Development and validation of a GC–MS method for the determination of ethylene glycol and diethylene glycol in pharmaceutical syrups

Abstract

Contamination of pharmaceutical syrups with ethylene glycol (EG) and diethylene glycol (DEG) has led to serious poisoning incidents worldwide, highlighting the necessity for strong regulations and reliable testing methods. A gas chromatography-mass spectrometry (GC-MS) method was developed and validated in accordance with ICH Q2(R2) guidelines. The method was developed for pharmaceutical syrup matrices and validated using multiple commercially available products, supporting its applicability for routine regulatory surveillance. It exhibited excellent linearity, with r² of 0.9991 for EG and 0.9963 for DEG. The detection (LOD) and quantification (LOQ) limits were 0.02% v/v and 0.06% v/v, respectively, below the regulatory limit of 0.10% set by the National Pharmaceutical Regulatory Agency (NPRA), Ministry of Health Malaysia. Accuracy ranged from 94.2% to 108.4%, and the %RSD for intermediate precision was 1.4% for EG and 2.5% for DEG, confirming the method’s reproducibility. When the validated method was applied to 51 pharmaceutical syrup samples available in the Malaysian market as part of post-market surveillance, 13 samples were detected to have EG. Of these, 11 samples had EG concentrations between the LOD and the LOQ, while two samples (PS-13 and PS-26) had measurable levels at 0.09% v/v. DEG was not detected in any sample. Therefore, this study emphasises the importance of regulatory surveillance, supported by validated analytical testing, to ensure the quality and safety of pharmaceutical syrups. Strengthening local surveillance mechanisms will further strengthen NPRA’s regulatory capacity and safeguard public health.

Graphical Abstract

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s40360-026-01184-2.

Introduction

Contamination of pharmaceutical products poses a serious threat to public health, often stemming from inadequate quality control, supply chain failures, or adulteration. One of the most critical contamination issues involves toxic substances such as ethylene glycol (EG) and diethylene glycol (DEG) [1–3]. EG and DEG are toxic industrial chemicals primarily used as antifreeze agents, solvents, and intermediates within chemical manufacturing [4]. Despite their intended industrial applications, EG and DEG have repeatedly been implicated in poisoning cases due to their unintended or deliberate presence in pharmaceutical products. Contamination typically occurs through the substitution or adulteration of pharmaceutical excipients, such as glycerine and propylene glycol, which are commonly used in liquid medications [5].

The metabolism of EG and DEG in the human body is central to their toxicity. Both compounds are primarily metabolised in the liver by the enzyme alcohol dehydrogenase (ADH), which converts EG into glycolaldehyde and DEG into 2-hydroxyethoxyacetaldehyde. These intermediates are further oxidised by aldehyde dehydrogenase to glycolic acid (from EG) and 2-hydroxyethoxyacetic acid (from DEG), respectively. Glycolic acid is then metabolised into oxalic acid, which can crystallise in the kidneys, leading to renal failure, while 2-hydroxyethoxyacetic acid contributes to metabolic acidosis and organ damage [6]. The accumulation of these toxic metabolites impairs cellular energy production, causes severe acidosis, and damages vital organs, ultimately resulting in potentially fatal outcomes if left untreated.

DEG poisoning has been linked to several major pharmaceutical contamination incidents worldwide, highlighting the critical consequences of inadequate quality control. The 1937 Elixir Sulfanilamide disaster in the United States, which caused more than 100 deaths, led to the enactment of the 1938 Federal Food, Drug, and Cosmetic Act [7]. Subsequent outbreaks included DEG-contaminated paracetamol elixirs in Bangladesh in 1990, which resulted in fatal renal failure among children [8], and contaminated medications in Haiti during 1995–1996 that caused widespread acute renal failure and high mortality among affected children [9]. More recently, DEG and EG contamination in paediatric cough and fever syrups was linked to acute kidney injury outbreaks in The Gambia and Indonesia in 2022, resulting in significant child mortality [10, 11]. These recurring incidents emphasise the requirement for stringent regulatory oversight and robust pharmaceutical quality control measures.

