Usefulness of 2-[18F]-fluoro-2-deoxy-d-glucose–Positron Emission Tomography/Computed Tomography for Staging and Evaluation of Treatment Response in IgG4-Related Disease: A Retrospective Multicenter Study

Authors


Abstract

Objective

To evaluate the usefulness of 2-[18F]-fluoro-2-deoxy-d-glucose–positron emission tomography/computed tomography (FDG-PET/CT) in IgG4-related disease (IgG4-RD) for the staging of the disease and the followup under treatment.

Methods

All patients included in the French IgG4-RD registry who underwent ≥1 FDG-PET/CT scan were included in the study. Clinical, biologic, pathologic, radiologic, and FDG-PET/CT qualitative and quantitative findings were retrospectively collected and analyzed.

Results

Twenty-one patients were included in the study and 46 FDG-PET/CT examinations were evaluated. At either diagnosis or relapse, all evaluated patients presented abnormal 18F-FDG uptake in typical IgG4-RD localizations. In most cases, FDG-PET/CT was more sensitive than conventional imaging to detect organ involvement, especially in arteries, salivary glands, and lymph nodes. In few cases (small-sized lesions and brain or kidney contiguous lesions), false-negative results were noted. Evaluation before and after treatment showed in most cases a good correlation of FDG-PET/CT results with treatment response and disease activity.

Conclusion

This large retrospective study shows that FDG-PET/CT imaging is useful for the staging of IgG4-RD. Moreover, FDG-PET/CT is useful to assess the response to treatment during followup.

INTRODUCTION

IgG4-related disease (IgG4-RD) recently has been recognized as a new disease characterized by diffuse or focal organ enlargement, with mass-forming or nodular/thickened lesions in various organs ([1]). Numerous tissues can be affected either synchronously or metachronously, the most frequent of which are the pancreas, salivary or lacrimal glands, lymph nodes, kidneys, and retroperitoneum ([2, 3]). Other organs or tissues can be involved, and most patients present with more than 2 organs involved at diagnosis ([3, 4]). IgG4-RD is therefore considered as a multiple organ systemic disease. The disease can be distinguished from autoimmune or other inflammatory conditions as well as from sarcoidosis and malignant tumors by characteristic pathologic findings ([5]). In addition to a polyclonal lymphocytic and plasmacytic infiltration with predominance of IgG4-positive plasma cells, severe tissue fibrosis, with a peculiar “storiform” aspect, is a characteristic of the disease and can be responsible for tissue loss of function, exocrine and/or endocrine pancreatic insufficiency in more than 30% of cases with type 1 autoimmune pancreatitis (AIP; IgG4-related pancreatitis) ([6, 7]), and renal insufficiency in more than 70% of cases with IgG4-related tubulointerstitial nephritis ([8]). In most cases, IgG4 serum levels are increased above 135 mg/dl. However, this is not specific of the disease ([9, 10]), and some patients have normal serum IgG4 levels despite typical organ involvement and characteristic histology ([11]). Because of the risk of tissue fibrosis and organ loss of function, treatment is usually recommended. The rate of response to first-line treatment by steroids is very high ([12]), but relapses are frequent ([13, 14]), and some patients require second- or third-line therapies ([4]).

Therefore, the precise “staging” of active or inactive specific organ involvement is essential to improve treatment decisions. Clinical, biologic, and conventional radiologic evaluation is currently used in combination for this evaluation.

2-[18F]-fluoro-2-deoxy-d-glucose–positron emission tomography/computed tomography (FDG-PET/CT) is a noninvasive imaging technique commonly used in the field of clinical oncology allowing whole-body examination. 18F-FDG accumulation is also recognized as useful for diagnosis, characterization, and monitoring of response to therapy in patients with some inflammatory disorders ([15, 16]). In case reports or short series of AIP, it has been shown that pancreatic and extrapancreatic tissues involved with IgG4-RD show 18F-FDG uptake ([17-19]). For this reason, it has been suggested that FDG-PET/CT could be useful at diagnosis to distinguish IgG4-related pancreatitis from other pancreatic pathologies ([20]). Moreover, it has been suggested that this exploration could be useful for the evaluation of disease distribution ([21]), guiding minimally invasive tissue diagnosis to support the diagnosis of IgG4-RD ([22]), highlighting conditions for which steroid therapy is indicated, monitoring response to therapy ([23]), or detecting a relapse ([24]).

To better evaluate the contribution of FDG-PET/CT findings in IgG4-RD, we retrospectively analyzed the results at diagnosis and/or relapse, after treatment, and/or during followup in patients from the French IgG4-RD registry. Results of FDG-PET/CT were also compared to other radiologic findings.

