Clinical characteristics of patients with bone sarcoidosis

Clinical characteristics of patients with bone sarcoidosis Author’s Accepted Manuscript Clinical characteristics of patients with bone sarcoidosis Ying Zhou, Elyse E Lower, Huiping Li, Yolanda Farhey,[.]

Author’s Accepted Manuscript

Clinical characteristics of patients with bone

sarcoidosis

Ying Zhou, Elyse E Lower, Huiping Li, Yolanda

Farhey, Robert P Baughman

DOI: http://dx.doi.org/10.1016/j.semarthrit.2017.02.004

To appear in: Seminars in Arthritis and Rheumatism

Cite this article as: Ying Zhou, Elyse E Lower, Huiping Li, Yolanda Farhey and Robert P Baughman, Clinical characteristics of patients with bone sarcoidosis,

http://dx.doi.org/10.1016/j.semarthrit.2017.02.004

This is a PDF file of an unedited manuscript that has been accepted for publication As a service to our customers we are providing this early version of the manuscript The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

www.elsevier.com/locate/semarthrit

Clinical characteristics of patients with bone sarcoidosis

Abstract

Objective: To assess the clinical features, diagnosis, and treatment of bone

sarcoidosis in the United States

Methods: Patients with bone sarcoidosis were identified and matched to sarcoidosis

patients based on race, gender, and age Detailed characteristics were obtained by medical record review

Results: A total of 64 patients with bone sarcoidosis were enrolled in the study The female: male ratio was 1.46: 1 and the White to Black ratio was 3: 1 Thirty-eight (59.4%) of 64 patients had bone symptoms Compared to matched cases, bone sarcoidosis patients have more multi organ involvement and higher incidence with liver, spleen and extra thoracic lymph node involvement than controls (P <0.05) Spine was the most commonly affected bone in 44 (68.8%)

of patients, followed by pelvis (35.9%) and hands (15.6%) MRI and PET/CT scan was the common imaging technology, which performed in 36 patients and 32 patients respectively, and with 97.2% and 93.8% positive bone uptake Laboratory test indicated anemia was more common in bone sarcoidosis group than controls (P=0.044) Infliximab was more commonly used in bone sarcoidosis patients than controls (P=0.009)

Conclusion: Bone sarcoidosis was associated with multi organs affection, and high

frequency of liver, spleen or extra thoracic lymph node involvement Infliximab

should be considered in those patients with aggressive and refractory bone

sarcoidosis

Introduction

Sarcoidosis is a multisystem disease typified by noncaseating granuloma

formation Although lungs and mediastinal lymph nodes are most commonly affected, the disease can involve any organ system [1] Only 3-5% of sarcoidosis patients have known bone disease [2] Previous studies suggest that small bones of the hands and feet are the most frequent sites of bone involvement with spine, pelvic, skull and rib localizations rarely reported [2] However, axial bone involvement may be more common than previously reported, since earlier studies relied mostly on plain x-rays which may be less sensitive for axial bone lesions With the usage of Magnetic

Resonance Imaging (MRI) and Positron Emission Tomography - Computed

Tomography (PET/CT) scanning, abnormalities suspicious for bone involvement are more frequently encountered However, since the differential diagnosis of these suspect lesions also includes malignancy and infection, a biopsy may be required for confirmation In 2014 the criteria for sarcoidosis organ involvement, including the bone, was updated by the World Association of Sarcoidosis and Other Granulomatous disease (WASOG) [3] By including newer diagnostic criteria for organ involvement and diagnostic techniques, this instrument should more precisely assess organ disease This study analyzes a large cohort of bone sarcoidosis patients compared with a case control group to better evaluate the prevalence of bone involvement in sarcoidosis and

to define treatment strategies

granulomas in either bone or other organs, after exclusion of other known causes of granulomatosis, as defined by the joint statement of the American Thoracic Society (ATS), European Respiratory Society (ERS) and WASOG [1] Using the WASOG instrument classification system, bone sarcoidosis was considered present when the criteria of definite (bone biopsy positive), highly probable, or probable bone

involvement was satisfied [3] The study was approved by the University of

Cincinnati Institutional Review board

Bone involvement includes both axial and appendicular sites Axial bone was defined as spine, pelvis, rib, skull, scapular, sternum, and clavicular bone while appendicular bone was defined as femur, tibia, fibula, feet, humerus, ulna, radius, and

hands The spine was further divided into cervical, thoracic, lumbar and sacral

divisions, while pelvis assessment includes ilium, ischium, and pubis

Extra bone organ involvement was also defined according to WASOG criteria [3] The involved organs assessed included bone marrow, lung, eye, skin, neurologic, renal, liver, spleen, extra thoracic lymph node, heart, sinus, parotid and muscle along with calcium metabolism Chest radiography was classified according to the modified Scadding staging system [4]: Stage 0 = no adenopathy or infiltrates; Stage 1= hilar and mediastinal adenopathy alone; Stage II = adenopathy and pulmonary infiltrates; Stage III = pulmonary infiltrates alone; and Stage IV = pulmonary fibrosis

