbidity and mortality (1, 2). Thus, the detection of the Staphylococcus aureus carrier state in patients with indwelling catheters and neutrophil counts < 4000/lal helps identify subjects at risk of suffering subsequent or concurrent infection. Our data show that especially AIDS patients are at risk of acquiring septicemia and should be the main group in whom colonization surveillance cultures are performed. Concomitant antibiotic therapy with for instance rifampicin can modify colonization and possibly lower the incidence of septicemia. A similar relationship between colonization and subsequent morbidity was demonstrated in patients undergoing hemodialysis or CAPD (4-6, 11), and it was shown by phage typing and antibiotic sensitivity pattern that strains from the anterior nares and a strain causing infection were similar (4). In our study the antibiotic susceptibility pattern was used as an additional phenotypic marker for identification of isolates. Although stringent evidence of the identity of strains of blood or nasal origin was not obtained by phage typing, restriction enzyme analysis or DNA hybridization, like other authors we are

Eur. J. Cli.n. Microbiol. Infect. Dis,

also of the opinion that the nasopharynx is the source of organisms which subsequently colonize and cause infection at other body sites (6). In all eight cases of septicemia an i.v. catheter, a skin lesion or i.v. drug abuse was present prior to infection, suggesting that these risk factors in combination with the staphylococcal carrier state made the infection possible. Similar arguments implicating such predisposing risk factors were raised by Jacobson et al. (1). It is widely accepted that the predominant reservoir of Staphylococcus aureus in nature is the human being (12). About 10-15 % of normal, non-hospital-associated individuals carry Staphylococcus aureus on the mucous membranes of their noses and throats (12,13). Various studies have shown that drug addicts, diabetics on insulin therapy and patients treated with hemodialysis or CAPD all have an increased rate of nasal carriage of Staphylococcus aureus (4-6, 13, 14). Furthermore, colonization is associated with skin conditions characterized by small lesions such as dermatoses, particularly atopic dermatitis, which are heavily colonized. It has been postulated that the sequence of events starts with nasal carriage, followed by spread to other body sites and dermal breaks caused for instance by wounds or catheters (12). From these colonized sites staphylococci gain access to the blood stream and occasionally septicemia ensues. Nasopharyngeal colonization with Staphylococcus aureus is probably very often acquired in hospital. This assumption may be supported by our data which indicate that AIDS patients (with frequent hospitalization) had a higher carriage rate than HIV-positive subjects (without previous hospitalization) and by the fact that 16.7 % of the carriers acquired staphylococci in hospital. Ganesh et al. (7) demonstrated that carriage of Staphylococcus aureus was increased in patients with asymptomatic HIV-infection to 49 %. In this study we found a colonization rate of 37.5 % in non-advanced HIV-infection and provide more information on patients with full-blown AIDS in whom a colonization rate of 58.8 % was observed. The low CD4 count in AIDS patients is not a factor per se favoring staphylococcal colonization, which is more likely to be linked to frequent contact with the hospital environment. The results of our study suggest that staphylococcal colonization, the presence of indwelling intravascular catheters and granulocytopenia represent the major risk factors for staphylococcal septicemia.

Vol. 11, 1992

The high frequency of colonization with the risk of subsequent septicemia raises the question as to how colonization might be prevented. In hemodialysis patients five-day systemic treatment with rifampicin proved to be efficacious in reducing colonization (6). HIV-infected patients are usually given a large number of drugs for prophylaxis and treatment of opportunistic infections so that administration of another systemic drug such as rifampicin is not justified in this patient group. Studies of topical agents such as penicillin, vanCOmycin, bacitracin or gentamicin showed that these agents had only short-term effects or promoted emergence of resistance to antibiotics (15, 16). Recent reports indicate that topical rnupirocin is effective in reducing Staphylococcus aureus colonization and infection (17-19). As mupirocin is well tolerated, further studies should be undertaken to test the efficacy of mupirocin in HIV-infected subjects, especially AIDS patients with risk factors for staphylococcal infection, with the aim of reducing the rate of Staphylococcus aureus septicemia. Furthermore, since most of Our septicemia cases were catheter associated hygiene measures relating to indwelling catheters should have high priority in the prevention of Staphylococcus aureus infection.