Box 1. Significance & Innovations

  • To date, this is the largest retrospective series of 2-[18F]-fluoro-2-deoxy-d-glucose–positron emission tomography/computed tomography (FDG-PET/CT) evaluations in IgG4-related disease (IgG4-RD).
  • We show that FDG-PET/CT is a useful imaging tool for “staging” of IgG4-RD before treatment at initial diagnosis or during a relapse.
  • FDG-PET/CT seems to be more sensitive than conventional imaging (ultrasound, CT, magnetic resonance imaging) to detect active localizations of IgG4-RD, especially for aortic/arterial, salivary, and nodal involvements.
  • FDG-PET/CT correlates with disease activity and could therefore be proposed to evaluate patients in future clinical trials.

PATIENTS AND METHODS

Patients

All patients from the French registry for IgG4-RD were selected for the study. Patients included in the registry responded to possible, probable, or definite IgG4-RD in accordance with the comprehensive diagnostic criteria proposed by Umehara et al ([25]). All patients from the registry who underwent whole-body FDG-PET/CT imaging at least once were retained for the study, and data were retrospectively collected from their medical records. FDG-PET/CT qualitative findings (number of sites showing a pathologic uptake of 18F-FDG) and semiquantitative findings (maximum of the standard uptake value [SUVmax]) were systematically recorded. Conventional radiologic imaging (ultrasonography [US], contrast-enhanced CT, and magnetic resonance imaging [MRI]) and biologic findings (serum IgG4 and C-reactive protein levels) of patients were also recorded for comparison with FDG-PET/CT results.

Methods

FDG-PET/CT was performed using an integrated scanner in different centers throughout France according to national and European guidelines ([26]). For each evaluation, the patients underwent ≥4 hours of fasting and good oral hydration before the study. Euglycemia was documented before injection administration by measuring blood glucose level. Dose of 18F-FDG was adapted to the system and the patient's weight. Patients rested during a 45–60-minute uptake period. CT transmission data acquisition was performed in order to correct the attenuation and to anatomically localize the pathologic uptake of 18F-FDG. Attenuation-corrected images were reconstructed using an iterative algorithm. The SUV, a semiquantitative index of 18F-FDG uptake in tissue, was computed as follows: SUV = PET activity/(injected dose/body weight), where 18F-FDG concentration is expressed in mg/ml.

FDG-PET/CT findings at the initial diagnosis or at a relapse before any treatment were analyzed for the usefulness to detect systemic lesions of IgG4-RD (IgG4-RD staging). All sites with abnormal 18F-FDG uptake as retained by the record, except those related to physiologic uptake, were retained as potential IgG4-RD organ involvement. Pathologic analysis for confirmation of IgG4-RD was not available for all sites with 18F-FDG uptake. Relapse was considered when a patient presented with clinical, biologic (increase of serum IgG4 levels), and radiologic findings typical of active IgG4-RD.

FDG-PET/CT findings at diagnosis were compared with results after treatment in individual patients to study the correlation with disease activity and the usefulness to evaluate treatment response. FDG-PET/CT findings were compared to results of other concomitant radiologic findings (US, CT, or MRI) to search for discrepancies. Other radiologic evaluations were performed no more than 1 month before or after FDG-PET/CT evaluation. Estimation of sensitivity for the FDG-PET/CT and for CT was evaluated by using any abnormal finding from clinical examination or imaging tests as the reference standard.

RESULTS

Patients' characteristics

Twenty-one patients among the 78 from the French IgG4-RD registry had an available FDG-PET/CT evaluation and were included in this study. The median age at the first FDG-PET/CT evaluation was 62 years (mean 61.1 years, range 21–83 years). There were 16 men and 5 women (sex ratio 3:2). Six patients had diabetes mellitus, with a controlled normal serum glucose level in all patients at the time of the FDG-PET/CT scan. IgG4-related organ involvement presented by patients is shown in Table 1. Patients presented with locoregional or diffuse lymph node enlargement (n = 17), salivary gland involvement (n = 12), type 1 AIP (n = 9), renal involvement (n = 8, with pseudotumoral presentation in 3), retroperitoneal fibrosis (n = 7), lung involvement (n = 5), lacrimal gland or orbital involvement (n = 4), sclerosing cholangitis (n = 4), arterial involvement (n = 3, with aortic involvement in 3, carotidian involvement in 2, and iliac involvement in 1), and pachymeningitis, mesenteric inflammatory pseudotumor, prostatitis, and hypophysitis (n = 1 each). The mean number of organs involved per patient determined by complete clinical, radiologic, and FDG-PET/CT findings was 3.5 (median 4, range 1–6). Serum IgG4 levels were elevated (>1.35 gm/liter) in 17 patients and normal in 4 patients (mean 7.48 gm/liter, range 0.198–47.35). Conventional histology was available in all patients, with typical morphopathologic findings of IgG4-RD in all patients, as shown in Table 1. Nine patients (43%) presented with typical clinical involvement, serum IgG4 elevation, and positive IgG4 immunostaining, and 8 patients (38%) presented with typical clinical involvement, serum IgG4 elevation, and morphopathologic findings, but no available immunostaining. Four patients (19%) presented with typical clinical involvement, normal serum IgG4, typical morphopathologic findings, and positive immunostaining.

