Do We Really Need That Pre-ACTH Cortisol? Large UK Study Challenges Standard Addison's Testing Protocol
New data suggests we might be able to skip the baseline sample in ACTH stim tests—and save clients money without sacrificing accuracy
If you've been practicing for any length of time, you've run countless ACTH stimulation tests. The protocol is burned into our brains: draw baseline cortisol, give synthetic ACTH, wait an hour, draw post-ACTH cortisol. But what if I told you that baseline sample might be unnecessary? A comprehensive study published in Frontiers in Veterinary Scienceanalyzing over 1,000 ACTH stimulation tests from UK veterinary practices suggests the pre-ACTH cortisol adds no diagnostic value when you're already running the full stimulation test.
Led by Armando C. Sánchez-Lara and colleagues at the University of Cambridge, this research challenges our standard approach to diagnosing hypoadrenocorticism and raises important questions about test efficiency, cost-effectiveness, and diagnostic stewardship.
The Standard Wisdom on Basal Cortisol
Current teaching tells us that a basal (pre-ACTH) cortisol >55 nmol/L (~2 μg/dL) effectively rules out hypoadrenocorticism (HA), allowing us to skip the full ACTH stimulation test in many cases. This cutoff has become widely accepted as a screening tool, particularly useful when Addison's disease is on your differential list but not at the top.
The logic is sound: if the adrenal glands are producing adequate cortisol at baseline, they're probably not failing. A high resting cortisol means you can save the client money by avoiding the expensive ACTH stimulation portion of the test.
But here's the catch: previous studies establishing this cutoff were predominantly from referral populations, where disease prevalence tends to be higher and case selection may be biased toward more obvious presentations. How well does this rule-out threshold perform in first-opinion practice, where we see undifferentiated cases and the prevalence of HA is presumably lower?
That's exactly what this study set out to determine.
The Study Design: Real-World Data at Scale
The researchers analyzed 1,017 ACTH stimulation tests performed between January 2019 and April 2023 at a large UK veterinary diagnostic laboratory. After excluding cases being tested for hypercortisolism (Cushing's disease), they had 878 cases to work with:
170 cases from a referral center (RC) with complete clinical histories
708 cases predominantly from first-opinion (FO) practices without detailed clinical information
This is a significant strength of the study—it captures real-world testing patterns across different practice settings. The FO cohort is particularly valuable because it reflects how most of us use ACTH stimulation testing: in general practice, on cases with varying degrees of clinical suspicion for Addison's.
All cortisol measurements were performed using radioimmunoassay (RIA), and hypoadrenocorticism was defined by the standard criterion: post-ACTH cortisol ≤55 nmol/L.
The Prevalence Question
Disease prevalence matters enormously when interpreting diagnostic tests, and this study revealed interesting differences:
Referral center (RC): 8.4% HA prevalence
First-opinion (FO): 4.4% HA prevalence
Combined (RC + FO): 5.1% HA prevalence
The higher prevalence in referral settings isn't surprising—these are pre-selected cases where clinicians had high enough suspicion to seek specialty consultation. The 4.4% prevalence in first-opinion practice is more representative of what most of us encounter when we run ACTH stim tests on our differentials.
This prevalence difference has important implications for test performance metrics, particularly positive and negative predictive values, which are prevalence-dependent.
Evaluating the Standard 55 nmol/L Cutoff
The researchers first evaluated the traditional basal cortisol cutoff of ≤55 nmol/L for identifying HA:
In the referral center cohort:
Sensitivity: 93% (catches most cases)
Specificity: 77% (lots of false positives)
Negative predictive value (NPV): 99% (great rule-out test)
Positive predictive value (PPV): 25% (poor rule-in test)
This confirms what we already knew: a basal cortisol >55 nmol/L is excellent for ruling out HA (99% NPV), but a value ≤55 nmol/L doesn't confirm the diagnosis (only 25% actually have HA).
The high sensitivity and NPV make biological sense. Dogs with failing adrenal glands generally can't maintain normal baseline cortisol levels. If they can, they probably don't have Addison's.
The Problem: Too Many False Positives
Here's where things get interesting. While the 55 nmol/L cutoff works well as a rule-out, its relatively low specificity (77%) means about one in four dogs with basal cortisol ≤55 nmol/L actually have normal adrenal function. These are false positives—dogs we'd unnecessarily subject to full ACTH stimulation testing based on the screening cutoff.
In clinical terms, this means if you use basal cortisol ≤55 nmol/L as your threshold for proceeding with ACTH stimulation, you're going to run a lot of unnecessary full tests on dogs that don't have HA.
