Trace-level quantification of levofloxacin NDSRIs using LC–TQ–MS/MS: a multi-metric sustainability and integrated toxicological–analytical framework

Abstract

Nitrosamine drug substance-related impurities (NDSRIs) have emerged as a critical regulatory concern due to their potential genotoxicity at trace levels, necessitating highly sensitive and scientifically robust analytical methodologies. Despite the widespread use of levofloxacin (LOX), no validated analytical method has been reported for the simultaneous determination of LOX-specific NDSRIs. Although several LC–MS/MS methods have recently been reported for NDSRIs associated with other pharmaceutical compounds, these studies primarily focused on compound-specific impurity monitoring and did not address levofloxacin-derived NDSRIs. Furthermore, most reported methodologies emphasize analytical sensitivity without integrating toxicological risk assessment and sustainability evaluation. In this study, a highly sensitive and selective LC–TQ–MS/MS method was developed and validated for the simultaneous trace-level quantification of N-nitroso methyl piperazine (N-NMP) and N-nitroso desmethyl levofloxacin (N-NDLOX) in LOX drug substance and tablet formulations. A systematic risk-based analytical development strategy was employed to define the Analytical Target Profile (ATP) and guide method development and optimization. Structural confirmation of the synthesized NDSRIs was achieved using advanced spectroscopic techniques, including ¹H–¹⁵N HMBC NMR. The developed method demonstrated excellent sensitivity, with limits of quantification of 0.6 ng mL⁻¹ for N-NMP and 1.0 ng mL⁻¹ for N-NDLOX, enabling reliable detection well below the proposed interim control limits. The method exhibited strong linearity (R² > 0.998), high accuracy (93–100%), and robust precision (CV < 10%), confirming its suitability for routine quality control applications. To support regulatory risk assessment, complementary (Q)-SAR analysis and carcinogenic potency categorization (CPCA) were integrated to establish scientifically justified interim control limits for LOX-associated NDSRIs. The sustainability of the developed method was further evaluated using a multi-metric framework incorporating AGREE, GAPI, RAPI, and the Environmental, Performance, and Practicality Index (EPPI). The results demonstrated a balanced analytical profile, with moderate environmental impact (AGREE score: 0.62; EI: 75.8%) and high practical applicability (PPI: 87.0%), resulting in an overall EPPI score of 81.4%. This study presents a unified analytical–toxicological–sustainability framework for NDSRI assessment, representing a shift from conventional method development toward performance-driven sustainable analytical design. The proposed method is suitable for regulatory implementation and provides a scalable strategy for controlling structurally complex nitrosamine impurities in pharmaceutical systems.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s13065-026-01872-4.

Introduction

Nitrosamine drug substance–related impurities (NDSRIs) have gained significant attention in the pharmaceutical industry due to their potential genotoxic and carcinogenic risks, even at trace exposure levels. Among this class, N-nitroso desmethyl impurities represent a particularly important and frequently encountered subgroup, arising from the nitrosation of desmethylated amine moieties that are structurally related to the parent Active Pharmaceutical Ingredient (API) [1–4]. Because these impurities retain the core molecular framework of the drug substance, they exhibit distinct physicochemical and toxicological characteristics compared with classical, low-molecular-weight nitrosamines, necessitating compound-specific risk evaluation and control strategies [5–7].

The regulatory focus on N-nitroso desmethyl impurities intensified following the detection of nitrosamines in several marketed pharmaceutical products, leading global health authorities to mandate comprehensive nitrosamine risk assessments in accordance with ICH M7(R2) guidelines[8]. Regulatory agencies such as the US Food and Drug Administration (FDA) [9] and the European Medicines Agency (EMA) [10] now require the systematic identification of potential NDSRIs, the justification of their formation pathways, and the establishment of stringent acceptable intake limits, often in the low-nanogram-per-day range. N-nitroso desmethyl impurities can be formed during API synthesis through reactions between desmethyl amines and nitrosating agents, introduced via contaminated reagents or solvents, or generated during formulation and storage under conducive chemical conditions [11–15]. These developments have created an urgent demand for analytical methodologies capable of reliable trace-level detection while meeting increasingly stringent regulatory expectations.

