Glenohumeral septic arthritis in intravenous drug users: 10-year retrospective analysis

INTRODUCTION

Septic arthritis is defined as inflammation of a joint due to infection. This can be caused by bacteria, viruses, or other kinds of pathogens.^[1] Septic arthritis can be a serious cause of morbidity and mortality, as delayed or inadequate treatment can lead to irreversible joint destruction or even death in 11% of patients. The case fatality rate is even higher in patients who have more than one joint affected by septic arthritis, reaching as high as 50%.^[2]

Septic arthritis is more common in elderly people or very young children. Any previous condition that damaged the joint also predisposes one to the development of infection.^[1] A prospective study of 7000 patients who were followed up for 3 years showed that risk factors for septic arthritis included being older than 80 years, diabetes, rheumatoid arthritis, recent joint surgery, immunosuppression, and skin infection.^[1,3] Another rapidly emerging risk factor for the development of septic arthritis is intravenous drug use.^[1,4]

Staphylococcus aureus is the most common causative organism of septic arthritis, followed by other Gram-positive cocci, including streptococci.^[2] Methicillin-resistant S. aureus (MRSA) infection is increasing, especially in patients with intravenous drug use,^[4] elderly and in patients with orthopedic procedure-related infection.^[5]

Patients with septic arthritis often present with pain, fever, and rigors.^[1]

The easiest way to diagnose septic arthritis is when bacteria are isolated from synovial fluid. However, in several patients with no bacteria isolated from the joint, bacteria were detected in blood culture in 11% and from other sources in 7%.^[1]

The role of imaging in the diagnosis and management of septic arthritis is not well defined. However, magnetic resonance imaging (MRI) can be very specific in identifying septic arthritis and osteomyelitis.^[6] Bone marrow changes followed by femoral head changes were found to be the most specific finding for hip septic arthritis, while synovial enhancement was found to be the most sensitive finding in a meta-analysis in children.^[7]

Treatment includes hospital admission, intravenous antibiotics, and joint aspiration, with refractory cases undergoing re-culture, consideration of alternate infectious foci, and further imaging to evaluate for osteomyelitis.^[1] It is important to remove the intra-articular pus from the joint, and for this, both arthroscopy and needle aspiration have similar outcomes.^[8,9]

The long-term mortality was closely similar in patients with synovial fluid culture-positive septic arthritis and synovial fluid culture-negative septic arthritis.^[1]

In our study, we are specifically looking at the cases of septic arthritis of the glenohumeral joint and assessing the association with intravenous drug use.

MATERIAL AND METHODS

RESULTS

Fifty-three shoulders with the diagnosis of glenohumeral septic arthritis on 50 patients met the inclusion criteria for our study [Table 1]. Out of these, 52 were native glenohumeral joints, and 1 was glenohumeral joint arthroplasty.

Demographics

37.7% (20/53) of cases reported recent intravenous drug use.

62.3% (33/53) of cases had no reported intravenous drug use. Out of the non-intravenous drug users, 14 cases report diabetes mellitus, 3 patients had known cancer, 2 were post recent shoulder surgeries, 1 had recent cat bite injury, 1 had a known infected central line, 1 had infected hardware, 1 had exposure to farm animals, and 1 had known HIV infection. 2 of the 14 cases with diabetes mellitus also had end-stage renal disease. 9 out of 53 cases had no known predisposing comorbidity.

There was a statistically significant difference in mean age between groups (ANOVA, p < 0.001). The mean age for intravenous drug users is 39.9 years, whereas the mean age for non-intravenous drug users is 60.82 years (p < 0.001) [Graph 1].

Close

Graph 1: Graph showing age ranges in intravenous drug users with glenohumeral septic arthritis versus non-users with glenohumeral septic arthritis.

Export to PPT

The overall male-to-female ratio is 1.5:1. The male-to-female ratio in intravenous drug users is 1.9:1, whereas it is 1.4:1 in non-intravenous drug users (p = 0.773).

In cases associated with intravenous drug use, the right shoulder was found to be 2.3 times more frequently infected 70% (14/20) than the left shoulder 30% (6/20) (p = 0.386). In non-intravenous drug users, both shoulders were equally affected.

70% (37/53) of all glenohumeral septic arthritis cases were acute. 15% (3/20) of cases associated with intravenous drug use were chronic, whereas 36% (12/33) of infections in non-intravenous drug users were chronic (p = 0.114).

DISCUSSION

Septic arthritis of the glenohumeral joint is rare. A study of 11 cases of glenohumeral septic arthritis from Israel reports that while systemic disorders such as liver disease, alcoholism, and malignancies were more commonly found as risk factors, intravenous drug use was a very insignificant risk factor, only found in <5% of patients.^[10] A review study including 97 patients with glenohumeral septic arthritis reports that around 27.8% of patients had intravenous drug use.^[11] In our study, we see a much higher percentage, with 37.7% of our glenohumeral septic arthritis patients having reported recent intravenous drug use.

Other studies have reported a mean age of 76 years in glenohumeral septic arthritis,^[10] but our study finds the overall mean age to be 52.9 years. More importantly, we find a much lower mean age of 39.9 years in cases associated with intravenous drug use with glenohumeral septic arthritis, compared to a mean age of 60.8 years in non-users. Hence, we conclude that younger patients are more affected by intravenous drug use-associated glenohumeral septic arthritis.

A slight male predilection of 1.5:1 has been reported in the past for glenohumeral septic arthritis,^[10] which is supported by our data, as we also find a male-to-female ratio of 1.5:1 in our whole cohort by using a Fisher’s exact test. However, in our intravenous drug use patients, we see a jump in the ratio with a male-to-female ratio of 1.9:1. In our non-user patients, we see a male-to-female ratio with a value of 1.4:1. The takeaway from these data is that overall, men are slightly predisposed to glenohumeral septic arthritis, with intravenous drug use worsening the odds for men.

