Predictive value of baseline serum angiotensin-converting enzyme (sACE) levels for steroid response in sarcoidosis

ABSTRACT

Background and Objective:

Sarcoidosis is a systemic granulomatous disease that primarily affects the lungs. Corticosteroids are the first-line therapy, but steroid resistance remains a major clinical challenge. This study evaluates baseline serum angiotensin-converting enzyme (sACE) levels as a potential biomarker for predicting steroid resistance in sarcoidosis.

Methods:

In this prospective single-centre study, 188 patients with biopsy-confirmed sarcoidosis had baseline sACE levels measured before corticosteroid initiation using the Kasahara colorimetric technique. Associations between sACE levels and treatment response were analysed using logistic regression and ROC analysis (SPSS v21; significance at P < 0.05).

Results:

The mean age of the cohort was 54 years, with 62% women. Common symptoms included cough, dyspnoea, fatigue, and arthralgia; hypercalcemia was seen in 11%. Elevated baseline sACE > 64 U/L was found in 41% of cases. ROC analysis yielded an (AUC) of 0.788 (95% CI: 0.685–0.891), with 83% sensitivity and 73% specificity for predicting steroid resistance. On multi-variate analysis, elevated sACE (P < 0.001), age > 54 years (P < 0.001), fatigue (P = 0.002), hypercalcemia (P = 0.019), and mediastinal lymph node involvement on CECT (P = 0.036) were independent predictors of poor steroid response.

Conclusion:

Elevated baseline sACE levels show potential as an adjunctive biomarker for identifying patients at risk of steroid resistance in sarcoidosis. However, given its moderate specificity and absence of longitudinal validation, sACE should be interpreted cautiously and in conjunction with clinical and radiological parameters. Multi-centre studies with longer follow-up are warranted to confirm these findings.

INTRODUCTION

Sarcoidosis is a multi-system granulomatous disease of unknown aetiology, most frequently involving the lungs and intrathoracic lymph nodes, though extrapulmonary sites such as the skin, eyes, liver, heart, and nervous system may also be affected.[1] Diagnosis relies on clinical, radiological, and histopathological correlation, with high-resolution computed tomography (HRCT) and serum biomarkers such as serum angiotensin-converting enzyme (sACE) often used to support disease assessment.[2,3]

Corticosteroids remain the cornerstone of sarcoidosis treatment. However, prolonged steroid therapy is associated with significant adverse effects, and a subset of patients exhibit suboptimal or incomplete responses despite adequate dosing.[4,5,6] The absence of standardized criteria for defining steroid resistance further complicates clinical decision-making, highlighting the need for objective, reproducible predictors of treatment response.[7,8]

Given the heterogeneous and unpredictable course of sarcoidosis, there is growing interest in biomarkers that can guide early therapeutic decisions and identify patients at risk of steroid resistance.[9,10] Among potential candidates, serum ACE has been studied for decades because it reflects granulomatous activity through its production by epithelioid cells within sarcoid granulomas.[11] Although earlier studies reported variable diagnostic sensitivity and specificity,[3] recent data suggest that elevated baseline sACE levels may correlate with active disease and poorer treatment outcomes.[12,13] Nonetheless, its utility is limited by lack of standardization and potential confounding from other granulomatous or inflammatory conditions such as tuberculosis and silicosis.[14]

Contemporary guidelines from the American Thoracic Society (ATS) and European Respiratory Society (ERS) advocate individualized, biomarker-supported approaches to sarcoidosis management but emphasize the absence of validated predictors of steroid response.[5,15] Against this backdrop, our study prospectively evaluates the predictive value of baseline sACE levels in identifying sarcoidosis patients at risk of steroid resistance. By focusing on pre-treatment biomarker assessment and integrating clinical and radiological correlates, we aim to improve early risk stratification and guide rational use of immunosuppressive therapy.

