Translate this page into:
What is the role of leukopenia in the assessment of septic arthritis?
, John Mickleyb, Karilyn T.M. Larkinc, Amy L. Speeckaerta,
∗Corresponding author: Amy L. Speeckaert. amy.speeckaert@osumc.edu
-
Received: ,
Accepted: ,
This article was originally published by Reed Elsevier India Pvt. Ltd. and was migrated to Scientific Scholar after the change of Publisher.
Abstract
Abstract
It is not well-understood how leukopenia affects the synovial white blood cell (WBC) and percent neutrophils (%PMNs) in the setting of septic arthritis. We sought to determine 1. Do synovial WBC and %PMNs differ between patients with culture positive septic arthritis with or without leukopenia? And 2. Are traditional thresholds of synovial fluid studies for accurately diagnosing septic arthritis still applicable in the leukopenic patient population?
A retrospective cohort study was performed at a single institution of 79 non-leukopenic and 11 leukopenic patients diagnosed with culture-positive septic arthritis. Demographic data, serum laboratory values, synovial laboratory values, and culture results were recorded. Significant differences in synovial laboratory values were evaluated using the Wilcoxon-Mann-Whitney test. Results are reported as median, interquartile range, and p values.
There was a significant difference in synovial WBC in leukopenic patients compared to non-leukopenic patients with culture positive septic arthritis (p = 0.01). No significant difference was found in the synovial %PMNs between two cohorts (p = 0.33).
Leukopenic patients with culture positive septic arthritis have significantly lower synovial WBCs compared to non-leukopenic patients. Traditional thresholds for synovial WBC are not reliable for excluding diagnosis of septic arthritis in leukopenic patients.
Keywords
Leukopenia
Neutropenia
Septic arthritis
Synovial white blood cell count
White blood cell count
WBC
CRP
PMN
ESR
1 Introduction
Septic arthritis is considered an orthopaedic emergency which can lead to dire consequences such joint destruction, systemic infection, and increased mortality.1–4 This diagnosis should be considered in any patient presenting with a painful, swollen joint with decreased range of motion.1,2,5,6 The gold standard for diagnosis remains bacterial growth from synovial fluid cultures, but these cultures can take up to 48 h. Therefore, many clinicians use presenting symptoms, physical exam, serum values and synovial analysis to guide initial treatment.1,2,5
Relevant serum laboratory values include serum white blood cell (WBC) count with differential, serum C-reactive protein (CRP), and serum erythrocyte sedimentation rate (ESR), which are typically elevated in the setting of acute septic arthritis.1,2,6–8 However, these studies alone are not diagnostic and arthrocentesis is performed to obtain synovial fluid for analysis. Laboratory studies for synovial fluid include synovial WBC count, percentage polymorphonuclear cells (%PMNs), gram stain, and culture.1,9–12 Various sensitivities and specificities have been demonstrated for these values, however WBC>50,000 cells/mm3 and %PMNs>90% are the traditionally accepted thresholds for diagnosing septic arthritis.1,2,6–14 There has been very limited research assessing if these threshold values can be utilized in immunocompromised or other at-risk populations.14,15 Furthermore, associations between levels of circulating serum white blood cells and resulting synovial WBC count and %PMNs in septic arthritis have not been explored.
In this study, we compared synovial laboratory values in patients with culture positive septic arthritis in leukopenic and non-leukopenic patients. We hypothesized that leukopenic patients, compared to non-leukopenic patients, would present with lower synovial WBC counts and percent PMNs in the setting of culture-positive septic arthritis. Secondarily, we sought to determine if traditional thresholds for synovial WBC counts and percent PMNs accurately diagnosed septic arthritis in leukopenic patients.
