Morning cortisol in immunotherapy: when to order it

A man being treated for melanoma with immunotherapy arrives for his fifth cycle saying he has been feeling tired. He has lost his appetite, feels dizzy when standing up, and spends much of the day lying down. The team attributes the picture to the cancer and to the treatment itself, which often leave people exhausted. Routine tests show slightly low sodium. Cortisol was drawn in the afternoon, along with the rest, and came back "borderline".

This scenario brings together almost every pitfall of the adrenal axis in immunotherapy. Checkpoint inhibitors can inflame the pituitary or the adrenal gland and reduce cortisol production. The symptoms are vague and overlap with those of cancer, the test was drawn at the wrong time, and the low sodium, an important clue, passed as a minor finding.

This text organizes the reasoning about morning cortisol in people receiving immunotherapy: how the axis is affected, when to order the test, how to read the result, which pitfalls to avoid, and what to do when the patient arrives critically ill.

Key points

• The European guideline recommends morning cortisol, TSH, free T4, glucose and electrolytes before each cycle (Husebye, 2022).
• In one study, only 40 out of every 100 cortisol measurements were drawn at the right time (Druce, 2022).
• Corticosteroids and opioids suppress ACTH and change how the test reads (Husebye, 2022).
• An ACTH within the normal range does not rule out adrenal insufficiency (Ishibashi, 2025).
• The deficiency can appear months after treatment ends (Wang, 2024).
• In the critically ill patient, treatment comes before confirmation.

Summary

• Why this is a reasoning scenario
• How immunotherapy affects the adrenal axis
• When to order morning cortisol
• How to read the result
• Pitfalls in interpretation
• Adrenal crisis and what cannot wait
• How NEXO helps train this reasoning
• Frequently asked questions
• Take-home messages

Why this is a reasoning scenario

Adrenal insufficiency from immunotherapy is a diagnosis that hides behind common symptoms. The European guideline notes that weakness, fatigue, sweating, tachycardia, weight change, nausea, vomiting and abdominal pain are associated with endocrine disorders, but they are also frequent in people undergoing cancer treatment (Husebye, 2022). When everything can be an effect of the cancer, nothing seems to call for investigation.

The risk is real. A meta-analysis found that checkpoint inhibitors increase the chance of severe adrenal insufficiency, with an odds ratio of 3.19, and of hypophysitis, with an odds ratio of 4.77, compared with control groups (Lu, 2019). Untreated cortisol deficiency can progress to adrenal crisis, a life-threatening emergency.

The test that helps, cortisol, looks simple, but it is easy to misread. It varies throughout the day, is affected by medications, and has thresholds that depend on the laboratory method. The reasoning must consider the time of collection, what the patient takes, and what the result really means.

There is also an attribution bias. The oncology team sees many tired patients, and tiredness becomes part of the landscape. The question that changes the outcome is simple: is this tiredness expected, or is there a treatable hormonal cause behind it?

How immunotherapy affects the adrenal axis

Cortisol is produced by the adrenal gland under the command of ACTH, released by the pituitary. Immunotherapy can affect either point, and the reasoning changes depending on the site.

Hypophysitis and isolated ACTH deficiency

The most common form is secondary adrenal insufficiency, caused by inflammation of the pituitary. The guideline describes that, with anti-CTLA-4, hypophysitis usually appears on average 10.5 weeks after the start, and with anti-PD-1 or anti-PD-L1, on average 27 weeks (Husebye, 2022).

The two forms are different. With anti-CTLA-4, the pituitary usually enlarges and several hormones may fall; with anti-PD-1, isolated ACTH deficiency is more common. In the form associated with anti-PD-1, hyponatremia is one of the most common findings, present in 63 out of every 100 patients, and MRI shows an abnormality in only 18 out of every 100 (Husebye, 2022).