MATERIALS AND METHODS

Patient selection and involvement

A total of 191 patients were initially enrolled. Three patients were excluded due to concurrent use of angiotensin-converting enzyme (ACE) inhibitors, which could confound sACE measurement. The final cohort thus comprised 188 biopsy-confirmed cases of sarcoidosis.

All patients provided written informed consent, and the study protocol was approved by the Institutional Ethics Committee. Data were collected prospectively using a standardized case record form to minimize observer bias.

Study definitions and outcome measures

Steroid responsiveness was defined as a clinically and/or radiologically meaningful improvement after corticosteroid therapy. Improvement included reduction in key symptoms (cough, dyspnoea, chest pain) and/or regression in imaging findings (e.g., reduction in lymphadenopathy or parenchymal involvement on HRCT).

Steroid resistance was defined as absence of clinical or radiological improvement despite an adequate corticosteroid regimen (typically 20–40 mg/day of prednisone or equivalent for ≥4–6 weeks), or recurrence of symptoms during tapering or maintenance phases. Objective functional improvement was assessed where available using pulmonary function tests (PFTs); an FVC gain ≥10% was considered clinically significant.

The primary outcome measure was the need for additional disease-modifying agents (DMAs) beyond corticosteroids. The cohort was divided into two pre-defined groups:

• Steroid-Responsive Group: Patients showing sustained clinical and radiological improvement on corticosteroid monotherapy, stable on tapering doses (5–15 mg/day), without relapse or need for additional therapy.

• Steroid-Resistant Group: Patients showing persistent symptoms or progression despite adequate corticosteroid therapy, requiring escalation to steroid-sparing agents such as methotrexate, azathioprine, mycophenolatemofetil, or biologics (e.g., TNF-α inhibitors).

All clinical and radiological evaluations were conducted in multi-disciplinary consensus; however, to reduce subjectivity, pre-defined response criteria were used, and imaging was reviewed independently by two pulmonologists and one radiologist blinded to sACE results.

Laboratory biomarker assessment

Venous blood samples were collected from all patients at baseline, before corticosteroid initiation, for the measurement of serum angiotensin-converting enzyme (sACE) levels. Estimation was performed using the Kasahara colorimetric method, a spectrophotometric assay employing p-hydroxyhippuryl-l-histidyl-L-leucine as the synthetic substrate. Upon hydrolysis by ACE, liberated hippuric acid is quantified colorimetrically, providing a reproducible estimate of sACE activity.[16] The institutional reference range was 8–65 U/L.

A threshold of >64 U/L was selected based on laboratory reference limits, prior published studies, and receiver operating characteristic (ROC) analysis in the present dataset. The ROC-derived optimal cutoff corresponded to the highest Youden’s Index, ensuring objective determination rather than arbitrary selection.

The primary focus of the study was on baseline sACE as a predictor of steroid response.

Although follow-up sACE levels were obtained in a subset of patients, they were not analysed longitudinally because of sample-size limitations and incomplete uniform sampling; this limitation has been explicitly acknowledged. The emphasis, therefore, remained on baseline sACE as an early, accessible biomarker for risk stratification.

Outcome assessment and response criteria

Outcome assessment combined clinical, radiological, functional, and laboratory parameters to classify patients as steroid-responsive or steroid-resistant through multi-disciplinary consensus using pre-defined criteria.

• Clinical Response:

• Improvement in cardinal symptoms (dyspnoea, cough, fatigue, and chest discomfort) as documented during follow-up visits. Symptom assessment was performed using standardized departmental follow-up templates that captured change in symptom severity and treatment response. Although formalized scoring tools (e.g., mMRC or Borg scales) were not uniformly used, the same structured symptom checklist was applied for all patients to ensure consistency. This pragmatic approach reflects real-world clinical practice, where formal activity indices such as Scadding staging or the Sarcoidosis Activity and Severity Index (SASI) are not routinely implemented.

• Radiological Response: Comparison of baseline and follow-up HRCT scans, with improvement defined as a reduction in parenchymal opacities, nodules, or lymphadenopathy. Imaging was independently reviewed by two pulmonologists and one radiologist blinded to sACE results, and discordant interpretations were resolved by consensus.