Pain was the most common presenting symptom. Other presenting symptoms include sepsis, swelling, fever, and drainage. Other studies have reported that low-grade fever is common in septic arthritis with rigors and a warm, swollen, tender joint.^[12,13] Something to keep in mind while evaluating septic arthritis is that joint pain is often blunted in the immunosuppressed patient.^[12] We conclude that while the classic presentation of red, hot, tender joint may suffice to diagnose the average patient, we must be extra vigilant in our immunosuppressed patients.

70% of our total cases had bacteremia. It was interesting to note that 95% of our intravenous drug use-associated cohort had bacteremia, compared to only 54.5% of non-users. MSSA bacteremia and MRSA account for 54% of bacteremia in our overall cohort, which is in keeping with previous literature. Staphylococcus bacteremia can cause metastatic infection in more than a third of cases and is associated with a 15–30% case fatality rate.^[14]

The most common pathogen isolated from joint fluid in our study is also S. aureus, with MSSA and MRSA being the most common. We note that a joint pathogen was only isolated in 38% of cases, which is in keeping with prior literature, where only 40% had a positive synovial fluid culture.^[15] Joint infection with MRSA has been shown to have a worse clinical outcome.^[12]

Only 11 of our total 53 cases grew the same organism on both blood culture and joint fluid culture, showing 20% of the blood culture results were identical to joint culture results. This contrasts with Daynes et al., 2016, where they found 91% of the blood culture results were identical to joint fluid culture results.^[16]

Imaging plays an important role in the diagnosis of septic arthritis, as well as providing guidance for diagnostic/therapeutic aspiration.^[17] However, the role of imaging is supportive rather than diagnostic, and must be interpreted alongside clinical, laboratory, and microbiological data^. Imaging is very important to find complications such as cartilage loss, abscess, embolic disease, and osteomyelitis.^[17] Plain radiographs are useful to evaluate cartilage and see if there is subchondral bone destruction depicted as subtle cortical irregularity [Figure 1]; however, these lack sensitivity.^[18] It is important to recognize that findings of septic arthritis on plain radiographs may be subtle. MRI would be the go-to choice in this case.^[19] MRI has a high specificity of 77% and sensitivity of 100% to detect septic arthritis.^[20] MRI findings in patients who have septic arthritis can start looking abnormal as early as 24 h after the start of infection.^[21]

Close

Figure 1: A 29-year-old male with recent intravenous drug use presents with septic arthritis of the glenohumeral joint. Plain radiograph showing axillary view of the right shoulder. Subtle cortical irregularity of bone at the humeral head (white arrow) is suggestive of osteomyelitis.

Export to PPT

Karchevsky et al. report on synovial enhancement; joint effusion and peri-synovial edema are commonly correlated with the clinical diagnosis of a septic joint. However, they do note that almost one-third of the patients with septic arthritis do not have a joint effusion, which may be as joint effusions are difficult to assess in small joints and often the effusion may escape into surrounding tissue if the capsule has a rent.^[22]

MRI can detect destruction of cartilage in these joints, where it is characterized by rough, fuzzy, and shedding articular cartilage with hyperintense T2 signal on both sides of the joint [Figure 2]. An important imaging consideration is that MRI is useful in evaluating the joint in septic arthritis as well as the surrounding soft tissues. One can also detect surrounding soft tissue infection, such as soft tissue abscess, which shows up as peripherally enhancing fluid collection on contrast-enhanced MR scan.^[22]

Close

Figure 2: A 29-year-old male with recent intravenous drug use presents with septic arthritis of the glenohumeral joint. Coronal T2 fat-saturated magnetic resonance image of the right shoulder demonstrating large joint effusion (white arrow); osteomyelitis of the humeral head and glenoid (white arrowheads); and intramuscular abscess in the supraspinatus muscle (black arrow). Please note that the acromioclavicular joint is also affected by an effusion and underlying osteomyelitis.

Export to PPT

MRI is also useful to evaluate osteomyelitis or bone infection.^[18,21-23] Karchevsky et al. also talk about abnormal bone marrow signal having a high association with osteomyelitis in these cases, especially if diffuse and seen on T1-weighted images.^[22]

Early and appropriate antibiotic administration is crucial in treating septic arthritis.^[24] Antibiotic courses of 3–4 weeks are usually sufficient for uncomplicated septic arthritis. Treatment may be extended to 6 weeks (about 1½ months) of antibiotic therapy if there is imaging data showing osteomyelitis.^[12]

If a patient has joint discomfort, reduced mobility or inability to bear weight, open surgery or arthroscopy for irrigation and drainage of purulent fluid is recommended. No significant differences were shown between open surgery versus arthroscopic surgery in the management of shoulder septic arthritis.^[25] Unfortunately, many authors report a high reoperation rate.^[25] Few patients get elective arthroplasty after septic shoulder arthritis.^[11]

Our study has several limitations. Being a single-center, retrospective study, the results may not be generalizable outside the Appalachian population. Our study is also limited by the number of cases available. Even though we have 53 cases of glenohumeral septic arthritis, a bigger sample size would increase the power of our study. We also acknowledge the limitation of underreporting bias or misclassification in ascertaining the history of intravenous drug use in our patients. This study also has shoulder-based analysis, with bilateral shoulders being treated as independent observations.

CONCLUSION

A younger age, the involvement of other joints and presence of other severe infections should prompt the radiologist to further inquire about possible intravenous drug use, which may not always be initially disclosed. Conversely, having a known history of intravenous drug use should prompt the radiologist to be on the lookout for other joint infections, endocarditis and septic emboli.