PET Scan Challenges Invasive Catheter Test for Diagnosing Curable Hypertension

Primary aldosteronism has long been described as the hidden epidemic of cardiovascular medicine. It is the most common form of secondary hypertension, driven by the uncontrolled production of the salt-retaining hormone aldosterone in the adrenal cortex, and it affects far more patients than most clinicians realise. Unlike ordinary high blood pressure, primary aldosteronism carries a heightened risk of heart and kidney damage that is independent of blood pressure readings themselves. Making matters more urgent, a significant share of cases are caused by a benign tumour on just one adrenal gland, which means the disease can often be cured outright by surgically removing that single gland. But before any scalpel is lifted, doctors must answer a deceptively difficult question: which adrenal gland, if either, is the rogue hormone factory? Getting that answer wrong can leave a patient with permanent adrenal insufficiency or leave the real culprit in place.

For decades, the gold standard for answering that question has been adrenal venous sampling, or AVS, a procedure in which an interventional radiologist threads catheters deep into the delicate veins that drain each adrenal gland and draws blood to compare hormone concentrations on the two sides. When it works, AVS is remarkably informative. But it is also technically demanding and invasive, and the right adrenal vein in particular is notoriously difficult to catheterise. Published failure rates for cannulation range from roughly 3.5 to 12 percent, and success depends heavily on the skill and experience of the operator. Inconclusive or failed studies are common enough that many centres cannot offer reliable lateralisation at all, and this limitation has driven an international search for a noninvasive alternative that could match the diagnostic power of catheter-based sampling without its risks and inconsistencies.

A candidate has now emerged from an unexpected corner of molecular imaging. Chemokine receptor CXCR4, a surface protein long studied in oncology and immunology, turns out to be abundantly expressed on aldosterone-producing lesions of the adrenal cortex. A radioactive tracer called gallium-68 pentixafor binds specifically to CXCR4, allowing positron emission tomography combined with computed tomography, or PET/CT, to photograph the hormone-producing tissue in three dimensions. A team at West China Hospital of Sichuan University has now published one of the most clinically grounded head-to-head evaluations of this technique against AVS, and their analysis adds an important twist: instead of simply asking whether the two tests agree with each other, the researchers measured whether each test predicted which patients were actually cured by surgery.

The retrospective cohort study, reported in Health Science Reports, enrolled patients with confirmed primary aldosteronism who underwent unilateral adrenalectomy at West China Hospital between February 2023 and June 2025. After applying strict exclusion criteria, 88 patients whose surgery was guided by AVS and 72 whose surgery was guided by CXCR4 PET/CT entered the final analysis. The two groups were well matched at baseline, with no significant differences in age, body mass index, blood pressure, kidney function, potassium levels, plasma aldosterone concentration, or the aldosterone-to-renin ratio used to confirm the diagnosis. All patients had visible adrenal nodules of at least five millimetres on contrast-enhanced CT, and every participant completed at least six months of postoperative follow-up, giving the investigators a real-world window into whether each imaging strategy translated into better health.

The technical logic of the two tests could hardly be more different. During AVS, performed under digital subtraction angiography guidance, patients received continuous adrenocorticotropic hormone infusion to stabilise hormone secretion, and blood was drawn from both adrenal veins and the inferior vena cava. Successful catheterisation required a cortisol ratio between the adrenal vein and the main trunk of greater than three, and unilateral disease was declared when the lateralisation index, a cortisol-corrected aldosterone ratio between the two glands, exceeded four. CXCR4 PET/CT, by contrast, required only a simple intravenous injection of three to five and a half millicuries of the tracer followed by imaging forty to sixty minutes later. A lesion was deemed positive if its maximum standardised uptake value exceeded 6.5, if its uptake relative to liver exceeded 2.5, or if its uptake relative to the opposite adrenal gland exceeded 2.4, and crucially the bright signal had to correspond to a discrete nodule on CT rather than diffuse glandular activity.