A systematic review gathered 123 patients with isolated ACTH deficiency due to immunotherapy (Wang, 2024). The diagnosis came after a median of 8 cycles, or 6 months of treatment, and 11 patients developed the condition 1 to 11 months after finishing immunotherapy (Wang, 2024). Fatigue and loss of appetite were the most common symptoms, and 88 patients had a normal pituitary MRI (Wang, 2024).

Primary adrenal insufficiency

Rarer, inflammation of the adrenal gland itself produces primary adrenal insufficiency. In it, cortisol falls and ACTH rises, because the pituitary tries to compensate. The American oncology guideline recommends measuring ACTH and cortisol in the morning and using the ratio between the two to separate the primary form from the secondary one (Brahmer, 2018).

The distinction has practical consequences. In the primary form, mineralocorticoid production can also fall, which leads to more marked hyperkalemia, dehydration and hypotension. In the secondary form, mineralocorticoid is usually preserved, and hyponatremia comes mainly from the effect of low cortisol on water. The text on urine volume and sodium in hyponatremia helps organize this reading.

Who is at higher risk

The combination of anti-PD-1 with anti-CTLA-4 is the highest-risk scenario. In a study of 373 patients, 7 of the 24 who received the combination developed secondary adrenal insufficiency, with a median of 8 weeks (Manaka, 2021). In 4 of these 7, the diagnosis came from regular ACTH and cortisol tests, when symptoms were still mild (Manaka, 2021). The same study described 2 patients, among 15 who continued on nivolumab alone after the combination, who developed isolated ACTH deficiency during the maintenance phase (Manaka, 2021). The risk, therefore, continues after anti-CTLA-4 leaves the regimen, and screening should continue during maintenance.

The European guideline adds that recovery from ACTH deficiency is extremely rare (Husebye, 2022). In most patients, glucocorticoid replacement will be permanent, which makes education about the disease a central part of care. Other immune-mediated events are discussed in the text on immune-mediated toxicity.

The symptoms of hypophysitis

The clinical presentation of immunotherapy-induced hypophysitis is nonspecific. A narrative review lists asthenia, lack of appetite, headache, vomiting, weight loss, low blood pressure, dizziness, reduced libido, hot flashes and, more rarely, visual changes (Di Stasi, 2025). Almost all of these symptoms also appear in advanced cancer or as an effect of other treatments.

Headache deserves special attention in those receiving anti-CTLA-4, because the enlarged pituitary can cause pain and, in rare cases, compress the visual pathways. The drop in libido and hot flashes point to gonadotropin deficiency and reinforce the suspicion that the entire pituitary was affected. Actively asking about these symptoms, which the patient rarely brings up on their own, changes the probability even before the exam.

When to order morning cortisol

The 2022 European guideline proposes a minimum set of tests before and during treatment. The set includes TSH, free T4, cortisol, glucose and electrolytes collected in the morning, between 8 and 9 a.m., with attention to any corticosteroid in use (Husebye, 2022). These tests should be repeated every 4 to 6 weeks, preferably before each cycle (Husebye, 2022).

In the more complete initial assessment, the guideline also includes ACTH, sex hormones, prolactin and glycated hemoglobin, collected in the morning (Husebye, 2022). This baseline allows values to be compared later, when a doubt arises.

Beyond scheduled screening, some situations call for cortisol right away:

• new or disproportionate fatigue, especially with loss of appetite;
• dizziness on standing, hypotension or a drop in blood pressure in someone who was hypertensive;
• low sodium without another clear explanation;
• nausea, vomiting or abdominal pain without a defined cause;
• new headache or visual change in someone receiving anti-CTLA-4;
• hypoglycemia in someone who does not use diabetes medication;
• any of these symptoms months after the end of immunotherapy.

Regular screening works. In a study of 265 patients, 22 had adrenal insufficiency: 8 were identified by screening and 14 only when they were already symptomatic, after a visit to the emergency department or a hospital admission (Druce, 2022). The authors attribute part of the failures to the inadequate timing of collection (Druce, 2022).

