Sarcoidosis is a granulomatous disease of unknown etiology affecting the lower respiratory tract in 90% of cases1,21. Involvement of the upper respiratory tract is admittedly uncommon and may assume various features4,22. Sinonasal involvement in sarcoidosis (SNS) has been estimated at between 0.7% and 6%9,17,18,24,25 but its actual incidence remains unclear for several reasons. First, in patients with known sarcoidosis, symptoms of SNS are nonspecific and may be caused by nonsarcoid rhinosinusitis18. Second, most physicians are not aware of such a manifestation of sarcoidosis and do not perform a proper local examination. Finally, in patients with previously unknown disease the diagnosis of isolated SNS is difficult since it may mimic other sinonasal (SN) granulomatous disorders including Wegener granulomatosis, tuberculosis, syphilis, and lepromatous leprosy. Indeed, deShazo and colleagues reviewed 50 reports of so-called SNS in the medical literature and identified only 9 patients who fulfilled strict diagnostic criteria5.
Although several case reports or small series have described SNS features, few studies have addressed the specific issues raised by SN involvement in the context of sarcoidosis. Some authors have emphasized that SN localization is an indicator of chronic sarcoidosis and is usually a recalcitrant form of the disease that requires aggressive systemic therapy3,11,12,24,25. Yet there is no published case-control study comparing sarcoid patients with and without SN involvement.
We present a retrospective single-center series of 20 consecutive patients with SNS. The aims of the study were 1) to describe the clinical and CT features of SNS, and 2) to determine whether the involvement of the SN tract is associated with a particular clinical phenotype of sarcoidosis in terms of demographics, respiratory and extrarespiratory manifestations, severity, and clinical course.
The study was retrospective and received institutional review board approval. The Department of Pneumology of Avicenne Hospital is highly specialized in sarcoidosis and has a large historical cohort. The patients were searched from a database in which are noted the main characteristics of the sarcoid patients, including the presence of SN involvement either at first admittance or during regular monitoring. Twenty patients with SNS were consecutively recruited between October 1993 and September 2000. Complete information was obtained from hospital and referring physician medical records and retrospectively evaluated by 2 of the authors (HN and FCA). Patients were included only if they met 3 criteria: 1) clinical and radiographic features compatible with a diagnosis of sarcoidosis, 2) histologic confirmation of noncaseating granuloma in the nasal or sinus tissue, and 3) exclusion of other causes of SN granulomatosis supported by histology of nasal and sinus tissue biopsies negative for fungus (using Grocott stain) and acid-fast bacilli, and no clinical evidence of another disorder associated with SN granulomatosis including Wegener granulomatosis, tuberculosis, aspergillosis, syphilis, and lepromatous leprosy.
Clinical, Laboratory, and Radiographic Investigations
At the time of SNS diagnosis, patients underwent physical examination including local examination of upper respiratory tract with nasal endoscopy (conducted by one of the authors [MO]), chest X-ray, pulmonary function tests, and serum angiotensin-converting enzyme (SACE) and antineutrophil cytoplasmic antibodies (ANCA) measured by ELISA. Most, but not all, had computed tomography (CT) of the paranasal sinuses. Chest radiographs were classified according to the recent statement on sarcoidosis: 0: normal; I: bilateral hilar lymphadenopathy with normal lung parenchyma; II: bilateral hilar lymphadenopathy with pulmonary infiltrates; III: pulmonary infiltrates without hilar lymphadenopathy; IV: pulmonary fibrosis/fibrocystic parenchymal changes7. Sinus CT scans were evaluated by 2 independent radiologists (PA, MB) who were blinded to patient information and dates of the scans. Sinus CT scans were graded using the Lund and Mackay scoring system13.
