17-OH Progesterone Testing With LC-MS: A Complete Clinical Guide to Accuracy, Ranges, and Workflow

17-OH progesterone testing with LC-MS (liquid chromatography–tandem mass spectrometry) measures the steroid 17-hydroxyprogesterone in serum or dried blood spots using chromatographic separation followed by mass-selective detection. This method delivers the specificity needed to distinguish 17-OHP from structurally similar steroids, making it the preferred approach for diagnosing and monitoring congenital adrenal hyperplasia (CAH). Results are reported in ng/dL or nmol/L, with typical adult reference ranges near 20 to 200 ng/dL depending on follicular versus luteal phase. Testing turnaround usually runs 2 to 5 days.

Why Conventional Immunoassays Fail Where 17-OH Progesterone LC-MS Succeeds

Immunoassays identify 17-OHP with an antibody. That antibody does not bind one molecule; it binds a chemical neighborhood. In serum, that neighborhood contains cross-reactants whose concentrations can exceed 17-OHP itself in certain patient groups.

The dominant cross-reactant in newborns is 17-hydroxypregnenolone sulfate, a fetal-zone steroid produced in high quantities during the first weeks of life. A second trigger is 11-deoxycortisol, which accumulates in patients treated with metyrapone, and 21-deoxycortisol, which accumulates in untreated CAH. Immunoassay platforms report these as 17-OHP, inflating results by 30 to 400 percent in published newborn screening comparisons.

LC-MS resolves the problem at the physical level. A liquid chromatograph first separates molecules by retention time on a reverse-phase column. Only then does the mass spectrometer count ions at a specific mass-to-charge transition. A cross-reactant with a different retention time or a different fragmentation pattern is simply not counted.

In plain terms: immunoassays estimate, LC-MS identifies.

The practical consequence is a lower limit of quantification. Modern LC-MS/MS assays reach a lower limit of quantification around 10 to 20 ng/dL, while many immunoassays cannot reliably report below 50 to 100 ng/dL. For neonatal CAH screening, where the diagnostic cutoff sits near 200 ng/dL in term infants, that difference changes which samples get flagged.

The detection step also depends on the ionization method. Most clinical labs use positive-mode electrospray ionization, targeting the m/z 331.2 to 313.2 or 331.2 to 97.1 transitions. Deuterated internal standards (17-OHP-d8) correct for extraction efficiency and matrix effects.

Quantified Accuracy: How Much LC-MS and Immunoassay Really Diverge

The gap between platforms is not anecdotal. It is measurable, and the size of the gap depends on the patient population.

In healthy adults with normal 17-OHP, immunoassay and LC-MS agree within roughly 15 to 25 percent. In neonates, the divergence is far larger. A widely cited comparison found that 60 to 70 percent of initially elevated newborn immunoassay results dropped below diagnostic thresholds on LC-MS/MS reanalysis. That is not a small calibration correction; it is a clinical reclassification.

For treated CAH patients, the divergence is smaller but still relevant. Many treated children run 17-OHP values between 100 and 1,000 ng/dL, a zone where immunoassay performance is acceptable but LC-MS discrimination between 500 and 800 ng/dL is more reliable. Tight titration of hydrocortisone targets a mid-range value, and 200 ng/dL of noise at that scale matters clinically.

Cost differences are real. A hospital- or reference-lab immunoassay runs roughly USD 10 to 30. LC-MS costs USD 40 to 120 per sample, with higher prices in send-out reference labs.

Here is the comparison framed so the trade-offs are explicit:

The choice is not automated. Where a result will trigger a treatment decision (newborn diagnosis, ambiguous genitalia, salt-wasting crisis), LC-MS is the defensible standard. For routine adult monitoring where values are comfortably within reference range, immunoassay remains acceptable.

Reference Ranges and Thresholds on LC-MS Platforms

A common mistake is treating a reference range as portable. LC-MS reference intervals are method-specific and age-specific. Adult values are not usable for neonates, and neonatal values depend on corrected gestational age.

Typical LC-MS-derived ranges are summarized below. These are reasonable published approximations, not a substitute for the reporting lab's own interval.

To reconcile the thresholds in a single sentence: the newborn screening cutoff of roughly 200 ng/dL in term infants sits close to the top of the healthy term range (250 ng/dL), which is why ambiguous results near 150–300 ng/dL require LC-MS confirmation rather than a repeat immunoassay.

The unit conversion is fixed: 1 ng/dL equals 0.0303 nmol/L. A result of 1,000 ng/dL is 30.3 nmol/L. Laboratories that report nmol/L use a cutoff near 6 nmol/L for neonates, which converts to roughly 200 ng/dL.

One non-obvious detail: in premature infants, the fetal adrenal zone produces 17-OHP physiologically. A preterm baby at 30 weeks corrected gestational age can show 17-OHP of 400 ng/dL on day 3 and be completely normal. The reflex to diagnose CAH on that number is a common source of avoidable treatment.

Dried Blood Spots vs. Serum: The Pre-Analytical Variables That Move Results

Sample type changes the result as much as the platform does.

Dried blood spot (DBS) testing is the backbone of newborn screening. A heel-prick capillary sample is spotted onto filter paper, dried, and shipped. DBS is less invasive and logistically superior. But DBS 17-OHP runs higher than serum by 10 to 30 percent, because capillary blood contains a different hematocrit fraction and because filter paper extraction introduces recovery bias. Every DBS assay must be validated against its own serum reference.