2 Patients and methods
2.1 Study population
A retrospective single institution chart review was performed after Institutional Review Board approval. Patients with presumed septic arthritis between the years of 2010–2018 who had an arthrocentesis and/or arthrotomy were identified. Inclusion criteria included (1) patients 18 and older, (2) suspicion of septic arthritis and (3) confirmed microorganism culture growth from an aspiration or arthrotomy sample (i.e., culture-positive septic arthritis). Exclusion criteria included post-operative infection or arthrotomy, joints with a prior arthroplasty, and indication for arthrotomy was non-infectious pathology (joint contracture, ganglion cyst, foreign body, etc). A total of 87 patients (79 patients without leukopenia and 8 patients with leukopenia) were identified. To ensure a broad enough search was performed, we expanded our search of leukopenic patients with septic arthritis to the years 2000–2018. This led to a total of 11 leukopenic patients with septic arthritis who were confirmed to have similar cohort characteristics to patients without leukopenia and with septic arthritis.
An a priori power analysis was performed, using a threshold of 10,000 cells/mm3 WBC for the synovial fluid aspirate for the minimum clinically important difference (MCID). This MCID was selected by the authors, since there is no established value for synovial fluid WBC. By assuming a larger standard deviation for synovial fluid WBC, as in the septic arthritis subset of Butler et al. (16,143 cells/mm3),15 our study with 90 patients would have 82.8% power, with an alpha = 5%, to detect a MCID.
Clinical data collected for the selected patients were demographics, joint affected, and pertinent diagnoses for leukopenic patients. Laboratory values recorded included serum WBC count and percent PMNs, serum absolute neutrophil count (ANC), synovial WBC count and percent PMNs, and microorganisms isolated from synovial cultures. Abnormal laboratory values were determined based on the normal reference range for our institution.
2.2 Definitions of cohorts
Eligible patients were then defined as being either “not-leukopenic” or “leukopenic” based on reference values from our institution's laboratory for serum WBC count: patients with serum WBC count <3700 cells/μL were defined as leukopenic.
All patient cases of leukopenia with associated immunosuppressive states and culture-positive septic arthritis were jointly reviewed by a fellowship-trained orthopaedic surgeon and a fellowship-trained hematologist-oncologist. This was done in order to limit classification bias for the two cohorts prior to comparison.
2.3 Statistical analysis
STATA 15.0, StataCorp (College Station, TX) was utilized to perform statistical analysis. Normal interval data was described using mean and standard deviation. Non-normal interval data was described using interquartile ranges (IQR). Categorical data was described using counts and porportions. Differences between variables of interest were then statistically tested with the independent t-test (normal and interval variables), chi square test (categorical data), and Wilcox-Mann-Whitney test (non-normal interval data). The normality of synovial WBC count and serum WBC were tested using the Shapiro-Wilk test. Hence, non-parametric (Spearman) correlation was calculated between synovial WBC and serum WBC for both the non-leukopenic and leukopenic patient cohorts in order to compare the relationship between these WBC values. Scatterplots of data and best-fit trendlines were generated and results of correlation testing were given by ρ (rank correlation coefficient) and p values. The significant level, alpha, was set to 0.05.
3 Results
3.1 Distribution of affected joints
In non-leukopenic patients, the predominant joint diagnosed with septic arthritis was the knee (46.8%). The glenohumeral joint was the second most common joint (17.7%) and hip the third most common joint (15.2%). In leukopenic patients, predominant joint diagnosed with septic arthritis was also the knee (54.6%) and accounted for the majority of cases (Table 1).
| Location of Joint Undergoing Aspiration or Arthrotomy | % of Joints with Septic Arthritis (Non-leukopenic) | % of Joints with Septic Arthritis (Leukopenic) | % of Joints with Septic Arthritis (All Patients) |
| Glenohumeral Joint | 14 (17.7%) | 1 (9.1%) | 15 (16.7%) |
| Acromioclavicular Joint | 4 (5.1%) | 1 (9.1%) | 5 (5.6%) |
| Elbow Joint | 1 (1.3%) | 0 (0.0%) | 1 (1.1%) |
| Wrist Joint | 5 (6.3%) | 1 (9.1%) | 6 (6.7%) |
| Hip Joint | 12 (15.2%) | 1 (9.1%) | 13 (14.4%) |
| Knee Joint | 37 (46.8%) | 6 (54.6%) | 43 (47.8%) |
| Ankle Joint | 6 (7.6%) | 1 (9.1%) | 7 (7.8%) |
| All Joints | 79 (100%) | 11 (100%) | 90 (100%) |
3.2 Diagnoses, laboratory, and culture findings in leukopenic patients
Eleven patients met the criteria for leukopenia; the remaining 79 patients were classified as non-leukopenic. Diagnoses pertinent to the serum WBC count finding of leukopenia in these 11 patients were summarized (Table 2). The most commonly isolated microorganism from synovial culture was Methicillin-resistant Staphylococcus aureus. It should be noted that only 2 of the 11 leukopenic patients would have been diagnosed with septic arthritis using a synovial WBC count threshold of 50,000 cells/μL, conferring a sensitivity of 18.2%.
