A case of delayed diagnosis in persistent Staphylococcus aureus bacteremia illustrating the importance of empiric imaging

Abstract

We present a case of persistent Staphylococcus aureus bacteremia (SAB) with delayed diagnosis of deep-seated asymptomatic abscesses leading to an unnecessarily prolonged hospital stay and significant morbidity. This case suggests that early empiric nuclear imaging might have led to earlier source control and, in turn, reduced morbidity. In addition, the case illustrates that even with a presumed superficial source of infection, one should still search thoroughly for additional and perhaps clinically occult secondary foci in patients with high-risk SAB.

Case Presentation

A 40-year old male with poorly controlled type 2 diabetes (HbA1c 14%) presented to the emergency department in diabetic ketoacidosis in the context of a recent left foot abscess. His medical history was significant for type 2 diabetes with medication noncompliance. No past or present injection drug use was reported, and no prosthetic material was present. He had suffered a mild traumatic injury to his foot 1 month prior to presentation but described only 1 day of symptoms of foot pain and swelling. On physical exam, he was febrile at 38.1°C with other vital signs within normal limits. Over the dorsum of his left foot was observed a small 2 × 2 cm area of erythema, swelling, and warmth that was tender to the touch. Physical examination was otherwise unremarkable. Initial blood cultures obtained shortly after presentation grew Staphylococcus aureus after 7 hours of incubation in all three vials. Both transthoracic and transesophageal echocardiogram were negative for endocarditis.

Within 72 hours, drug susceptibility testing confirmed infection with a methicillin-susceptible Staphylococcus aureus (MSSA) and empirically started vancomycin was changed to cefazolin 2 g dosed every 8 hours. Blood cultures were persistently positive over the first 9 days of admission. Magnetic resonance imaging (MRI) of the foot confirmed a multi-loculated subcutaneous abscess. This lesion appeared to increase in size despite antibiotic therapy and attempted percutaneous drainage by radiology. Surgical drainage of the abscess was eventually performed on day 8 of the admission. Following this, the patient defervesced and blood cultures promptly cleared. It is important to note that, except for foot pain, he was at this point otherwise asymptomatic, without additional localizing signs or symptoms on repeated evaluation.

On day 10 the patient began to develop right thigh pain and right leg weakness, which later progressed to bilateral weakness as well as urinary retention. An urgent spine MRI was performed, which revealed no apparent spinal column lesions but instead incidentally found a pelvic mass that could not be well characterized without further imaging. Subsequent computed tomography (CT) scan of the abdomen and pelvis revealed multiple abscesses: deep pelvic, peri-urethral, perianal, right thigh, and renal abscesses, as well as septic emboli in the lung bases. Around this time the patient also developed neck swelling with subsequent CT of the neck revealing a massive 18 × 9 × 12 cm abscess apparently arising from the paraspinal soft tissues (see Figure 1). All drainable collections were evacuated and grew MSSA. Given the apparent clinical failure of cefazolin and concern for cefazolin inoculum effect, antibiotic therapy was changed to intravenous cloxacillin, 2 g dosed intermittently every 4 hours (1).

Figure 1:

CT neck axial slices show an 18 × 9 × 12 cm abscess that drained over 600 mL of frank pus

One month into therapy, CT imaging documented near complete resolution of the pelvic abscesses and the previously identified pulmonary septic emboli. However, somewhat surprisingly the follow-up imaging studies revealed a new 5 × 10 cm abscess located deep to the scapula overlying the posterior chest wall. This lesion was also completely asymptomatic. Percutaneous drainage was arranged and the intravenous cloxacillin course extended to a total of 3 months. Follow-up imaging at 3 months revealed complete resolution of the intra-abdominal and thoracic abscesses. Although the paraspinal abscess had substantially improved by this time, it had not completely resolved, and ultimately an additional 3-month course of oral cloxacillin consolidation therapy was provided until a complete radiographic resolution could be documented. Table 1 summarizes the findings. The patient suffered reduced mobility requiring rehabilitation, and as a result had a total inpatient stay of 2.5 months.