Levofloxacin (LOX) is a clinically important fluoroquinolone antibiotic widely used to treat respiratory, urinary tract, and skin infections. As the biologically active L-isomer of ofloxacin, it inhibits DNA gyrase and topoisomerase IV, thereby disrupting bacterial DNA replication and repair. Its high oral bioavailability, predictable pharmacokinetics, and broad antimicrobial spectrum have contributed to its widespread clinical use [16–20]. Given its widespread global consumption and availability in multiple dosage forms, even trace-level contamination with genotoxic impurities, such as NDSRIs, represents a significant patient safety concern. In particular, N-nitroso desmethyl levofloxacin (N-NDLOX) and N-nitroso methyl piperazine (N-NMP) are structurally plausible impurities that may arise during manufacturing or storage, highlighting the need for highly sensitive and selective analytical approaches.

The formation of LOX-associated NDSRIs can be rationalized based on the structural features of levofloxacin and known nitrosation chemistry. N-NDLOX may arise through nitrosation of desmethyl levofloxacin (DLOX), a potential process-related or degradative intermediate containing a secondary amine functionality susceptible to reaction with nitrosating agents under favorable conditions. Similarly, N-NMP may originate from N-methyl piperazine (NMP), which can be present as a synthetic intermediate, impurity, or degradation-related fragment associated with the piperazine moiety of levofloxacin. Trace nitrosating species, such as nitrite impurities introduced through raw materials, process water, excipients, or environmental exposure, may facilitate the conversion of these amine-containing precursors into their corresponding nitrosamine derivatives during synthesis, formulation, or long-term storage. Although the exact formation pathways of LOX-related NDSRIs remain to be fully elucidated, these mechanisms are consistent with currently recognized nitrosamine formation pathways reported for pharmaceutical products.

Despite the increasing regulatory emphasis on nitrosamine control, no analytical method has been reported to date for the simultaneous trace-level quantification of LOX-specific NDSRIs in pharmaceutical formulations, particularly with integrated consideration of both analytical performance and sustainability. Existing methods for nitrosamine analysis are largely focused on small, volatile nitrosamines and often fail to address structurally complex NDSRIs derived from drug substances. Moreover, current analytical strategies predominantly emphasize sensitivity, with limited attention to environmental impact, resource efficiency, and method practicality, resulting in a fragmented approach to analytical method development [21–25].

In parallel, global initiatives such as the United Nations 2030 Agenda have reinforced the importance of sustainable scientific practices. Green Analytical Chemistry (GAC) provides a structured framework for minimizing environmental impact by reducing solvent consumption, simplifying workflows, and using safer reagents [26–31]. However, the exclusive focus on greenness may overlook critical aspects such as analytical performance and operational feasibility. To address this limitation, the concept of White Analytical Chemistry has emerged, emphasizing the balanced integration of environmental sustainability, analytical reliability, and practical applicability. In this context, the use of multi-metric assessment tools enables a more comprehensive and realistic evaluation of analytical methods.

In the present study, we report the development and validation of a highly sensitive LC–TQ–MS/MS method for the simultaneous determination of LOX-related NDSRIs (Fig. 1) in drug substance and tablet formulations. In addition to analytical method development, a novel integrated framework combining (Q)-SAR prediction, carcinogenic potency categorization (CPCA), and multi-metric sustainability assessment was employed to establish scientifically justified control strategies for LOX-NDSRIs. The sustainability profile of the method was evaluated using the Analytical GREEnness (AGREE), Green Analytical Procedure Index (GAPI), Red Analytical Performance Index (RAPI), and Environmental, Performance, and Practicality Index (EPPI) tools, enabling a holistic assessment aligned with the principles of White Analytical Chemistry [32–36].

Fig. 1.

Chemical structures of A LOX API, B N-NMP, and C N-NDLOX

Despite significant progress in Green Analytical Chemistry, the application of sustainable analytical approaches to nitrosamine determination remains challenging. The ultra-trace concentration levels at which nitrosamines must be monitored often necessitate the use of highly sensitive LC–MS/MS instrumentation, organic solvents, and energy-intensive analytical workflows. Consequently, many reported nitrosamine methods achieve the required analytical sensitivity but have sustainability limitations. Conversely, efforts to improve environmental performance may compromise analytical sensitivity, selectivity, or regulatory compliance. In addition, most sustainability assessments reported for analytical methods rely primarily on greenness metrics and do not adequately consider analytical performance or practical applicability. As a result, there remains a need for balanced evaluation strategies that simultaneously address environmental impact, analytical reliability, and operational feasibility, particularly for regulatory-driven analyses such as NDSRI determination.