In Malaysia, concerns have emerged about the potential contamination of pharmaceutical syrups with EG and DEG. Although no major poisoning incidents have been reported locally, the National Pharmaceutical Regulatory Agency (NPRA) has proactively implemented regulatory surveillance measures to mitigate risks and guarantee public safety. In routine quality control, evaluation of pharmaceutical products is primarily based on Certificates of Analysis (CoA) provided by manufacturers and standard pharmacopoeial testing of finished products and active pharmaceutical ingredients, whereas specific testing for EG and DEG is not routinely performed as part of excipient quality control. From an analytical perspective, this illustrates the need for a reliable, simple, and validated chromatographic method suitable for implementation in regulatory laboratories. Although gas chromatography (GC) and liquid chromatography (LC) methods have been reported for the determination of EG and DEG, their applicability in surveillance workflows varies with respect to matrix suitability, operational complexity, and integration into regulatory monitoring systems [12–14]. In this context, a fit-for-purpose GC-MS method was developed and validated for the simultaneous identification and quantification of EG and DEG in pharmaceutical syrups to support NPRA surveillance and strengthen post-market quality assurance.

Methodology

Chemicals and reagents

LC-MS-grade methanol was sourced from Merck, Darmstadt, Germany. EG (99.17%, CAS No: 107-21-1) and DEG (99.47%, CAS No: 111-46-6) were purchased from LGC Labor GmbH, Augsburg, Germany.

Standard and sample solutions preparation

A mixed stock solution of EG and DEG was prepared at 1 mg/mL in methanol and diluted to obtain working standard solutions at 6, 8, 10, 12, and 14 µg/mL. The study focused on a single matrix, pharmaceutical syrup. Six commercially available syrup products from local pharmacies were selected as representative blank matrices. These products were screened using the developed GC-MS method prior to validation and were confirmed to be free of detectable EG and DEG. For sample preparation, 1 mL of syrup was transferred into a 100 mL volumetric flask, mixed with 50 mL of methanol, and sonicated for 5 min. The volume was then adjusted to the mark with methanol, and the solution was filtered through a 0.45 μm nylon syringe filter before being transferred into a 2 mL vial for analysis.

Chromatography and mass spectrometry conditions

Analysis was carried out using a GC-MS system (GC-MS QP2010, Shimadzu Corporation, Japan) with helium (99.99% purity) as the carrier gas under linear velocity flow control mode at 30.0 cm/s. Separation was achieved using a BP20 (WAX) GC capillary column (30 m × 0.25 mm × 0.25 μm). The flow rate was maintained at 0.65 mL/min, with a system pressure of 39.3 kPa. A split injection mode with a 20:1 split ratio was used, with an injection volume of 1 µL. The injector temperature was maintained at 250 °C. The column temperature was programmed as in Table 1. The mass spectrometer was operated in electron ionisation (EI) mode at 70 eV. The ion source temperature was set at 230 °C, while the interface temperature was maintained at 240 °C to ensure efficient ion transfer. The scan range was set from m/z 25 to m/z 200, allowing comprehensive detection of target analytes. The identification of the target analytes was confirmed by comparison with the NIST 14 mass spectral library. Quantitative analysis was performed in selected ion monitoring (SIM) mode, with the specific mass-to-charge (m/z) fragments for each analyte shown in Table 2.

Table 1.

GC temperature ramping for the analysis

Oven ramp	Rate (ºC/min)	Final temperature (ºC)	Hold time (min)	Run time (min)
Initial		100	1
Ramp 1	10	130	7
Ramp 2	20	240	3
Post run	20	250	3	23

Table 2.

SIM-mode mass fragmentation for EG and DEG

Analyte	Acquisition time (min)	Quantifier ion (m/z)	Qualifier ion 1 (m/z)	Qualifier ion 2 (m/z)
EG	7.00–11.00	31	33	62
DEG	11.00–17.00	45	75	31

Validation

The validation phase included establishing the limit of detection (LOD), limit of quantification (LOQ), linearity, specificity and selectivity, accuracy, and precision of the method according to the International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use [15]. Sensitivity was established using the signal-to-noise (S/N) approach, with LOD and LOQ at 2 µg/mL (0.02%) and 6 µg/mL (0.06%), respectively. Linearity was examined across five concentration levels (6, 8, 10, 12, and 14 µg/mL) by plotting peak area against concentration and performing linear regression. Method specificity and selectivity were assessed by comparing chromatographic signals from blank sample solutions, spiked samples, and mixed standards. Accuracy was evaluated by spiking known concentrations (8, 10, and 12 µg/mL) into the sample matrix, achieving acceptable recoveries of 90%-110%. Precision was assessed using intermediate precision, with six independent replicates at 10 µg/mL, ensuring that the relative standard deviation (%RSD) remained ≤ 5%. A second analyst conducted the analysis on a different date to confirm reproducibility, following the same criteria.

