Often, trauma and inflammatory diseases affect the acromioclavicular (ACV) joint. However, septic arthritis is rarely seen in this localization. In Anglo-Saxon literature reports only 15 cases described to date in detail (Table 1).1–11 In most cases it occurs in patients with risk factors such as immunosuppression. Diagnosis is based on clinical data and isolation of the etiologic microorganism in1 microbiological cultures, although imaging tests, such as magnetic resonance, ultrasound or scintigraphy, may contribute to the demonstration of arthritis. Here, we describe the clinical, laboratory, imaging and therapeutic data in a series of 6 patients with septic arthritis of the ACV joint, diagnosed in a university hospital center.

The study was a retrospective review (1989–2012) of the medical records of patients with septic arthritis found in the database of diagnostic coding of the Rheumatology department of a university hospital, with a reference population of 800000. The Rheumatology department takes care of the diagnosis and treatment of the vast majority of non-prosthetic septic arthritis. Thus, in this period, 101 patients were diagnosed with pyogenic septic arthritis, 6 of whom (6.1%) had ACV joint involvement. All patients had a microbiological germ isolation in joint fluid obtained by arthrocentesis of the ACV joint. The following variables were collected: epidemiological, clinical manifestations, risk factors, imaging, treatment received and laboratory tests such as complete blood count (Coulter), erythrocyte sedimentation rate, C reactive protein (CRP by nephelometry), biochemistry, blood and joint fluid culture.

In this period 101 patients with pyogenic septic arthritis were diagnosed at the Rheumatology department. Six patients had ACV septic arthritis (6%): five male and one female, mean age at diagnosis of 51.2 years (range 46–73). All patients had some risk factor and the most common was chronic renal failure, present in 4 patients (66.7%). In all cases there was a site of entry and in 2 patients it was an intravascular device (33.3%). One patient had a history of gout and ankylosing spondylitis, while the rest did not have a prior arthropathy (Table 2). The median time from onset of symptoms to diagnosis was 6 days (range 2–10).

All patients had swelling and localized pain in the affected joint and pain on ACV abduction and flexion of the shoulder in extreme movements (Fig. 1). Four patients had fever as the initial manifestation, all showing leukocytosis and elevated acute phase reactants: the average C-reactive protein was 105±22U/L and an erythrocyte sedimentation rate of 85±19mm/h. Arthritis was monoarticular in 3 patients and polyarticular in 3. Bilateral involvement occurred in 2 cases.

Fig. 1.

Photograph of a patient suffering from septic arthritis of the ACV joint. There is a certain swelling and cutaneous erythema in the upper left shoulder.

(0.1MB).

To confirm the clinical diagnosis of ACV arthritis, ultrasonography was performed in 3 cases, finding abscessed collections in soft tissues around the ACV joint. In 3 cases, scintigraphy was especially useful in patients with polyarticular affection (Fig. 2). We were unable to perform imaging in one of the patients due to clinical instability. Plain radiographs showed no erosions in any of the cases. MRI was not performed in any of them.

Fig. 2.

Tc99 scan which revealed both acromioclavicular joints increased uptake in a patient with septic polyarthritis.

(0.1MB).

ACV joint arthrocentesis was performed in all cases, with isolation of the causal pathogen through a joint fluid culture. The most frequently isolated organism was Staphylococcus aureus (S. aureus) (4 patients, 6.7%). In the other 2 cases the isolated organism was a streptococcus (one Streptococcus pneumoniae, one Streptococcus agalactiae). All patients had positive blood cultures due to hematogenous spread.

All patients received intravenous antibiotic treatment for at least 3 weeks and oral antibiotics to complete a minimum of six weeks, with the exception of patients who died early. Patients initially received empirical antibiotic therapy, which subsequently was adapted to the result of the antibiogram. Antibiotic treatment is detailed in Table 2.

Articular lavage were performed in 2 patients and another required surgical drainage of the ACV joint. Another patient, the most severely immunosuppressed, died 9 days after the diagnosis of septic shock, despite having monoarticular involvement. Another patient died nine days from hypovolemic shock caused by a hematoma of rectus abdominis sheath, in the context of anticoagulation for acute myocardial infarction. The cases with polyarticular disease presented a more torpid evolution, but eventually recovered completely. There was evidence of greater diagnostic delay in monoarticular cases.

