Measuring content uniformity in low-dose tablets using near-infrared transmission analysis

Significance of the topic

Near-infrared (NIR) transmission analysis offers a rapid, non-destructive approach to quantify active pharmaceutical ingredient (API) content across whole tablets. For low-dose formulations—where API represents only a few percent of tablet mass—accurate content uniformity testing is crucial for patient safety, regulatory compliance (USP <905>), and efficient manufacturing release. Transmission FT-NIR interrogates the entire tablet volume, providing a practical advantage over surface-limited reflection measurements for bulk homogeneity assessment.

Objectives and study overview

This application study evaluated the capability of a Thermo Scientific Antaris FT-NIR analyzer to predict API concentration in low-dose compressed tablets (~5 mg label claim; ~2% w/w in a 250 mg tablet). The goals were to develop and validate a calibration model spanning below and above label claim (synthetic samples from ~50% to ~150% label claim), compare the FT-NIR predictions against a validated HPLC reference, and determine whether the method meets USP <905> content uniformity acceptance criteria.

Methodology

The calibration strategy combined 149 calibration standards (synthetic and production-derived) with an independent validation set of 49 standards covering approximately 50%–150% label claim (~2.5 mg to 7.2 mg API). HPLC served as the reference method for traceability. Spectral acquisition used transmission FT-NIR with preprocessing and chemometric modeling as follows:

• Pretreatment: second derivative spectra with Norris smoothing (segment length = 11, gap = 10).
• Spectral region: ca. 8650 cm-1 to 8880 cm-1.
• Scattering correction: none (constant pathlength treatment).
• Multivariate model: Partial Least Squares (PLS) regression; model used 4 latent factors selected based on principal component spectra and PRESS behaviour.

Repeatability was assessed by measuring the same tablets multiple times and evaluating percent relative standard deviation (RSD). The study also considered sample presentation effects (tablet orientation, embossing) on spectral predictions.

Instrumentation

The experiments were performed on a Thermo Scientific Antaris Fourier Transform Near-Infrared (FT-NIR) Analyzer (original Antaris model). The report notes an updated Antaris II model is available offering higher speed and performance.

Main results and discussion

Key performance metrics demonstrate strong agreement between FT-NIR predictions and HPLC reference values:

• Pearson correlation coefficient: 0.9816, indicating a very strong linear relationship between NIR-predicted and HPLC-measured API content.
• Root Mean Square Error of Calibration (RMSEC): 0.168 (reported as ~3.3% of label claim).
• Root Mean Square Error of Prediction (RMSEP): 0.149 (~2.9% of label claim).
• Model complexity: 4 PLS factors; no significant spectral outliers detected (Chauvenet criterion).
• Repeatability: RSD < 1.0% for repeated measurements on individual tablets when presentation was held constant.

The PRESS plot and residual analyses supported the chosen model complexity and indicated robust predictive behaviour across the 50%–150% calibration range. The study highlights that transmission sampling typically yields better accuracy and precision for whole-tablet content uniformity compared with reflection, which samples only near-surface layers (~first 500 microns). However, tablet orientation, embossing, and stamping can introduce variability; controlling presentation or incorporating orientation variability into the sampling strategy is necessary to ensure method robustness.

Benefits and practical applications of the method

The validated FT-NIR transmission method offers several practical advantages for QC and PAT workflows:

• Non-destructive, rapid analysis enabling high-throughput screening of individual tablets.
• Minimal sample preparation and shorter operator training compared with titration or HPLC.
• Capability to probe the entire tablet volume, improving detection of API inhomogeneity in low-dose products.
• Quantitative performance sufficient to meet USP <905> criteria for the studied batch, supporting release decisions.
• Potential for simultaneous collection of transmission and reflection data on Antaris instruments to capture both bulk and coating information.

Future trends and potential applications

Areas likely to expand or improve application of FT-NIR for content uniformity include:

• Integration into in-line and at-line PAT systems for real-time monitoring during tablet production.
• Improved instrument designs (higher throughput, increased sensitivity) and broader spectral coverage to enhance low-concentration detection.
• Advanced chemometrics and machine learning models to improve calibration transferability between instruments and to better accommodate presentation variability.
• Hybrid approaches combining transmission NIR with imaging modalities (hyperspectral imaging) to map within-tablet API distribution.
• Regulatory acceptance growth of spectroscopic PAT methods combined with validated reference methods to reduce the reliance on destructive assays for routine QC.

Conclusion

The study demonstrates that transmission FT-NIR on an Antaris FT-NIR analyzer can accurately and precisely predict API content in low-dose tablets when an appropriate calibration set (50%–150% label claim) and validated chemometric model are used. The method achieved strong correlation with HPLC, low prediction errors (RMSEP ~2.9% label claim), and excellent repeatability under controlled presentation. Transmission FT-NIR is therefore a viable, efficient tool for content uniformity testing in manufacturing and QC environments, provided attention is paid to sample presentation and calibration design.

References

• Randle S., Hirsch J. Application note: Measuring content uniformity in low-dose tablets using near-infrared transmission analysis. Thermo Fisher Scientific, AN51389 0522 (2022).
• United States Pharmacopeia (USP) Chapter 905: Uniformity of Dosage Units.
• United States Pharmacopeia (USP) Chapter 1119: Near Infrared Spectroscopy.