Near Infrared (NIR) Analysis of Tablets Containing Two Active Ingredients

Robert Mattes1, Denise Root1, Ed Brunson2, Suresh Potharaju 2, James Johnson2, Almoazen Hassan 2, 1FOSS NIRSystems, Inc., 2University of Tennessee, Department of Pharmaceutical Sciences
Abstract Near infrared (NIR) analysis of tablets provides a fast, accurate means of monitoring tablets during production that meets with PAT initiatives. Dosage forms with multiple active pharmaceutical ingredients (APIs) pose greater analytical difficulties than single API dosage forms. This study evaluates NIR analysis of tablets containing 4 mg Chlorpheniramine maleate (CPM) and 10 mg phenylephrine HCL per dosage. The tablets were produced with nominal batch levels to include a range of +/- 20% of each ingredient by the University of Tennessee Pharmaceutical Sciences Department. Ten tablets from each batch level, for a total of ninety tablets, were tested and HPLC was run on each batch. Partial least squares regression models were developed for both APIs and predictive capability of the models was analyzed. Introduction Near-infrared spectroscopy (NIRS) is an analytical technique based on absorption measured in the near-infrared region of the electromagnetic spectrum that is between the visible and the mid-infrared. The fundamental absorption bands of functional groups occur in the mid-infrared and are very strong. The overtone absorptions of these fundamental bands occur in the near infrared (NIR) spectral region and allow direct measurement without sample preparation due to the relative weakness of absorption. The OH, CH, NH and SH bonds have the strongest overtone absorbance in the NIR region. There has been considerable interest in the ability to test solid dosage form samples more frequently than the ten per batch specified by the US Pharmacopeia monograph on content uniformity (CU). Due to concerns of the European Union for better statistically based sampling and the FDA initiative of process analytical technology (PAT) for better understanding and monitoring production, interest has increased in utilizing NIR for tablet assay and content uniformity testing. NIR can be used as a rapid at-line analysis method to obtain processing feedback nearly real-time during a tableting campaign. Transmission NIRS through the tablet has been preferred to reflectance NIRS due to heterogeneity within tablets. Reflection NIRS technique may be used for coating analysis, but for bulk tablet analysis, transmission NIRS technique may yield more consistent results. Laboratory methods for tablet assay and content uniformity are usually time consuming since they are routinely done by HPLC, which requires lengthy calibration runs, mixing of buffers and procurement and disposal of solvents. Analyzing ten tablets for content uniformity could take hours and the results may not be available to the tablet press operators or for batch release for many days or even weeks after the tablets are compressed. Statistical process control (SPC) techniques can be applied while measuring the tablets with NIR real-time during tableting so that assay and CU problems can be detected before they go beyond acceptable limits. Experimental Twelve batches of tablets, 0.25 diameter with a nominal weight of100 mg, were formulated with a label claim of 10.0 mg PE-HCl and 4.0 mg CPM. The CPM and PE-HCl were varied to plus and minus 20% of this label claim in a fractional factorial design of experiment (DOE). The lactose excipient was varied to maintain constant weight. See Table I below. The tablets were compressed on an Elizabeth-Hata tablet press (HT-AP 18 SS-U/I rotary tablet press, Elizabeth Hata International, Inc., North Huntingdon, PA).
TableI.SamplesforNIR/HPLCAnalysis FormulaCompositions(mg/tab) 1 2 3 4 5 6 7 8 9 Formulas
Item 1 2 3 4 5 Material N1(PE) PEHCl 10.00 CPM Lactose DTG MCC MgSt 64.75 25.00 0.25 100.00 N1 (CPM) 4.00 70.75 25.00 0.25 100.00 74.75 25.00 0.25 100.00 Placebo T1 8.00 3.20 63.55 25.00 0.25 100.00 T2 8.00 4.80 61.95 25.00 0.25 100.00 T3 9.00 3.60 62.15 25.00 0.25 100.00 T4 9.00 3.40 62.35 25.00 0.25 100.00 T5 10.00 4.00 60.75 25.00 0.25 100.00 T6 11.00 3.60 60.15 25.00 0.25 100.00

Figure 3. Second derivative math pretreatment of calibration spectra. Figure 1. Tablets in NIR instrument in transmission mode. Figure 2. Raw spectra of calibration samples.

Figure 4. PRESS plot of PLS factors used to predict CPM concentration. 5 factors used.

Figure 5. PRESS plot of PLS factors used to predict PE-HCl concentration. 6 factors used.