Patients with SNS were compared with control subjects who had sarcoidosis but no SN involvement. Sarcoid controls were part of the historical cohort of the pneumology department of Avicenne Hospital. Each patient with SNS was matched with 2 controls for the date of admittance in our institution. In our institution SN symptoms are systematically searched with attention. Among the 40 controls, 4 had SN symptoms (stuffiness: n = 1; stuffiness and rhinorrhea: n = 2; epistaxis: n = 1) but negative nasal biopsy.
Assessment of Outcome
Patients with SNS and sarcoid controls were managed with regular assessment, at time intervals depending on their clinical status. Recovery was defined by the complete resolution of all sarcoidosis symptoms and no disease recurrence for at least 6 months without any treatment.
Analysis was performed using the Statview version 5.0 statistical package (SAS Institute, Cary, NC). Results are expressed as mean ± standard deviation or median (range). Groups were compared by unpaired Student t test, Mann-Whitney test, or chi-square test when appropriate. The probability of recovery from sarcoidosis was estimated by the Kaplan-Meier method and compared by the logrank test.
Among the 1267 patients with sarcoidosis seen in our center between October 1993 and September 2000, 20 patients fulfilled the inclusion criteria, which provided an estimation of SN involvement at 1.6%.
The study population included 7 men and 13 women with a mean age of 32 ± 9 years. Nine patients were black and 11 were white. One patient with SNS was exposed to wood dust and 1 to asbestos. There was no case of familial sarcoidosis. Time interval between the first SN symptoms and the diagnosis of sarcoidosis varied widely, ranging from 9 years before to 13 years after. SN symptoms occurred in the course of prior sarcoidosis in 8 patients (40%), whereas they preceded disease diagnosis in 12 patients (60%). Among these 12 patients, 4 initially presented with strictly isolated SN symptoms and 8 had other associated signs related to sarcoidosis. The mean delay between the first SN symptoms and the histologic confirmation of SNS was 3.8 years (±2.57 yr) in the whole population and was particularly lengthy in the cases with previously unknown sarcoidosis (5.7 ± 2.51 yr). Histologic diagnosis of SNS was made based on endoscopic nasal or sinus mucosa biopsy in 13 patients (65%) and on surgical specimen in 7 patients (35%).
Clinical features of patients with SNS are shown in Table 1. At presentation, the most common SN symptoms were stuffiness (defined as a subjective feeling of decreased airflow) (90%), anosmia (70%), and rhinorrhea (70%). Epistaxis was present in only 30%. Local examination of upper respiratory tract was constantly abnormal. Main changes in the nasal mucosa consisted of hypertrophy (75%) and purplish coloring with pale yellowish nodules (50%) (Figure 1), which were mostly seen on the septum or inferior turbinates. Septal perforation was noted in only 1 case at the time of SNS diagnosis but developed in 2 other patients in the course of the disease. A disfiguring saddle nose deformity was observed in 2 patients.
Ninety percent of patients with SNS had extrarespiratory manifestations during follow-up, two-thirds of which were severe. The most frequent severe extrarespiratory localizations were central nervous system (CNS) (n = 3), eye (n = 3), chronic intrahepatic cholestasis (n = 3), heart (n = 2), and hypercalcemia (n = 1). CNS involvement was pituitary lesion with diabetes insipidus in 1 case, encephalic space-occupying mass and retrobulbar optic neuritis in 1 case, and both meningoencephalic and hypothalamic sarcoidosis in the remaining 1 case. Lupus pernio was associated with SNS in 10 patients (50%). Initial chest X-ray was normal in 4 patients.