Table 1. Clinical, pathologic, and biologic characteristics of patients*
PatientSexClinical involvementHistologic findingsaIgG4, gm/literbCRP, mg/liter
  1. CRP = C-reactive protein; LNs = lymph nodes; hpf = high-power field; ratio = IgG4+ plasma cell:IgG+ plasma cell ratio; AIP = autoimmune pancreatitis; TIN = tubulointerstitial nephritis; RPF = retroperitoneal fibrosis.
  2. aTissues analyzed for histologic features. All tissues analyzed and indicated here showed typical histologic features of IgG4-RD, including in all patients dense lymphoplasmacytic infiltrate and fibrosis, typically with a storiform pattern (except in LNs and salivary glands); obliterative phlebitis and/or increased numbers of eosinophils were seen in some cases and granuloma or prominent neutrophilic infiltrate was absent in all cases.
  3. bSerum IgG4 levels are given at diagnosis before treatment.
  4. cTissues with available immunostaining (IS).
  5. dElevated serum IgG4 level (>1.35 gm/liter).
1MRight parotidian sclerosing sialadenitis, regional LNsParotidc0.1980
  LNc  
  IS: >50 cells/hpf, ratio 50%  
2MPachymeningitisMeningesc3.11d5
  IS: >10 cells/hpf  
3MAIP, TIN, LNs, sialadenitis, pleural and lung involvementPancreasc21.9d34
  Kidneyc  
  LNc  
  IS: >10 cells/hpf, 50 cells/hpf and ratio 40%, 80 cells/hpf and ratio 50%, respectively  
4MTIN, renal pseudotumor, LNs, interstitial pneumonitisLN47.35d5
  Kidney  
5MSclerosing cholangitis, RPF, LNs, lacrimal and submandibular glandsLiverc0.5273
  RPFc  
  LNc  
  Lacrimal gland  
  IS: 10 cells/hpf, 36 cells/hpf, and 36 cells/hpf, respectively  
6MAIP, sialadenitisPancreasc2.9d<5
  IS: numerous IgG4+ plasma cells, not quantified  
7MRPF, pulmonary nodulesRPF3.44d40
8MRPF, mesenteric pseudotumor, LNs, pulmonary involvementRPF13d10
9MRPF, aortitis, sialadenitis, LNsSalivary gland1.4d53
10FAIP, LNs, interstitial pneumonitisPancreas6.4d38
  LN  
11MOrbital pseudotumor, LNs, AIP, renal pseudotumorPancreas0.52<4
  LNc  
  IS: ratio >90%  
12MAIP, LNs, RPF, sialadenitisPancreas4.08d82
  LN  
  Salivary gland  
13FAIP; lacrimal, parotidian, and submandibular glands; LNsPancreas5.14d24
  Parotid  
  LNc  
  IS: ratio 30%  
14MRPF, LNs, arterial aneurysms (aorta, iliac arteries, carotid)LNc2.5d20
  IS: >150 cells/hpf, ratio 60%  
15MTIN, LNs, sialadenitis, AIP, sclerosing cholangitis, prostateKidneyc16.5d4
  Salivary gland  
  IS: >10 cells/hpf, ratio >50%  
16MTIN, LNs, sialadenitis, AIP, RPFKidneyc17.5d6
  Salivary glandc  
  LNc  
  IS: ratios 32.5%, 44%, and 57%, respectively  
17FTIN, LNs, sialadenitisSalivary gland18.8d<5
  LNc  
  IS: 40 cells/hpf, ratio 50%  
18FSclerosing cholangitis, LNsLiver1.9d60
  Biliary tree  
  LNc  
  IS: 100 cells/hpf  
19FAortitis, carotid, LNsLNc0.5968
  IS: ratio 30%  
20MAIP, sclerosing cholangitis, orbital pseudotumor, sialadenitis, TINSalivary gland1.96d64
  Pancreas  
  Biliary tree  
  LN  
21MTIN, LNs, sialadenitis, hypophysitisKidney20.6d5
  Salivary gland  

FDG-PET/CT evaluation for organ involvement in IgG4-RD

A total of 46 FDG-PET/CT examinations for 21 patients were retrospectively collected. FDG-PET/CT was performed either at diagnosis before treatment (n = 10 patients) or during the followup of the disease (n = 11 patients). In this last group, FDG-PET/CT was performed at the time of IgG4-RD relapse either before retreatment (n = 9 patients) or under treatment (n = 2 patients). Some of the patients had several FDG-PET/CT evaluations under treatment for IgG4-RD, accounting for the total of 46 examinations.