This work, therefore, extends beyond conventional method development by integrating toxicological risk assessment and sustainability evaluation into a unified framework, providing a scalable, regulatory-relevant strategy for controlling structurally complex nitrosamine impurities in pharmaceutical systems.

Experimental section

Standards and reagents

The reference compounds, including levofloxacin drug substance (LOX, purity 99.79%), desmethyl levofloxacin (DLOX, purity 99.32%), and N-methyl piperazine (NMP), were obtained as complimentary samples from Synpure Laboratories (Hyderabad, India). In the absence of commercially available standards for LOX-associated NDSRIs, N-NMP and N-NDLOX were synthesized in-house and structurally confirmed by spectroscopy to ensure analytical reliability. LC–MS grade solvents, including acetonitrile, methanol, ammonium formate, formic acid, and trifluoroacetic acid, were procured from Honeywell (Mumbai, India). The selection of LC–MS-grade reagents was essential to minimize background noise and achieve high sensitivity for trace-level detection. High-purity water was obtained from a Milli-Q purification system. Commercially available LOX tablet formulations (500 mg strength) from three manufacturers were procured to evaluate the method's applicability across diverse matrices. All solvents and reagents used for chromatographic analysis were of LC–MS grade, while chemicals employed for impurity synthesis were of analytical reagent grade and used as received from commercial suppliers. Supplier information for the principal standards and analytical reagents is provided in this section.

Analytical instrumentation and software tools

Chromatographic separation and detection were performed using an Acquity H-Class UPLC system coupled with an Xevo TQ-XS triple quadrupole mass spectrometer (Waters Corporation, Milford, MA, USA). The system was equipped with a quaternary solvent manager and a flow-through needle autosampler to ensure precise solvent delivery and minimize carryover during trace-level nitrosamine analysis. Mass spectrometric detection was performed in multiple reaction monitoring (MRM) mode to achieve high sensitivity and selectivity for quantifying LOX-associated NDSRIs. Structural characterization of the synthesized impurities was performed using a Bruker Avance III 500 MHz nuclear magnetic resonance (NMR) spectrometer (Bruker BioSpin GmbH, Rheinstetten, Germany). Data acquisition, instrument control, chromatographic processing, peak integration, calibration curve generation, and quantitative analysis were performed using MassLynx and TargetLynx software (Waters Corporation, Milford, MA, USA). In silico mutagenicity assessment was conducted using DEREK Nexus and SARAH Nexus software (Lhasa Limited, Leeds, United Kingdom) in accordance with ICH M7(R2) recommendations. Sustainability assessments were performed using the AGREE calculator software and the published methodologies for GAPI, RAPI, and EPPI evaluation to assess the environmental, analytical performance, and practical applicability characteristics of the developed method.

Mobile phase solutions preparation

The mobile phase system was designed to achieve optimal ionization efficiency and chromatographic performance for structurally distinct NDSRIs. Mobile phase A consisted of 0.1% formic acid in water, while mobile phase B was acetonitrile. Formic acid was selected as an acidic modifier to enhance protonation efficiency under positive electrospray ionization conditions, thereby improving sensitivity and signal stability. The mobile phases were degassed by ultrasonication to eliminate dissolved gases and ensure consistent chromatographic performance.

The diluent and impurity standard stock solution preparation

A diluent comprising acetonitrile and 0.1% formic acid in water (60:40, v/v) was selected based on solubility compatibility with LOX and its NDSRIs, as well as to maintain consistency with the mobile phase composition. This approach minimized solvent mismatch effects and ensured stable peak shapes during injection. Standard stock solutions were prepared by dissolving accurately weighed quantities of LOX-NDSRIs in the diluent, then sonicating to ensure complete dissolution. Serial dilutions were performed to obtain working concentrations suitable for trace-level quantification, ensuring linear response within the analytical range.