The timing of collection

Cortisol is highest in the early morning and falls over the course of the day. A normal value at 8 a.m. may look low at 4 p.m., and a low value in the afternoon may be physiological. In the study of 265 patients, of the 1301 cortisol measurements, only 40 in every 100 were collected between 6 and 10 a.m., and the others could not be interpreted (Druce, 2022).

The authors acknowledge that screening faces challenges: the incidence is low, the presentation is insidious, the test depends on the timing and is affected by external corticosteroids (Druce, 2022). They therefore propose an algorithm with the correct collection procedure and the interpretation of cortisol, and argue that the team should be trained to use it (Druce, 2022).

This is a problem of organization, not of knowledge. In many services, routine tests are collected at the time of the infusion, which may be in the afternoon. Adjusting the workflow so that cortisol is collected early, even if on another day, turns a useless test into a test that guides management.

How to read the result

The reading combines the cortisol value, ACTH and the context. The thresholds vary according to the assay used by the laboratory, and the interpretation must take the method into account.

Most used thresholds

The emergency guidance from the British endocrinology society proposes that a cortisol above 450 nmol/L (16.3 µg/dL), collected before any glucocorticoid, rules out deficiency (Higham, 2018). In a patient with few symptoms, a 9 a.m. cortisol below 200 nmol/L (7.2 µg/dL), or a random one below 100 nmol/L (3.6 µg/dL), indicates starting hydrocortisone (Higham, 2018).

The European guideline on endocrine effects of immunotherapy notes that the cutoff depends on the assay, and that immunoassays usually give readings slightly higher than tandem mass spectrometry (Husebye, 2022). In the stimulation test, the same guideline considers a cortisol above 485 nmol/L (17.6 µg/dL) at 60 minutes, measured by mass spectrometry, to be normal (Husebye, 2022). The values in this text are therefore a guide, and the report is valid with the reference of the laboratory that issued it.

Newer assays read cortisol differently. A study with new-generation assays, in 95 patients evaluated with a stimulation test, found that a morning cortisol of 306 nmol/L (11.1 µg/dL) had a specificity of 93 in every 100 for predicting a normal response to the test (Shrestha, 2026). The authors note that, with these assays, the old thresholds generate false positives and unnecessary tests, and recommend that each service know the thresholds of the assay it uses (Shrestha, 2026).

Another study, with 344 patients, proposed that a morning cortisol above 332 nmol/L (12.0 µg/dL) identifies adequate reserve and may make the stimulation test unnecessary (Almazrouei, 2025). Values below 96 nmol/L (3.5 µg/dL) corresponded to an insufficient response to the test (Almazrouei, 2025).

The table summarizes the reading in three bands, with the caveat that the numbers depend on the assay.

Band	Example threshold	Reading	Next step
High	Above 450 nmol/L (16.3 µg/dL) (Higham, 2018)	Deficiency unlikely	Continue screening
Intermediate	Between the assay thresholds	Indeterminate	ACTH stimulation test
Low	Below 200 nmol/L (7.2 µg/dL) at 9 a.m. (Higham, 2018)	Deficiency likely	ACTH, treat and call endocrinology
Any, critically ill patient	Without waiting for the result	Treat as deficient	Collect and treat at the same time

The role of ACTH

ACTH helps locate the lesion. Low cortisol with high ACTH suggests the primary form; low cortisol with low or inappropriately normal ACTH suggests the secondary form (Brahmer, 2018). The guideline recommends a stimulation test with synthetic ACTH when the results are indeterminate (Brahmer, 2018).

ACTH within the normal range is misleading. In one study, 14 in every 100 patients with adrenal insufficiency due to immunotherapy had preserved ACTH, and all of those who underwent the stimulation test showed reduced adrenal reserve (Ishibashi, 2025). A normal ACTH with low cortisol should be read as inappropriate.