Sinus CT and Laboratory Findings
CT scans were not performed systematically in patients with SNS, particularly during the first 2 years of the study. Patients who had been investigated with CT did not differ significantly from patients without CT in terms of gender, age, stuffiness, anterior or posterior rhinorrhea, anosmia, crusting, epistaxis, facial pain, mucosal hypertrophy, purplish mucosa with granulations, nasal polyps, or septal perforation. Fifteen patients underwent CT scan. The sarcoid lesions identified with CT are summarized in Table 2. All tested patients had evidence of sinusal involvement, mucosal hyperplasia, or filling. Maxillary and anterior ethmoidal sinus involvement was the most frequent asconfirmed by the Lund and Mackay scoring system. Multiple small mucosal nodules from 2 to 3 mm of size were observed on the septum and the turbinates in 3 cases (20%). Bone perforation or lyses was observed in 7 cases (47%), mainly on the nasal septum (n = 6) and the inferior turbinates (n = 4) but also on the hard palate (n = 2) and the nasal bone (n = 1) (Figure 2).
SACE was measured in 19 patients and was increased at presentation in most of them (73.7%). Yet, only 1 of the 4 patients with initially strictly isolated SN symptoms had elevated SACE. All included patients had negative ANCA.
Clinical Course and Treatment
Patients with SNS were followed for a median of 9.2 years (range, 1.2-38.1 yr) after the diagnosis of sarcoidosis and 7.3 years (range, 0.1-19 yr) after the diagnosis of SNS. At the end of follow-up, 2 patients had recovered from all sarcoidosis symptoms. No patient died during follow-up.
Local treatment including alkaline nasal douche and topical corticosteroids never achieved satisfactory control of SN symptoms. Corticosteroids were never administered by intranasal injection. All patients with SNS needed systemic treatment because of SN symptoms alone (n = 3, 15%) or in association with another sarcoidosis localization (n = 17, 85%). Two patients received hydroxychloroquine as the sole therapy: in 1 case time observation was short (1 mo) and in the other therapy was continued until last visit (45 mo) because of SN symptoms. Eighteen patients (90%) received oral corticosteroids. Patients always required high doses (between 30 and 60 mg of prednisone daily) of corticosteroids to control SN symptoms and long-term treatment for a median time of 82 months (range, 22-161 mo). Prednisone maintenance dosage was 10.5 ± 6 mg daily. It depended exclusively on SN involvement in 7 patients (38%), on SN involvement in addition to another localization in 6 patients (34%), and on another localization while SN was controlled in 5 patients (28%). As a result, corticosteroid side effects were frequent (50% of treated patients). In 13 cases (72%), corticosteroids were associated with at least 1 other drug that was used as a corticosteroid-sparing treatment, because of steroid-resistant sarcoidosis, and/or because of severe corticosteroid side effects (hydroxychloroquine: n = 12, methotrexate: n = 8, thalidomide: n = 4, cyclophosphamide: n = 2, azathioprine: n = 2, pentoxifylline: n = 4, cyclosporine: n = 1, and colchicine: n = 1).
Seven patients underwent SN surgery: 3 patients before the diagnosis of SNS and the remaining 4 after. SNS recurred in all patients, and SN symptoms worsened after surgery in 1 case. One additional patient underwent plastic surgery of the nose while SNS was inactive with a long-term therapy of 20 mg daily of prednisone, with good result.
Comparing Patients with SNS and Sarcoid Controls
Results are summarized in Tables 3 and 4. Groups were statistically similar in terms of demographics, smoking habits, and occupational or environmental exposure, although black patients tended to be more frequent in the SNS group.
Clinical features were significantly different between the 2 groups for the presence of extrarespiratory localization, their number, and their severity (see Table 3). The frequency of each extrarespiratory manifestation was not significantly different between the 2 groups except lupus pernio, which was higher in patients with SNS (50% vs. 0%, p < 0.001). Chest X-rays, pulmonary function tests, and SACE were statistically similar between the 2 groups (see Table 3).
Comparisons between the SNS group and the control group regarding treatment revealed a difference in the rate of patients requiring systemic treatment (100% vs. 57.7%, respectively, p < 0.001) and in the duration of treatment (82 mo [range, 22-161 mo] vs. 22 mo [range, 14-73 mo], respectively, p < 0.0001). The probability of recovering from sarcoidosis was significantly lower in patients with SNS compared with controls (0% vs. 51.7% at 5 yr, 6.2% vs. 55.7% at 10 yr, and 21.9% vs. 66.8% at 15 yr, p = 0.0006).