Application of the validated method to the surveillance samples

Through its post-market surveillance programme, NPRA received 51 pharmaceutical syrup samples (PS-01 to PS-51) from various Product Registration Holders (PRHs). All samples were submitted in their original commercial packaging and underwent an initial compliance review of their labels and package inserts before being stored under controlled conditions. Following method validation, the samples were analysed by GC-MS to detect contamination with EG and DEG. The concentrations of EG and DEG in the tested surveillance syrup samples (% v/v) were calculated using the following formula (1), accounting for the 100-fold dilution (1 mL of syrup diluted to 100 mL with diluent):

1

where  C is the measured concentration (µg/mL), the total solution volume (mL), the sample volume taken (mL),  P the purity of the standard (decimal),  DF the dilution factor, and  D the density (g/mL). The factor converts µg to g and scales to a percentage.

Results

Development and validation of GC-MS method for EG and DEG

The developed GC-MS method successfully separated EG and DEG within 23 min and was further validated for LOD, LOQ, linearity, specificity, accuracy, and precision, thereby demonstrating compliance with ICH Q2(R2).

System suitability

System suitability testing confirmed acceptable chromatographic performance for the analysis of EG and DEG (Fig. 1). Parameters, including theoretical plate number, tailing factor, and %RSD, were within acceptable limits, indicating reliable performance (Table 3).

Fig. 1.

Chromatograms for system suitability testing (n = 6)

Table 3.

System suitability parameters for the GC-MS analysis in EG and DEG detection

Parameter	EG	DEG	Acceptance criteria
Theoretical plate (N)	41,517	597,944	> 2000
Tailing factor (Tf)	1.670	1.667	< 2
%RSD for peak area (%)	3.145	4.853	< 5

LOD and LOQ

The sensitivity of the GC-MS method was evaluated by determining the LOD and LOQ for EG and DEG using the S/N ratio approach, with standard solutions prepared in methanol (Table 4).

Table 4.

Validation of LOD and LOQ values

Analyte	LOD (2 µg/mL)		LOQ (6 µg/mL)
	Signal-to-noise ratio (S/N)
	Mean (n = 3)	%RSD	Mean (n = 10)	%RSD
EG	49.88	1.59	169.88	2.66
DEG	55.27	1.93	200.02	3.48

Linearity and range

Linearity of the GC-MS method for EG and DEG was confirmed across the tested concentration range (Fig. 2). Both correlation coefficient values exceeded the acceptance criterion (r² > 0.995), exhibiting excellent linearity and a consistent detector response.

Fig. 2.

Linear regression analysis and coefficient of determination for (a) EG and (b) DEG calibration curves

Specificity

The specificity of the gas chromatographic method was assessed to confirm its capacity to precisely identify and quantify EG and DEG in pharmaceutical syrup matrices (Fig. 3). This evaluation was carried out across multiple commercial syrup products from different brands and formulations, representing variations within the same syrup matrix type. Representative NIST 14 mass spectral library matches of the target analytes, EG and DEG, are shown in Fig. 4, confirming the reliability of compound identification based on spectral similarity and characteristic fragmentation patterns. In addition, ion fragmentation profiles (m/z values) and relative ion intensities (%) for EG and DEG supported the specificity of the GC–MS method, demonstrating consistent, unambiguous analyte identification across these products (Table 5).

Fig. 3.

Representative total ion chromatogram of (a) unspiked sample solution, (b) spiked sample solution (10 µg/mL), and (c) mixed standard solution (10 µg/mL), showing the specificity of the GC-MS method

Fig. 4.

Representative NIST 14 mass spectral library match of the target. (a) EG and (b) DEG

Table 5.

Relative ion intensity (%) of fragment ions (m/z) for EG and DEG

		EG		DEG
Ion fragmentation (m/z)	Quantifier ion	31		45
	Qualifier ion	33	62	75	31
Relative ion intensity (%)		39.70	5.59	31.54	7.30

Accuracy

The GC-MS method was evaluated through recovery studies at three concentrations (8, 10, and 12 µg/mL) of EG and DEG in six commercially available pharmaceutical syrups. The reported recoveries are the averages across the different syrup formulations (Table 6).

Table 6.