| Subject Number | AffectedJoint | Serum WBC (x103 cells/μL) | ANC (x103 cells/μL) | Synovial WBC (cells/μL) | Synovial %PMNs (cells/μL) | Diagnosis | Microorganism(s) in Culture |
| 1 | Knee | 0.8 | 0.77 | 8900 | 88 | Acute myeloid leukemia, Graft versus host disease | Methicillin-resistant Staphylococcus epidermidis |
| 2 | Acromio-clavicular | 0.9 | 0.20 | 4222 | 92 | Lung neuroendocrine cancer (on carboplatin and etoposide) | Methicillin-resistant Staphylococcus aureus, Cutibacterium acnes |
| 3 | Hip | 0.8 | 0.22 | 10,080 | 97 | Acute lymphocytic leukemia (on methotrexate, mercaptopurine, and vincristine) | Staphylococcus epidermidis |
| 4 | Knee | 0.7 | 0.31 | 13 | 2 | Acute myeloid leukemia | Candida guillermondii |
| 5 | Knee | 0.6 | 0.53 | 7268 | 98 | Multiple myeloma (on velcade, melphalan, thalidomide, and prednisone) | Klebsiella pneumoniae |
| 6 | Wrist | 2.2 | 0.95 | 8700 | 94 | Hepatitis C liver cirrhosis | Pseudomonas aeruginosa |
| 7 | Knee | 3.2 | 1.15 | 24,000 | 68 | Status post orthotopic liver transplantation | Enterococcus faecalis |
| 8 | Knee | 3.3 | 1.98 | 104,000 | 89 | Chronic prednisone use | Serratia marcescens |
| 9 | Gleno-humeral | 2.9 | 2.00 | 63,510 | 97 | Severe sepsis | Methicillin-resistant Staphylococcus aureus |
| 10 | Knee | 3.2 | 2.18 | 13 | 30 | Acquired immunodeficiency syndrome due to human immunodeficiency virus | Methicillin-sensitive Staphylococcus aureus |
| 11 | Ankle | 2.4 | 2.28 | 35,800 | 97 | Status post kidney and pancreas transplantation with lymphoproliferative disorder | Methicillin-sensitive Staphylococcus aureus |
Comparison of our two cohorts confirmed similarity between the groups with respect to age, sex, serum ESR, and CRP (Table 3). As expected, these two groups differed significantly with regard to serum WBC count. They did not, however, vary significantly with regard to synovial percent PMNs.
| Synovial Fluid Study | Non-Leukopenic Patients | Leukopenic Patients | P Value |
| White Blood Cell Count (cells/μl) | 48,325 (IQR: 18,953–113,135) | 8900 (IQR: 4222–35,800) | 0.01 |
| Percent PMNs (%) | 94% (IQR: 90%–97%) | 92% (IQR: 68%–97%) | 0.33 |
3.3 Serum WBC count interpretation
There was a significant difference (p < 0.0001) between median serum WBC for non-leukopenic (median:12.77, interquartile range: 9.97–17.08) and leukopenic (median: 2.2, interquartile range: 0.8–3.2) patients (Fig. 1A). Similarly, median serum %PMNs differed significantly (p = 0.0174) between non-leukopenic (median: 80%, interquartile range: 74%–88%) and leukopenic (median: 60%, interquartile range: 36%–88%) patients (Fig. 1B).