LOX-API sample preparation

Sample preparation was strategically designed to minimize matrix interference while maintaining simplicity and reproducibility. LOX drug substance samples were dissolved directly in diluent at a concentration of 10 mg/mL, followed by sonication to ensure complete extraction of NDSRIs. The solution was filtered through a 0.22 µm PVDF membrane to remove particulate matter without adsorptive loss of analytes. For tablet formulations, powdered samples were extracted using the same diluent system, then sonicated and centrifuged to separate insoluble excipients. This simplified preparation approach eliminates the need for complex extraction or clean-up steps, thereby reducing potential analyte loss and improving method practicality. The use of PVDF filters was validated to ensure negligible adsorption of LOX and its NDSRIs, supporting accurate quantification.

Results and discussion

Owing to the lack of commercial and pharmacopeial standards for N-NMP and N-NDLOX, we synthesized and characterized these compounds using advanced spectroscopic techniques.

General procedure for synthesis of both N-nitroso compounds

A solution of amine compounds (NMP & Desmethyl LOX) (1.0 eq each) in THF (10 mL) was cooled to 0 °C and stirred. To this, aqueous NaNO₂ (1.2 eq, 5 mL) was added dropwise, followed by AcOH (1.0 eq). After stirring at RT for 12 h, the mixture was diluted with ice water and extracted with EtOAc. The combined organic layers were dried over Na₂SO₄ and chromatographed on silica gel (EtOAc/hexane) to afford N-nitroso derivatives 1a–2a. The synthetic scheme is shown in Fig. 2.

Fig. 2.

Synthetic scheme of A N-NMP, and B N-NDLOX

Structural characterization of N-NMP

Compound 1a: Yield-93%; ESI–MS [M + H]⁺ 130.12. ¹H-NMR (500 MHz, dmso-d6): δ (ppm) 4.18 (t, J = 5 Hz,2H), 3.72 (t, J = 5 Hz,2H), 2.52 (t, J = 5 Hz,2H), 2.27 (t, J = 5 Hz,2H), 2.22 (s, 3H) 4 (Fig. S1).

Structural characterization of N-NDLOX

Compound 2a: Yield-85%; ESI–MS [M + H]⁺ 377.36. ¹H-NMR (500 MHz, dmso-d6): δ(ppm) 15.16 (s, 1H), 9.00 (s, 1H), 7.65 (d, J = 12 Hz,1H), 4.95 (q, J = 7 Hz,1H), 4.62 (d, J = 1.5 Hz,1H), 4.60 (d, J = 1 Hz,1H), 4.41 (t, J = 2.5 Hz,2H), 3.90 (t, J = 5.5 Hz,2H), 3.55 (t, J = 5.5 Hz,2H), 3.29 (t, J = 5.5 Hz,2H), 1.47 (d, J = 6.5 Hz,3H) (Fig. S2).

Since conventional ¹H and ¹³C NMR analyses did not exhibit signals characteristic of the nitroso group, additional spectroscopic investigations were required. Accordingly, 2D ¹⁵N–¹H HMBC spectroscopy was utilized to probe long-range nitrogen–hydrogen interactions. The resulting spectra exhibited distinct ¹⁵N–¹H correlations, providing unequivocal evidence for the incorporation of the nitroso group within the molecular structures of both LOX-NDSRIs (Fig. 3).

Fig. 3.

¹H-¹⁵N HMBC NMR spectra of A N-NMP, and B N-NDLOX

Establishing an interim control threshold for LOX-NDSRIs via CPCA (carcinogenic potency categorization approach) methodology

To address the challenges of establishing acceptable intake limits for nitrosamines, the EMA introduced the CPCA in July 2023. This approach categorizes nitrosamines by carcinogenic potency, facilitating AI determination, particularly when compound-specific toxicological data are limited. NDSRI compounds are classified according to their α-hydrogen count and assigned to one of five potency categories, each associated with an AI value between 18 and 26.5 and 1500 ng/day (CDER Nitrosamine Impurity Acceptable Intake Limits FDA).