Recovery (%) of EG and DEG in pharmaceutical syrup samples spiked at 8, 10, and 12 µg/mL

Analyte	Spiked concentration (µg/mL)	Recovery (%)*	95% CI (±)
EG	8	94.2 ± 2.61	6.49
	10	101.0 ± 1.59	3.95
	12	108.4 ± 0.12	0.30
DEG	8	103.5 ± 2.32	5.77
	10	107.1 ± 0.99	2.45
	12	106.8 ± 1.69	4.20

*Tabulated values are mean + SD (n = 3)

Precision

Intermediate precision of the GC-MS method for EG and DEG at 10 µg/mL was determined by two analysts (Table 7).

Table 7.

Intermediate precision for EG and DEG at 10 µg/mL

Analyst	Replicate	Recovery (%)
		EG	DEG
1	1	108	101
	2	106	100
	3	107	103
	4	110	100
	5	108	102
	6	107	109
2	7	106	101
	8	103	103
	9	107	104
	10	108	102
	11	106	101
	12	107	104
Mean		107	102
SD		1.5351	2.6104
%RSD		1.4345	2.5498
95% CI		0.9754	1.6586

Application of the validated method

The validated GC-MS method was successfully applied to the analysis of 51 pharmaceutical syrup surveillance samples. A summary of the surveillance findings is presented in Table 8, while the complete dataset, including sample information, active ingredients, manufacturing dates, and individual analytical results, is provided in Supplementary Material (Table S1).

Table 8.

Summary of EG and DEG analysis in 51 pharmaceutical syrup surveillance samples

Analytical finding	EG, n (%)	DEG, n (%)
Not detected (< LOD)	38 (74.5)	51 (100.0)
Detected between LOD and LOQ	11 (21.6)	0 (0.0)
Quantified (≥ LOQ)	2 (3.9)	0 (0.0)
Concentration detected	0.09% v/v (n = 2)	Not detected

LOD: 0.02% v/v, LOQ: 0.06% v/v

Discussion

The present study contributes to this global public health objective by developing and validating a GC-MS method for the simultaneous quantification of EG and DEG in pharmaceutical syrup products. The method development phase prioritised achieving efficient separation, minimal matrix interference, and adequate sensitivity to detect potentially harmful levels of EG and DEG within international permissible limits. The method validation adhered strictly to the International Council for Harmonisation (ICH) Harmonised Guideline: Validation of Analytical Procedures Q2(R2), encompassing system suitability, LOD, LOQ, specificity, linearity, accuracy, and precision. The validated method was subsequently applied to surveillance samples of pharmaceutical syrups.

Regulatory oversight

Regulatory agencies worldwide have set stringent limits on EG and DEG contamination in pharmaceutical products. In Malaysia, the National Pharmaceutical Regulatory Agency (NPRA) adopts the internationally recognised permissible limit of 0.10%, consistent with guidelines from the WHO, the U.S. FDA, the European Commission, and the Indian Pharmacopoeia Commission. GC-based methods remain the standard analytical approach for detecting these contaminants [16–19].

Malaysia implements both pre- and post-market surveillance strategies to monitor contamination with EG and DEG. Pre-market controls include evaluating CoAs from raw material suppliers, while post-market surveillance involves routine sampling, laboratory verification, and GC-MS testing. NPRA also conducts inspections of syrup manufacturers to ensure compliance with Good Manufacturing Practice (GMP) standards [20]. Similar approaches are adopted internationally. The U.S. FDA applies risk-based sampling focused on manufacturers with prior contamination records, whereas the European Medicines Agency incorporates real-time monitoring and international regulatory collaboration [21, 22]. In India, the Pharmacovigilance Programme of India has expanded regional centres to strengthen adverse drug reaction (ADR) reporting and drug safety monitoring nationwide [23, 24].

The findings of this study are interpreted within the current regulatory framework and, more importantly, inform the next steps in NPRA’s surveillance strategy. Although 13 of 51 samples (25.5%) contained EG, all detected concentrations remained well below the regulatory limit. From a compliance perspective, none of the tested products would trigger immediate regulatory action, such as a product recall, as all results fall within acceptable safety limits. However, in a regulatory surveillance context, the detection of EG, particularly in samples with concentrations approaching the limit (0.09%), would be treated as a signal for intensified monitoring rather than enforcement. In practice, NPRA would prioritise follow-up investigations focusing on the affected product batches and their associated manufacturers, with particular attention to high-risk excipients such as propylene glycol and glycerol. This would include targeted retesting of retained samples, expanded batch sampling, and verification of supplier CoAs to trace potential upstream sources of contamination within the supply chain. The absence of DEG in all samples further suggests that the current contamination risk in the surveillance samples analysed is low and not indicative of the severe adulteration scenarios reported in past poisoning outbreaks. Nevertheless, the presence of EG reinforces the value of sensitive analytical screening in detecting sub-threshold contamination that may otherwise remain undetected under routine compliance testing alone.