• Pulmonary Function Testing: In a subset of patients where spirometry was feasible, a ≥10% increase in forced vital capacity (FVC) was considered an objective marker of response. The number and characteristics of this subset are described in the results section to ensure transparency.

• Laboratory Assessment: Baseline sACE levels served as the principal biomarker for analysis. Persistently elevated values at 4–6 weeks were noted only descriptively and not incorporated into statistical models due to incomplete data capture.

• DMA escalation was guided by pre-defined multi-disciplinary consensus criteria incorporating objective clinical, functional, and radiologic parameters to minimize subjectivity

Although full blinding across all assessments was not feasible, the structured, pre-defined criteria and independent radiologic review were implemented to minimize observer bias and enhance reproducibility.

Statistical analysis

Descriptive statistics were used to summarize demographic and clinical characteristics. Continuous variables were presented as mean ± standard deviation (SD) for normally distributed data and as median with interquartile range (IQR) for non-normally distributed data. Categorical variables were expressed as frequencies and percentages.

Comparative analyses were conducted using appropriate statistical tests. The Chi-square test or Fisher’s exact test was used to compare categorical variables, as applicable. For continuous variables, independent t-tests or Mann–Whitney U tests were employed based on data distribution. Odds ratios (ORs) with 95% confidence intervals (CIs) were calculated using univariate logistic regression to identify variables associated with the need for disease-modifying agents (DMAs).

Variables that were statistically significant in univariate analysis (P < 0.05) and deemed clinically relevant were included in the multi-variate binary logistic regression model. Multi-collinearity was assessed using the variance inflation factor (VIF), with all included variables showing VIF <2, confirming no significant collinearity. Model calibration was evaluated using the Hosmer–Lemeshow goodness-of-fit test, with a P value >0.05 considered acceptable.

Internal model validation was performed using a random 70:30 and subsequently a 60:40 training–validation split approach. Logistic regression coefficients were derived from the training set and tested on the validation subset to assess reproducibility. Predicted probabilities were compared between steroid-responsive and steroid-resistant groups using independent t-tests, and model discrimination was further evaluated using receiver operating characteristic (ROC) curves.

The area under the ROC curve (AUC), sensitivity, specificity, and 95% CI were calculated for both training and validation cohorts. The optimal sACE threshold for predicting steroid resistance was determined using Youden’s Index, and classification accuracy was assessed using confusion matrices.

Given the limited number of events in the validation subset, formal bootstrapping was not feasible in SPSS version 21. To minimize overfitting risk, the model’s stability was evaluated through repeated random splits, ensuring consistency of effect direction and statistical significance across iterations.

All analyses were conducted using SPSS Statistics software (version 21.0; IBM Corp., Armonk, NY, USA). A two-tailed P value of <0.05 was considered statistically significant.

RESULTS

A total of 191 patients with biopsy-confirmed sarcoidosis were screened, of whom 3 were excluded due to concurrent use of ACE inhibitors, which could confound serum angiotensin-converting enzyme (sACE) levels. The final analysis included 188 patients.

The mean age of participants was 53 ± 13.8 years, and 116 (61.7%) were female. The majority of patients (116, 61.7%) had pulmonary involvement, while 18 (9.6%) each had gastrointestinal involvement or uveitis. Other extrapulmonary manifestations included generalized lymphadenopathy (16, 8.5%), erythema nodosum (6, 3.2%), salivary gland involvement (6, 3.2%), cardiac involvement (4, 2.1%), renal disease (2, 1.1%), and facial nerve palsy (2, 1.1%).

Clinical characteristics

The most common presenting symptoms were cough (106, 56%), dyspnoea (105, 55%), fatigue (23, 12%), fever (36, 19%), chest pain (31, 16%), and arthralgia (48, 26%).

Seventeen (9%) patients were asymptomatic at presentation, detected incidentally during evaluation for other causes.