Could a Lower Cutoff Work Better?
The researchers explored whether lowering the basal cortisol cutoff to ≤22 nmol/L might improve diagnostic performance:
In the referral center cohort:
Sensitivity: 93% (unchanged—still catches the same cases)
Specificity: 92% (much better—fewer false positives)
NPV: 99% (unchanged—still excellent rule-out)
PPV: 50% (improved—half of positives actually have HA)
This lower cutoff improved specificity dramatically while maintaining sensitivity. Essentially, it better identified which dogs truly needed the full ACTH stimulation test.
However, when applied to the first-opinion cohort or the combined dataset, the ≤22 nmol/L cutoff showed slightly reduced sensitivity (90-91%), meaning it might miss a few more cases compared to the 55 nmol/L threshold.
The Game-Changer: Post-ACTH Cortisol Alone
Here's where the study gets really interesting. The researchers used receiver operating characteristic (ROC) analysis to evaluate diagnostic performance:
Basal cortisol area under curve (AUC): 0.93-0.97 (excellent)
Post-ACTH cortisol AUC: 1.0 (perfect)
That's right—post-ACTH cortisol showed perfect diagnostic accuracy across all datasets. When you know the post-ACTH cortisol value, the pre-ACTH value provides zero additional diagnostic information.
Think about what this means: if you're running a full ACTH stimulation test anyway, the baseline cortisol measurement is redundant. The post-ACTH value alone perfectly discriminates between dogs with and without hypoadrenocorticism.
Why This Matters for Your Practice
This finding has significant practical implications:
1. Cost Savings
ACTH stimulation tests are expensive—running two cortisol measurements instead of one adds unnecessary cost. In practices where clients are price-sensitive or when running ACTH stims on multiple differentials, eliminating the baseline measurement could make testing more accessible.
The cost savings might seem modest per test, but they add up across a busy practice, and any reduction in diagnostic costs without compromising accuracy benefits both clients and patient care.
2. Simplified Protocols
Fewer samples mean less handling, less chance of sample mix-up, simplified submission paperwork, and reduced labor costs. These operational efficiencies matter, especially in high-volume practices.
3. No Loss of Diagnostic Accuracy
This is the critical point: you're not sacrificing anything diagnostically. The post-ACTH cortisol ≤55 nmol/L criterion provides perfect discrimination between HA and normal adrenal function. The baseline value genuinely doesn't add information when you have the post-ACTH result.
But What About Using Basal Cortisol as a Screening Test?
Here's where clinical judgment comes in. The study confirms that basal cortisol >55 nmol/L is indeed an excellent screening tool to avoid running full ACTH stimulation tests:
When to use basal cortisol screening:
HA is on your differential but not strongly suspected
You want to rule out HA before pursuing other diagnostics
Cost is a significant client concern
You're working up vague signs (weakness, GI signs, lethargy) where HA is possible but many other differentials exist
A basal cortisol >55 nmol/L effectively rules out HA (99% NPV), allowing you to confidently pursue other diagnostic directions without the expense of ACTH stimulation.
When to skip straight to ACTH stimulation:
HA is strongly suspected based on clinical presentation
You need a definitive answer regardless of baseline cortisol
The patient is critically ill and needs rapid diagnosis
You're in a referral setting where cases are pre-selected
In these scenarios, just run the post-ACTH sample. Skip the baseline measurement—it's not adding value.
The Clinical Decision Tree
Based on this study, here's a rational approach to hypoadrenocorticism testing:
Scenario 1: HA on differential but not top suspect
Run resting cortisol as screening test
If >55 nmol/L: HA effectively ruled out, pursue other differentials
If ≤55 nmol/L: Proceed to ACTH stimulation (post-ACTH sample only)
Scenario 2: HA strongly suspected
Skip resting cortisol entirely
Run ACTH stimulation test (post-ACTH sample only)
Interpret based on post-ACTH cortisol ≤55 nmol/L cutoff
Scenario 3: Critical patient, immediate decision needed
If patient stable enough: run ACTH stim (post-ACTH only)
If unstable: treat empirically while awaiting results
This approach maximizes diagnostic efficiency while minimizing unnecessary testing and cost.
Important Caveats and Considerations
Assay Differences Matter
This study used radioimmunoassay (RIA) for cortisol measurement. Different assays (chemiluminescent immunoassays, for example) may have different performance characteristics and cutoffs. The specific 55 nmol/L threshold might not translate perfectly across all laboratory platforms.