To further minimise contamination risks, NPRA requires pharmaceutical products to comply with international standards, including WHO, USP, Ph. Eur., and BP guidelines. Post-market monitoring also encompasses ADR reporting through the Malaysian Adverse Drug Reaction Monitoring Programme (MADRAC), as well as rapid regulatory actions, such as product recalls and public alerts, when contamination is identified [25–27]. In addition, NPRA collaborates with international organisations, including WHO and ASEAN regulatory networks, to strengthen surveillance and respond effectively to emerging contamination risks [28, 29]. Collectively, these regulatory efforts highlight the growing importance of continuous surveillance and reliable analytical methods for safeguarding public health and maintaining pharmaceutical product quality, particularly following recent poisoning outbreaks linked to contaminated pharmaceutical syrups [30].

System suitability and sensitivity

System suitability is a critical parameter for ensuring the reliability and performance of an analytical instrument before sample analysis. It is a quality-control measure that verifies the instrument meets predefined acceptance criteria using standard reference materials [31]. This study assessed system suitability by evaluating system precision and analysing the consistency of repeated injections from the same vial. The results showed average theoretical plate numbers (N) of 41,517 and 597,944 for EG and DEG, respectively, indicating efficient chromatographic separation. The tailing factors (T_f) were 1.670 for EG and 1.667 for DEG, suggesting acceptable peak symmetry. The %RSD values for EG and DEG were 3.145% and 4.853%, respectively. These findings confirm the system’s suitability for the intended analysis, as values fall within acceptable limits and demonstrate high consistency across injections.

The sensitivity of the developed GC-MS method was assessed by determining the LOD and LOQ for EG and DEG. These parameters are essential for evaluating the capability of the method to detect and quantify low concentrations of EG and DEG in the syrup matrix. The LOD is the lowest concentration at which EG and DEG can be detected, but not necessarily quantified. The LOQ is the minimum concentration at which EG and DEG can be quantified with acceptable accuracy and precision. In this study, the LOD and LOQ values for EG and DEG were determined using a signal-to-noise (S/N) ratio approach, with LOD set at approximately 3 and LOQ at approximately 10. The results indicate that the LOD for EG was 2 µg/mL, with a mean S/N ratio of 49.88 and a %RSD of 1.59% across three replicate measurements (n = 3). Similarly, the LOD for DEG was 2 µg/mL, with a slightly higher mean S/N ratio of 55.27 and a %RSD of 1.93%, suggesting that the method provides consistent, reproducible detection at low analyte concentrations.

For quantification, the LOQs for both EG and DEG were set at 6 µg/mL to ensure reliable measurement with adequate precision. The mean S/N ratio for EG at this concentration was 169.88, with a %RSD of 2.66% (n = 10), demonstrating high reproducibility. Likewise, DEG exhibited a mean S/N ratio of 200.02 at the LOQ, with a %RSD of 3.48%, indicating that the method can provide consistent and precise quantification at this concentration. The %RSD values for both LOD and LOQ were within acceptable limits (< 5%), confirming the feasibility of the analytical method. The slightly higher %RSD for DEG than for EG suggests minor variations in detector response, which may be attributed to differences in analyte volatility or ionisation efficiency during analysis [32, 33]. Nevertheless, the developed GC-MS method demonstrated excellent sensitivity for detecting and quantifying EG and DEG at low levels, confirming its suitability for regulatory and quality-control applications in pharmaceutical syrup formulations.

Linearity of detector response

The linearity of the developed GC-MS method was assessed to evaluate the relationship between analyte concentration and detector response for EG and DEG. Linearity is a fundamental validation parameter that ensures the method produces reliable, proportional responses over a defined concentration range, a requirement for accurate quantification [34]. The regression equations for EG and DEG were y = 53921x – 62966.6 and y = 88610.65x – 91137.7, respectively. The r² values were 0.9991 for EG and 0.9963 for DEG, indicating excellent linearity between concentration and measured peak area. These values confirm that the method exhibits high linearity across the tested concentration range. The linearity achieved in this study is comparable to that of previously reported GC-based methods for determining EG and DEG in pharmaceutical syrup matrices [2, 35].

