電腦斷層在HLA-B27血清陰性僵直性脊椎炎病患的診斷可靠性––209個病人的回溯性研究

Diagnostic validity of computed tomography for HLA-B27 negative ankylosing spondylitis: a retrospective study of 209 patients

目的：影像中呈現薦腸關節炎是僵直性脊椎炎診斷的重要特徵。然而，由於判讀影像的觀點和品質因不同人而有極大的差異，因此單使用X光片來判斷薦腸關節炎的程度，是非常不容易的。再者，當病患的HLA-B27血清呈現為陰性時，要做出正確的診斷就更是困難。過去的研究報告指出使用X光片來診斷薦腸關節炎容易低估薦腸關節炎的發生率，並且電腦斷層可以有效地幫助診斷僵直性脊椎炎。然而目前並沒有其它研究，探討關於電腦斷層在HLA-B27血清陰性的僵直性脊椎炎病患診斷上的應用。方法：我們回顧了從2009年到2013年，這五年內本院所有的薦腸關節電腦斷層報告，並查閱這些做過薦腸關節電腦斷層的病患同時期的薦腸關節X光報告。我們總共蒐集到544個病患，因為下背痛超過三個月而來求診，並且接受薦腸關節電腦斷層檢查。當中，有209個病患HLA-B27血清呈現為陰性，而其中有204個病患同時有可取得的薦腸關節X光報告。另外，我們從病歷記載中針對病人的臨床幾項特徵，加以統計分析。結果：總共408個薦腸關節接受X光及電腦斷層檢查，兩種檢查方式都確認有薦腸關節炎的比例為20.1%。然而，有67.6%的病人使用電腦斷層診斷薦腸關節炎的嚴重度高過X光診斷薦腸關節炎的嚴重度，12.3%的病人電腦斷層診斷薦腸關節炎的嚴重度低於X光診斷薦腸關節炎的嚴重度。根據電腦斷層診斷薦腸關節炎的結果，78.9%的病人符合the modified New York criteria，可是根據X光診斷的結果只有26%的病人可以符合。原本X光診斷不符合僵直性脊椎炎診斷的病人，經進一步電腦斷層檢查後，有57.3%病人達到診斷標準。結論：薦腸關節電腦斷層是一個敏感性高、有助於診斷僵直性脊椎炎的影像工具。我們建議患慢性發炎性背痛的病患，若其HLA-B27呈陰性，且X光的診斷無法確認，臨床上仍強烈懷疑僵直性脊椎炎，使用薦腸關節電腦斷層檢查可以提供快速且精確的診斷。

Objective: Radiographic sacroiliitis is considered the hallmark of ankylosing spondylitis (AS). However, large interobserver variations make plain radiographic diagnosis of sacroiliitis notoriously difficult. The heterogeneity of HLA-B27 negative AS makes its diagnosis even more challenging. Several reports have shown that sacroiliitis tends to be underestimated using radiography, and that computed tomography (CT) facilitates AS diagnosis in patients with suspected spondyloarthritis. However, no studies have stressed the clinical utility of CT, particularly in the diagnosis of HLA-B27(-) AS. Methods: We retrospectively evaluated 209 HLA-B27(-) patients with chronic lower back pain and suspected AS. Radiography and CT reports were examined and outcomes were compared. Results: Among 408 sacroiliac (SI) joints examined using both radiography and CT, there was agreement between the two methods in the sacroiliitis grading of 82 (20.1%) SI joints. However, sacroiliitis grade using CT was higher in 276 (67.6%) SI joints and lower in 50 (12.3%) SI joints. CT evaluation of SI joints showed that 78.9% of patients met the radiographic sacroiliitis criteria of the modified New York criteria, while only 26% of patients satisfied the criteria for plain radiography. Surprisingly, 117 patients (57.3%), who did not meet the modified New York AS diagnosis criteria for plain radiography met the criteria for CT. Conclusion: CT was sensitive and useful in providing evidence for the diagnosis of AS. We suggest that HLA-B27(-) patients with equivocal plain radiography results and chronic inflammatory lower back pain be examined using CT.