To evaluate the performance of FDG-PET/CT to detect IgG4-RD organ involvement, we first analyzed FDG-PET/CT performed at diagnosis or at the time of a relapse in patients without treatment for IgG4-RD (n = 19). The results are shown in Table 2. Only one of these patients (patient 9) received 10 mg/day of prednisone for polyarthritis at the time of the FDG-PET/CT evaluation. Abnormal 18F-FDG uptake was noticed for all patients. In most patients (n = 16 of 19), >1 organ showed abnormal 18F-FDG uptake. Abnormal 18F-FDG uptake was observed in lymph nodes (n = 14), salivary glands (n = 6, parotid and/or submandibular), lungs and/or pleura (n = 6), kidneys (n = 5), aorta and/or other arteries (n = 4), pancreas (n = 4), retroperitoneal fibrosis (n = 2), and mesenteric inflammatory pseudotumor, breast, biliary tract, liver, and prostate (n = 1 each). SUVmax was not always available on FDG-PET/CT records. The median SUVmax was 4 (range 1.4–7.4). The highest SUVmax values were observed with retroperitoneum, lung, and nodal involvement. 18F-FDG abnormal uptake corresponded to sites usually associated with IgG4-RD organ involvement. In case of an “unusual” location, pathologic documentation confirmed IgG4-RD involvement (on liver biopsy for patient 18). Breast 18F-FDG uptake in patient 17, however, was not confirmed by pathologic analysis. Organs clinically involved and symptomatic are shown in Table 2 (lymph nodes [n = 6], salivary glands [n = 3], pancreas [n = 3], retroperitoneal fibrosis [n = 2], and meninges, lung, liver, breast, and lacrimal glands [n = 1 each]). Clinically involved organs were hypermetabolic on FDG-PET/CT in 89% of patients (n = 17 of 19). Furthermore, hypermetabolic sites of clinically asymptomatic localizations on FDG-PET/CT were detected in 17 of 19 patients. These were represented by 31 involvements of organs (lymph nodes [n = 11], lungs and/or pleura [n = 5], kidneys [n = 4], aorta and arteries [n = 4], salivary glands [n = 3], and pancreas, mesenteric inflammatory pseudotumor, biliary tract, and prostate [n = 1 each]).

Table 2. Results of FDG-PET/CT in patients without treatment and results of clinical, CT, and MRI radiologic findings*
PatientDisease stateaOrgan or tissues involved with IgG4-related disease
With 18F-FDG uptakeWithout 18F-FDG uptake
  1. FDG-PET/CT = 2-[18F]-fluoro-2-deoxy-d-glucose–positron emission tomography/computed tomography; MRI = magnetic resonance imaging; LNs = lymph nodes; RPF = retroperitoneal fibrosis.
  2. aAt the time of FDG-PET/CT.
  3. bOrgan involvement clinically symptomatic at the time of PET/CT.
1RelapseMRI+ 
 Right parotidb 
 LNs 
2RelapseCT−MRI+
 Aortic crossMeningesb
3RelapseCT+ 
 LNsb 
 Bilateral pleura 
 Lungs 
 Kidneys 
 CT− 
 Left parotid 
4RelapseCT+ 
 Kidneys 
 LNsb 
 Left basal lung 
5RelapseCT+ 
 RPFb 
 LNs 
 CT− 
 Diffuse bone marrow 
 Thymus 
7RelapseCT+ and MRI+ 
 RPFb 
8RelapseCT+CT+
 Mesenteric pseudotumorRPF
9DiagnosisCT not available 
 Abdominal and thoracic aorta 
 Axillar LNs 
 Right pleura 
10DiagnosisCT+ 
 Pancreasb 
 Celiac and mediastinal LNs 
 Lung nodulesb 
11DiagnosisCT−CT+
 Bilateral parotidLacrimal/orbitalb
 CT+Kidney
 PancreasLung
 LNsbMRI+
  Lacrimal/orbital
13DiagnosisCT+ and MRI+ 
 Cervicalb 
 Mediastinal LNs 
 Retroperitoneal LNsb 
 MRI+ and CT− 
 Bilateral parotidsb 
 CT− 
 Bilateral submandibular glands 
14DiagnosisCT+ 
 Periaortic 
 Right primitive iliac artery 
 CT− 
 Left carotidian aneurysm 
15RelapseCT not available 
 Axillar, mediastinal, and bilateral iliac LNs 
 Bilateral kidneys 
 Pancreasb 
 Biliary tract 
 Prostate 
16DiagnosisCT+CT− and MRI+
 Bilateral lungsRPF
 Bilateral mediastinal LNs 
 Left axillar LNsb 
17DiagnosisCT+CT+
 Multiple LNsbKidney
 CT− 
 Right breastb 
18 CT+ 
  Multiple mesenteric LNs 
  Liverb 
  CT− 
  Subcarinal LNs 
  Right lung apex 
19DiagnosisCT− 
 Left carotid 
 Ascendant aorta 
 CT+ 
 Abdominal aorta 
 Celiomesenteric and lateroaortic LNs 
20DiagnosisCT+ 
 Bilateral kidneys 
 Submaxillary glandsb 
 Right parotidb 
 Pancreasb 
 Cervical, mediastinal, and hilar LNs 
21DiagnosisCT+ 
 Subclavicular and mediastinal LNs 
 CT− and MRI− 
 Bilateral kidneys 
 CT− 
 Bilateral parotids 