The calibration curve shows well-distributed data points along the regression line, with minimal deviation from linearity. This indicates that the instrumental response remains stable and consistent across the analyte concentration range, further supporting the efficacy of the developed method. The slightly lower r² for DEG compared with EG may reflect minor variations in detector response, possibly due to differences in molecular structure, volatility, or ionisation efficiency [36]. However, both analytes achieved r² values above the commonly accepted threshold of 0.995, confirming the suitability of the developed method for quantitative analysis [37].

Specificity of the GC-MS method

An integral aspect of this evaluation was assessing chromatographic separation using the retention times of the target analytes, ensuring that no interfering peaks from excipients or other matrix components were present [38]. The absence of interference is essential to validate the method’s ability to distinguish EG and DEG from potential co-eluting substances commonly found in syrup formulations. A blank syrup sample was analysed systematically to assess specificity, and the resulting chromatogram showed no detectable peaks at the retention times of EG and DEG. This finding confirmed that the matrix did not contribute endogenous interference, establishing a baseline for subsequent analyses. Conversely, the chromatographic profiles of the spiked syrup sample and the mixed standard solution showed well-resolved, distinct peaks at their expected retention times. The consistency of these peaks across standard solutions and spiked samples reinforced the method’s selectivity in detecting and quantifying EG and DEG, with no interference from matrix components.

Furthermore, representative NIST 14 mass spectral library matches demonstrated high spectral similarity between the detected analytes and the reference spectra for EG and DEG, providing additional confirmation of analyte identity. The ion fragmentation profiles also supported method specificity. EG exhibited a predominant fragment at m/z 33 (39.7% relative ion intensity of fragment ions (m/z) for EG), with a secondary fragment at m/z 62 (5.59%), while DEG was characterised by a dominant fragment at m/z 75 (31.54%) alongside m/z 31 (7.30%). These distinct fragmentation patterns, together with reference-ion-ratio-matched reference spectra and NIST 14 library matching, met ion-ratio criteria for confirmation, thereby supporting the unambiguous identification and quantification of EG and DEG in pharmaceutical syrup matrices.

Accuracy of EG and DEG quantification

Accuracy is the extent to which the results of an analytical procedure approach the true value, with minimal influence from matrix effects or instrument variability [39]. The accuracy of the developed GC-MS method was evaluated by determining the percentage recovery of EG and DEG at three quality control levels (low, medium, and high), i.e., 8, 10, and 12 µg/mL [40]. The recovery results indicate that the method provided consistent quantification across all tested concentrations. For EG, the mean recovery ranged from 94.2% to 108.4%, with 95% confidence intervals (CIs) of 0.30%–6.49%. DEG exhibited recoveries of 103.5%–107.1%, with 95% CIs ranging from 2.45% to 5.77%. The observed recovery values fall within the acceptable range of 90%–110%, indicating that the method provides accurate quantification with minimal systematic error. Compared to the previous method [2], which showed more variable recoveries (81.3%–116.8%) within the acceptable range of 80%–120%, the current method demonstrated a narrower recovery range of 90%–110%, suggesting improved consistency.

At the lower concentration of 8 µg/mL, EG recovery was slightly below 100%, suggesting minor losses during sample preparation or slight matrix effects. Recovery values increased at 10 and 12 µg/mL, while remaining within acceptable limits of variation. In contrast, DEG recovery remained consistently above 100% across all spiked concentrations, which may be attributed to enhanced extraction efficiency or minor variations in detector response. Despite these differences, the SD and CI values were within acceptable limits, supporting the method’s accuracy. Overall, the high recovery rates across all tested concentrations indicate that the method reliably quantifies the analytes without considerable loss or interference.

Precision and reproducibility

Precision is an integral parameter in method validation, reflecting the consistency of results when the analytical procedure is repeated under identical experimental conditions [41]. In this study, the intermediate precision of the GC-MS method was evaluated by analysing EG and DEG at a target concentration of 10 µg/mL across multiple replicates performed by two analysts. The results provide insight into the method’s consistency and overall analytical performance. The %RSD for EG is was? 1.4345%, indicating minimal variability across the 12 replicates. Similarly, the %RSD for DEG is was? 2.5498%. Both %RSD values were within the commonly accepted threshold of ≤ 5%, indicating excellent repeatability and reproducibility. Thus, the data show that the developed GC-MS method provides high precision across different analytical conditions. The low %RSD values confirm that the method is reliable for quantifying EG and DEG in pharmaceutical syrup matrices. Such precision is paramount for regulatory and quality control purposes, as it ensures that the method consistently detects and quantifies these analytes with minimal variability, thereby enhancing confidence in the analytical data.