The pituitary image

Pituitary MRI is mandatory when symptoms are suspicious and remains the first-choice imaging test in hypophysitis (Di Stasi, 2025). In a study of 66 patients with pituitary deficiency, the initial MRI showed gland enlargement in 25 out of 100 and empty sella in 8.3 out of 100 (Angelousi, 2026).

The European guideline notes that a normal MRI does not rule out hypophysitis (Husebye, 2022). Imaging helps confirm and exclude other causes, such as metastasis, but the diagnosis is biochemical. The text on what the adrenal can produce reviews the hormonal physiology of the gland.

Pitfalls in interpretation

Several situations common in oncology distort cortisol and ACTH. Recognizing them avoids both the missed diagnosis and unnecessary treatment:

• collection outside the morning window, which makes the value uninterpretable;
• use of glucocorticoid as an antiemetic, for edema or to treat another immune-mediated event, which suppresses the axis;
• use of opioids, which also suppress ACTH (Husebye, 2022);
• ACTH within the normal range with low cortisol, which is still a deficiency;
• thresholds from one assay applied to another;
• starting levothyroxine before assessing cortisol.

Corticosteroid in use

External glucocorticoid changes everything. The European guideline calls for attention to any recent or current corticosteroid when reading cortisol (Husebye, 2022). Some corticosteroids show up in the test and raise the measured value; others, such as dexamethasone, suppress the axis and go unnoticed by most assays (Brahmer, 2018).

The British guidance considers it crucial to obtain a good history of recent glucocorticoid use to interpret the result (Higham, 2018). It is also worth asking about less-remembered routes, such as inhaled, topical and joint injections. Those who received a corticosteroid to treat another immune-mediated event may have axis suppression that adds to the pituitary injury.

Thyroid and adrenal together

When immunotherapy affects more than one gland, the order of treatment matters. The European guideline recommends excluding adrenal insufficiency before starting levothyroxine, because levothyroxine can precipitate an adrenal crisis in those with untreated insufficiency (Husebye, 2022). A low or normal TSH with low free T4 suggests pituitary injury and calls for cortisol (Husebye, 2022).

Adrenal crisis and what cannot wait

Adrenal crisis is the most serious complication. The estimate is 8 crises per 100 patient-years among people with adrenal insufficiency, and infections are the main triggering factor (Dineen, 2019). Surgery, trauma, forgetting the medication, intense physical activity and psychological stress can also precipitate the condition (Dineen, 2019).

The clinical picture of the crisis is broad. The review describes hypotension and hypovolemia as the predominant findings, especially in the primary form, but notes that patients often arrive with nonspecific symptoms, such as lack of appetite, nausea, vomiting, abdominal pain, fatigue, lethargy, fever or altered consciousness (Dineen, 2019). On tests, hyponatremia, hypoglycemia, hypercalcemia, anemia and acute kidney injury may appear, and hyperkalemia suggests the primary form (Dineen, 2019).

In the severely ill patient, the reasoning reverses. The British guidance recommends that every patient on immunotherapy who arrives severely ill be treated as if they had cortisol deficiency until proven otherwise (Higham, 2018). Cortisol and ACTH should be collected before the glucocorticoid, but collection cannot delay treatment.

The American guideline suggests that, when the diagnosis is not yet clear and a stimulation test will be needed later, the choice of glucocorticoid for the emergency takes this into account, since dexamethasone is not measured by most assays (Brahmer, 2018). The decision should be shared with endocrinology.

The picture of the crisis can be mistaken for sepsis, dehydration or disease progression. Hypotension that responds poorly to volume, vomiting, abdominal pain, confusion and hyponatremia form a pattern that should bring the adrenal to mind. The text on perfusion beyond pressure helps assess the unstable patient.

Prevention

Prevention relies on patient education (Dineen, 2019). Anyone with adrenal insufficiency should know how to adjust the glucocorticoid on sick days, when to seek care and to carry a steroid dependency alert card (Dineen, 2019). The European guideline reinforces the emergency card for all patients on replacement (Husebye, 2022).