To our knowledge, the present study is the first case-control series comparing patients with SNS to patients without SN involvement. The results suggest that SN involvement is a recalcitrant manifestation of sarcoidosis that is associated with more frequent and severe involvement of vital organs, with a particularly long-standing course. However, after prolonged and heavy systemic treatment, our patients were stabilized and none died from sarcoidosis or related complications.
Although admittedly a rare localization of sarcoidosis, the actual frequency of SNS is difficult to appraise accurately. Published data differ depending on whether SNS is defined according to symptoms, nasal endoscopy, sinus CT, or histopathology. Histologically proven SNS has been estimated between 0.7% and 6%9,17,18,24,25. On the basis of strict diagnostic criteria similar to those proposed by deShazo et al5, we found a prevalence of 1.6%.
In our experience as well as in the literature17,20, SN involvement occurred either before systemic involvement or during the course of previous sarcoidosis within a wide time interval, but symptoms were mostly present at the onset of sarcoidosis in patients with florid disease. Yet, in 20% of our patients the diagnosis of sarcoidosis was delayed because SN symptoms were initially isolated. In such cases with isolated chronic rhinosinusitis unresponsive to local treatment, biopsy should be systematically performed since SACE is rarely increased and local examination and sinus CT are nonspecific, although suggestive. Of course, the measure of ANCA, which was always negative in our patients, is helpful to rule out the possibility of Wegener granulomatosis, the trickiest differential diagnosis of SNS.
SN involvement may vary from an almost quiet to a very disabling manifestation of sarcoidosis. Symptoms of SNS are nonspecific but anosmia was particularly frequent in our population, affecting 70% of cases, which has not been emphasized in previous series3,5,6,17,24. We found a high percentage of purulent rhinorrhea, as did Wilson et al24. Nasal endoscopy was always abnormal in our population and suggestive although not specific. The most frequent pattern was hypertrophic or purplish coloring of the mucosa with pale yellowish nodules, which are believed to represent the sarcoid granulomas. As noted by other authors, the lesions predominated on the septum and inferior turbinates6,24.
SNS CT has been rarely described in the literature2,3,5,12. Three groups of abnormalities have been identified at CT by Bourjat and colleagues2: 1) nodules, mucosal hyperplasia, and filling; 2) lysis; and 3) atrophy. The nodules were considered the most evocative sign of sarcoidosis2. They are about 3 mm, numerous, localized on the nasal septum and the lower turbinates, and clearly visible on the axial slices. However, they were rare in the current series. The most characteristic lesions were medial osteo and chondrolytic lesions: septal perforation, atrophy of the skeleton of the lower turbinates, lysis of the nasal bones. The lytic lesions were always associated with a total or subtotal filling of the majority of the sinuses. They may reflect advanced irreversible SN lesions.
Demographic features and respiratory involvement did not differ significantly between sarcoid patients with or without SN involvement. There was no history of exposure to aero-contaminants that could have favored SN localization, and histopathology never revealed a foreign body-type reaction within granulomas. Among extrarespiratory localizations we found a highly positive association between SNS and lupus pernio, which was present in 50% of our cases. Lupus pernio is a chronic persistent violaceous skin lesion with a predilection for the nose, cheeks, and ears. This manifestation has been recognized as an indicator of chronic multisystem sarcoidosis by James and colleagues8. The association between SNS and lupus pernio has been described previously8,20,24 with a high frequency, similar to that in the current study. Neville et al20 reported a series of 34 patients in which 17 had evidence of upper respiratory tract sarcoidosis and 26 had lupus pernio. In 9 patients, these features coexisted. The pathogeny of such an association remains unclear but may result from contiguous extension of granulomas. Although the CNS was affected in 15% of our patients with SNS, the difference compared with controls did not reach the level of significance, perhaps because of the small number of patients. Indeed, it has been suggested that CNS sarcoidosis may sometimes be a consequence of perineural spread from SNS16,19, which involves areas of exceptional vascularity such as the hypothalamus, optic nerve, and pituitary gland14. It is noteworthy that these areas were the site of sarcoidosis in our 3 cases with SNS and CNS involvement. Another explanation may be the common embryogenic origin of both pituitary gland and SN mucosa.
