Development and validation of model for sparing adrenal venous sampling in diagnosing unilateral primary aldosteronism

Multimodal integration of radiology and pathology signatures for distinguishing between aldosterone-producing adenomas and nonfunctional adrenal adenomas

OBJECTIVE

To develop and validate a nomogram combining radiomics and pathology features to distinguish between aldosterone-producing adenomas (APAs) and nonfunctional adrenal adenomas (NF-AAs).

METHODS

Consecutive patients diagnosed with adrenal adenomas via computed tomography (CT) or pathologic analysis between January 2011 and November 2022 were eligible for inclusion in this retrospective study. CT images and hematoxylin & eosin-stained slides were used for annotation and feature extraction. The selected radiomics and pathology features were used to develop a risk model using various machine learning models, and the area under the receiver operating characteristic curve (AUC) was determined to evaluate diagnostic performance. The predicted results from radiomics and pathology features were combined and visualized using a nomogram.

RESULTS

A total of 211 patients (APAs, n = 59; NF-AAs, n = 152) were included in this study, with patients randomly divided into either the training set or the testing set at a ratio of 8:2. The ExtraTrees model yielded a sensitivity of 0.818, a specificity of 0.733, and an accuracy of 0.756 (AUC = 0.817; 95% confidence interval [CI]: 0.675-0.958) in the radiomics testing set and a sensitivity of 0.999, a specificity of 0.842, and an accuracy of 0.867 (AUC = 0.905, 95% CI: 0.792-1.000) in the pathology testing set. A nomogram combining radiomics and pathology features demonstrated a strong performance (AUC = 0.912; 95% CI: 0.807-1.000).

CONCLUSION

A nomogram combining radiomics and pathology features demonstrated strong predictive accuracy and discrimination capability. This model may help clinicians to distinguish between APAs and NF-AAs.

Treatment decision based on unilateral index from nonadrenocorticotropic hormone-stimulated and adrenocorticotropic hormone-stimulated adrenal vein sampling in primary aldosteronism

OBJECTIVE

Adrenal venous sampling (AVS) is recommended for identifying the subtype of primary aldosteronism before making a surgical treatment decision, but failed cannulation of one adrenal vein is common. To evaluate whether using results of one adrenal vein during AVS could accurately predict unilateral primary aldosteronism.

METHODS

A retrospective study was conducted in primary aldosteronism patients receiving bilaterally or unilaterally successful AVS. The aldosterone-cortisol ratio from the adrenal vein divided by the aldosterone-cortisol ratio from the inferior vena cava (IVC) was calculated as the AV/IVC index.

RESULTS

The study examined 455 patients with primary aldosteronism, including 347 patients with unilateral primary aldosteronism. Among them, 250 and 125 patients received non- adrenocorticotropic hormone (ACTH) and ACTH-stimulated AVS, respectively, and 80 patients received both forms of AVS. Under non-ACTH-stimulated AVS, AUC of the AV/IVC index to diagnose ipsilateral and contralateral primary aldosteronism were 0.778 and 0.924, respectively. The specificity was 100% for both, with sensitivities of 5 and 26%, respectively, when using cutoffs of 17.05 to diagnose ipsilateral primary aldosteronism and 0.15 to diagnose contralateral primary aldosteronism. When using cutoffs of 3.60 and 0.70, the specificity decreased, but if combined with CT results (ipsilateral or contralateral adrenal nodules larger than 10 mm), the specificity could be maintained at 99%, with sensitivities of 33 and 45%, respectively. Under ACTH-stimulated AVS, the AV/IVC index showed similar accuracy to diagnose ipsilateral and contralateral primary aldosteronism.

CONCLUSION

The unilateral AV/IVC index can be used to diagnose unilateral primary aldosteronism during AVS. Combining CT results can increase the accuracy further.

Screening for primary aldosteronism on and off interfering medications

OBJECTIVE

To determine whether antihypertensives will affect diagnostic accuracy of the aldosterone-to-renin ratio (ARR) to an extent that is clinically relevant.

