Limitations of Using Imaging Diagnosis for Psoas Abscess in Its Early Stage

Transcription

1 ORIGINAL ARTICLE Limitations of Using Imaging Diagnosis for Psoas scess in Its Early Stage Toshihiko Takada 1-3, Kazuhiko Terada 1, Hideki Kajiwara 1,2 and Yoshiyuki Ohira 2 stract Objective Patients diagnosed with psoas abscess have a high mortality rate. The major cause of its poor prognosis is delayed treatment. Therefore, making a correct diagnosis rapidly is important. Both computed tomography (CT) and magnetic resonance imaging (MRI) are considered to be the gold standards as imaging modalities that have a high sensitivity for detecting psoas abscess. There have been few reports regarding the limitations of these methods, but psoas abscess in its early stage may go undetected by CT and MRI. Methods Detection of psoas abscess by CT and MRI was investigated in the present study through a retrospective review of 15 patients in whom psoas abscess was diagnosed during a course of ten years at our hospital. Results In all patients, psoas abscess was diagnosed by at least a plain CT, enhanced CT, and/or plain MRI. The interval between the onset of symptoms and diagnosis was 20.9±17.9 days (mean ± standard deviation). In three patients, repeat imaging identified a psoas abscess, whereas initial imaging failed to detect it. The overall sensitivity of plain CT, enhanced CT, and plain MRI for psoas abscess was 78%, 86%, and 88%, respectively. From six days after the onset of symptoms, the sensitivity of each modality was 100%, while the sensitivity from day one to five days was only 33%, 50%, and 50%, respectively. Conclusion Although CT and MRI are considered to be gold standard modalities for diagnosing psoas abscess, both methods can fail to notice this condition in its early stage. Key words: psoas abscess, imaging diagnosis, sensitivity (Intern Med 54: , 2015) () Introduction Since psoas abscess was first described by Mynter in 1881 (1), it has been recognized as an infrequent, insidious, easily overlooked, and potentially life-threatening condition (2). Its detection has increased with the widespread availability of computed tomography (CT) (3), prior to which most cases were diagnosed at autopsy (4). However, psoas abscess is still regarded as rare, even in recent reports (5). Psoas abscess is classified as either primary or secondary (6). Primary psoas abscess transpires from the haematogenous or lymphogenous spread of an infection from a distant site, while secondary psoas abscess is caused by the direct extension of an infection from adjacent structures. The mortality rate of patients with primary psoas abscess is relatively low at 2.4%. In contrast, the mortality rate rises to 18.9% for secondary abscess and approaches 100% in untreated patients (3). The primary cause of death is delayed or inadequate treatment (3); therefore, it is essential to make a correct diagnosis of psoas abscess early after its onset and to provide the appropriate treatment. Psoas abscess is often misdiagnosed because patients tend to have nonspecific symptoms (5). For example, the hyperextension of the hip may cause pain due to the stretching of the affected psoas muscle, but this finding is not specific to psoas abscess (7). Its clinical presentation is variable, with the classical signs and symptoms of fever, back pain, and a limp only being present in 30% of patients (6). CT and Department of General Medicine, Kimitsu Chuo Hospital, Japan, Department of General Medicine, Chiba University Hospital, Japan and Department of Healthcare Epidemiology, School of Public Health in the Graduate School of Medicine, Kyoto University, Japan Received for publication January 9, 2015; Accepted for publication February 23, 2015 Correspondence to Dr. Toshihiko Takada, ttakada@water.ocn.ne.jp 2589