Results of the current study clearly show that patients with SNS have a more severe illness in terms of number and severity of extrarespiratory localizations. This probably explains the poorer outcome of these patients compared with sarcoid controls, and their significantly lower probability of remission and higher requirement of systemic treatment. These findings have been mentioned by other authors1,3,11,12,23-25, but to our knowledge this is the first case-control study to ascertain them.
Although SNS is known to be a difficult problem in sarcoidosis, its therapeutic management has never been clearly codified. Krespi et al proposed a staging system to categorize the severity and sites of involvement and to guide the aggressiveness of therapy11. This approach is interesting but limited by 2 issues: first, the patients were recruited with SN symptoms as their primary manifestation of sarcoidosis and the extent of systemic disease was not well documented; second, the period of follow-up was rather short (mean of 20 mo). In our experience, local treatment was always inadequate; systemic therapy was always required, often with a high dosage of corticosteroids. The opportunity of sparing treatments should be discussed precociously, since patients with SNS have a chronic course of sarcoidosis and need prolonged therapy, which frequently leads to corticosteroid adverse effects. Importantly, the threshold level of corticosteroids to control sarcoidosis was dictated by SN involvement in 72% of our patients, while a complete resolution of SN symptoms was obtained in only 28% of them.
Surgery for SNS is controversial. Short-term results may help reduce symptoms, but surgery alone does not eradicate the disease or prevent recurrence15. Moreover, several authors have emphasized that the surgical procedure may be dangerous. Neville et al20 noted that submucous resection may provoke nasal septal or palatal perforation. Accordingly, 1 of our patients who underwent surgery experienced severe intra- or postoperative complications. Taken together, these results suggest that surgery should be avoided in patients with active SNS. However, 2 recent studies10,12 advocated the role of endoscopic sinus surgery in patients with nasal obstruction or chronic sinusitis due to anatomic blockage from SNS. In such a context, surgery may allow for improved SN hygiene through office endoscopic debridement, irrigation, and topical application of medicines.
Limitations of the current study are as follows: 1) this is a single-center and retrospective analysis with possible biases including the estimated prevalence of SN involvement and the validity of the control group. However, all patients with sarcoidosis seen in our department are systematically asked about the presence of SN symptoms at each visit; patients with SN symptoms undergo endoscopic examination by an experienced physician (MO) and, if abnormal, guided biopsies of the nasal or sinus mucosa. 2) We focused only on patients with histologically proven SN involvement, and therefore we may have selected patients with a more severe disease. In fact, some authors have suggested that SN mucosal biopsy may be negative in patients with authentic SNS6,12,24, and that patients with a positive biopsy had a worse prognosis12. However, we believe that the best and only reliable diagnostic criterion for SNS is histologic evidence, since there is no clinical or radiologic-specific sign of SNS.
We conclude that, although uncommon, SNS is associated with a local and general severity of sarcoidosis including disease dissemination, serious localizations, and absence of recovery. It is a recalcitrant localization that mostly requires high doses of systemic corticosteroids and represents a difficult therapeutic challenge. However, after a long and difficult period with systemic treatments, patients may finally be controlled. Further multi-center, prospective, observational studies are needed to evaluate the efficacy of specific medications such as methotrexate or thalidomide and the role of surgery.