3.4 Synovial WBC count interpretation
The median synovial WBC count was 48,325 (interquartile range: 18,953–113,135) cells/μl for non-leukopenic patients and 8900 (interquartile range: 4222–35,800) cells/μl for leukopenic patients (Fig. 2A). The median synovial %PMNs was 94% (interquartile range: 90%–97%) for non-leukopenic patients and 92% (interquartile range: 68%–97%) in leukopenic patients (Fig. 2B). There was a significant difference in synovial WBC count in leukopenic patients with septic arthritis compared to non-leukopenic patients with septic arthritis (Table 3). No significant difference, however, was found in the synovial %PMNs between the two cohorts (p = 0.33).
3.5 Association between synovial and serum WBC
For non-leukopenic patients, no significant correlation between synovial WBC and serum WBC was seen (ρ = −0.1223, p = 0.3279) and nearly all serum WBC values were less than 30 × 103 cells/uL (Fig. 3A). For leukopenic patients, 1 patient was observed to have a synovial WBC above 100,000 cells/uL but remained leukopenic for the corresponding serum WBC. In leukopenic patients, a trend was seen between increasing serum WBC and synovial WBC, but did not reach significance (ρ = 0.5103, p = 0.1088).
4 Discussion
The likelihood of septic arthritis increases as synovial WBC increases.1,2,6,9–12 The commonly accepted threshold of synovial WBC >50,000 cells/μL has a likelihood ratio of 7.7 while synovial %PMNs >90% is has a likelihood radio of 3.41. While these values are commonly applied to immunocompetent patients, it remains unclear if they accurately predict septic arthritis in leukopenic patients who have lower numbers of circulating leukocytes. Butler et al. demonstrated that the sensitivity and specificity for synovial WBC>50,000 cells/μL were 0.55 and 0.97, respectively, for immunocompromised patients presenting with septic arthritis.15 The sensitivity and specificity of synovial WBC>25,000 cells/μL were 0.70 and 0.89, respectively. These results suggest that immunocompromised patients should have a lower threshold value for these synovial laboratory values.15 Bell et al. evaluated 16 immunocompromised patients with culture positive septic arthritis and found the average synovial WBC was 29,000 cells/mm3. If the traditional synovial WBC cut off of 50,000 cells/mm3 was used, seven out of 10 patients with septic arthritis would have been.16 These studies suggest that traditional threshold values for septic arthritis cannot be used in immunocompromised patients. Interestingly, our median synovial WBC in non-leukopenic patient cohort was less than 50,000 cells/mm3 (48,325 cells/mm3). The patient populations in these studies were heterogenous with various diagnoses leading to non-quantifiable levels of immune dysfunction. Neither study used laboratory diagnostic criteria to diagnose an immunocompromised status or evaluated diagnostic criteria for septic arthritis in leukopenic patients.
Our study adds to the limited body of knowledge that evaluates the accuracy of diagnosing septic arthritis in leukopenic patients. Leukopenic patients had significantly lower synovial WBC compared to non-leukopenic presenting with septic arthritis with a median synovial WBC well below the accepted threshold of 50,000 cells/mm3. The traditional threshold of >50,000 synovial WBC would have missed all but 2 cases of septic arthritis. A threshold of 20,000 would have only captured 3 of 11 patients. Synovial WBC increased with serum WBC, but this trend was not found to be significant. While median synovial %PMNs was lower in leukopenic patients, this difference did not reach significance.
Since neutrophils are the primary cell initiating the immune response to septic arthritis,13,17 and neutropenia is a well described and accepted overall risk for infections,18,19 we hoped to evaluate patients with the diagnosis of neutropenia, defined as an absolute neutrophil count (ANC) < 1 × 103 cells/μl.20–22 Only six of our patients met the criteria of neutropenia leading our study to be underpowered to detect differences in this particular subset. However, all six patients that met the criteria of neutropenia had low synovial WBC, with the highest value being 10,080 cells/μl (Table 2). Of the five patients who did not meet the criteria of neutropenia, synovial WBC >20,000 was seen in all but one patient (subject number 10). It should be noted that this subject had the diagnosis of acquired immunodeficiency syndrome (AIDs), which has been shown not only to affect neutrophil count but also impaired neutrophil function and chemotaxis.23 This could contribute to the extremely low synovial WBC in this subject. Multi-centered studies with larger patient populations are necessary to evaluate the effect of neutropenia on synovial WBC and %PMN in patients with culture positive septic arthritis.