Six months after surgery, the outcomes were strikingly similar between the two groups. Overall, 93.1 percent of all 160 patients achieved biochemical success, meaning their aldosterone and renin values normalised, and 82.5 percent achieved clinical success in terms of blood pressure and potassium control. Complete biochemical success occurred in 84.1 percent of the AVS group and 93.1 percent of the CXCR4 group, a difference that did not reach statistical significance, and complete clinical success rates of 40.9 percent versus 30.3 percent were likewise statistically indistinguishable. Hormone profiles, potassium levels, and complication rates at one and six months after surgery were essentially identical between the groups. In other words, patients whose operations were guided by a painless PET scan fared just as well as those whose operations were guided by an invasive catheter procedure.

The sharpest test of the new imaging came from a subgroup of sixteen patients who underwent both examinations before surgery. Using AVS as the reference standard, CXCR4 PET/CT achieved a sensitivity of 70 percent, a specificity of 83.3 percent, and an overall accuracy of 75 percent. The positive predictive value was a strong 87.5 percent, meaning that when the scan pointed to a gland, it was usually right, but the negative predictive value of 62.5 percent was considerably weaker. Agreement between the modalities was moderate, with a Cohen’s kappa of 0.54. When the investigators instead used the surgically confirmed cure as the ultimate truth, AVS correctly identified the culprit gland in twelve of sixteen cases, an accuracy of 75 percent, while CXCR4 PET/CT was correct in ten of sixteen, or 62.5 percent, with no statistically significant difference between the two.

One particularly illuminating case exposed a subtler vulnerability of the new technique. A 62-year-old man with long-standing hypertension carried three nodules on his left adrenal gland and one on his right. AVS indicated left-sided oversecretion, and the PET scan largely agreed: two of the three left nodules glowed with intensely elevated uptake, while the third left nodule and the right-sided nodule showed low signal. Yet when the resected tissue was examined under the microscope, all three left nodules proved to be aldosterone-producing adenomas on immunohistochemical staining for the aldosterone synthase enzyme CYP11B2, even the one the scan had ignored. The patient was nonetheless cured by removing the left gland, but the discordance revealed that CXCR4 receptor density does not always map perfectly onto actual hormone production at the level of individual lesions, particularly in glands studded with multiple nodules of varying biology.

The authors are careful to frame their findings honestly. The study was not designed as an equivalence or non-inferiority trial, and with modest sample sizes in both arms, the absence of statistical differences may simply reflect insufficient statistical power rather than true equivalence. The retrospective design introduces selection bias, since younger patients appeared more likely to choose the noninvasive scan, and routine pathology relied on standard staining rather than systematic immunohistochemistry. The conclusions also apply only to patients whose nodules are visible on CT; the value of CXCR4 PET/CT in CT-negative disease or in microadenomas smaller than five millimetres remains unknown. The team calls for prospective, multicentre studies to validate standardised quantification thresholds and to test hybrid strategies that combine molecular imaging with selective sampling.

Even with those caveats, the clinical implications are considerable. For patients with a clear solitary nodule, or for those in whom catheterisation is infeasible or has failed, CXCR4 PET/CT now offers evidence-based, radiation-guided lateralisation with a high positive predictive value and outcomes that match catheter-based care. For patients with complex multinodular glands or borderline scan findings, AVS retains an essential role, and the most powerful diagnostic strategy may be one that weaves the two together, using the noninvasive scan as a first line and reserving catheters for the ambiguous cases. What was once a purely mechanical, invasive question about which adrenal vein carries the hormone flood can increasingly be answered with a molecular photograph, and for the millions of hypertensive patients whose curable disease hides in plain sight, that shift could mean faster diagnosis, safer surgery, and a genuine chance at a medication-free life.

Subject of Research: Comparison of adrenal venous sampling and CXCR4-targeted 68Ga-pentixafor PET/CT for diagnosing the unilateral subtype of primary aldosteronism

Article Title: Primary Aldosteronism Subtype Diagnosis: Adrenal Venous Sampling vs. CXCR4 PET CT

Article References: Primary Aldosteronism Subtype Diagnosis: Adrenal Venous Sampling vs. CXCR4 PET CT. (n.d.). https://doi.org/10.1002/edm2.70254

Image Credits: AI Generated

DOI: 10.1002/edm2.70254

Keywords: primary aldosteronism, hypertension, adrenal venous sampling, CXCR4, PET/CT, 68Ga-pentixafor, adrenalectomy, aldosterone, molecular imaging, PASO criteria, CYP11B2, adrenal nodules