FDG-PET/CT evaluation for treatment response and disease activity in IgG4-RD

To evaluate the performance of FDG-PET/CT to assess the response to treatment, we compared results of FDG-PET/CT before and after treatment in 12 treatment sequences (11 patients: 3, 4, 7, 8, 9, 11, 14, 16, 18, 19, and 21) (Table 3). The median delay between initiation of treatment and the second FDG-PET/CT evaluation was 5.5 months (range 1–24 months). In one patient (patient 3), FDG-PET/CT was performed at month 1 because of clinical worsening despite treatment. Treatments given to patients for IgG4-RD are shown in Table 3. Complete normalization of 18F-FDG uptake was observed in 5 treatment sequences (5 patients) after a mean delay of 7.6 months (range 3–24 months). Patients initially presented nodal, arterial, lung, and kidney involvement (n = 2 each) and salivary gland or retroperitoneal involvement (n = 1 each). When available, serum IgG4 levels decreased under treatment in all of these patients and completely normalized in 2 patients. In 2 patients, 18F-FDG uptake was unchanged under treatment: in patient 9 (PET/CT at month 9 of treatment) for aortic and pleural involvement and in patient 16 (PET/CT at month 14 of treatment) for lung and nodal involvements. These 2 patients received rituximab alone as treatment for IgG4-RD. Despite stable 18F-FDG uptake on FDG-PET/CT, serum IgG4 levels decreased and normalized after treatment. These 2 patients were considered as treatment responders by the physicians. In 3 other patients (patient 8 at month 8, patient 11 at month 12, and patient 19 at month 5), control FDG-PET/CT findings after treatment gave conflicting results. Complete improvement of 18F-FDG uptake was noted for some organs (mesenteric inflammatory pseudotumor for patient 8, pancreas for patient 11, and aorta and carotid for patient 19) and was associated with either new sites showing 18F-FDG uptake (mediastinum and pulmonary hilum in patient 8 and lung nodules in patient 11) or an increase of SUVmax in some previous sites (lymph nodes in patient 19). In patient 19, a new pathologic analysis of a hypermetabolic lymph node confirmed typical IgG4-RD histology. At month 11 of treatment, a new FDG-PET/CT scan showed a complete disappearance of nodal 18F-FDG uptake. In patient 8, new mediastinal and bilateral hilar hypermetabolism was associated with a relapse a few months later with pulmonary involvement (Figure 1). In the 2 remaining patients (patients 3 and 18), an increase of 18F-FDG uptake despite treatment was observed. Patient 3 had a new FDG-PET/CT evaluation because of clinical and biologic worsening under treatment with prednisone and rituximab. This patient had a typical history of IgG4-RD with histologically documented pancreas involvement (type 1 AIP with fibrosis, lymphoplasmacytic infiltrate, and >10 IgG4-positive plasma cells per high-power field [hpf] diagnosed 10 years before), kidney involvement (tubulointerstitial nephritis with fibrosis, lymphoplasmacytic interstitial infiltration, and >50 IgG4-positive plasma cells per hpf with an IgG4-positive plasma cell:IgG-positive plasma cell ratio of 40%), and lymph node involvement (lymphoid hyperplasia with plasma cell infiltration, 80 IgG4-positive plasma cells per hpf, and a ratio of 50%). Lymphocyte and plasma cell infiltrates in these 3 tissues were polyclonal (polyclonal κ:λ light chain ratio). The new evaluation led to the diagnosis of a B cell lymphoma in this patient. In the other patient (patient 18), human herpesvirus 8–negative Castleman's disease was finally retained because of associated systemic symptoms and severe biologic inflammatory syndrome without response to steroids and other immunosuppressive treatments except to tocilizumab. In this case, pathologic examination of liver and lymph node biopsy samples was in accordance with pathologic criteria for IgG4-RD (storiform fibrosis with polyclonal lymphoplasmacytic infiltrate in the liver and biliary tract, follicular hyperplasia, interfollicular plasmacytosis, and Castleman's-like lymph nodes with polyclonal characteristics, and 100 IgG4-positive plasma cells per hpf).