DBS is also sensitive to hematocrit. A newborn with hematocrit of 65 percent (common in the first week) spreads the blood differently across the filter than one with 40 percent, and the measured 17-OHP shifts. In the range of hematocrit between 30 and 65 percent, this can account for a 15 to 25 percent effect.

Serum requires a venipuncture, volume of 200 to 500 microliters, and prompt processing. Whole blood at room temperature loses measurable 17-OHP within 24 hours if the sample sits. Serum separated and frozen at −20°C is stable for at least 30 days; at −80°C, for months.

Timing of collection matters for cycling women. In the luteal phase, 17-OHP rises into the 50 to 300 ng/dL range, and a follicular-phase sample is required for the standard non-classic CAH evaluation.

In plain terms: when a result surprises you, the first question is not which assay was used but when and how the sample was drawn.

Non-Classic CAH and the ACTH Stimulation Problem

Non-classic congenital adrenal hyperplasia (NCCAH) is a partial 21-hydroxylase deficiency. Basal 17-OHP in NCCAH overlaps with healthy carriers of a CYP21A2 mutation, so a random basal value cannot reliably separate the two.

The traditional ACTh stimulation test (250 micrograms of cosyntropin IV) gives a 60-minute post-stimulation value. A post-stimulation 17-OHP above 1,000 to 1,500 ng/dL (roughly 30 to 45 nmol/L) supports NCCAH, while values below 1,000 ng/dL suggest carrier status or normal. The exact cutoff depends on the reference lab and is debated in the guideline literature.

Here LC-MS earns its place. At post-stimulation values around 1,000 ng/dL, immunoassay cross-reactivity can push a normal result into the NCCAH range. Genotyping with a CYP21A2 panel remains the confirmatory step, and the combination of LC-MS biochemistry plus genotyping is the current standard of care.

A common misconception is that basal 17-OHP normalizes the diagnosis. In classic salt-wasting CAH, basal 17-OHP can exceed 10,000 ng/dL, which is why suspected salt-wasting cases need immediate cortisol and electrolyte management while confirmatory testing is pending. In NCCAH, basal values often sit between 200 and 1,500 ng/dL, overlapping both normal and affected populations.

The counterintuitive point: a "high" NCCAH result on immunoassay is the least trustworthy single number in endocrine testing, and the diagnostic weight should shift to the LC-MS post-stimulation result and the CYP21A2 genotype.

Practical Workflow: What to Actually Do

The steps below are the ones that change outcomes in real practice.

• For suspected neonatal CAH, obtain an LC-MS/MS serum 17-OHP and simultaneous electrolytes, glucose, renin, and cortisol. Do not wait for the reference-lab result before treating a salt-wasting clinical picture.
• For term newborns with a screening result between 100 and 300 ng/dL, request LC-MS confirmation before additional workup. Most of these resolve to normal.
• For premature infants under 34 weeks corrected gestational age, compare the result to gestational-age-specific intervals before considering treatment; elevations of 2 to 3 times the adult upper limit are often physiologic.
• For adult women with suspected NCCAH, collect samples before 10 a.m. in the follicular phase, and pre-arrange genotyping if the LC-MS result is between 200 and 1,500 ng/dL.
• For treated CAH children, monitor with the same LC-MS platform and the same collection time at every visit. Cross-platform comparison is a larger source of variation than day-to-day biology.
• For any 17-OHP result that conflicts with the clinical picture, repeat the LC-MS test before changing therapy.

The consistent use of one platform is not a bureaucratic preference. Published monitoring data show that switching an individual child between immunoassay and LC-MS can shift apparent values by 100 to 300 ng/dL at mid-range concentrations, which is larger than the effect of a dose change at many maintenance levels.

Timeline: From Sample Draw to Clinical Decision

The full testing timeline matters most in the neonatal setting, where hours count.

In the nursery, a 2 to 5 day turnaround means the initial diagnostic decision is usually made on clinical grounds and electrolytes, not on the 17-OHP. That is exactly why the assay's specificity matters: a false-positive confirmatory result arriving on day 4 can trigger an unnecessary workup, and a false-negative one can delay treatment.

Limitations and When LC-MS Is Still Not Enough

LC-MS removes cross-reactivity, but it does not remove biology. Four limitations deserve explicit attention.

First, 17-OHP is pulsatile and stress-sensitive. Corticotropin release from pain or illness raises 17-OHP within minutes, and a single acute-illness sample is not a baseline.

Second, LC-MS measures total 17-OHP in serum, not free hormone. Changes in binding proteins, especially in pregnancy or during estrogen therapy, can shift total values without changing free levels.

Third, a borderline LC-MS result does not by itself diagnose CAH. Confirmation requires genotyping or the ACTH stimulation test, and the decision to treat depends on clinical and biochemical aggregates.

Fourth, matrix effects in LC-MS are real. Ion suppression from co-eluting phospholipids can reduce sensitivity, and clinical labs must validate against a certified reference material and include internal standards. A poorly validated LC-MS assay can perform worse than a well-validated immunoassay. In plain terms, the platform's theoretical superiority assumes a competent, validated method.

Finally, the long-term clinical benefit of LC-MS over immunoassay in monitoring non-classic adult disease is not established by large outcome trials. The mechanism argument is strong and the cohort data are supportive, but a monitoring-driven improvement in outcomes remains plausible rather than proven. Responsible interpretation treats LC-MS as the better measurement tool, not as a distinct clinical intervention with its own survival data.