Bacterial infectious arthritis is monoarticular in 85% of cases and most commonly affects the knee. In the absence of trauma or joint manipulation, hematogenous spread is the most common cause. For this reason, the large joints, which are more vascularized, have more risk of septic arthritis.12

Septic arthritis of the ACV joint is a rare entity, although rapidly destructive. It usually occurs in immunosuppressed patients or those with a discontinuity of defense barriers, but even in this group of patients it is unusual. However, the diagnosis cannot be excluded in healthy patients.1 Its low incidence explains the limited number of patients seen in a single hospital over a period of 22 years. For this reason, the largest global series includes 4 patients and we have found a total of 15 described in the literature, most of them reported as isolated cases.1–11 We did not include cases in which there was a germ isolated from blood cultures or cultures of joint fluid. Shoulder arthritis in large series of septic arthritis is common, around 5%–10%. These sets differ sternoclavicular arthritis, but ACV arthritis is recognized in many cases along with the glenohumeral joint, so its true incidence cannot be known.12–21 The diagnostic coding system used in our hospital allowed us to differentiate glenohumeral septic arthritis to that located on the ACV joint, so we were able to justify the 6% incidence described in the present series. Possibly, in many published series of septic arthritis, ACV infectious arthritis is included in those located on the shoulder.

The demographic characteristics of our series are similar to those described by Bossert et al.1 The higher percentage of males is noteworthy. The mean age at diagnosis was also 50–60 years.

Typically, its clinical presentation is characterized by swelling and pain over the joint, but in many cases the shoulder range of motion is preserved, although it may be painful in extreme movements. Thorough clinical inspection is essential for diagnosis. Its course is acute and early diagnosis occurs often. They are serious infections and hematogenous dissemination and polyarticular involvement are poor prognostic factors. We should not forget this location in the presence of septic arthritis from other locations, as it probably often goes unnoticed in the course of septic arthritis in more striking locations. The polytropic incidence of septic arthritis in our series was 29%, and yet, in the case of ACV, half of them had more than one affected joint.22

Arthrocentesis should be performed early in the disease and lead to a definitive diagnosis. It is essential to perform blood cultures and joint fluid culture. S. aureus is the most frequently isolated germ, followed by various types of streptococci. These results agree with those of other published cases and the studies of septic arthritis of other joints.1–11 There is a high percentage of positive blood cultures, indicating that there is usually hematogenous spread. Notably, in this series, it was present in all cases. Therefore, it is important to interrogate the patient about a possible gateway and consider the concomitant existence of infective endocarditis. It is noteworthy that in this series there were no patients addicted to intravenous drugs, but our hospital is a referral center for patients suffering from human immunodeficiency virus infection.

Antibiotic treatment is the same as in septic arthritis in another location and it is recommended that it initially be given intravenously. It is advised that the duration of antibiotic last not less than 4 weeks. Sometimes medical treatment by itself may be ineffective and it may be necessary to conduct joint lavages or surgical drainage.12–21 Regarding imaging tests, simple X rays must be carried out but changes can be observed late, such as erosions. Ultrasound and MRI allow earlier diagnosis and assess the local extension of the infection. Ultrasound will show an increase in joint space with a loosening of the capsule and it will likewise allow us to make guided punctures to obtain joint fluid. Magnetic resonance has greater definition of tissue, it detects earlier damage, is sensitive and specific, and is not operator-dependent. In patients with polyarticular or paucisymptomatic, scintigraphy may help with the diagnosis.23–25 The main limitations of this study include its retrospective nature and the limited number of patients, as well as the low incidence of this disease.

In conclusion, septic arthritis of the ACV joint is a rare entity, although it may be underdiagnosed. It generally affects immunosuppressed individuals, but cannot be excluded in healthy patients. Since this is a small joint, there is a high risk of fast joint destruction, so early diagnosis is crucial. Although arthrocentesis remains the diagnostic test of choice, the procedure can be difficult due to the characteristics of the joint. Therefore, imaging tests such as ultrasound or magnetic resonance imaging may be useful. Hematogenous spread is practically constant, as may be the existence of a entry gateway. Polyarticular involvement and immunosuppression are poor prognostic factors. Antibiotic treatment should be started early.

The authors declare no conflicts of interest.