1. Baughman RP, Lower EE, du Bois RM. Sarcoidosis. Lancet
2. Bourjat P, Braun J. [Sinonasal sarcoidosis: CT evaluation]. J Radiol
3. Braun JJ, Gentine A, Pauli G. Sinonasal sarcoidosis: review and report of fifteen cases. Laryngoscope
4. Dash GI, Kimmelman CP. Head and neck manifestations of sarcoidosis. Laryngoscope
5. deShazo RD, O'Brien MM, Justice WK, Pitcock J. Diagnostic criteria for sarcoidosis of the sinuses. J Allergy Clin Immunol
6. Fergie N, Jones NS, Havlat MF. The nasal manifestations of sarcoidosis: a review and report of eight cases. J Laryngol Otol
7. Hunninghake GW, Costabel U, Ando M, Baughman R, Cordier JF, du Bois R, Eklund A, Kitaichi M, Lynch J, Rizzato G, Rose C, Selroos O, Semenzato G, Sharma OP. ATS/ERS/WASOG statement on sarcoidosis. American Thoracic Society/European Respiratory Society/World Association of Sarcoidosis and Other Granulomatous Disorders. Sarcoidosis Vasc Diffuse Lung Dis
8. James DG. Dermatological aspects of sarcoidosis. Q J Med
9. James DG, Barter S, Jash D, MacKinnon DM, Carstairs LS. Sarcoidosis of the upper respiratory tract (SURT). J Laryngol Otol
10. Kay DJ, Har-El G. The role of endoscopic sinus surgery in chronic sinonasal sarcoidosis. Am J Rhinol
11. Krespi YP, Kuriloff DB, Aner M. Sarcoidosis of the sinonasal tract: a new staging system. Otolaryngol Head Neck Surg
12. Long CM, Smith TL, Loehrl TA, Komorowski RA, Toohill RJ. Sinonasal disease in patients with sarcoidosis. Am J Rhinol
13. Lund VJ, Kennedy DW. Staging for rhinosinusitis. Otolaryngol Head Neck Surg
14. Maillard AA, Goepfert H. Nasal and paranasal sarcoidosis. Arch Otolaryngol
15. Marks SC, Goodman RS. Surgical management of nasal and sinus sarcoidosis. Otolaryngol Head Neck Surg
16. Mazziotti S, Gaeta M, Blandino A, Vinci S, Pandolfo I. Perineural spread in a case of sinonasal sarcoidosis: case report. AJNR Am J Neuroradiol
17. McCaffrey TV, McDonald TJ. Sarcoidosis of the nose and paranasal sinuses. Laryngoscope
18. Meybeck A, Just N, Heurtebise F, Chanez P, Crampette L, Darras J, Wallaert B. [Involvement of the nasal sinuses in sarcoidosis. A prospective study of 63 patients]. Rev Mal Respir
19. Milgrim LM, Cohen AF. Nonpulmonary sinonasal sarcoidosis with intracranial invasion. Otolaryngol Head Neck Surg
20. Neville E, Mills RG, Jash DK, Mackinnon DM, Carstairs LS, James DG. Sarcoidosis of the upper respiratory tract and its association with lupus pernio. Thorax
21. Newman LS, Rose CS, Maier LA. Sarcoidosis. N Engl J Med
22. Shah UK, White JA, Gooey JE, Hybels RL. Otolaryngologic manifestations of sarcoidosis: presentation and diagnosis. Laryngoscope
23. Takada K, Ina Y, Noda M, Sato T, Yamamoto M, Morishita M. The clinical course and prognosis of patients with severe, moderate or mild sarcoidosis. J Clin Epidemiol
24. Wilson R, Lund V, Sweatman M, Mackay IS, Mitchell DN. Upper respiratory tract involvement in sarcoidosis and its management. Eur Respir J
25. Zeitlin JF, Tami TA, Baughman R, Winget D. Nasal and sinus manifestations of sarcoidosis. Am J Rhinol