Key Words. Sarcoidosis. Prognosis. Spirometry. Chest radiograph. Dyspnea.

Received: 17 February 2003
Accepted After Revision: 3 July 2003
Correspondence and Reprints: Marc A. Judson, M.D.
Professor of Medicine
Division of Pulmonary and Critical Care Medicine
96 Johnathan Lucas Street,
Charleston, SC 29475
Tel: (843) 792-3162.
Fax: (843) 792-0732
Email: judsonma@MUSC.edu

Sarcoidosis is a multi-system granulomatous disease of unknown cause. The disease has a variable clinical course from an asymptomatic state to a progressive life- threatening illness [1]. The prognosis of Sarcoidosis is affected by clinical [2-4], radiographic [5-8], genetic [9], and racial [4] factors. Corticosteroid therapy may also affect the prognosis of Sarcoidosis [6,10,11].

The National Heart, Lung, and Blood Institutes’ - sponsored study entitled A Case Control Etiologic Study of Sarcoidosis (ACCESS) examined patients with Sarcoidosis at 10 centers within the United States [12]. A cohort of Sarcoidosis patients who were enrolled in ACCESS was reexamined two years after study entry to assess their clinical status. This report describes the two-year prognosis of the follow-up ACCESS cohort.

Methods

We reassessed a cohort of 215 consecutive ACCESS cases 2 years after initial enrollment. Because clinical data was collected in the ACCESS study for only 3 years, two-year follow-up data could only be obtained from subjects enrolled in the first year of ACCESS, and therefore only these subjects were eligible for enrollment. All cases enrolled were required to have had a tissue biopsy demonstrating noncaseating granulomas within six months of enrollment. We excluded known causes of granulomatous inflammation by demonstration of negative stains of the biopsy specimens for mycobacterial and fungal organisms and negative polarized light examination for particulate matter. As part of ACCESS, all cases had data collected concerning their demographic characteristics; socioeconomic status, spirometry, chest radiograph results, extent of extrapulmonary involvement with Sarcoidosis [13], and a 5 point dyspnea scale modified from Watters et al [14]. Chest radiographs were classified by Scadding stages: 0- normal chest radiograph, 1- bilateral hilar adenopathy with normal lung parenchyma, 2- bilateral hilar adenopathy with pulmonary infiltrates, 3- pulmonary infiltrates without hilar adenopathy, 4-  pulmonary fibrosis/fibrocystic parenchymal changes [15].

For each patient in the cohort, we defined "improved" and "worse" at the follow-up measurement as listed in Table 1. Measurements that did not meet criteria for improvement or worsening were defined as "unchanged." The following characteristics were evaluated that might confound the prediction of outcome: ever smoked (yes, no), currently smoking (yes, no), annual family income ($5,000-  $9,999, $10,000 - $19,999, $20,000 - $29,999, $30,000 - $39,999, $40,000 $49,000, >$50,000), erythema nodosum (yes, no), splenic involvement (yes, no), use of corticosteroids at ACCESS entry (yes, no), use of corticosteroids at 2 year follow-up (yes, no), health insurance limited (yes, no), could not get health care (yes, no), seen by a specialist (yes, no), missed appointments (yes, no), takes medication (all of the time, not all of the time), could not get medication (yes, no), feels medication can control Sarcoidosis (very confident, somewhat confident, not confident, not taking medication).

Statistical Methods: The statistical elements of change were categorized into a trichotomous outcome representing the patients' condition as: improved, unchanged, or worsened. These categorical outcomes were compared to categorical demographic and clinical characteristics that represented the patients' initial state. The distribution of improvement is presented using point estimates of the three percentages representing the categories of improved, unchanged, and worsened. All comparisons were accomplished using p- values from Chi Square Statistic. Critical p- values were considered 0.05 and were reported from the SAS statistical software.