Table 3. Results of FDG-PET/CT before and after treatment in 12 treatment sequences*
PatientIgG4, gm/literSites with abnormal 18F-FDG uptakeTreatment
BeforeAfterBeforeAfter
  1. FDG-PET/CT = 2-[18F]-fluoro-2-deoxy-d-glucose–positron emission tomography/computed tomography; LNs = lymph nodes; CS = corticosteroids; AZA = azathioprine; NA = not available; MTX = methotrexate; RTX = rituximab; IPT = inflammatory pseudotumor; SUV = standard uptake value; HT = hepatic transplantation; TAC = tacrolimus; MMF = mycophenolate mofetil.
447.350.98Kidney; cervical, submandibular, mediastinal, axillar, lomboaortic, iliac, and inguinal LNs; lungMo 3: 0 (physiologic fixation)CS
73.440.76RetroperitoneumMo 3: 0 (physiologic fixation)CS + AZA
8b15.73.14Bilateral lungsMo 5: 0 (physiologic fixation)CS + AZA
142.5NAPeriaortic and right iliac artery, left carotidian aneurysmMo 3: 0 (physiologic fixation)CS + MTX
2120.61.98Kidneys, subclavicular and mediastinal LNs, parotidsMo 24: 0 (physiologic fixation)CS + AZA
91.40.74Abdominal and thoracic aorta, axillar LNs, right pleuraMo 9: stable aortic and pleura hypermetabolismRTX
1617.50.56Bilateral lungs; bilateral, mediastinal, and left axillar LNsMo 14: stable lung and LN hypermetabolismRTX
8a134.79Mesenteric IPTMo 8: mesenteric IPT: 0, but new mediastinal and bilateral hilar hypermetabolismCS
110.523.49Parotids, corpocaudal pancreasMo 12: pancreas: 0, parotids: stable, lungs: new hypermetabolismCS + MTX
190.59NALeft carotid; ascendant and abdominal aorta; coronary stomachic, celiomesenteric, and lateroaortic LNsMo 5: decrease for abdominal aorta, increase for LN SUVCS
321.94.62Multiple LNs; pleura, lung, and kidneys; left parotidMo 1: SUVmax increase of all previous sitesCS + RTX
181.9NAMultiple mesenteric and subcarinal LNs, liver, right lung apexMo 6: mediastinal and mesenteric LNs, 2 intraperitoneal masses, diffuse osteomedullary, liver lesionsAfter HT, CS + TAC + MMF
Figure 1.

Representative 2-[18F]-fluoro-2-deoxy-d-glucose (18F-FDG)–positron emission tomography/computed tomography before and after treatment in 2 treatment sequences in patient 8. A, Mesenteric inflammatory pseudotumor with avid 18F-FDG uptake (arrow). B, At month 8 (corticosteroids stopped since 15 days), complete disappearance of mesenteric 18F-FDG uptake (arrow) and appearance of new right hilar (solid arrow), right mediastinal, and left hilar 18F-FDG uptake. C, At month 12, clinical relapse with bilateral lung 18F-FDG uptake (arrows). D, At month 17, under treatment with steroids and azathioprine, complete disappearance of pulmonary 18F-FDG uptake (arrows).

Comparison of FDG-PET/CT and conventional radiology

For 31 FDG-PET/CT evaluations, other radiologic analysis (US, CT, or MRI) realized at the same time was available for comparison. In 14 (45.2%) of 31 cases, FDG-PET/CT found the same number and sites of organ involvement as other radiologic procedures. In the other 17 cases (54.8%), the comparison of FDG-PET/CT with other radiologic findings gave conflicting results. Most of these (11 of 17) presented with abnormal 18F-FDG uptake without abnormal radiologic or clinical findings. The organs involved were the aorta (n = 2), large-size arteries (n = 2), parotids (n = 4), pulmonary hilar lymph nodes (n = 2), kidneys (n = 1), and breast (n = 1).

In 8 cases, organ involvement detected by radiologic examination either by MRI or CT was not associated with 18F-FDG uptake on PET/CT examination. These discrepancies were observed for a small-sized (16 × 6 mm) spinal IgG4-related pachymeningitis detected on MRI (patient 2) in a patient considered to have active disease; for lacrimal gland involvement in patient 11 despite bilateral lacrimal and orbital infiltration on CT and MRI in a patient considered to have active disease before treatment; for kidney involvement (patients 11 and 17) despite renal bilateral nodular infiltration on CT, where both patients had active disease before treatment; for lung involvement (patient 11), with few pulmonary micronodules noted on CT without 18F-FDG uptake; and for retroperitoneal fibrosis in 4 cases. In 3 of the 4 latter cases, the lack of 18F-FDG uptake despite radiologic abnormalities was associated with inactive disease after a treatment course. Inactive disease was retained by a physician in regard to global clinical, biologic, and radiologic findings at the time of FDG-PET/CT. For 1 patient, active disease was retained, but the patient presented with only a very localized periureteral infiltration on CT findings.

