Sarcoidosis diagnostic scores have ‘good to excellent’ discriminatory ability

Key takeaways:

• Sarcoidosis diagnostic scores can identify sarcoidosis among alternative diseases.
• A score of less than 3 in the clinical test and less than 5 in the biopsy test meant a low likelihood for sarcoidosis.

The ability to discern sarcoidosis from other diagnoses was “good to excellent” in two sarcoidosis diagnostic scores, according to results published in Annals of the American Thoracic Society.

However, both scores performed either poorly or fairly in differentiating sarcoidosis from noninfectious granulomatous diseases, according to researchers.

“This multicontinental study confirms that both sarcoidosis diagnostic score (SDS) Clinical and SDS Biopsy can reliably distinguish sarcoidosis from alternative diagnoses in most cases but could fail in some groups,” Florence Jeny, MD, PhD, of the department of respiratory diseases at Sorbonne University and at Avicenne Hospital, both in Paris, and colleagues wrote. “These findings held regardless of the prevalence of TB in patient populations and regardless of the patient demographics. The SDS Clinical and SDS Biopsy reliably differentiated infectious granulomatous diseases from sarcoidosis but was not reliable in cases of noninfectious granulomatous diseases.”

In a large, multicontinental study of nine different centers, Jeny and colleagues analyzed 1,041 patients with biopsy-confirmed sarcoidosis and 1,035 patients without sarcoidosis to validate the SDS and evaluate how well it can distinguish sarcoidosis from other diagnoses such as interstitial lung diseases, COPD/asthma and infectious/noninfectious granulomatous diseases.

Researchers gave each patient a score using a modified World Association of Sarcoidosis and Other Granulomatous Disorders organ assessment instrument based on the existence of granuloma on biopsy and both highly probable and least probable criteria related to sarcoidosis.

The presence of biopsy findings in relation to SDS was tested in SDS biopsy, and an SDS that factors in no findings from a biopsy was tested in SDS clinical.

To test the ability of both scores to differentiate sarcoidosis from various diagnoses and then compare them, researchers used area under the curve (AUC).

Researchers found “good to excellent” ability to distinguish sarcoidosis among other diagnoses with SDS clinical (AUC, 0.888; 95% CI, 0.874-0.902) and SDS biopsy (AUC, 0.979; 95% CI, 0.973-0.985).

Notably, the clinical score had better discriminatory ability in identifying these patients in low TB prevalence centers (AUC, 0.912; 95% CI, 0.895-0.929) rather than high TB prevalence centers (AUC, 0.854; 95% CI, 0.832-0.876; P < .001). This was not seen with SDS biopsy, according to researchers.

Further, the ability to differentiate sarcoidosis from other diagnoses was better in women (AUC, 0.901; 95% CI, 0.883-0.919) than men (AUC, 0.863; 95% CI, 0.84-0.887; P = .01) when using the clinical score, and researchers even observed higher sensitivity (94.6% vs. 92.2%) and specificity (75.8% vs. 68.9%) in women in an SDS clinical test with a score of at least 3.

Looking specifically at granulomatous diseases, researchers identified 135 cases of infectious diseases, including TB, non-TB mycobacterial and fungal infections, as well as 54 cases of noninfectious diseases including chronic beryllium disease (CBD), common variable immunodeficiency (CVID), antineutrophil cytoplasmic antibody (ANCA)-related vasculitis and hypersensitivity pneumonitis (HP) in the total cohort.

In terms of telling noninfectious granulomatous diseases from sarcoidosis, researchers found poor discriminatory ability in SDS clinical (AUC, 0.684; 95% CI, 0.602-0.766) and fair ability in SDS biopsy (AUC, 0.754; 95% CI, 0.673-0.835). The SDS clinical test with a score of at least 3 and the SDS biopsy with a score of at least 6 both demonstrated specificity under 50%.

For both noninfectious diseases of CBD and CVID, researchers deemed that the scores cannot discern sarcoidosis from them because of high SDS clinical and biopsy. However, when telling HP from sarcoidosis, SDS clinical (AUC, 0.912; 95% CI, 0.84-0.984) and SDS biopsy (AUC, 0.971; 95% CI, 0.919-1) showed high AUCs that are suggestive of “excellent” discriminatory ability.

In terms of distinguishing sarcoidosis from mycobacterial/fungal infections, researchers found “good to excellent” AUCs with SDS clinical (AUC, 0.867; 95% CI, 0.836-0.899) and SDS biopsy (AUC, 0.969; 95% CI, 0.952-0.987).

Researchers also assessed patients with lymphoma (n = 100) to test both scores in their ability to discern sarcoidosis from this disease and found “fair to good” abilities in SDS clinical (AUC, 0.716; 95% CI, 0.661-0.77) and SDS biopsy (AUC, 0.98; 95% CI, 0.969-0.99).

Jeny and colleagues proposed one diagnostic algorithm for each score after accounting for the above data; however, these are not applicable to those with CBD, ANCA-related vasculitis or CVID.

For SDS clinical without biopsy, researchers found that a sarcoidosis diagnosis could be accepted with a score greater than 9, and a lower probability of sarcoidosis was found with a score less than 3, which was evident by the difference in patients with this score who had sarcoidosis (6.2%) and those who did not have sarcoidosis (72.2%).

For scores between these numbers, researchers found that sarcoidosis can be confirmed if granulomas consistent with the disease are observed on the biopsy.

In terms of SDS biopsy, a score greater than 11 meant sarcoidosis, whereas sarcoidosis was doubtful with a score less than 5. Clinicians must evaluate fungal/mycobacterial infections or malignancy causes for scores that fall between these numbers regardless of sarcoidosis specificity, according to researchers.

“We found cutoff values for the SDS Clinical and SDS Biopsy that allowed the diagnosis of sarcoidosis to be safely confirmed or rejected in most cases except for CBD, granulomatosis associated with CVID, or ANCA-related vasculitis,” Jeny and colleagues wrote. “Our work thus insists on the importance of a careful interrogation and diagnostic approach in search of an occupational exposure to beryllium or repeated infections/hypogammaglobulinemia or ANCA positivity.”