Analytical application for the determination of EG and DEG in pharmaceutical syrup surveillance samples

The developed and validated GC-MS method was subsequently applied to pharmaceutical syrup surveillance products, with 51 samples systematically analysed for EG and DEG. The surveillance samples were selected as part of an initial nationwide post-method-development screening exercise conducted by NPRA, encompassing pharmaceutical syrups manufactured in Malaysia. The samples represented different manufacturers and formulations available in the local market, reflecting NPRA’s proactive surveillance approach in response to international concerns about contamination of pharmaceutical syrups with EG and DEG [42].

All analyses were conducted using the validated method, with LODs and LOQs of 0.02% and 0.06%, respectively, for both analytes. From the 51 samples tested (Fig. 5), 38 samples (74.5%) showed no detectable levels of either EG or DEG, remaining below the LOD of 0.02%. These findings indicate no contamination within the analytical sensitivity range. However, 13 samples (25.5%) showed detectable levels of EG contamination. Among these, 11 were below the LOQ (i.e., between LOD and LOQ) while two samples had quantified EG concentrations of 0.09%, slightly above the LOQ. Although still below the maximum allowable limit, the presence of quantifiable EG raises safety concerns, particularly in paediatric formulations or in the context of cumulative exposure [43]. Notably, DEG was not detected in any of the 51 samples, suggesting that the observed contamination pattern was limited to EG.

Fig. 5.

Prevalence of (a) EG and (b) DEG contamination in pharmaceutical syrup surveillance samples (n = 51)

The detection of EG in a subset of samples may reasonably be attributed to multiple sources across the pharmaceutical supply chain. EG is structurally similar to several commonly used pharmaceutical excipients and may arise as an unintended contaminant in ingredients such as propylene glycol or glycerol, particularly when these are sourced from non-pharmaceutical-grade or inadequately purified industrial-grade materials. Historical contamination incidents have also shown that inadequate quality control of solvent-grade excipients remains a critical risk factor for the presence of EG and DEG in liquid oral formulations [12]. In addition, contamination may occur during manufacturing through shared equipment, insufficient cleaning validation, or cross-contamination between production batches [44]. Supply chain complexity, including variability in excipient sourcing, storage conditions, and transport, may further contribute to low-level contamination or impurity carryover [45, 46].

Although the levels observed in this study were low, the persistence of EG highlights the need for stringent testing of incoming raw materials and continuous monitoring of high-risk excipients in syrup formulations. Although the method was validated and tested with finished pharmaceutical syrup products in this study, its demonstrated sensitivity and accuracy also make it suitable for quality-control analysis of raw materials. This versatility enhances its utility in pre- and post-market surveillance, particularly in regulatory laboratories where high-throughput, accurate screening is imperative.

Limitations and recommendations

Selected analytical and methodological limitations in the present study could be addressed to improve the method’s reliability. First, weighted regression modelling was not evaluated during calibration curve development, despite its established role in addressing heteroscedasticity commonly observed in chromatographic calibration data [47]. In the present study, the calibration curve exhibited a negative y-intercept, suggesting potential systematic bias and possible signal suppression, particularly at lower concentrations. This behaviour indicates that the ordinary least squares model may not fully describe the concentration–response relationship and could lead to underestimation of the true analyte concentration [48]. Accordingly, applying weighted regression models (e.g., 1/x or 1/x² weighting) may improve residual distribution and minimise low-concentration bias, thereby enhancing calibration accuracy and model robustness.

Second, no internal standard was employed in the quantitative analysis. The absence of an internal standard is conventionally considered a limitation, as it may not fully account for variability arising from sample preparation, injection-volume fluctuations, instrumental drift, and matrix effects [49–52]. However, the impact of this omission is mitigated by the overall performance of the validated method. The assay demonstrated good accuracy and precision across all quality control levels, and a consistent relationship was observed between analyte responses in blank pharmaceutical syrup matrices and neat solutions, suggesting limited matrix-related signal distortion within this formulation class. In addition, the syrup dosage form generally presents a more compositionally controlled matrix than crude botanical preparations, thereby reducing the influence of matrix-related analytical variability and the reliance on internal standardisation. While the inclusion of isotopically labelled or structurally similar internal standards would further enhance analytical robustness and is recommended for future method optimisation, the current method demonstrates acceptable performance characteristics. Overall, the validation data support the suitability of the approach for its intended regulatory surveillance application.