Common comorbidities included diabetes mellitus (62, 33%), hypertension (55, 29%), and hypothyroidism (22, 12%).

Radiological findings and hypercalcemia

Radiological staging, based on the Scadding classification, showed that most patients had Stage I or Stage II disease, reflecting early or moderate pulmonary involvement. Specifically, Stage I disease was seen in 59 patients (31.4%), Stage II in 54 (28.7%), Stage III in 6 (8.5%), and Stage IV in 6 (3.2%), while 53 patients (28.2%) had extrapulmonary sarcoidosis (Stage 0). Thus, 71.8% of patients had Scadding Stages I–IV, whereas 28.2% had Stage 0 or extrapulmonary-only disease not classifiable under the Scadding system.

Hypercalcemia was noted in 20 patients (10.6%), as summarized in Table 1.

Table 1.

Baseline characteristics of the study population (n^†=188)

Characteristic	n (%)/Mean±SD
Age (years)	53±13.8
Gender
Male	72 (38.3)
Female	116 (61.7)
Smokers	18 (9.6)
Previous ATT exposure	25 (13.3)
Comorbidities
Diabetes mellitus	62 (33.0)
Hypertension	55 (29.3)
Hypothyroidism	22 (11.7)
Hypercalcemia	20 (10.6)
Mantoux test
Positive	25 (13.3)
Negative	133 (76.7)
Chest X-ray (Scadding stage)
Stage 0	53 (28.2)
Stage I	59 (31.4)
Stage II	54 (28.7)
Stage III	16 (8.5)
Stage IV	6 (3.2)
CECT chest findings
Mediastinal lymphadenopathy	89 (47.3)
Ground-glass opacities/interstitial thickening	52 (27.7)
Fibrosis	12 (6.4)
Mode of tissue diagnosis
Conventional TBNA	6 (3.1)
EBUS-TBNA	72 (38.3)
EBB	4 (2.1)
TBLB	34 (18.1)
Extrapulmonary biopsy	72 (38.3)
ANA positive	27 (14.4)
Baseline sACE >64 U/L	77 (41.0)
Required immunomodulators	47 (25.0)

ATT: anti-tuberculosis therapy; CECT: contrast-enhanced computed tomography; TBNA: transbronchial needle aspiration; EBUS-TBNA: endobronchial ultrasound-guided TBNA; EBB: endobronchial biopsy; TBLB: transbronchial lung biopsy; ANA: anti-nuclear antibody; sACE: serum angiotensin-converting enzyme. ^†n=total number of samples

Need for immunomodulatory therapy

A total of 47 patients (25%) required additional disease-modifying agents (DMAs), while 141 patients (75%) responded adequately to corticosteroid monotherapy. Among those requiring immunomodulatory therapy, 32 (27.6%) were female and 15 (20.8%) were male. In contrast, among patients who did not require additional therapy, 84 (72.4%) were female and 57 (79.2%) were male.

Gender distribution differed significantly between the two groups (P < 0.05), with a higher proportion of females among those requiring DMAs.

Serum ACE levels

Serum ACE (sACE) levels were non-normally distributed (Shapiro–Wilk P < 0.05). The median sACE level for the cohort was 48 IU/L (IQR 28–92), with values ranging from 1 to 490 IU/L. Elevated sACE levels (>64 IU/L) were observed in 77 patients (41%) [Figure 1].

Figure 1.

Comparison of serum angiotensin-converting enzyme levels in the patients with need for immunomodulators and patients without need for immunomodulators

Patients with elevated baseline sACE were significantly more likely to require additional immunomodulatory therapy compared with those having normal sACE values (P < 0.001). This association was further evaluated through univariate and multi-variate logistic regression analyses to determine the independent predictive value of sACE for steroid resistance.