Given the retrospective nature of this study, there are a number of limitations. Not all data was available for analysis in every patient and this limited the ability to control for potential confounding variables. Additionally, some patients prior to 2010 did not have complete demographic data, medical histories, or laboratory values as the encounters pre-dated the institutions electronic medical record. The diagnosis of culture-positive septic arthritis in the presence of leukopenia is exceedingly rare with only 11 cases identified over an 18 year period, limiting the power of the study to perform more detailed sub-group analyses. Furthermore, only patients from a single institution were included in this study; in order to evaluate the validity of these findings, a multi-centered study would be helpful.
5 Conclusion
In summary, we found that synovial WBC is significantly lower in leukopenic patients presenting with culture proven septic arthritis compared to non-leukopenic patients. Even the most conservative diagnostic threshold of synovial WBCs of >20,000 cells/mm3 cannot accurately diagnose septic arthritis in the leukopenic population. Ongoing studies are evaluating which patient factors and laboratory values are predictive of septic arthritis in this population.
Authors contribution
CRB, RS, KTML and ALS designed the study. DL and JM performed the chart review and compiled data. CRB drafted the manuscript. RS, KTML, DL, JM, and ALS provided important revisions of intellectual content. ALS served as corresponding author. All authors are responsible for the accuracy of the manuscript.
Patient consent/Ethical committee approval
This study was approved by our Institutional Review Board. No patient consent was required for a retrospective review.
Disclosures
None of the authors report any conflicts of interest. No funding was received for this study.
References
- Management of septic arthritis: a systematic review. Postgrad Med. 2008;84:265-270.
- [Google Scholar]
- A prospective 2-year study of 75 patients with adult-onset septic arthritis. Rheumatology. 2001;40:24-30.
- [Google Scholar]
- Clinical features and outcome of septic arthritis in a single UK Health District 1982-1991. Ann Rheum Dis. 1999;58:214-219.
- [Google Scholar]
- Evidence-based diagnostics: adult septic arthritis. Acad Emerg Med. 2011;18:781-796.
- [Google Scholar]
- Evaluation and management of septic arthritis and its mimics in the emergency department. West J Emerg Med. 2019;20:331-341.
- [Google Scholar]
- Diagnostic utility of laboratory tests in septic arthritis. Emerg Med J. 2007;24:75-77.
- [Google Scholar]
- Laboratory tests in adults with monoarticular arthritis: can they rule out a septic joint? Acad Emerg Med. 2004;11:276-280.
- [Google Scholar]
- Bacterial or crystal-associated arthritis? Discriminating ability of serum inflammatory markers. Scand J Infect Dis. 1998;30:591-596.
- [Google Scholar]
- Synovial fluid leukocytosis in bacterial arthritis vs. reactive arthritis and rheumatoid arthritis in the adult knee. Scand J Rheumatol. 1992;21:283-288.
- [Google Scholar]
- Diagnosing joint infections: synovial fluid differential is more sensitive than white blood cell count. Eur J Orthop Surg Traumatol. 2014;24:1469-1474.
- [Google Scholar]
- Early diagnosis of septic arthritis in immunocompromised patients. J Orthop Sci. 2018;23:542-545.
- [Google Scholar]
- Septic arthritis in immunosuppressed patients: do laboratory values help? J Am Acad Orthop Surg Glob Res Rev. 2020;4
- [Google Scholar]
- Quantitative relationships between circulating leukocytes and infection in patients with acute leukemia. Ann Intern Med. 1966;64:328-340.
- [Google Scholar]
- 2002 guidelines for the use of antimicrobial agents in neutropenic patients with cancer. Clin Infect Dis. 2002;34:730-751.
- [Google Scholar]
- Febrile neutropenia in hematologic malignancies. Curr Hematol Malig Rep. 2013;8:370-378.
- [Google Scholar]
- Evaluation and management of patients with isolated neutropenia. Semin Hematol. 2013;50:198-206.
- [Google Scholar]
- Impaired neutrophil function in patients with AIDS or AIDS-related complex: a comprehensive evaluation. J Infect Dis. 1988;158:1268-1276.
- [Google Scholar]