Results

A total of 251 cases of the 740 ACCESS cases were enrolled in the first year of ACCESS and therefore asked to return for the two-year follow-up visit and gave their consent. Three of these cases were determined to be ineligible when the eligibility of all cases was reviewed during the second year of ACCESS enrollment. Of the remaining 248 cases, 215 completed the two-year follow-up visit and are the subject of this report. There were no significant demographic differences between the patients who returned for the two-year follow-up and those who did not (data not shown).

Table 1: Definitions of change in clinical status compared to enrollment

FVC: Forced vital capacity. FEV1: Forced expiratory volume in 1 second, CXR: Chest radiograph., *Neither improved or worse

Table II lists the demographic data concerning the cohort. Of the 215 patients, all had repeat evaluations of organ involvement and completed another dyspnea scale. One hundred and ninety-three had repeat spirometry, and 205 had follow-up chest radiographs. The mean length of time between ACCESS enrollment and the follow-up visit was 25.2 + 0.1 months with a median of 25 months.

Table II: Patient characteristics, N=215

Table III shows the change in the clinical parameters of FVC, FEV1, Scadding stage of chest radiograph, and dyspnea scale between ACCESS enrollment and the two-year follow-up. For each measurement, approximately twice as many patients had improvement as had worsening. FVC, FEV1, Scadding stage, and the dyspnea scale remained unchanged over the two-year period in the majority of the patients.
A total of 11 of 176 (6%) with Scadding stage 0, 1, or 2 chest radiographs progressed to stage 3 or 4 over the two-year follow-up period (data not shown).

Table IV shows the relationship of the change in FVC to the change in FEV1 over the two-year follow-up. Changes in FVC and FEV1 were discordant (better vs worse) in no patients and were concordant in 80.3% (155/193).

There was no effect of gender or age on FVC, FEV1 Scadding stage or dyspnea scale (data not shown), but African Americans had less improvement in FVC compared to Caucasians (Table V). Patients with worsening dyspnea over the two years were more likely to have one or more new organs involved with Sarcoidosis (p = 0.013) (Table VI).

Several patient characteristics (see methods section) were analyzed and some of these were associated with prognosis (Table Vll). Patients with erythema nodosum at presentation were more likely

Table IV: Comparison of FEV1 outcomes to FVC outcomes

to have improvement in the chest radiograph at follow-up (p = 0.007). Patients with a lower annual family income were more likely to worsen with respect to dyspnea (p = 0.01) and more likely to have new organ involvement at follow-up (p= 0.045).

The relationships of corticosteroid use to FVC and dyspnea were complicated. Patients taking corticosteroids at follow-up tended to be improved or worsened with respect to FVC and dyspnea more often than those not taking corticosteroids (Table VII). Ninety-five out of 205 (46%) patients were not taking corticosteroids at enrollment and 75 of these 95 (79%) never received corticosteroids over the two-year follow-up. One hundred and ten of the 205 (54%) patients were receiving corticosteroids at enrollment and 52 (47%) were maintained on corticosteroids throughout the two years of follow-up.

Fifty of the 215, (23% a) of patients, developed one or more new organs involved with Sarcoidosis over the two-year follow-up period. These fifty patients developed a total of 67 new organs involved with Sarcoidosis (Table VIII). The skin was the most frequent new organ involved with Sarcoidosis over the two-year follow up. Table IX shows the development of new organ involvement in terms of race, sex, and age. No consistent patterns are observed, although African Americans may have had more frequent development of skin involvement than Caucasians.

Table V: Change in disease status by race

•p value from a Chi square or F-test statistic comparing

distributions among the 2 groups.

Table X shows that development of new organ involvement in the follow-up period was more common in African Americans compared to Caucasians and in those with extra-thoracic organ involvement at baseline. However, development of new organ involvement was not affected by gender or age.

Discussion

The prognosis of Sarcoidosis is variable and is difficult to predict in an individual patient. Several studies have found associations between various clinical, biochemical, and genetic factors and the prognosis of Sarcoidosis. Most of these studies have involved relatively homogeneous populations of one racial or ethnic group. In ACCESS, we examined a relative large population of Sarcoidosis patients in the United States with a diverse mix of socioeconomic status and ages, an adequate representation of both genders, and the Caucasian and African American races [16]. The cohort of 215 ACCESS patients who underwent two-year follow-up was similarly diverse.