ROC curve analysis

Receiver operating characteristic (ROC) curve analysis demonstrated that a serum ACE level of 64 U/L was the optimal threshold for predicting steroid resistance in sarcoidosis. The area under the curve (AUC) was 0.788 (95% CI: 0.685–0.891; P < 0.001), indicating good discriminative ability. This cut-off yielded a sensitivity of 83%, specificity of 73%, positive predictive value (PPV) of 50.6%, and negative predictive value (NPV) of 92.8%. The corresponding diagnostic odds ratio (DOR) was 11.3, with positive and negative likelihood ratios (LR+ =3.07; LR– =0.23). The high NPV suggests that lower sACE levels are particularly useful for ruling out steroid resistance, while elevated levels warrant closer monitoring or early escalation of therapy [Figure 2].

Figure 2.

Receiver operating characteristic (ROC) curve showing the diagnostic performance of baseline serum ACE levels for predicting steroid resistance in sarcoidosis. Area under the curve (AUC) = 0.788 (95% CI: 0.685–0.891; P < 0.001); Sensitivity = 83%; Specificity = 73%; Positive Predictive Value (PPV) =50.6%; Negative Predictive Value (NPV) =92.8%

Factors associated with steroid resistance

Univariate analysis demonstrated that several clinical and biochemical variables were significantly associated with steroid resistance, including elevated serum ACE levels, fatigue, hypercalcemia, thoracic involvement, and ground-glass opacities on HRCT. In addition, female sex and older age (>54 years) showed trends toward association with treatment resistance.

In multi-variate logistic regression analysis, elevated serum ACE levels (>64 U/L) (P < 0.001), fatigue (P = 0.002), hypercalcemia (P = 0.019), age >54 years (P < 0.001), and mediastinal lymph-node involvement on CECT (P = 0.036) remained independent predictors of steroid resistance, whereas thoracic involvement lost significance (P = 0.157) [Table 2].

Table 2.

Predictors of steroid resistance in sarcoidosis (univariate and multi-variate logistic regression)

Variable	Univariate OR (95% CI)	P (Univariate)	Adjusted OR (95% CI)	P (Multi-variate)
Elevated sACE (> 64 U/L)	8.91 (4.21 – 18.88)	< 0.001	54.39 (11.05 – 268.28)	<0.001
Fatigue	3.29 (1.34 – 8.06)	0.007	14.34 (2.55 – 80.55)	0.002
Hypercalcaemia	3.51 (1.36 – 9.08)	0.007	5.52 (1.33 – 22.99)	0.019
Age >54 years	2.47 (1.21 – 5.05)	0.013	4.28 (1.76 – 10.38)	<0.001
Female sex	2.02 (1.03 – 3.96)	0.041	2.58 (0.98 – 6.79)	0.055
CECT – Mediastinal LN involvement	—	—	10.56 (1.16 – 95.77)	0.036
Thoracic involvement	2.16 (1.04 – 4.51)	0.038	—	0.157
Dyspnoea	—	—	0.26 (0.07 – 1.02)	0.054
Salivary gland involvement	3.14 (0.98 – 110.40)	0.052	—	0.052
Erythema nodosum	0.59 (0.07 – 5.19)	0.632	—	0.917
Cardiac involvement	1.00 (0.10 – 9.85)	1.000	—	0.361

Although female sex retained a modest effect size (adjusted OR 2.58, P = 0.055), it did not achieve statistical significance after adjustment for biochemical and radiological factors.

These findings suggest that both biochemical markers (sACE, calcium) and host factors (age, sex) interact with disease phenotype to influence steroid responsiveness, emphasizing the multi-factorial nature of treatment resistance in sarcoidosis.

Note: Internal validation using a 70–30 and 60–40 split-sample approach was attempted; however, smaller subsets yielded unstable probability estimates, and bootstrapping-based validation is planned in future larger cohorts.

Comorbidities and confounding factors

Among the 188 patients, 62 (33%) had diabetes mellitus, 55 (29.3%) had hypertension, and 22 (11.7%) had hypothyroidism. A higher proportion of patients with diabetes or hypertension demonstrated elevated sACE levels and a greater likelihood of requiring additional immunomodulatory therapy [Table 3].