2 magnetic resonance imaging (MRI) are considered to be gold standard imaging modalities with a high sensitivity for the detection of psoas abscess (7). Although there is little literature about the shortcomings of these methods, there is a published case report that suggests plain CT has limitations in the early diagnosis of psoas abscess (5). We had recently reported a patient whose psoas abscess could not be detected by initial enhanced CT or MRI (8). Accordingly, we were prompted to perform the present study to investigate the possibility that neither CT nor MRI may be able to detect psoas abscess in its early stage. Study design Materials and Methods This study was performed at Kimitsu Chuo Hospital, a teaching hospital with 661 beds in Chiba Prefecture, Japan. We reviewed patients with psoas abscess over a ten-year period between April 2004 and March The diagnosis of psoas abscess was confirmed by the orthopedic team. We excluded patients in whom the timing of the onset of symptoms related to psoas abscess was unknown. The following information was retrospectively obtained from the patients medical records: the results of imaging studies (as interpreted by the orthopedic team and radiologists), the interval between the onset of symptoms and performance of diagnostic imaging, the time between onset and confirmation of the diagnosis, patient s sex, age, past medical history, clinical presentation, etiology of psoas abscess, results of bacteriological examination, treatment, and outcome. CT scans were mainly conducted using a 64-channel multidetector CT scanner (LightSpeed VCT; GE Medical Systems, Milwaukee, USA). For an enhanced CT, a contrast medium [100 ml iopamidol (Oipalomin 300); Konica Minolta, Tokyo, Japan; 1.5 ml/s; delay, 100 seconds] was used. In addition, 1.5-T MR scanners (Signa HDxt; GE Medical Systems) with a 8-channel cervical-thoracic-lumbar spine array coil were mainly used for the MRI scanning. The routine MRI scans were with the following parameters: (1) spin-echo T1- weighted [repetition time (TR), 500 milliseconds (ms); echo time (TE), 9.8 ms;, echo train length (ETL) 3; and matrix, ], (2) spin-echo T2-weighted (TR, 4,020 ms; TE, 95 ms; ETL 19 and matrix, ). All sequences were acquired with the following parameters: field of view, 32 cm; layer thickness, 5 mm. Approval for this study was given by the research ethics committee of Kimitsu Chuo Hospital. Statistical analysis The sensitivity of each imaging modality was calculated. Afterwards, the sensitivities of each imaging modality during the early period (within five days after the onset of symptoms) were compared to those during the late period (six days or more after the onset). Results Nineteen patients with psoas abscess were identified by this retrospective analysis, among whom four patients were excluded because the time of the onset of symptoms related to the abscess was unknown. In the remaining 15 patients, there were eight male patients. Their mean age was 63.3± 14.4 (mean ± standard deviation). All cases were of unilateral psoas abscess (seven right-sided, eight left sided). There were six primary abscesses, four secondary to identifiable causes, and five with unknown etiologies. A plain CT was completed 18 times, an enhanced CT was completed 14 times, and a plain MRI was completed 8 times (Table 1). The imaging studies were assessed by both the orthopedic team and by radiologists; they were in complete agreement over the findings of the studies being conducted. The interval between the onset of symptoms and diagnosis had a mean of 20.9±17.9 (mean ± standard deviation) days. In three patients (No. 4, 12, and 13 in Table 1), the initial imaging failed to detect the psoas abscess (Figure) despite its high pre-test probability, which was estimated by the clinical manifestation. One of these three patients (No. 4) was the only subject in the study who expired. The overall sensitivity of the imaging modalities for the diagnosis of psoas abscess was 78% for plain CT, 86% for enhanced CT, and 88% for plain MRI. When imaging was performed at six days or more after the onset of symptoms, the sensitivity was 100% for all three modalities. In comparison, the imaging sensitivity decreased to 33% for plain CT, 50% for enhanced CT, and 50% for plain MRI when imaging was performed up to five days after the onset (Table 2). Discussion Although CT and MRI are considered to be the gold standard methods, known for their high sensitivity in diagnosing psoas abscess, the findings of this retrospective study suggested that both modalities can overlook an abscess in its early stage. Psoas abscess is often classified as pyomyositis, which is defined as a bacterial infection of skeletal muscle that usually manifests as a single or multiple abscesses (9). Pyomyositis can be divided into three clinical stages. In the early stage, there may not be a discrete abscess, but the muscle has a wooden or hard rubber feel on palpation. Aspiration will not yield pus at this time (10). The present investigation identified three patients with no detectable abscess from initial imaging studies, although their clinical manifestations were compatible with psoas abscess. Their lesions might have been at an early stage of the inflammatory process, before the formation of a clinically apparent abscess. When we reviewed the imaging studies of these three patients, there were some subtle signs of inflammation, including an unclear border of the psoas muscle in patient No

3 Table 1. Clinical Features of the 15 Patients with Psoas scess. Patient No. Year Sex Age F 82 Past Medical History Lumbar spondylosis Clinical presentation Left buttock pain Diagnosis Etiology Bacterial pathogen Plain CT Enhanced CT Plain MRI Vertebral osteomyelitis Secondary F 68 - Left hip pain unknown M 56 - Right thigh pain, lumbar pain Right PA Vertebral osteomyelitis GBS (pus) Unknown No culture F 74 RA Rihgt lumbar pain Right PA Primary MSSA (blood, pus) M M M 30 - Right lumbar pain, swelling Left thigh, lumbar pain Right inguinal pain, swelling Right PA Penetration of colon cancer Lung abscess The day diagnosis confirmed Positive (Day 7) Positive (Day 8) 7 Treatment Outcome Positive (Day 5) Positive (Day4) 4 Secondary MSSA (blood) Positive (Day 26) Positive (Day 24) 24 Secondary M 63 HT, AP Right lumbar pain Right PA unknown F F 56 Left buttock, thigh pain Right inguinal, thigh pain M 56 - Left hip pain M 55 DM, CRF on HD F 86 Hepatocellular carcinoma Left lower abdominal pain Left hip pain Streptococcus constellatus E.coli (blood, pus) Negative (Day 2) Positive (Day 10) Positive (Day 10) Positive (Day20) Positive (Day20) 20 Primary MSSA (blood) Positive (Day 30) Positive (Day 30) Positive (Day 30) 30 Right PA unknown Tb (pus) Positive (Day 40) Positive (Day 40) 40 Primary Right PA primary Septic arthritis of left hip primary Proteus.mirabilis (pus) S.pneumonia (pus) Eikenella corrodens (pus) MSSA (pus) Primary MSSA (blood) Pyelonephritis M 62 Lung cancer Left lumbar pain unkonown F 56 Cervical cancer Right buttock pain Right PA unknown PA: psoas abscess, : antibiotics, MSSA: methicillin-sensitive S. aureus Secondary E.coli (blood, urine) E.coli (pus) Unknown Blood, pus culture Positive (Day 61) Positive (Day 61) Positive (Day 61) 61 Positive (Day 8) Positive (Day 8) 8 Positive (Day 16) Positive (Day 16) Positive (Day 16) 16 Positive (Day 56) Positive (Day 56) Positive (Day 56) 56 Positive (Day 3) Negative (Day1) Negative (Day4) Positive (Day 3) Negative (Day1) Positive (Day4) Positive (Day 20) Positive (Day 20) 20 Positive (Day 10) Positive (Day 11) Positive (Day 11) Death 3 4 (plain CT on day two) and a slight enlargement of the muscle in patient No. 13 (plain CT on day four). Neither the orthopedic team nor the radiologists who initially assessed these images interpreted the changes as positive for psoas abscess, although they suspected the diagnosis. Thus, these changes may represent very early signs of psoas abscess, but detecting such subtle changes is not always possible. Therefore, when psoas abscess is strongly suspected and early im- 2591