Laboratory test results captured included total white blood cell (WBC),

hemoglobin (Hb), platelet count (PLT), alanine aminotransferase (ALT), aspartate transaminase (AST), total bilirubin (Bili), alkaline phosphatase (ALP) and serum calcium level All the laboratory data was extracted near the time of diagnosis with the normal reference ranges provided by the laboratory system of the University of Cincinnati The paired groups with both laboratory data of bone sarcoidosis case and its matched case were extracted for analysis All patients were seen by either Dr RPB

or Dr EEL Recommendations for anti-inflammatory treatment for bone symptoms were made using a standard protocol [5] Prescribed drugs including corticosteroids

(prednisone), cytotoxic drugs, and other biologics were recorded as current, past, or never

Two sarcoidosis control groups were identified: one group consists of all patients seen in the clinic without bone or bone marrow involvement (1738 patients) and the second group consists of matched case controls for each patient with bone sarcoidosis

By definition, the matched case control patient, who was matched for race, gender, and age, was identified from the total sarcoidosis clinic population and could not have bone or bone marrow involvement

Statistics: The differences of organ involvement between the two groups were

analyzed by 2

test corrected by continuity or Fisher’s exact test if needed The strength of association was expressed by odds ratio (OR) with 95% confidence

interval (95% CI) Mean values of other laboratory items were calculated and

compared between the two groups using Student’s t test Number with other abnormal laboratory values was calculated as well, and 2

test was performed The difference of medication usage between the two groups was analyzed by 2

test and corrected by continuity or Fisher’s exact test if needed Statistical analysis was performed using SPSS 19.0 package software with a value of P < 0.05 considered statistically

significant

Results

A total of 64 (3.4%) of the total 1802 sarcoidosis patients seen during the past six years at the University of Cincinnati Medical Center had bone involvement These 64 patients were compared to a control group of 64 sarcoidosis patients matched for race, gender, and age but without evidence for bone or bone marrow sarcoidosis

involvement All 128 patients were diagnosed with biopsy confirmed noncaseating granulomas in either bone or other organs Demographic data of patients with bone sarcoidosis are summarized in Table 1 In the bone sarcoidosis group, the female: male gender ratio was 1.46: 1 and the white to black race ratio was 3: 1 This gender and age group was similar to the entire clinic population: female:male: 1185/553; white: black:other: 1032/686/20; and median age of 44 years old), p>0.05 However, bone sarcoidosis was more common in white patients (P = 0.016) Thirty-eight of 64 (59.4%) of bone involved patients experienced bone symptoms including pain,

reduced exercise capacity, numbness or swelling

Table 2 summarized the organ involvement for bone sarcoidosis and control patients Thoracic disease was the most commonly involved extra-bone/bone marrow organ reported with 89.1% involvement for bone sarcoidosis patients and 92.2% in controls

No patients in the study had identified sarcoidosis muscle involvement The bone involved sarcoidosis patients experienced significantly higher incidence of liver,

spleen and extra thoracic lymph node involvement (26.6%, 20.3%, and 29.7%

respectively) compared to controls (P <0.05 for all) As noted in Figure 1, multi-organ involvement was much more common in the bone afflicted population compared to controls (Chi square=68.059, p<0.0001).Three or more organ involvement was found

in 61 (95.3%) of bone sarcoidosis patients versus 18 (28.1%) in controls (Chi

square=58.329, p<0.0001)

Fifty six of 64 (87.5%) of the bone involved patients experienced axial bone

involvement including spine, pelvis, rib, skull, scapular, sternum and clavicle, while only 19 (29.7%) demonstrated appendicular bone involvement of the hands, femur or humerus Eleven of 19 patients with appendicular bone involvement also displayed axial bone involvement Table 3 summarizes the bone involvement of all 64 cases As can be seen the spine was the most commonly affected bone in 44 (68.8%) of patients followed by pelvis (35.9%) and hands (15.6%) Overall, 35 (54.7%) patients

witnessed only single site bone involvement while 29 patients had two or more sites involved Hand involvement was identified in 6 of 16 black patients which was significantly more prevalent than in white patients (4 of 48 patients, P=0.017) There was no significant difference in other bone sites between blacks and whites In four of the patients with hand involvement, a skin lesion overlying the bone lesion was noted The frequency of this finding is not known, since the presence or absence of this finding was not recorded in all patients

According to WASOG criteria, all 64 patients met the criteria with definite (biopsy confirmed), “highly probable”, or “probable” involvement Bone or bone marrow biopsy was performed in 42 (65.6%) patients with evidence of noncaseating epithelioid cell granulomas The biopsy site was determined by bone symptoms or detected by imaging Table 4 demonstrates the image evaluation of bone involvement This includes fifty-three patients who had either MRI or CT/PET imaging All but two patients underwent diagnostic imaging for specific bone complaints

MRI and PET/CT scanning were the most common imaging modalities employed for bone evaluation Thirty-six and 32 patients underwent single modality testing with positive findings reported in 97.2% and 93.8% by MRI and PET-CT, respectively A total of 53 (82.8%) patients underwent both PET/CT and/or MRI Of these, 52