Discussion Figure 2 shows the raw NIR spectra from the calibration set. By taking the second derivative of the spectra, as shown in Figure 3, the baseline was normalized and the spectral features were enhanced. Smoothing was done on the derivative with a segment of 10 and a gap of 0. HPLC analysis of individual tablets was not completed in time for this modeling and the batch nominal values were modeled in their place. Partial least squares (PLS) regression was used to develop the prediction model which uses principal component analysis and is a variation of principle component regression (PCR). The correct number of principal components or factors was determined by the Vision software supplied with the instrument by determining where the predicted residual error sum of squares (PRESS) reaches a minimum. Figure 4 shows the PRESS plot for the CPM PLS model selecting 5 factors. Figure 6 is a plot of the principle component loadings for factors two through five. The loadings plots for the first factor were very similar for the two actives due to strong aromatic C-H stretch absorbance in both and were therefore left out for better comparison. The resulting CPM model had an R2 value of 0.9606 and a standard error of calibration (SEC) of 0.3172. The one-left-out cross validation demonstrated good predictability with a standard error of cross validation (SECV) of 0.3301. Figure 7 shows the NIR predicted CPM level versus the batch nominal values. The large variance from batch nominal value can be observed in the vertical dispersion. Twenty-five percent of the total tablets scanned were randomly left out of the calibration for prediction model validation. Figure 8 shows the NIR predictions of the validation set versus the batch nominal values for the CPM active ingredient. The standard error of prediction (SEP) was 0.3613 for the validation samples. Figure 5 is a plot of the PRESS leading to a model with 6 factors for PE-HCl. The resulting model had an R2 value of 0.9960 and a standard error of calibration (SEC) of 0.2750. Figure 10 shows the NIR predicted PE-HCl level versus the batch nominal values. Figure 11 shows the NIR predictions of the validation set versus the batch nominal values for the PE-HCL active. The standard error of prediction (SEP) was 0.2448 for the validation samples. Table II is the repeatability results for a separate set of T-5 label claim tablets containing 4.0 mg CPM and 10.0 mg PE-HCl. The relative standard deviation (RSD) for CPM active of these 5 tablets was 0.8806 %. The RSD for the repeatability for the same tablets for 10.0 PE-HCl was 0.7921 % (Table III). The Vision software has a convenient routine analysis method for calculating content uniformity automatically and produces a 21 CFR part 11 compliant report. Table IV shows the content uniformity results for 4.0 mg CPM in T-5 label claim tablets. Table V shows the content uniformity results for 10.0 mg PE-HCl in the same tablets.

Figure 6. PLS loadings spectra showing where CPM is highly correlated with spectral data.

Figure 7. Calibration set. NIR predicted versus CPM value. R2 = 0.9606, SEC= 0.3172, SECV= 0.3301.

Figure 8. Validation set. 25% of tablets were left out of the calibration set for CPM model validation. SEP=0.3613.

Figure 9. PLS loadings spectra showing Where PE-HCl is highly correlated with spectral data. Table II. CPM repeatability.
Tablet NIR Prediction CPM Batch Nominal

Figure 10. Calibration set. NIR predicted versus PE-HCl value. R2 = 0.9960, SEC= 0.275, SECV=0.2991. Table III. PE-HCl repeatability.
Tablet NIR Prediction PE-HCl Batch Nominal

Figure 11. Validation set. 25% of tablets were left out of the calibration set for PE-HCl model validation. SEP=0.2448. Table IV. Content Uniformity CPM RSD=3.0%
Label Sample Claim 1 2 3 4 5 6 7 8 9 10 Test %of Result Target P/F 4.233 105.832Pass 4.505 112.614Pass 4.155 103.865Pass 4.284 107.098Pass 4.197 104.925Pass 4.264 106.601Pass 4.529 113.236Pass 4.208 105.197Pass 4.312 107.808Pass 4.234 105.862Pass

Tablet #18 Tablet #18 Tablet #18 Tablet #18 Tablet #18 RSD%

4.2333 4.2839 4.2641 4.3123 4.3279 0.8806

4.0 4.0 4.0 4.0 4.0

Tablet #18 Tablet #18 Tablet #18 Tablet #18 Tablet #18 RSD %

10.5458 10.5099 10.6275 10.6307 10.7257 0.7921

10.0 10.0 10.0 10.0 10.0

4 4 4 4 4 4 4 4 4 4

10
T7 11.00 4.40 59.35 25.00 0.25 100.00

11
T8 12.00 3.20 59.55 25.00 0.25 100.00

12
T9 12.00 4.80 57.95 25.00 0.25 100.00

FiveLevelFractionalFactorialDesignFormulas

Table V. Content uniformity PE-HCl RSD=1.0% Conclusion This study shows that near infrared assay for the determination of content uniformity of tablets provides a fast and accurate means of monitoring tablets that could be used for production that is in step with the FDA Process Analytical Technology (PAT) initiatives. The data showed promising results that could relieve laboratory workload and bring analysis closer to real-time for process monitoring. Ten tablets could be analyzed in less than three minutes. The HPLC results were not available at the time of analysis. The true HPLC values would demonstrate lower standard error.
Label Sample Claim 1 2 3 4 5 6 7 8 9 10 Test Result %of Target P/F

10 10.546 105.458Pass 10 10.442 104.418Pass 10 10.385 103.849Pass 10 10.51 105.099Pass 10 10.372 103.722Pass 10 10.627 106.275Pass 10 10.668 106.677Pass 10 10.5 104.999Pass 10 10.631 106.307Pass 10 10.547 105.471Pass

Totalmg/tab

The NIR instrument used in the study was an XDS MasterLab (FOSS NIRSystems, Laurel, MD) capable of automatically measuring multiple tablets positioned in a special tray (Figure 1). The tray used for this study had 31 positions and was loaded to scan ten tablets from each batch. The ten tablets were scanned in less than three minutes, taking a reference spectrum before scanning each batch. Spectra were collected in the transmission mode from 800 nm to 1650 nm with 0.5 nm data intervals and 32 scans were co-added to produce a single spectrum.