Table 1:

Summary of infectious foci

Focus of infection	Date identified	Date of documented resolution
L foot SC abscess	Day 1	Day 10
Deep pelvic abscesses (peri-urethral, deep pelvis abutting bladder, perianal)	Day 13	Day 30
R thigh SC abscess	Day 13	Day 30
R renal abscess	Day 13	Day 30
Septic emboli in lungs	Day 13	Day 30
Chest wall abscess	Day 30	3 months
Paraspinal abscess	Day 13	6 months

L = left; R = right; SC = subcutaneous

Discussion

There is extensive literature on metastatic Staphylococcus aureus infection and the clinical predictors of complicated bacteremia (2–4). High-risk SAB is often associated with some or all of the following factors: persistently positive blood cultures 48 to 96 hours after initial result, persistent fevers 72 hours into effective therapy, community-acquired onset of bacteremia, time to detection in automated blood cultures of less than 12 hours, and the presence of skin findings suggestive of disseminated infection (2,4). The most well-established of these factors is persistently positive blood cultures, which is strongly associated with complications such as vertebral osteomyelitis, endocarditis, and deep-seated abscesses (2). Despite knowledge of these predictors, many cohort studies highlight the difficulty in detecting deep infectious foci at the time of first positive blood culture in a given patient in part due to clinically occult disseminated disease (2,5–7). Indeed, some cohorts estimate between one-third and one-half of patients with persistent Staphylococcus aureus bacteremia (SAB) will have asymptomatic metastatic foci of infection (8,9). Thus, empiric imaging may be useful in the management of complicated SAB, especially when blood cultures remain persistently positive. Berrevoets et al found a reduction in 3-month mortality with early positron emission tomography/computed tomography (PET/CT) imaging on patients with complicated SAB in their retrospective analysis (10). In their study, the results of the PET scan altered treatment in 75% of cases, usually by demonstrating a focus requiring operative drainage or a prolonged course of antibiotics. The utility of various nuclear medicine studies to detect different types of deep infectious processes has been reviewed (11).

In our case, we believe that the lack of early imaging delayed discovery of the drainable collections, which complicated and prolonged the antibiotic treatment course. For instance, our patient’s chest wall abscess was only discovered upon follow-up CT imaging of the chest 1 month into treatment. Multiple deep pelvic and renal collections were also only incidentally found on imaging intended to investigate the vertebral column. A whole-body PET/CT or white blood cell (WBC) scan performed early in the presentation, after recognition of persistent SAB, would have identified these asymptomatic deep infectious foci sooner and thus may have allowed for prompt drainage and perhaps a shortened antibiotic course.

In our case, the presence of a superficial subcutaneous abscess on the foot may have led to a reluctance to search further for other occult sources of infection, under the assumption that the source of bacteremia had already been identified. However, in Berrevoets’ analysis of high-risk SAB patients already committed to 6 weeks of antibiotics due to a documented source, 50% had additional foci (10). This case highlights the prudence of early empiric imaging in high-risk SAB patients even in the presence of a known source, due to the high likelihood of clinically silent disease.

Finally, we acknowledge the potential financial and logistical implications of empiric imaging for all patients with complicated SAB and concede that cost-effectiveness has not been demonstrated. However, for patients with SAB and several clinical predictors of complicated infection, in particular, persistent bacteremia despite effective therapy, an enhanced search for disseminated disease using nuclear medicine imaging might be indicated. A cost-effectiveness analysis of such an approach may be a worthwhile area of future study.

Conclusion

We have presented a case of persistent Staphylococcus aureus bacteremia that may have benefitted from earlier recognition of deep clinically silent infectious foci. In our case, both morbidity and duration of treatment were affected, perhaps due to the presence of a known source which deterred further investigation. We hope this serves as an example of the harmful consequences of delayed diagnosis and increases vigilance in searching for clinically silent disease in patients with high-risk SAB through serial exams and/or empiric imaging.

Competing Interests:

The authors have nothing to disclose.

Ethics Approval:

N/A

Informed Consent:

N/A

Registry and the Registration No. of the Study/Trial:

N/A

Animal Studies:

N/A

Funding:

No funding was received for this work.

Peer Review:

This article has been peer reviewed.