(98.1%) demonstrated positive bone findings by either PET/CT or MRI imaging Figure 2 demonstrates a positive PET-CT from a patient with multi organ

involvement including bone; whereas, Figures 3 and 4 reveal spine involvement by MRI Ten patients underwent plain film x-ray examination because of symptoms, and

in eight of 10 appendicular bone involvement was noted However, PET-CT was positive in the other two patients

Figure 4 summarizes the results of imaging in the bone sarcoidosis patients Bone involvement was detected in the 26 asymptomatic patients by MRI and or PET/CT scan Of the 38 symptomatic patients, bone disease was detected by MRI and PET/CT

in 27 patients In the remaining 11, bone disease was detected by conventional

radiographs in six and CT scan in five For the symptomatic patients, bone pain was the presenting or a major manifestation

Laboratory testing within six months of the diagnosis of bone involvement was available in 52 patients The mean total white blood count in bone involved

sarcoidosis patients was 6500 + 2310 cells/µL (mean + standard deviation) and was significantly lower than the paired control population (7600 + 2870 cells/µL, p

=0.048) There was no significant differences noted between hemoglobin levels or platelet counts Likewise values for liver and renal function and serum calcium were similar between the groups except for a slightly higher level of AST reported in the bone sarcoidosis group (34 + 30.2 U/L) versus control group (25 + 15.3 U/L,

p=0.008) The median values for both groups were still within the normal range for our laboratory of 10-35 U/L

Table 5 shows the medication usage for both the bone involved and control patients In the bone sarcoidosis group, five patients never received anti-inflammatory therapy; whereas, 59 patients were current or previous users of prescription anti-

inflammatory therapy Prednisone was the most commonly used medication in the group with 31 patients current users and 20 patients previous users Methotrexate was prescribed in 35 (54.7%) patients, hydroxychloroquine in 24 (37.5%), and infliximab

in 15 (23.5%) Infliximab was more commonly administered in the bone group compared to controls, (P=0.009) However, there was no prescribing differences between those with or without bone symptoms (data not shown)

Discussion

Bone sarcoidosis is a relatively rare organ manifestation with the frequency of bone involvement ranging from 3% to 13% [2] In this single institution study of more than 1800 sarcoidosis patients Sixty-four (3.6%) patients were confirmed to have bone involvement Interestingly, although white patients were three times more likely than blacks to have bone sarcoidosis, black patients experience more hand bone involvement A previous report suggested that peripheral bone lesions were more prevalent in black patients [2] In contrast, another study demonstrated that 95% of bone involved patients were white [6] Obviously, this raises the possibility of a racial predilection for prevalence and specific bone site of granulomatous disease

It is very likely that bone sarcoidosis may be underestimated In our study, nearly half of bone sarcoidosis patients were asymptomatic or presented with mild

nonspecific symptoms Although PET/CT scanning was useful in detecting

asymptomatic bone lesions, it is not routinely performed in sarcoidosis In a

respective study of 94 sarcoidosis patients, fluorodeoxyglucose (FDG) enhancement suggesting bone or bone marrow involvement was identified in 34% of patients [7] Typical radiographic features including trabecular pattern, osteolysis or cysts/punched out lesions were considered as “highly probable” bone sarcoidosis according to the WASOG criteria The presence of positive PET/CT, MRI, or gallium-67 bone

imaging classified lesions as “probable” bone sarcoidosis In the current study,

PET/CT scan or MRI had a 98.1% sensitivity MRI was usually performed to evaluate specific bone symptoms, while PET/CT scan was ordered for a more systemic

evaluation of granulomatous disease In this study, there were no features on either MRI or PET/CT which could distinguish sarcoidosis and malignancy Because these two techniques have limited specificity in distinguishing between sarcoidosis and metastatic lesions [8-10], bone or bone marrow biopsy was frequently performed in our series Non-caseating granulomas were identified in all 35 of the 52 patients with positive PET/CT or MRI who underwent biopsy This observation suggests that bone sarcoidosis should be considered in sarcoidosis patients with bone lesions detected by either PET/CT or MRI, and biopsy may not be required for patients with typical imaging patterns In contrast, CT scanning appears insensitive for detecting small bone sarcoid lesions in the hands and feet [8] In our study, positive CT imaging was seen in only 71.4% of 14 bone patients detected by other modalities Although most patients in this bone affected group have evidence of intrathoracic sarcoidosis on chest imaging, normal chest x-rays and CT scans have also been obtained from

patients with documented spine involvement In the presence of negative chest

roentgenographic findings, MRI or PET/CT scanning should be considered for the evaluation of sarcoidosis patients with unexplained bone complaints [11]

Symptoms of bone sarcoidosis include pain, reduced exercise capacity,

numbness, swelling or distortion of fingers Often pain is a prominent feature of axial skeletal sarcoidosis, and it can be the initial site of disease Swelling and hand

distortion are common features of appendiceal bone involvement Bone symptoms can often identify sites for image guided or blind biopsy However, in this study 26 of

64 (40.6%) patients were asymptomatic, and the bone lesions were incidentally