Harmonization of IGF1 immunoassay methods using an LC-MS/MS method and associated normative dataset.

This study demonstrates that commutable reference materials derived from healthy donor serum, calibrated against an LC-MS/MS method, can successfully harmonize diverse IGF1 immunoassays and enable the establishment of unified age-related reference intervals.

The Gist

Imagine you are trying to measure the height of a growing child. If one doctor uses a ruler marked in inches, another uses a tape measure marked in centimeters, and a third uses a laser that sometimes gets confused by the child's hair, you will get three different numbers for the same child. You wouldn't know if the child is actually growing or if the tools are just broken.

This is exactly the problem doctors faced with IGF1, a crucial protein in our blood that helps us grow, repair muscles, and keep our bones strong. For years, different laboratories used different "rulers" (immunoassay machines) to measure IGF1. Even though they all claimed to be calibrated against the same international standard, they gave wildly different results. A patient might be told they have a deficiency in one hospital and be perfectly fine in another, leading to confusion and potentially wrong treatments.

This paper is the story of how a team of scientists fixed this "ruler problem" and built a universal measuring tape.

The Problem: The "Broken Rulers"

The scientists found that the different machines (like the Siemens Immulite, Roche Cobas, and DiaSorin Liaison) were like people speaking different dialects. Even when looking at the same blood sample, they reported different IGF1 levels. Sometimes the difference was huge—up to 60%!

To make matters worse, because the machines disagreed, every lab had to create its own "normal" range. So, what was considered "normal" for a 40-year-old man in one lab might be considered "abnormal" in the next. This made it impossible to compare patient results across the country.

The Solution: The "Master Reference"

The team decided to create a set of Reference Materials (RMs). Think of these as gold-standard calibration weights or perfectly cut wooden blocks of known sizes.

1. Making the Standards: They took blood from healthy donors and created four different samples with known, precise amounts of IGF1.
2. The "Truth" Machine: To know the exact amount of IGF1 in these samples, they used a high-tech method called LC-MS/MS. You can think of this as a super-precise scale that counts the actual molecules, rather than guessing based on chemical reactions like the other machines do.
3. The Test: They tested these "gold-standard" blocks on all the different hospital machines to see if the machines could read them correctly.

The Discovery: One Machine Was Out of Step

Most of the machines read the "gold-standard" blocks perfectly. They were commutable, meaning they behaved with the test samples exactly the same way they behaved with real patient blood.

However, one machine (the Siemens Immulite) had a glitch. It read the lowest-concentration block incorrectly. It was like a ruler that was accurate for long distances but got the first few inches wrong. The scientists realized this machine needed a special adjustment (recalibration) using only the blocks it could read correctly.

The Result: A Universal Language

Once they used these "gold-standard" blocks to recalibrate the machines, something magical happened. The different machines started agreeing with each other.

• Before: The results were scattered like a messy pile of puzzle pieces that didn't fit.
• After: The results lined up perfectly, like a neat row of soldiers.

The variation (the "noise" in the data) dropped by about 42% to 62%. This means the measurements became much more reliable.

The New "Normal"

With the machines now speaking the same language, the team could finally create one single set of "normal" ranges for the whole country. They combined data from thousands of healthy people (from the Lifelines biobank) to draw a new map of what IGF1 levels should look like for men and women of every age.

Now, a doctor in Groningen and a doctor in Utrecht can look at the same patient's result and know exactly what it means, without needing to translate between different lab systems.

Why This Matters

• Better Diagnosis: Children with growth disorders and adults with hormone issues can finally get the right diagnosis, no matter which hospital they visit.
• Cost Savings: Labs don't need to spend money creating their own reference ranges anymore.
• Fairness: It stops the confusion where a patient is treated differently just because they went to a different building.

In short, this paper took a chaotic situation where everyone was using different rulers and gave everyone a single, perfect measuring tape. Now, when we talk about growth and health, we are all finally on the same page.

Technical Summary

1. Problem Statement

Despite all current Insulin-like Growth Factor 1 (IGF1) immunoassays (IAs) being calibrated against the same WHO International Standard (02/254), significant inter-assay variability persists. This leads to:

• Divergent Clinical Decisions: Results can differ by a factor of 1.5 to 2 between different commercial platforms, causing inconsistent diagnoses for growth disorders and metabolic conditions.
• Method-Dependent Reference Intervals: Laboratories currently rely on manufacturer-specific or locally derived reference intervals, which are not interchangeable across methods.
• Lack of Standardization: The absence of commutable reference materials (RMs) and a definitive reference method has prevented true standardization. Previous attempts using a single harmonization sample were insufficient for IGF1 compared to Growth Hormone (GH).