Table 3.

Stratification of common comorbidities by serum ACE level and steroid response (n=188)

Comorbidity	Present, n (%)	sACE> 64 U/L, n (%)	Required DMA, n (%)
Diabetes mellitus	62 (33.0)	26 (42)	18 (39)
Hypertension	55 (29.3)	20 (36)	16 (34)
Hypothyroidism	22 (11.7)	9 (41)	7 (32)

sACE=serum angiotensin-converting enzyme; DMA=disease-modifying agent

However, on multi-variate logistic regression, these comorbidities did not remain independent predictors of steroid resistance. Their inclusion in the model did not substantially alter the significance of key variables such as sACE > 64 U/L, fatigue, and hypercalcemia, suggesting limited confounding in the present cohort.

While formal sensitivity analyses excluding these comorbid subgroups were not feasible due to sample size constraints, such analyses are recommended in larger multi-centre studies to better define the independent contribution of metabolic and endocrine factors to treatment resistance in sarcoidosis.

DISCUSSION

This prospective study evaluated the predictive value of serum angiotensin-converting enzyme (sACE) in determining steroid responsiveness among patients with sarcoidosis. We observed that elevated baseline sACE levels (>64 U/L) were significantly associated with subsequent need for additional immunomodulatory therapy, suggesting that sACE may serve as a useful adjunctive indicator of treatment resistance. The diagnostic performance of sACE in our cohort, with a sensitivity of 83% and specificity of 73%, aligns with several prior reports demonstrating its moderate accuracy as a biochemical correlate of granulomatous activity rather than a stand-alone prognostic tool.[14]

Nevertheless, the clinical utility of sACE remains debated. Druyan et al.[14] recently reported that sACE has low specificity and poor positive predictive value, emphasizing its limited reliability as a solitary biomarker of disease activity. Their cohort showed that while the negative predictive value was high (94.6%), elevated ACE levels correlated poorly with disease progression. This reinforces our interpretation that normal or low sACE levels are reassuring, but elevated levels must be contextualized with clinical and radiologic findings rather than interpreted deterministically.

Interpretation of findings

The diagnostic performance of serum ACE in our study showed a high negative predictive value (92.8%), indicating that normal or low sACE levels reliably identified patients likely to respond to corticosteroid therapy. In contrast, the positive predictive value (50.6%) was modest, suggesting that elevated sACE levels alone may not definitively predict steroid resistance. Thus, while high sACE should prompt closer clinical monitoring or early escalation, low values are particularly useful as a “rule-out” biomarker for steroid resistance.

In our multi-variate model, elevated sACE, fatigue, hypercalcemia, older age (>54 years), and mediastinal lymph-node involvement emerged as independent predictors of steroid resistance, while female sex showed a nonsignificant trend. These associations highlight the multi-factorial nature of treatment response, where biochemical, demographic, and imaging features interact.

Fatigue has long been recognized as a surrogate of systemic inflammation and chronic immune activation in sarcoidosis.[17] Hypercalcemia, driven by unregulated 1-α-hydroxylase activity within granulomas, also reflects disease burden and macrophage activation.[18] Both correlate with treatment refractoriness, suggesting that biochemical markers of granuloma metabolism may signal a more active or disseminated disease phenotype.

Genetic determinants of ACE activity have also been examined. Moller et al.[19] demonstrated that ACE I/D polymorphisms did not correlate with sarcoidosis susceptibility, prognosis, or treatment response in a Danish cohort, implying that inter-individual differences in sACE are largely non-genetic or secondary to disease state. Our data are consistent with these findings, showing no significant relationship between genotype surrogates and steroid response.

Serum ACE as a biomarker and its integration with imaging

While sACE remains a readily available biochemical marker, its non-specificity limits its independent predictive power. Levels may fluctuate with comorbidities such as diabetes, thyroid disease, or chronic liver dysfunction, and can be influenced by demographic and genetic heterogeneity.