The structural assessment of the identified NDSRIs indicated that both N-NMP and the N-NDLOX possess two α-hydrogens each (2,2). According to the CPCA scoring system, these impurities were assigned potency scores of 4 and 3 for N-NDLOX and N-NMP, respectively. Consequently, both were classified in CPCA potency category 3, which denotes a lower carcinogenic concern than categories 1 and 2. A detailed summary of the categorization outcome for LOX-related NDSRIs is presented in Fig. 4. To establish an acceptable interim control limit for these impurities, the CPCA-recommended equation was applied, wherein the control limit (µg/g) is derived from the ratio of the acceptable intake (expressed in ng/day) to the maximum daily dose (MDD, expressed in mg) of the drug substance. For LOX, an MDD of 750 mg was considered [37, 38]. On this basis, the calculated interim control limit is < 2 µg/g for N-NDLOX and 0.5 µg/g for N-NMP (equivalent to 1500 ng/750 mg for N-NDLOX and 400 ng/750 mg for N-NMP). This specification defines a scientifically supported threshold consistent with regulatory guidance to safeguard patient safety.

Fig. 4.

Flow chart of CPCA approach for A N-NMP, and B N-NDLOX

The interim control limit was calculated according to the CPCA approach using the following relationship:

Control limit (µg/g or ppm) = acceptable intake (ng/day) / MDD (mg/day).

For N-NMP, the acceptable intake assigned to CPCA Category 3 was 400 ng/day. Considering the maximum daily dose of levofloxacin as 750 mg/day, the interim control limit was calculated as:

400 ng/day ÷ 750 mg/day = 0.53 µg/g.

which was conservatively rounded to 0.5 µg/g.

For N-NDLOX, the corresponding AI value was 1500 ng/day. Applying the same MDD:

1500 ng/day ÷ 750 mg/day = 2.0 µg/g.

Thus, the interim control limit for N-NDLOX was established as 2 µg/g.

(Q)- SAR data prediction

The (Q)-SAR approach plays a crucial role in regulatory decision-making by efficiently classifying chemical substances into predefined risk categories (as outlined in the ICH M7(R2) guideline). It enables the rapid identification of potentially hazardous compounds, thereby supporting safety evaluations and ensuring regulatory compliance in pharmaceutical development. LOX, N-NMP, and N-NDLOX were submitted to DEREK, and predictions were generated using the ICH classification model. The anticipated results, summarized in Table 1, reveal that both statistical and expert rule-based (Q)-SAR predictions for the NDSRIs were positive, classifying them as Class 3 and indicating potential mutagenicity. The integration of computational (Q)-SAR tools with CPCA-guided potency categorization and specification limit determination for LOX-NDSRIs represents a critical advancement in promoting a more scientifically robust framework for pharmaceutical risk assessment. These efforts contribute directly to achieving SDG 3, particularly Targets 3.8 and 3.d, by supporting universal access to safe medicines and enhancing health risk management capabilities.

Table 1.

Summary of (Q)-SAR DEREK software genotoxicity/mutagenicity assessment

Structure	ICH M7 class	Coherent of concern	Derek prediction	Sarah prediction	Experimental data	Similar to API	Overall in Silico
	Class-3	Yes (N-nitroso compound)			Carc: Unspecified Ames: Unspecified	Alert(s) not found in API	Positive
	Class-3	Yes (N-nitroso compound)			Carc: Unspecified Ames: Unspecified	Alert(s) not found in API	Positive

The mutagenicity assessment was performed using complementary expert rule-based (DEREK Nexus) and statistical (SARAH Nexus) (Q)-SAR platforms in accordance with ICH M7(R2) principles. Both LOX-associated NDSRIs were identified as nitrosamine-containing structures and classified as Class 3 impurities, with concordant positive outcomes from the in silico assessment. The positive classification was primarily associated with the presence of the nitrosamine structural feature, a recognized alert for potential mutagenicity. These predictions were used as supportive evidence for toxicological risk assessment and CPCA-based specification setting.

Analytical target profile and method requirements

In accordance with the general concepts of ICH Q14, [39] the development of the analytical procedure began with defining the Analytical Target Profile (ATP), which established the method's intended purpose, target analytes, analytical matrix, and required sensitivity, selectivity, accuracy, precision, and regulatory performance requirements. The ATP served as the foundation for method development and validation activities. The complete ATP for this investigation is presented in Table 2.

Table 2.