Always know what assay your reference laboratory uses and what their validated cutoffs are. Don't assume cutoffs from one lab apply to another.
Atypical Addison's
This study focuses on typical hypoadrenocorticism with both glucocorticoid and mineralocorticoid deficiency. Atypical HA (glucocorticoid deficiency alone) represents about 10-30% of cases and may have different testing considerations, though the post-ACTH cortisol criterion should still apply.
Clinical Context Still Matters
No test exists in a vacuum. ACTH stimulation testing should always be interpreted alongside clinical signs, electrolyte abnormalities (hyponatremia, hyperkalemia), azotemia, and other laboratory findings. A dog with classic clinical and laboratory features of Addison's doesn't need ACTH testing—treat first, test later if needed.
Conversely, a mildly low post-ACTH cortisol in a dog with normal electrolytes and vague signs might warrant repeat testing or investigation for other causes.
The Sick Euthyroid Problem
Critically ill dogs from any cause can have low baseline cortisol due to "sick euthyroid syndrome" or inadequate adrenal reserve in the face of severe illness. This is why post-ACTH cortisol is so valuable—it assesses reserve capacity, not just baseline function.
A sick dog with low baseline cortisol but normal post-ACTH response doesn't have primary hypoadrenocorticism, even though baseline cortisol might suggest it.
What About Dogs Already on Glucocorticoids?
This study doesn't address the common clinical scenario where dogs have received glucocorticoids prior to testing. Exogenous steroids suppress the HPA axis and can affect both baseline and post-ACTH cortisol measurements.
Standard recommendations remain:
Avoid prednisone/prednisolone for at least 24-48 hours before testing
Dexamethasone has minimal cross-reactivity with most cortisol assays and can be used for stabilization
If testing can't be delayed, proceed anyway—severely low post-ACTH cortisol despite recent steroid administration is highly suggestive of HA
Implications for Teaching and Protocols
This study suggests we should update our teaching and standard protocols. Many veterinary textbooks and laboratory submission forms still list ACTH stimulation testing as requiring both pre- and post-ACTH samples. This research provides evidence to simplify:
Updated ACTH stimulation protocol:
Give synthetic ACTH (cosyntropin or tetracosactide)
Collect blood sample 60 minutes post-injection
Measure cortisol
Interpret: ≤55 nmol/L (~2 μg/dL) = hypoadrenocorticism
That's it. No baseline sample needed.
The Counterargument: Why Might We Keep Measuring Baseline?
Despite this compelling data, there are potential arguments for maintaining baseline cortisol measurement:
Quality control: Having both measurements allows identification of sample handling errors or unexpected results
Documentation: Some clinicians value having baseline data for medical records or second opinions
Academic interest: Teaching hospitals may want complete data sets for research
Client communication: Some clients appreciate seeing "before and after" numbers
These are valid considerations, but they're not diagnostic reasons. They're about documentation, quality assurance, and communication—not about improving diagnostic accuracy.
The Bottom Line
This large, real-world dataset from UK veterinary practices provides strong evidence for reconsidering how we approach ACTH stimulation testing for canine hypoadrenocorticism:
Key findings:
Basal cortisol >55 nmol/L is an excellent screening test to rule out HA in both referral and first-opinion settings (99% NPV)
Post-ACTH cortisol alone provides perfect diagnostic accuracy (AUC 1.0)
Pre-ACTH cortisol adds no diagnostic value when post-ACTH cortisol is measured
Omitting baseline cortisol reduces cost and simplifies testing without compromising diagnostic performance
Clinical recommendations:
Use resting cortisol (>55 nmol/L) as a screening tool when HA is possible but not strongly suspected
When running ACTH stimulation tests, skip the baseline sample and measure only post-ACTH cortisol
Interpret post-ACTH cortisol ≤55 nmol/L as diagnostic for hypoadrenocorticism
This isn't about cutting corners—it's about practicing evidence-based medicine and eliminating redundant testing that doesn't improve patient care. When research clearly demonstrates that a component of our standard protocol adds no diagnostic value, we should be willing to update our approach.
Your clients will appreciate the cost savings, your staff will appreciate the simplified protocol, and your patients will receive the same diagnostic accuracy with one less needle stick. That's a win all around.
Full Text Link: Reassessing the role of basal cortisol in ACTH stimulation testing for canine hypoadrenocorticism: insights from a large UK referral and first-opinion dataset - Frontiers in Veterinary Science, PMCID: PMC12571564