4 Intern Med 54: , 2015 Patient 4 Plain CT (Day 2) Plain CT (Day 10) Enhanced CT (Day 10) Plain CT (Day 3) Enhanced CT (Day 3) Patient 12 Enhanced CT (Day 1) Patient 13 Enhanced CT (Day 1) Plain CT (Day 4) Enhanced CT (Day 4) Figure. Computed tomography findings in a patient with psoas abscess. The abscess could not be detected by initial imaging in patient No. 4 (plain CT on day two), in patient No. 12 (enhanced CT on day one, plain CT on day four), and in patient No. 13 (enhanced CT on day one). Repeat imaging revealed the abscess (arrows). Table 2. Sensitivity of Imaging Studies for Diagnosis of Psoas scess. Within 5 days of the onset Six days or more after the onset Plain CT 33% (2/6) 100% (12/12) 78% (14/18) Enhanced CT 50% (2/4) 100% (10/10) 86% (12/14) Plain MRI 50% (1/2) 100% (6/6) 88% (7/8) Imaging modality Overall aging studies fail to confirm the diagnosis, repeating the investigation after several days should be considered. Our study has several limitations. First, the sample size was small because psoas abscess is not a common condition and the study was conducted in only one hospital. A small sample size limits its generalizability; in particular, the value of sensitivity may be inaccurate. The variable in each modality and in each period was also small, thereby limiting the reliability of our comparative analysis. In addition, the small sample size restricted us from investigating factors, other than time period, for false studies. Clarifying risk factors for diagnostic delay (e.g. patients characteristics, type of psoas abscess and pathogen, location of abscess) could allow physicians to further estimate the possibility of false studies. Therefore, these factors should be further investigated. Second, imaging modalities in our hospital were standard, but not the most updated ones. The newest medical imaging modalities could exhibit advanced performance. Third, only T1 and T2 weighted images were available for our patients due to the lack of time for these unscheduled studies. Some techniques (e.g. fat suppression) for MRI could have yielded higher sensitivity (11). Finally, since the psoas abscesses with unknown timelines were eliminated we do not know how these patients imaging results at diagnosis could have affected the clinical results. Despite these limitations, this is the first study to suggest the possibility that CT and MRI could fail to detect psoas abscess in its early stage, even though they are considered to be gold standard modalities. Because the major cause of death in patients with psoas abscess is delayed or inadequate treatment (3), the prompt initiation of antibiotic therapy without waiting for the confirmation of its diagnosis might be recommended in these circumstances. The authors state that they have no Conflict of Interest (COI). Acknowledgement We thank Wakako Kaneko Mikami for her insightful advice from the perspective of a radiologist. 2592

5 References 1. Mynter H. Acute psoitis. Buffalo Med Surg J 21: , Lai YC, Lin PC, Wang WS, et al. An update on psoas muscle abscess: An 8-year experience and review of literature. Int J Gerontol 5: 75-79, Gruenwald I, rahamson J, Cohen O. Psoas abscess: case report and review of the literature. J Urol 147: , Altemeier WA, Alexander JW. Retroperitoneal abscess. Arch Surg 83: , Mückley T, Schütz T, Kirschner M, Potulski M, Hofmann G, Bühren V. Psoas abscess: the spine as a primary source of infection. Spine 28: E106-E113, Chern CH, Hu SC, Kao WF, Tsai J, Yen D, Lee CH. Psoas abscess: making an early diagnosis in the ED. Am J Emerg Med 15: 83-88, Mallick IH, Thoufeeq MH, Rajendran TP. Iliopsoas abscesses. Postgrad Med J 80: , Takada T, Terada K, Kajiwara H, Ikusaka M. Imaging- psoas abscess. Lancet 383: 280, Crum NF. Bacterial pyomyositis in the United States. Am J Med 117: , Chiedozi LC. Pyomyositis. Review of 205 cases in 112 patients. Am J Surg 137: , Bley TA, Wieben O, François CJ, Brittain JH, Reeder SB. Fat and water magnetic resonance imaging. J Magn Reson Imaging 31: 4-18, The Japanese Society of Internal Medicine