Analytical target profile (ATP)

Intended purpose Investigation of the content of Levofloxacin-NDSRIs in drug substance and tablet formulations

Link to Critical Quality Attribute (Trace-level detection and quantification of LOX-NDSRIs) The analytical method should be capable of quantifying LOX-NDSRIs, 1) a sensitivity that can meet the 10% of the AI value 2) specificity, report range, accuracy, and precision, which can be detected in DS and tablet formulations

Performance characteristics	Acceptance criteria	Rationale
Specificity	Absence of interference in the blank sample matrix Q to q ratio for test samples within ± 20% tolerance	The general recommendation by the ICH Q2 and Q14 guidelines Recommendation in ICH Q2(R2) Annex II
Sensitivity	The quantitation limit (QL) of 1 µg/g	The proposed necessitated the QL of 1 µg/g
Matrix effect	Q-to-q ratio for test samples within ± 20% tolerance	Recommendation in ICH Q2(R2) Annex II
Accuracy	80−120% for QL limit	The general recommendation by the ICH Q2 and Q14 guidelines
Precision	80−120% CV ≤ 10 for QL limit	The general recommendation by the ICH Q2 and Q14 guidelines
Report Range	0.06–0.75 µg/g for N-NMP 0.1–3 µg/g for N-NDLOX	The AI value is 400, the interim control limit for N-NMP is 0.5 µg/g (AI/MDD = 400 ng/750 mg) The AI value is 1500, the interim control limit for N-NDLOX is 2 µg/g (AI/MDD = 1500 ng/750 mg)

In this investigation, the primary objective was to quantify trace levels of LOX-NDSRIs using state-of-the-art, highly sensitive detection techniques. MRM, a technique known for its exceptional specificity and sensitivity, was employed because of its use in drug development workflows. LC-TQ-MS/MS with MRM detection was used to analyze LOX-NDSRIs in LOX formulations. The sample preparation methodology was carefully modified based on the solubility profiles of the impurities, the LOX API, and excipients to ensure efficient extraction and accurate chromatographic separation of the target NDSRIs.

Assessment of matrix interference

To ensure analytical reliability, early-stage risk assessment addresses uncertainties, particularly matrix effects, by refining sample prep and instrumental controls. In this investigation, the effects were evaluated using the quantifier-to-qualifier ion ratio (the ratio of the peak area of a qualifier ion to that of the quantifier ion for a given analyte-Q-to-q ratio) strategy, in accordance with ICH Q2(R2) Annex II.

The average ratio was calculated from six independently prepared calibration standards. The following expression was applied to estimate the deviation of the Q-to-q ratio for individual sample solutions.

Q-to-q ratio deviation (%) = ((Average Q-to-q ratio-Sample Q-to-q ratio)) / (Average Q-to-q ratio) X 100.

Analytical procedure development and optimization

Improving the sample preparation process and filter evaluation study

Reliable trace-level quantification of NDSRIs requires a sample preparation strategy that minimizes matrix effects while ensuring high recovery and reproducibility. The sample preparation approach was therefore designed to balance analytical efficiency with procedural simplicity. The solubility characteristics of LOX and its associated NDSRIs were first evaluated to identify a suitable extraction medium. LOX is freely soluble in aqueous media and moderately soluble in organic solvents, whereas NDSRIs exhibit differential polarity due to the presence of nitroso functionalities. Based on these considerations, a binary solvent system comprising 0.1% formic acid in water and acetonitrile (40:60, v/v) was selected to simultaneously solubilize the parent drug and impurities while maintaining compatibility with LC–MS analysis.

Sonication time was systematically optimized to maximize extraction efficiency without compromising analyte stability. A duration of 15 min was found to provide quantitative recovery (> 98%) for both NDSRIs with minimal variability. Extended sonication did not result in detectable degradation, confirming the analyte's stability under the applied conditions. To ensure robustness of the sample preparation procedure, a filter-compatibility study was conducted to assess potential analyte adsorption. Comparative analysis of filtered and unfiltered samples showed no significant difference in peak areas, confirming that the PVDF membrane filter does not contribute to analyte loss. The resulting procedure eliminates the need for complex extraction or cleanup steps, thereby reducing procedural variability and enhancing the method's practicality.