Using any abnormal finding from clinical examination or imaging tests as the reference standard, the sensitivity for the FDG-PET/CT and the CT to detect IgG4-RD organ involvement was 83% and 73%, respectively. Specificity could not be evaluated accurately in this study.

DISCUSSION

IgG4-RD is a polyclonal lymphoproliferative disease associated with tissue fibrosis and subsequent possible loss of organ function ([5, 25]). The cause and the epidemiology of this emerging disease are largely unknown. Diagnostic criteria have been proposed, and recently a consensus statement on pathologic characteristics of the disease was published ([5]). Treatment approaches are based on results of clinical studies in type 1 AIP (IgG4-related pancreatitis) ([27-29]) or in few case series of other IgG4-RD manifestations ([4, 30]). To assess the diagnosis and eliminate other inflammatory or malignant conditions, clinical, biologic, and pathologic findings are necessary. An increase in serum IgG4 of greater than 135 mg/dl can be observed in several other conditions and has not been shown to be specific to the disease ([9, 10, 31]). Therefore, it is important to assess typical clinical or radiologic organ involvement and to obtain pathologic analysis from a tissue biopsy. 67Ga scintigraphy has been reported to be useful in AIP, showing pancreatic, but also hilar and salivary gland, uptakes. 67Ga uptakes correlated with activity of the disease and IgG4 levels ([32]). In some case reports, it has been shown already that tissue infiltration by inflammatory cells and lymphoplasmacytes is associated with an increased uptake of 18F-FDG on PET/CT in IgG4-RD ([17-23]). Therefore, FDG-PET/CT could help the diagnosis by showing a characteristic pattern of organ involvement. However, this abnormal uptake is not specific and will not differentiate IgG4-RD from the diagnosis of malignant lymphoma, vasculitis, sarcoidosis, idiopathic retroperitoneal fibrosis ([33]), or other inflammatory conditions. For example, lymph node enlargement is very frequent in IgG4-RD and is associated with 18F-FDG uptake as in several malignant conditions. However, when facing the diagnosis of an abnormal pancreas, the pattern of extrapancreatic 18F-FDG uptake, especially on salivary glands, has been shown to help distinguish type 1 AIP (IgG4-related pancreatitis) from other pancreatic autoimmune or malignant diseases ([20]). Therefore, in some cases, the pattern of 18F-FDG uptake on PET/CT has been shown to support the diagnosis of IgG4-RD ([24]). In our patients, we could not always confirm by pathologic analysis that all of the sites associated with 18F-FDG uptake were truly associated with active disease. However, all of these patients presented the comprehensive diagnostic criteria of IgG4-RD with characteristic pathologic findings. Moreover, 18F-FDG uptake was noticed on usual locations of IgG4-RD such as lymph nodes, salivary and lacrimal glands, arteries, kidneys, pancreas, and retroperitoneum. All of the patients in the study presented with abnormal 18F-FDG uptake at the evaluation before treatment. This was noticed in more than one organ in 84% of cases. Therefore, our study confirms that FDG-PET/CT is helpful for the pretreatment evaluation of patients with IgG4-RD.

The staging of organ involvement is important in patients with IgG4-RD for several reasons. Most of the patients present with more than one organ involved ([4, 34]), and some of these organ involvements can be associated with poor clinical manifestations, for example, as in aortic or lymph node involvement. In such cases, clinical evaluation or conventional radiologic findings without whole-body acquisition can fail to detect them. When we compared FDG-PET/CT evaluation with conventional radiologic findings, conflicting results were present in more than half of cases. This was mostly related to abnormal 18F-FDG uptake on FDG-PET/CT without radiologic abnormalities. Moreover, this was observed on the aorta, large-size arteries, parotids, lymph nodes, and kidneys, which represent usual locations of IgG4-RD lymphoplasmacytic infiltration. These findings suggest that FDG-PET/CT could detect early involvement of IgG4-RD and is more sensitive than conventional radiology. The presence of involvement of multiple organs at diagnosis has an impact on treatment strategies because it is associated with an increased risk of relapse ([14, 19, 35, 36]). In addition to the precise staging of the disease at diagnosis, FDG-PET/CT uptake can be helpful to guide tissue biopsy for pathologic analysis ([22, 24]).