Although end-stage pulmonary Sarcoidosis usually develops over one or two decades [17], this assessment of the clinical status of the patient at two years has important prognostic implications. In one

Table VI: Relationship or changes in clinical measurements

cohort of 193 Sarcoidosis patients [2], only 22% had persistent disease two years after diagnosis, and only an additional 5% of this cohort had disease resolution over the subsequent three years. Therefore most patients with persistent disease at two years were unlikely to have resolution of Sarcoidosis.

We found that more than 80% of patients with Sarcoidosis have improvement or stability of spirometry, chest radiographic findings, and dyspnea over the two-year period after diagnosis.
The resolution was not statistically related to whether the patients did or did not receive therapy. These data are similar to other studies in homogeneous populations of Sarcoidosis patients. Romer [5] found that only 13% (27/210) of Danish Sarcoidosis patients had radiographic progression over a mean of 5.5 years. Other authors have noted a similar good prognosis. Chappell and coworkers [81] showed that 75% (112/150) of patients with Sarcoidosis referred to a single medical consultant in New South Wales had resolution of stage I or 11 chest radiographs over a two-year period. Nagai and colleagues [18] showed that 68% of chest radiographs of Japanese Sarcoidosis patients cleared within 3 years. Mana and coworkers [2] found that only 22%a (35/193) had persistence of Sarcoidosis at two years as assessed by clinical features, chest radiographic findings, spirometric

Table VII: Patient characteristics associated with clinical measurements

changes, gallium scanning, and serum Angiotensin Converting Enzyme levels. One limitation of this radiographic analysis is that only a small number of patients progressed to Scadding stage 3 or 4 chest radiographs that are associated with a worse symptoms and prognosis. However this was to be expected, as the length of follow-up was limited to two years. Another limitation of the ACCESS analysis of chest radiographs is that the initial and follow-up chest radiographs were not directly compared. It is possible that significant chest radiographic changes may occur without a change in Scadding stage. However, we suspect that they would have observed frank worsening of a significant number of chest radiographs by Scadding stage if the radiographic findings of this cohort truly worsened.

Data concerning the natural course of spirometry and pulmonary symptoms in Sarcoidosis is sparse. Similar to our patients, more than 70% of Japanese (18, 19), and Spanish [2] patients had resolution of pulmonary symptoms within two years. This is in contrast to a study of Finnish patients [19] in which more than 75% had persistent symptoms 2 years after diagnosis. Our data concerning spirometry are also similar to studies of more homogeneous populations. We found that 88% of our subjects had


Table VIII: Organ involvement for all patients who were seen at follow-up, N = 215

improved or stable spirometry after two years, which is similar to the finding of 18% (29/160) Spanish Sarcoidosis patients having worsening or a persistently abnormal FVC two years after diagnosis [20].

We found that African Americans had worse outcomes in FVC over two years compared to Caucasians. There were no other racial, gender, or age (< 40 versus > 40 years) differences in terms of FEV1, chest radiographic findings, and level of dyspnea. African American patients are known to have a worse prognosis of Sarcoidosis compared to Caucasians (more frequent relapse, more extrapulmonary involvement, worse lung perfusion on chest radiograph) [21,22], although there are limited data specifically concerning the outcome of spirometry. Johns and colleagues [23] found that FVC was decreased in African American Sarcoidosis patients compared to Caucasians followed an average of 5 years after diagnosis, but no analysis of the change in spirometry was performed.

We found that African Americans were more likely to develop new organ involvement with Sarcoidosis over two years compared to Caucasians. This is consistent with a previous study [23] that found that African Americans had more extra-thoracic disease than Caucasians at 5 years, although the time of onset of extra-thoracic involvement was not reported. We also found that extra-thoracic organ involvement at baseline (within 6 months of diagnosis) was associated with an increased likelihood of new organ involvement over two years. New organ involvement was not affected by gender or age. We found no significant patterns of new organ involvement, but this may have been because of inadequate sample size. It is important to point out that organ involvement is different than severity of involvement. ACCESS did not assess the severity of organ involvement with Sarcoidosis with the exception of the lung where measurements of spirometry, chest radiographic findings, and level of dyspnea were made.