The oncology team plays an important role at this stage. The patient spends more time with oncology than with endocrinology, and it is in the infusion room that many questions come up. Making sure the patient knows what to do in case of fever, vomiting or surgery reduces the risk of a crisis. The text on the risks of immunotherapy before surgery shows another moment when this information matters.

Back to the patient from the opening

The man with melanoma has fatigue, loss of appetite, postural dizziness and low sodium, and the cortisol was drawn in the afternoon. The first step is to repeat the cortisol with ACTH in the morning, before any glucocorticoid, and to ask about recent steroids and opioids. If he is unstable, treatment comes first, with collection beforehand.

If the morning cortisol comes back low and the ACTH low or normal, the hypothesis is secondary deficiency, probably hypophysitis or isolated ACTH deficiency. MRI helps complete the assessment, but a normal result does not change management. If the sodium corrects and the energy returns after replacement, the hypothesis gains clinical strength, beyond the biochemistry. Endocrinology joins the follow-up, and the patient leaves with guidance for sick days and an alert card.

Follow-up does not end with the diagnosis. The European guideline describes recovery of ACTH production after immunotherapy-induced hypophysitis as very rare, which makes replacement, in most cases, a permanent treatment (Husebye, 2022). The team should therefore explain to the patient that replacement continues even after immunotherapy ends. Stopping cancer treatment also does not end the risk of new cases, since some diagnoses appear months after the last infusion (Wang, 2024). Anyone who completed immunotherapy and returns with fatigue, loss of appetite or low sodium deserves the same question about cortisol. At follow-up visits, reviewing the alert card and the guidance for sick days keeps prevention alive (Dineen, 2019).

How NEXO helps train this reasoning

In NEXO, the case of the day is fictional, with a real clinical presentation and up to six progressive clues, and you propose the diagnosis by searching for the ICD-10 code.

The limits on using the game are described on the NEXO economy page. If you research the teaching of clinical reasoning, the page for researchers explains how to get in touch. Professionals who write or review cases find information under curation and under creators, and institutions can learn about the proposal on the partnerships page. Training differential diagnosis with clues and searching for the right ICD-10 code complete this practice. The reading on fever and other effects in melanoma therapy broadens the context.

Frequently asked questions

When should cortisol be ordered in someone receiving immunotherapy?

The European guideline recommends morning cortisol, with TSH, free T4, glucose and electrolytes, before each cycle or every 4 to 6 weeks (Husebye, 2022).

Why does the collection time matter?

Cortisol falls over the course of the day. In one study, 60 out of every 100 measurements were collected outside the morning window and could not be interpreted (Druce, 2022).

Does a normal ACTH rule out adrenal insufficiency?

Barely. In 14 out of every 100 patients with adrenal insufficiency due to immunotherapy, ACTH was in the normal range (Ishibashi, 2025).

Does a normal MRI exclude hypophysitis?

It rules out little. In a review of 123 patients with isolated ACTH deficiency, 88 had a normal MRI (Wang, 2024).

Can adrenal insufficiency appear after treatment ends?

It can. Eleven patients developed isolated ACTH deficiency 1 to 11 months after finishing immunotherapy (Wang, 2024).

What should be done with the severely ill patient?

Treat as cortisol deficient until proven otherwise, drawing cortisol and ACTH before the glucocorticoid without delaying treatment (Higham, 2018).

Take-home messages

• Checkpoint inhibitors can cause secondary or primary adrenal insufficiency, with symptoms that are confused with those of cancer.
• Cortisol should be drawn in the morning, and a value drawn in the afternoon rarely guides management.
• Corticosteroids and opioids suppress the axis and must be part of how the test is read.
• A normal ACTH and a normal MRI are far from ruling out the diagnosis.
• The combination of anti-PD-1 with anti-CTLA-4 is the highest-risk scenario.
• In the severely ill patient, the glucocorticoid comes before confirmation, with a prior draw.