To date, the activity of the disease is evaluated by physicians on clinical, biologic, and radiologic evaluation without a well-recognized method to score the disease. IgG4 serum levels have been shown to vary with disease activity ([30, 31]), however, without a systematic clear correlation with evolution ([35]). Moreover, in some patients, serum IgG4 levels are normal at diagnosis ([11, 37]). An IgG4-RD responder index has been proposed recently ([38]), based on physician subjective evaluation of several IgG4-RD locations. In this setting, our results of FDG-PET/CT comparison before and after treatment during followup are very encouraging. In most cases (7 of 12), 18F-FDG uptake finally disappeared after 3–24 months (median 5 months) from treatment initiation. Two more patients were stable for 18F-FDG uptake without worsening, and were considered clinically and biologically as responders. Interestingly, these 2 patients were treated only by rituximab without corticosteroids. Only 2 patients clearly presented with an increase of 18F-FDG uptake on the second evaluation under treatment and in both cases, a diagnosis of a malignant lymphoproliferative disease was finally retained. From these results and previous reports of FDG-PET/CT normalization under treatment ([19, 23]), we suggest that patients with IgG4-RD should be evaluated at diagnosis and during followup by FDG-PET/CT together with clinical and biologic evaluation. It must be further evaluated if abnormal uptake after or during treatment is associated with an increased rate of early or long-term relapse after treatment discontinuation. Because abnormal 18F-FDG uptake is not specific for IgG4-RD and also can be observed in several other inflammatory, malignant, or even infectious disorders, these results always should be analyzed by keeping in mind these differential diagnoses and balanced by clinical, biologic, and other radiologic findings.

Conversely, the comparison with other radiologic findings in patients with IgG4-RD showed a lack of sensitivity of FDG-PET/CT in few patients. In some patients with active disease, abnormal findings on CT or MRI were not associated with 18F-FDG uptake on PET/CT. This was noted in some cases of small-sized pachymeningitis and small-sized lung nodules, a case of parotid involvement, another case with lacrimal gland involvement, and more surprisingly, in 2 cases of kidney involvement with nodular infiltration on a CT scan. These false-negative results could account for 2 principal causes: first, the spatial resolution of FDG-PET/CT with failure to detect millimetric lesions, and second, the physiologic fixation of 18F-FDG, especially in the brain and kidneys, with failure to detect lesions contiguous to these organs. However, this was a rare event, and in most other cases, a negative result of FDG-PET/CT with abnormal findings on CT or MRI was related to nonactive disease, especially in the context of retroperitoneal fibrosis that has already been documented ([33, 39]).

IgG4-RD relapses frequently after treatment discontinuation ([13, 35]). Relapses usually occur shortly ([13, 14]), but sometimes they occur after a very long period of time ([40]). The best strategy for the followup of individual patients with IgG4-RD has yet to be defined, but FDG-PET/CT combined with clinical and biologic evaluation can be proposed in regard to our retrospective study and previous reports ([19, 23, 24, 39]).

Finally, some authors have reported an increased risk of malignancies during IgG4-RD evolution ([41]). A significant incidence was found in our study, with 2 patients presenting a malignant lymphoproliferative disease. In this context, FDG-PET/CT could represent an important modality for the followup of these patients and detection of malignancy, especially in nonresponders to usual treatments.

In this retrospective study of 21 patients with IgG4-RD and 46 FDG-PET/CT evaluations, we showed that FDG-PET/CT is a very interesting procedure at diagnosis and during followup of the disease. At diagnosis, 18F-FDG uptake is observed in involved organs with the advantage of whole-body functional evaluation. 18F-FDG uptake is in most cases correlated with the activity of the disease and improves after treatment. Our results also suggest that FDG-PET/CT is more sensitive than conventional radiologic imaging to detect IgG4-RD involvement. More prospective studies are needed in this rare disease to better assess the sensitivity and the specificity of this imaging procedure and to define the optimal disease activity score.

AUTHOR CONTRIBUTIONS

All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Ebbo had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study conception and design. Ebbo, Schleinitz.

Acquisition of data. Ebbo, Grados, Guedj, Gobert, Colavolpe, Zaidan, Masseau, Bernard, Berthelot, Morel, Lifermann, Palat, Haroche, Mariette, Godeau, Bernit, Costedoat-Chalumeau, Papo, Hamidou, Harlé, Schleinitz.

Analysis and interpretation of data. Ebbo, Grados, Guedj, Gobert, Colavolpe, Haroche, Schleinitz.

Acknowledgments

The authors would like to thank Professor Laurent Daniel, Dr. Bruno Chetaille, Professor Jean-François Mosnier, and Dr. Christiane Copie for reviewing histologic specimens of the patients.

Ancillary