METHODS

Confirmatory tests were used to confirm or exclude PA diagnosis. Area under the receiver operating characteristic curve (AUC), specificity and sensitivity of ARR performance in different conditions were calculated.

RESULTS

208 PA and 78 essential hypertension (EH), and 125 PA and 206 EH patients, were included in the retrospective and prospective cohort, respectively. AUC of ARR on interfering medications was comparable to ARR off interfering medications (retrospective: 0.82 vs. 0.87, p = 0.20; prospective: 0.78 vs. 0.84, p = 0.07). At a threshold of 20 pg/μIU, the sensitivity of ARR on interfering medications was lower (11.1-23.2%) while the specificity was higher (10.2-15.2%) than ARR off interfering medications. However, when the ARR threshold on interfering medications was lowered to 10 pg/μIU, both the sensitivity (retrospective: 0.91 vs. 0.90, p = 0.61; prospective: 0.86 vs. 0.82, p = 0.39) and specificity (retrospective: 0.49 vs. 0.59, p = 0.20; prospective: 0.58 vs. 0.66, p = 0.10) were comparable to the ARR threshold off interfering medications.

CONCLUSION

Using ARR to screen for PA whilst taking interfering antihypertensive drugs is feasible in most cases, but the ARR threshold needs to be reduced.

TRIAL REGISTRATION

ClinicalTrials.gov identifier: NCT04991961.

Identifying primary aldosteronism patients who require adrenal venous sampling: a multi-center study

Adrenal venous sampling (AVS) is crucial for subtyping primary aldosteronism (PA) to explore the possibility of curing hypertension. Because AVS availability is limited, efforts have been made to develop strategies to bypass it. However, it has so far proven unsuccessful in applying clinical practice, partly due to heterogeneity and missing values of the cohorts. For this purpose, we retrospectively assessed 210 PA cases from three institutions where segment-selective AVS, which is more accurate and sensitive for detecting PA cases with surgical indications, was available. A machine learning-based classification model featuring a new cross-center domain adaptation capability was developed. The model identified 102 patients with PA who benefited from surgery in the present cohort. A new data imputation technique was used to address cross-center heterogeneity, making a common prediction model applicable across multiple cohorts. Logistic regression demonstrated higher accuracy than Random Forest and Deep Learning [(0.89, 0.86) vs. (0.84, 0.84), (0.82, 0.84) for surgical or medical indications in terms of f-score]. A derived integrated flowchart revealed that 35.2% of PA cases required AVS with 94.1% accuracy. The present model enabled us to reduce the burden of AVS on patients who would benefit the most.

Validation of Existing Clinical Prediction Tools for Primary Aldosteronism Subtyping

The new clinical prediction score (SCORE) has been recently proposed for primary aldosteronism (PA) subtyping prior to adrenal vein sampling (AVS). This study aimed to compare that SCORE with previously published scores and their validation using a cohort of patients at our center who had had positive SIT confirming PA and had been diagnosed with either bilateral PA according to AVS or unilateral PA if biochemically cured after an adrenalectomy. Final diagnoses were used to evaluate the diagnostic performance of the proposed clinical prediction tools. Only Kamemura's model (with a maximum score of 4 points) and Kobayashi's score (with a maximum score of 12 points) reached 100% reliability for prediction of bilateral PA; however, with sensitivity of only 3%. On the other hand, the values of SCORE = 3 (with sensitivity of 48%), the SPACE score ≥18 (with sensitivity of 35%), the Kobayashi's score ≤2 (with sensitivity of 28%), and the Kocjan's score = 3 (with sensitivity of 28%) were able to predict unilateral PA with 100% probability. Furthermore, Umakoshi's and Young's models both reached 100% reliability for a unilateral PA with score = 4 and both predictive factors together respectively; however, the sensitivity was lower compared with previous models; 4% and 14%, respectively. None of the clinical prediction tools applied to our cohort predicted unilateral and bilateral subtypes together with the expected high diagnostic performance, and therefore can only be used for precisely defined cases.