Want to train your reasoning in the face of fatigue and low sodium? Play the case of the day on NEXO.

References

1. Almazrouei R, Alshamsi J, Abdelsalhen M, et al. Diagnostic performance of morning basal serum cortisol for assessing adrenal reserve: a single-center experience from the United Arab Emirates. Ther Adv Endocrinol Metab. 2025. PMID 40351888. DOI 10.1177/20420188251335180.
2. Angelousi A, Asimakopoulou S, Bourgioti C, et al. Magnetic resonance imaging features of hypophysitis in patients with cancer treated with immune checkpoint inhibitors. J Neuroendocrinol. 2026. PMID 41243162. DOI 10.1111/jne.70116.
3. Brahmer JR, Lacchetti C, Schneider BJ, et al. Management of Immune-Related Adverse Events in Patients Treated With Immune Checkpoint Inhibitor Therapy: American Society of Clinical Oncology Clinical Practice Guideline. J Clin Oncol. 2018. PMID 29442540. DOI 10.1200/JCO.2017.77.6385.
4. Di Stasi V, La Sala D, Cozzi R, et al. Immunotherapy-Related Hypophysitis: A Narrative Review. Cancers (Basel). 2025. PMID 39941803. DOI 10.3390/cancers17030436.
5. Dineen R, Thompson CJ, Sherlock M. Adrenal crisis: prevention and management in adult patients. Ther Adv Endocrinol Metab. 2019. PMID 31223468. DOI 10.1177/2042018819848218.
6. Druce I, Tawagi K, Shaw JLV, et al. Routine Screening for Central and Primary Adrenal Insufficiency during Immune-Checkpoint Inhibitor Therapy: An Endocrinology Perspective for Oncologists. Curr Oncol. 2022. PMID 35877230. DOI 10.3390/curroncol29070370.
7. Higham CE, Olsson-Brown A, Carroll P, et al. Society for Endocrinology Endocrine Emergency Guidance: Acute management of the endocrine complications of checkpoint inhibitor therapy. Endocr Connect. 2018. PMID 29930025. DOI 10.1530/EC-18-0068.
8. Husebye ES, Castinetti F, Criseno S, et al. Endocrine-related adverse conditions in patients receiving immune checkpoint inhibition: an ESE clinical practice guideline. Eur J Endocrinol. 2022. PMID 36149449. DOI 10.1530/EJE-22-0689.
9. Ishibashi Y, Baba R, Okada A, et al. Normal adrenocorticotropic hormone levels do not exclude adrenal insufficiency during immune checkpoint inhibitor therapy: evidence from clinical, steroid, and structural analyses. Front Endocrinol (Lausanne). 2025. PMID 41189622. DOI 10.3389/fendo.2025.1683546.
10. Lu J, Li L, Lan Y, et al. Immune checkpoint inhibitor-associated pituitary-adrenal dysfunction: A systematic review and meta-analysis. Cancer Med. 2019. PMID 31679184. DOI 10.1002/cam4.2661.
11. Manaka K, Sato J, Takeuchi M, et al. Immune checkpoint inhibitor combination therapies very frequently induce secondary adrenal insufficiency. Sci Rep. 2021. PMID 34078988. DOI 10.1038/s41598-021-91032-6.
12. Shrestha E, Albert SG, Ghimire A, et al. Newer Generation Immunoassays in Screening for Adrenal Insufficiency: Comparison of Baseline Morning Cortisol Levels With Cosyntropin Stimulation Testing. AACE Endocrinol Diabetes. 2026. PMID 42221401. DOI 10.1016/j.aed.2026.03.001.
13. Wang F, Shi X, Yu X, et al. Immune checkpoint inhibitor-induced isolated adrenocorticotropic hormone deficiency: a systematic review. Front Endocrinol (Lausanne). 2024. PMID 38318292. DOI 10.3389/fendo.2024.1326684.