References 1. Jacobson MA, Gellermann H, Chambers H: Staphylococcus aureus bacteremia and recurrent staphylococcal infection in patients with acquired inununodeficiency syndrome and AIDS-related complex. American Journal of Medicine 1988, 85: 172-176. 2. Witt D J, Craven DE, McCabe WR: Bacterial infections in adult patients with the acquired immune deficiency syndrome (AIDS) and AIDS-related complex. American Journal of Medicine 1987, 82: 900-906. 3. Whimbey E, Gold JWM, Poisky B, Dryjanskl J, Hawkins C, Blevins A, Brannon P, Kiehn TE, Brown AE, Armstrong D: Bacteremia and fungemia in patients with the acquired immunodeficlency syndrome. Annals of Internal Medicine I986, 104: 511-514. 4. Luzar MA, Coles GA, Failer B, Slingeneyer A, Dab GD, Briat C, Wone C, Knefati Y, Kessler M, Peluso F: Staphylococcus aureus nasal carriage and infection in patients on continuous ambulatory peritoneal dialysis. New England Journal of Medicine 1990, 322: 505-509. 5. I)avies SJ, Ogg CS, Cameron JS, Poston S, Noble WC: Staphylococcus aureus nasal carriage, exit-site infection and catheter loss in patients treated with continuous ambulatory peritoneal dialysis, Peritoneal Dialysis International 1989, 9: 61--64.

989

6. Yu VL, Goetz A, Wagener M, Smith PB, Ribs JD, Hanchett J, Zuravleff J J: Staphylococcus aureus nasal carriage and infection in patients on hemodialysis. New England Journal of Medicine 1986, 315: 91-96. 7. Ganesh R, Castle D, McGibbon D, Phillips I, Bradbeer C: Staphylococcal carriage and HIV infection. Lancet 1989, ii: 558. 8. Lane HC, Masur H, Edgar LC, Whalen G, ]Rook AH, Fauci AS: Abnormalities of B-cell activation and immunoregulation in patients with the acquired immunodeficiency syndrome. New England Journal of Medicine 1983, 309: 453-458. 9. Ellis M, Gupta S, Galant S, Hakim S, VandeVen C, Toy C, Cairo MS: Impaired neutrophil function in patients with AIDS or AIDS-related complex: a comprehensive evaluation. Journal of Infectious Diseases 1988, 158: 1268-1276. 10. Horsburgh CR: Mycobacterium avium complex infection in the acquired immunodeficiency syndrome. New England Journal of Medicine 1991, 324: 1332-1338. 11. Pignatari A, Pfaller M, Hollis R, Sesso R, Leme i, Herwaldt L: Staphylococcus aureus colonization and infection in patients on continuous ambulatory peritoneal dialysis. Journal of Clinical Microbiology 1990, 28: 1898-1902. 12. Sheagren JN: Staphylococcus aureus. The persistent pathogen. New England Journal of Medicine 1984, 310: 1368-1373, 13. Tuazon CU, Perez A, Kishaba T, Sheagren JN: Staphylococcus aureus among insulin-injecting diabetic patients. An increased carrier rate. Journal of the American Medical Association 1975, 231: 1272. 14. Chow JW, Yu VL: Staphylococcus aureus nasal carriage in hemodialysis patients. Its role in infection and approaches to prophylaxis. Archives of Internal Medicine 1989, 149: 1258--1262. 15. Bryan CS, Wilson RS, Meade P, Sill LG: Topical antibiotic ointments for staphylococcal nasal carriers: survey of current practices and comparison of bacitracin and vancomycin ointments. Infection Control 1980, 1: 153-156. 16. Williams JD, Waltho CA, Ayliffe GA, Lowbury EJ: Trials of five antibacterial creams in the control of nasal carriage of Staphylococcus aureus. Antimicrobial Agents and Chemotherapy 1975, 7: 294--297. 17. Reagan DR, Doebbeling BN, Pfaller MA, Sheetz CT, Houston AK, Hollis RJ, Wenzel RP: Elimination of coincident Staphylococcus aureus nasal and hand carriage with intranasal application of mupirocin calcium ointment. Annals of Internal Medicine 1991, 114: 101106. 18. Boelaert JR, DeSmedt RA, DeBaere YA, Godard CA, Matthys EG, Schurgers ML, Daneels RF, Gordts BZ, VanLanduyt HW: The influence of calcium mupirocin nasal ointment on the incidence of Staphylococcus aureus infections in haemodialysis patients. Neplu'ology, Dialysis, Transplantation 1989, 4: 278-281. 19. Hill RLR, Duckworth G J, Casewell MW: Elimination of nasal carriage of methicillin-resistant Staphylococcus aureus with mupirocin during a hospital outbreak. Journal of Antimicrobial Chemotherapy 1988, 22: 377384.