Recent immunogenetic insights, however, offer partial explanation for this variability. Rossides et al.[20] reported that patients carrying HLA-DRB1 × 04 more often had elevated sACE levels, whereas those with HLA-DRB1 × 03 typically showed normal values; these alleles were also linked with extrapulmonary manifestations. This suggests that the magnitude of sACE elevation may reflect underlying HLA-driven immunopathology and granuloma burden, potentially accounting for inter-ethnic differences in biomarker performance.

Concurrently, imaging biomarkers have expanded the landscape of activity assessment. Evolving evidence suggests that ¹⁸F-FDG PET/CT can assist in evaluating therapeutic response. Kawai et al.[16] demonstrated that patients with high baseline standardized uptake values (SUVs) exhibited more rapid and pronounced regression of disease following immunosuppressive therapy, underscoring PET’s potential for early response monitoring. Because both PET and high-resolution CT are expensive and carry radiation exposure, serum ACE measurements offer a pragmatic, low-cost adjunct for tracking disease activity, particularly in resource-limited settings.

Taken together, these data position sACE not as a definitive marker, but as a complementary biochemical index that, when combined with imaging and clinical findings, can enhance accuracy of disease-activity assessment.

Relevance to clinical practice

From a practical standpoint, baseline sACE testing can support early therapeutic planning. Patients with high sACE should undergo closer follow-up and multi-disciplinary review, while those with normal or low levels can often continue corticosteroid monotherapy with confidence of good response. Integration of sACE with age, fatigue, hypercalcemia, and radiologic patterns—particularly mediastinal lymphadenopathy and fibrosis—yields a more nuanced risk profile.

Combining sACE with other biomarkers such as soluble IL-2 receptor, chitotriosidase, or KL-6 could further improve predictive precision.[21] Ultimately, a multi-modal algorithm incorporating biochemical, imaging, and immunologic markers is likely to outperform any single parameter.

Context, limitations, and future directions

Our study provides real-world, biopsy-confirmed data from an Indian cohort characterized by a high prevalence of metabolic comorbidities—factors often under-represented in global sarcoidosis datasets. While sACE showed significant association with steroid resistance, it remains a supportive rather than decisive biomarker.

Limitations include the single-centre design, modest sample size, lack of serial sACE measurement, and absence of prospective PET correlation. Future research should explore combined sACE-PET models, serial biomarker kinetics, and integration of HLA typing into predictive frameworks. Validation of such composite indices across multi-ethnic cohorts could ultimately yield a standardized, evidence-based tool for anticipating treatment response.

In conclusion, serum ACE is best regarded as an adjunctive marker that complements—but does not replace—clinical and imaging evaluation. When interpreted alongside demographic, biochemical, and radiologic indicators, it can assist in the early identification of patients at risk for steroid resistance and guide individualized, steroid-sparing strategies.

Limitations and impact of comorbidities

Interpretation of sACE as a predictive biomarker must consider the confounding influence of comorbidities that can independently elevate enzyme levels. Conditions such as diabetes mellitus, hyperthyroidism, and granulomatous infections are known to raise serum ACE and may reduce its specificity for sarcoidosis. In our cohort, diabetes and hypertension were present in approximately one-third of patients; although neither emerged as independent predictors of steroid resistance, they may still influence biomarker expression and treatment response.

From a methodological standpoint, this single-centre study with a modest event count and high background prevalence of tuberculosis (13.3% with prior ATT or Mantoux positivity) limits generalizability. Regional epidemiologic overlap with other granulomatous diseases could also influence baseline sACE levels. The absence of comparator groups (healthy or non-sarcoid granulomatous controls) further restricts assessment of biomarker specificity.

Although internal validation using 70–30 and 60–40 split-sample approaches was attempted, smaller subsets yielded unstable probabilities, underscoring the need for larger multi-centre datasets or bootstrapping-based validation to confirm model robustness. Future research should stratify biomarker performance by comorbidity burden, incorporate appropriate control groups, and apply formal validation techniques to enhance reliability and external validity (r1, r2).