Third, although matrix effects were acknowledged qualitatively, they were not quantitatively assessed. Matrix effects are a recognised consideration in GC–MS analysis, as excipient composition can influence analyte ionisation efficiency, recovery, and detector response [53]. Pharmaceutical syrups typically have well-defined excipient systems; however, matrix complexity can still vary depending on formulation additives and flavouring agents, which may contribute to matrix-associated signal variability. Future studies could further strengthen method characterisation by incorporating a quantitative assessment of matrix effects, for example, through matrix-matched calibration across representative formulations, to better capture inter-product variability. It is also relevant to note that the electron ionisation (EI) technique employed in this study is comparatively robust and less prone to matrix-induced ion suppression or enhancement effects than softer ionisation approaches such as chemical ionisation [54]. As a result, any unquantified matrix effects are expected to be limited in magnitude and, where present, would most likely manifest as conservative bias rather than false-positive inflation. Accordingly, this limitation is not expected to materially affect the identification of products exceeding regulatory thresholds or the overall interpretation of surveillance findings.

Some contextual limitations should also be considered when interpreting the findings. The study was confined to a single liquid dosage form (syrup); therefore, the method’s applicability to other pharmaceutical formulations, such as suspensions, emulsions, and elixirs, was not evaluated. In addition, no direct exposure assessment of EG and DEG was conducted; hence, the findings are limited to analytical detection and do not extend to toxicological exposure or risk quantification in affected populations.

Complementary to these analytical refinements, additional measures, such as external proficiency testing to ensure inter-laboratory consistency [55], the development of multi-residue screening approaches to broaden detection of excipient adulterants [56], and the adoption of more sensitive and selective analytical technologies [57], may further support analytical performance. Improved data integration systems and continued collaboration among pharmaceutical, healthcare, and scientific stakeholders may also enhance the overall effectiveness of post-market pharmaceutical safety monitoring [58–62] (Fig. 6).

Fig. 6.

Prospective framework for enhancing EG and DEG surveillance using GC-MS

Conclusion

To conclude, this study offers an evaluative overview of Malaysia’s regulatory approach to controlling EG and DEG in pharmaceutical syrups. A validated GC-MS method is presented for the sensitive and specific detection of EG and DEG in these products, supporting efforts to enhance product safety. This analytical approach proved suitable for routine quality control applications, enabling the reliable identification of EG and DEG contaminants.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

ADH

Alcohol dehydrogenase

ADR

Adverse drug reaction

AGREE

Analytical GREEnness

ASEAN

Association of Southeast Asian Nations

BP

British Pharmacopoeia

CIs

Confidence intervals

CoA

Certificates of Analysis

DEG

Diethylene glycol

EG

Ethylene glycol

EI

Electron ionisation

GC

Gas chromatography

GMP

Good Manufacturing Practice

ICH

International Council for Harmonisation

LC

Liquid chromatography

LOD

Limit of detection

LOQ

Limit of quantification

MADRAC

Malaysian Adverse Drug Reaction Monitoring Program

MS

Mass spectrometry

NPRA

National Pharmaceutical Regulatory Agency

Ph. Eur

European Pharmacopoeia

PRHs

Product registration holders

S/N

Signal-to-noise

SIM

Selected ion monitoring

USP

United States Pharmacopoeia

WHO

World Health Organization

Author contributions

Muhammad Amirul Amil: Writing – review & editing, Writing – original draft, Visualization, Methodology, Formal analysis, Data curation, Conceptualization. Muhammad Shahariz Mohamad Adzib: Writing – review & editing, Project administration, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Nur Baizura Bujang: Writing – review & editing, Validation, Supervision, Methodology, Conceptualization. Kok Zheng Gan: Writing – review & editing, Methodology, Conceptualization. Sara Khalida Jamal Khair: Writing – review & editing, Methodology, Conceptualization. Malissa Wani Mohammad Amini: Investigation. Ahmad Zariq Syafiq Zaire: Investigation. Ahmad Yusri Mohd Yusop: Writing – review & editing, Validation, Supervision, Methodology, Funding acquisition, Conceptualization.

Funding

This research received no external funding.

Data availability

The data supporting the findings of this study are not publicly available due to regulatory restrictions, but may be obtained from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

The study protocol was reviewed by the Medical Research and Ethics Committee (MREC) of the Ministry of Health Malaysia and granted an exemption, as the study did not involve human subjects or interventions (NMRR ID-25-02211-WXU).

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.