Portal de Periódicos da CAPES

Etiology of Osteomyelitis Complicating Sickle Cell Disease

1998; American Academy of Pediatrics; Volume: 101; Issue: 2 Linguagem: Inglês

10.1542/peds.101.2.296

1098-4275

Mark W Burnett, James W. Bass, Bruce A. Cook,

Hematological disorders and diagnostics

The most common cause of osteomyelitis isStaphylococcus aureus.12 However, it is well-known that in children with sickle cell disease (SCD) who develop osteomyelitis, infection is often attributable toSalmonella. A controversy that arises is whether S aureus is the most common cause of infection in these children overall, or does Salmonella actually predominate. The authors of the chapter on osteomyelitis and septic arthritis in the current edition of a major pediatric text state that "Salmonella osteomyelitis tends to occur more often in children with hemoglobinopathies, although even in this group, S aureus remains the predominant pathogen."1 The authors of the chapter on osteomyelitis and septic arthritis in the current edition of a major pediatric infectious diseases text state that "seventy percent of all lesions or blood cultures in children with hemoglobinopathies and presumed osteomyelitis yieldSalmonella microorganisms, 10% contain S aureus, and aerobic Gram-negative rods are isolated in 7%."2 To support this statement, a single 1981 reference is cited that summarizes the results of nine reports previous to this time.3 No references on this subject subsequent to this time are listed although some studies published since 1981 have foundS aureus to be the most common cause of osteomyelitis in children with SCD.4-6 Apparently, the controversy persists. We sought to resolve this issue in a comprehensive review, the results of which are the subject of this report.A review of the world's literature cross-referencing osteomyelitis and SCD was conducted using Medline and a bibliographic search back to 1959. Articles were included only if they listed the total cases in the series, with a breakdown into groups according to etiology. Case reports and other studies that did not provide this information were not included in this review.Nine reports of series worldwide dating from 1975 to 1996 that met our criteria for review were found in the search.3-11Results of the analysis of these reports are shown in the Table. The report by Givner et al3 reviews all series reported before 1981. Of the nine series reviewed by Givner, eight were from medical facilities in the United States, and one was from Nigeria. Of 68 patients in the Givner review, 50 cases were attributable to Salmonella and 7 cases were attributable to S aureus, a ratio of 7:1. Excluding the Givner review, eight reports were after 1981. They total 137 patients; 62 were attributable to Salmonella and 45 to S aureus, a ratio of 1.4:1. Overall, the ratio is 2.2:1.This review shows that Salmonella species are the most common cause of osteomyelitis in patients with SCD worldwide. The relative incidence is over twice that of S aureus; however,S aureus was the second most common cause of osteomyelitis in this review. Interestingly, reports before 1981 showed a significantly higher incidence of Salmonella as causative agents in SCD patients with osteomyelitis than in subsequent reports. During these two time periods the ratio of Salmonella toS aureus changed from 7:1 to 1.4:1.It has been speculated that the overrepresentation ofSalmonella in these patients may be attributable to high prevalence of animal-associated Salmonella in the environment of patients in developing countries, where many of these reports originate. This does not appear to be the answer as the relative incidence of Salmonella osteomyelitis compared withS aureus is higher in the United States than in the developing countries. When only those reports from the United States are analyzed, there are 80 total patients; 56 (70%) were attributable to Salmonella and 11 (14%) to S aureus, a ratio of 5.1:1.The major organisms that made up the "other organisms" listed were mostly Gram-negative bacilli, 37 of 41 being Gram-negative enteric bacilli. These organisms ranked almost equal to S aureus in this analysis. When these organisms are grouped withSalmonella, the ratio of Gram-negative enteric bacilli toS aureus increases to 2.9:1. A hypothesis that might explain these observations is that intravascular sickling of the bowel probably occurs, as it does in many tissues of patients with SCD. This may lead to episodes of patchy ischemic infarction and superficial devitalization of the bowel, permitting transient mucosal barrier breakdown. This would predispose to repeated episodes of bacteremia in these patients, who may have clinical but undiagnosedSalmonella enteritis, as well as those who may have subclinical Salmonella enteric infections. This may also explain why these patients have such a high incidence of osteomyelitis attributable to other Gram-negative enteric bacilli. In a recent review of 135 normal children with osteomyelitis, 75 (55%) had positive cultures from blood or bone; only 2 (2.6%) of these 75 were Gram-negative enteric bacilli.12 In contrast,Salmonella plus other Gram-negative enteric bacilli were responsible for infection in 149 (73%) of our SCD patients with osteomyelitis.We conclude that Salmonella is the most common cause of osteomyelitis in patients with SCD, both in developing countries and in developed countries, and that its relative incidence is more than twice that of S aureus. We also conclude that other Gram-negative enteric bacilli are unusually common causes of osteomyelitis in SCD patients, and we offer a plausible explanation for this occurrence.