Future perspectives

Larger, multi-centre studies are warranted to validate these findings and to evaluate the long-term outcomes of patients stratified by baseline and serial sACE levels. Tracking dynamic changes in sACE during therapy may further clarify its role as a response-monitoring biomarker rather than a static predictor. Combining sACE with other promising serum biomarkers—such as soluble interleukin-2 receptor (sIL-2R), chitotriosidase, or KL-6—could enhance predictive precision and provide a more comprehensive assessment of inflammatory activity (r1, r2).[21,22] Developing standardized protocols for measurement and interpretation will be critical to ensure reproducibility across centres. Ultimately, incorporating sACE as part of a structured, multi-modal approach may help reduce unnecessary corticosteroid exposure, enable timely escalation to steroid-sparing therapy, and optimize patient outcomes.

Another limitation is the absence of a control or comparator group, such as patients with other granulomatous disorders (e.g., tuberculosis, hypersensitivity pneumonitis, or silicosis). Including such groups would have allowed a more direct assessment of the specificity of serum ACE levels for sarcoidosis. Given the known overlap in sACE elevation across granulomatous diseases, our results should be interpreted with caution and viewed as internally comparative rather than disease-specific. Future multi-centre studies incorporating healthy and disease controls are warranted to clarify diagnostic and predictive boundaries.

Implications for practice

Integrating sACE testing into clinical workflows could support early therapeutic decision-making by identifying patients who are less likely to respond to corticosteroids alone. However, the moderate specificity observed in this and prior studies underscores the importance of cautious interpretation. sACE should serve as an adjunctive tool—complementing, not replacing, clinical judgment and multi-disciplinary evaluation. When interpreted alongside clinical features, radiological findings, and other laboratory markers, sACE measurement can contribute meaningfully to individualized, evidence-based management of sarcoidosis.

CONCLUSIONS

This prospective study suggests that elevated serum angiotensin-converting enzyme (sACE) levels (>64 U/L) are associated with a higher likelihood of steroid resistance in sarcoidosis, particularly when accompanied by fatigue, hypercalcemia, or older age. While these findings underscore the potential role of sACE in identifying patients who may require additional immunosuppressive therapy, its moderate specificity and overlap with other granulomatous or metabolic conditions limit its standalone clinical use.

Therefore, sACE should be interpreted as part of a broader, multi-modal framework that integrates clinical, radiological, and laboratory parameters. Larger multi-centre studies with serial biomarker tracking and standardized validation methods are needed to confirm these findings and define the precise role of sACE in guiding personalized, steroid-sparing strategies for sarcoidosis management.

Abbreviations

The following abbreviations are used in this manuscript:

sACE Serum Angiotensin-Converting Enzyme

ILD Interstitial Lung Disease

GCs Glucocorticoids

ATS American Thoracic Society

ERS European Respiratory Society

WASOG World Association for Sarcoidosis and Other Granulomatous Disorders

QoL Quality of Life

FDA Food and Drug Administration

HRCT High-Resolution Computed Tomography

TBNA Transbronchial Needle Aspiration

EBUS-TBNA Endobronchial Ultrasound-Guided Transbronchial Needle Aspiration

EBB Endobronchial Biopsy

TBLB Transbronchial Lung Biopsy

ANA Anti-nuclear Antibody

ATT Anti-Tuberculosis Treatment

CECT Contrast-Enhanced Computed Tomography

sACE Serum Angiotensin-Converting Enzyme

Ethics approval and consent to participate

The study was conducted in accordance with the Declaration of Helsinki, and the protocol was approved by the Ethics Committee of Amrita Institute of Medical Sciences and Research, Kochi (ECASM-AIMS-2022-109) on 07/06/2022.

Conflicts of interest

There are no conflicts of interest.

Funding Statement

Nil.