Table IX: Type of Organ Involvement By Age, race, and sex

Table X: Number of patients with and without new organ involvement at two-year follow-up for specified demographic subgroups

We found that there was an excellent concordance between changes in FVC and FEV1. FVC and FEV1 were discordant (worse versus better) in none (0/l93) of the patients. This suggests that using FVC alone is unlikely to result in a misclassification in or change in pulmonary function over two years. However, changes in FVC alone were inadequate to describe the change in pulmonary status of our patients, as changes in chest radiographic findings or the level of dyspnea often but did not always move in the same direction as FVC (Table VI).

Erythema nodosum at presentation was associated with greater improvement in Scadding radiographic stage than the absence of erythema nodosum at presentation. This has been observed in other studies [2, 3, 19, 24].

Lower annual family income was associated with less improvement in dyspnea and more organ involvement over the two-year follow-up. This may suggest that socioeconomic factors play a role in the clinical outcome of Sarcoidosis.

More patients with improved FVC or dyspnea at follow-up were taking corticosteroids than not, and more patients with a worse FVC or dyspnea at follow-up were taking corticosteroids than not. Similarly more patients with improved dyspnea were seeing a sub-specialist than not, and more patients with worse dyspnea were seeing a sub-specialist at follow-up than not.

The use of corticosteroids may improve acute pulmonary Sarcoidosis in the short-term [25,26], resulting in a steroid-responsive group with an improved clinical status. However, patients with relatively refractory disease (a steroid unresponsive group or frequent relapse group) usually receive corticosteroids for a prolonged period and may be a group with a worse clinical outcome. We did identify a subgroup of 52/205 (25%) subjects from this cohort who were maintained on corticosteroids for the duration of the two-year follow-up, and we suspect that most of these individuals represent the refractory cases.
In addition, the use of corticosteroids may promote relapse of Sarcoidosis when the medication is discontinued or tapered that might adversely affect the clinical course [6,10]. This may also explain our findings concerning sub-specialists in that the expertise of sub-specialists may lessen patients' dyspnea, but sub-specialists might also be referred patients with refractory disease who develop worsening dyspnea. ACCESS did not obtain data concerning the dose of corticosteroids that was used by patients, which may also affect the rate of relapse [11] Some data was obtained in ACCESS concerning the discontinuation or addition of corticosteroid therapy during the two-year follow-up, but this will be the subject of a subsequent report focusing specifically on therapy.

In conclusion, data from this heterogeneous United States Sarcoidosis population indicate that Sarcoidosis tends to improve or remain stable in the majority of patients. African Americans tend to have a worse prognosis by spirometry and organ involvement than Caucasians. Erythema nodosum at presentation is associated with a good outcome. Lower annual family income is associated with a worse outcome. The presence of extra-pulmonary organ involvement at presentation is a risk factor for developing new organ involvement over two years. FVC and FEVI had a similar trend over 2 years. Changes in chest radiographs and dyspnea usually, but not always, parallel spirometric changes. Therefore assessment of change in pulmonary status at two years requires an independent assessment of spirometry, chest radiographic changes, and degree of dyspnea. The interaction of corticosteroid therapy with two-year outcome is complex.

References

1 Hunninghake GW, Costabel U, Ando M et al: Statement on Sarcoidosis, Am J  Respir Crit Car Med  1999; : 160: .736 4- 55.

2 Mana J, Salazar A, . Manresa F: Clinical factors predicting the persistence of activity in Sarcoidosis; a multivariate analysis of 193 cases. Respiration 1994; 61: 219.25.

3 Neville E, Walker AN, James DG: Prognostic factors predicting the outcome of Sarcoidosis: an analysis of 818 patients. Q J Med 1983; 208:525-33.