2. Methodology

The study employed a multi-phase approach involving sample collection, method comparison, and statistical harmonization:

• Reference Materials (RMs): Four new RMs were prepared from pooled serum of healthy blood bank donors, spanning a concentration range of 7 to 32 nmol/L.
• Study Populations:
  • Commutability Study: 42 patient samples (leftover clinical samples) and 30 healthy volunteers.
  • Reference Interval Dataset: 1,520 samples from the Lifelines biobank (healthy volunteers, ages 8–94) and 1,599 recalculated data points from a previous study (698 men, 901 women).
• Analytical Methods:
  • Immunoassays (IAs): Four commercial platforms were tested: Diasorin Liaison, Roche Cobas, IDS iSYS, and Siemens Immulite 2000.
  • Reference Method: A validated Liquid Chromatography-Tandem Mass Spectrometry (LC-MS/MS) method, calibrated against WHO IS 02/254, served as the anchor.
• Experimental Design:
  • Commutability Testing: RMs and clinical samples were measured in duplicate across all methods following IFCC Working Group protocols.
  • Recalibration: Regression curves were established between the IAs and the LC-MS/MS method using the commutable RMs. Patient and volunteer results were then recalculated to align with the LC-MS/MS scale.
  • Reference Interval Calculation: The LMS method (Cole and Green) was used to generate age- and sex-specific reference intervals (Z-scores) based on the harmonized LC-MS/MS data.

3. Key Contributions

• Development of Commutable RMs: Successfully created four serum-based RMs that are commutable across three major immunoassay platforms (Liaison, Cobas, iSYS) and LC-MS/MS.
• Harmonization Protocol: Demonstrated a practical workflow for harmonizing disparate IAs to a single reference scale (LC-MS/MS) using matrix-matched RMs, rather than relying on a single standard.
• Normative Dataset: Established a robust, age- and sex-specific reference dataset for IGF1 based on the LC-MS/MS method, covering a wide age range (8–94 years).
• Quantification of Variability: Provided concrete data on the magnitude of bias between IAs and LC-MS/MS and the effectiveness of recalibration.

4. Key Results

• Commutability: Three of the four RMs were commutable across all tested methods. However, RM1 (the lowest concentration, ~7 nmol/L) was found to be non-commutable specifically with the Siemens Immulite assay, likely due to matrix effects or antibody specificity issues at low concentrations.
• Systematic Bias: LC-MS/MS results were consistently 0% to 60% lower than those obtained by immunoassays. This suggests IAs may overestimate IGF1 due to cross-reactivity with variants, degradation products, or incomplete dissociation of binding proteins.
• Impact of Recalibration:
  • Applying the new calibration curves based on the RMs significantly reduced inter-method variability.
  • Standard Error of Estimate (SEE) decreased by 42% for patient samples and 62% for healthy volunteer samples.
  • Regression lines between methods became more parallel and aligned with the identity line.
• Reference Intervals:
  • Harmonized data allowed for the creation of universal reference intervals.
  • A statistical comparison (Mann-Whitney U test) showed no significant difference between the recalculated historical male data and new Lifelines data ($p=0.061$).
  • A significant difference was found in females ($p=0.01$), particularly in the 40–60 age range, potentially due to underlying conditions (e.g., insulin resistance) in the historical patient cohort or sample skewing.
• Outliers: Several patient samples showed extreme discrepancies between IAs and LC-MS/MS, likely due to heterophilic antibodies or variable IGF-binding protein concentrations, highlighting the limitations of IAs in complex matrices.

5. Significance and Conclusion

• Clinical Utility: The study proves that harmonizing IGF1 methods using commutable RMs allows different laboratories to use a single set of age- and sex-specific reference intervals. This eliminates the need for method-specific intervals, reduces diagnostic errors, and prevents patient misclassification.
• Cost and Efficiency: Universal reference intervals save laboratories the cost and time of establishing local reference ranges.
• Future Direction: The authors advocate for manufacturers to adopt these commutable RMs for standardization. While a true reference method (LC-MS/MS) is not yet routine in all labs, using it as an anchor via RMs is a viable interim solution to achieve global harmonization.
• Caveat: The non-commutability of the lowest concentration RM with the Immulite assay indicates that method-specific limitations still exist, particularly at the lower end of the measurement range, necessitating continued vigilance in method validation.