4 Israel HL, Karlin P, Menduke H. Factors affecting outcome of sarcoidosis Ann  NY Acad  Sci 1986; 465: 609.18,.

5 Romer FK: Presentation of Sarcoidosis and outcome of pulmonary changes. Dan
Med Bull 1982; 29: 27-32.  '.

6 Reich JM: Mortality of intra-thoracic Sarcoidosis in referral vs population-based servings: influence of stage, ethnicity, and corticosteroid. Chest 2002: 121: 32-9.

7 Piettinalho A, . Tukiainen P, Haahtela T, Persson T. et al: Early treatment of stage II sarcoidosis improves 5-year pulmonary function. Chest 2002; 121:24-31.

8 Chappell AG, Cheung NY, Hatchings. HA. Sarcoidosis:; a long-term follow-up study. Sarc Vasc Diff Lung Dis 2000; 17: 167-73.

9 Berlin M, Fogdell-Hahn A., Olerup O, Elkund A, Grunewald J: HLA-DR predicts the prognosis in Scandinavian patients with pulmonary Sarcoidosis. Am J Respir Crit Care Med 1997: 156: 1601-.5.

10 Gottlieb JE, Israel HL, Steiner RM, Triolo J, Patrick H: Outcome in Sarcoidosis. The relationship of relapse to corticosteroid therapy. Chest  1997:111:623-31.

11 Rizzato G,. Montemurro L, Colombo P: The late follow-up of chronic sarcoidosis patients previously treated with corticosteroids Sarc Vasc Diff Lung Dis 1999:15: 52-8.

12 ACCESS Research Group. Design of a case-controlled etiologic study of  Sarcoidosis (ACCESS). J Clin Epidemiol 1988:41:56570.

Georgetown University Medical Center: Henry Yeager Jr.; David L. Rabin; Susan Stein

Case Western Reserve University-Henry Ford Health Sciences Center: Michael C. Iannuzzi; Benjamin Rybicki; Marcie Major; Mary Maliarik; John Popovich Jr.

Johns Hopkins University School of Medicine: David R. Moller; Carol J. Johns*; Cynthia Rand; Joanne Steimel,

Medical University of South Carolina: Marc A. Judson; Susan D'Alessandro, Nancy Heister; Theresa Johnson; Daniel T. Lackland; Janardan Pandey; Steven Sahn; Charlie strange

Mount Sinai Medical Center. Alvin S. Teirstein; Louis DePalo: Sheldon Brown: Marvin Lesser; Maria L. Padilla; Marilyn Marshall

National Jewish Medical and Research Center. Lee S. Newman; Cecile Rose; Juli Bernard

University of Cincinnati Medical Center: Robert P. Baughman; Elyse E. Lower; Donna B. Winget

University of lower College of Medicine: Geoffrey McLennan; Gary Hunninghake; Chuck Dayton; Julia Stutzman

University of Pennsylvania and MCP-Hahnemann University Medical Centers: Milton D. Rossman; Eddy A. Bresnitz; Ronald Danielle, Jackie Regovich; William Sexauer

National Heart, Lung, and Blood Institute: Robert Musson; Joanne Deshler; Paul Sortie; Margaret Wit.

Study Chairman: Reuben Cherniack,

Study Co-Chairman: Lee Newman

Clinical Coordinating Center: CLINICAL TRIALS & SURVEYS Corp.: Garrett L. Knatterud, Michael L. Terrill, Brace NY. Thompson, Kathleen Brown, Margaret Frederick, Frances LoPresti, Patricia Wilkins, Martha Canner, Judy Dotson.

Data Safety and Monitoring Board: William Martin (Chair); Takamaru Ashikaga; David B. Coultas; Gerald S. Davis; Fred Gifford; James J. Schlesselman; Diane Stover; Robert Musson (Executive Secretary)

Ex Officio: Reuben Cherniack; Genell L. Knatterud: Lee Newman

*Deceased.