Chromatographic separation method development

The chromatographic method development was guided by the need to achieve efficient separation of structurally related NDSRIs from the parent LOX under trace-level detection conditions. A key challenge in this system arises from the polarity mismatch between LOX and its corresponding nitrosamine impurities, which necessitates careful selection of stationary phase chemistry to ensure adequate retention and resolution. Initial method development was carried out using a mixed standard solution (1 mg/mL) to evaluate chromatographic behavior under controlled conditions. During this phase, the LC effluent was diverted to waste to prevent contamination of the mass spectrometer source. A systematic screening approach was adopted, evaluating multiple reversed-phase C18 columns with varying dimensions and particle characteristics, as well as different mobile phase compositions.

Early trials using an Agilent Porosil C18 column (150 × 4.6 mm, 2.7 μm) under gradient conditions resulted in distorted peak shapes and inadequate resolution between LOX and its NDSRIs, indicating insufficient selectivity of the stationary phase. Similarly, the Waters Cortecs T3 C18 column (100 × 4.6 mm, 2.7 μm) led to early elution of analytes with poor retention, particularly for N-NMP, highlighting the limitations of conventional reversed-phase systems for moderately polar nitrosamine derivatives. In contrast, the Waters XBridge BEH C18 column (150 × 2.1 mm, 2.5 μm) provided significantly improved chromatographic performance, offering enhanced retention and resolution for both LOX and its associated NDSRIs. The improved separation is attributed to the hybrid particle technology, which facilitates better interaction with analytes of intermediate polarity.

Mobile phase optimization was performed to ensure compatibility with electrospray ionization while maintaining chromatographic efficiency. Several volatile buffer systems, including 0.1% trifluoroacetic acid, ammonium acetate, and formic acid, were evaluated. Among these, 0.1% formic acid was selected as the optimal modifier, as it enhanced protonation efficiency under positive ESI conditions while maintaining sharp peak shapes and high signal intensity. Other buffers either resulted in ion suppression or compromised chromatographic performance. Gradient elution conditions were systematically optimized to achieve adequate separation within a reasonable analysis time. The final gradient program enabled clear resolution of N-NMP, LOX, and N-NDLOX with retention times of 3.75, 6.97, and 8.86 min, respectively. The optimized chromatographic conditions are summarized in Table 3, and representative chromatograms demonstrating specificity and separation are presented in Fig. 5 and Fig. S3.

Table 3.

Optimized UPLC chromatographic conditions

Parameter	Details
UPLC conditions
Separation mode	Reversed Phase
Instrument	Waters H-Class UPLC (Waters Corporation, USA) Auto Sampler Manager with a Flow-Through Needle (FTN- L20FTP352G) Quaternary Solvent Manager (QSM- L20QSP358A) PDA-photo diode array detector model (L20UPD127A)
Column	Waters XBridge column (BEH C18-150 mm × 2.1 mm, 2.5 μm)
Flow mode	Gradient
MP-A	0.1% FA in water
MP-B	Acetonitrile
Flow rate	0.4 mL/min
Needle wash	Acetonitrile: water (80:20) v/v
Gradient Program	Time (min)	%MP-A	%MP-B
	0.0	90	10
	1.0	90	10
	6.0	10	90
	11.0	10	90
	11.1	90	10
	12.0	90	10
Column oven temperature	40 °C
Sample cooler temperature	25 °C
Injection volume	10.0 µL
Wavelength	250 nm
Run time	12.0 min

Fig. 5.

Typical MRM chromatograms of A Diluent, B Placebo, C 5 ng of standard N-NMP, D Spiked 5 ng of N-NMP in tablet formulation at 10 mg/mL sample concentration, E 20 ng of N-NDLOX, and F Spiked 20 ng of N-NDLOX in tablet formulation at 10 mg/mL sample concentration

Column temperature was found to be a critical parameter influencing retention stability and peak shape. A temperature of 40 °C was selected to ensure reproducibility and minimize variability in retention time. The optimized method provided well-defined peak profiles with consistent retention behavior across multiple injections, confirming the robustness of the chromatographic system. Overall, the chromatographic method development was not based on empirical trial and error but was guided by a systematic evaluation of analyte properties, interactions with the stationary phase, and ionization compatibility. This approach enabled the establishment of a robust separation platform suitable for sensitive and selective LC–TQ–MS/MS analysis of LOX-related NDSRIs. The chromatographic behavior and relative retention characteristics of LOX and the synthesized NDSRIs under the optimized conditions are illustrated in the representative overlay UV chromatograms shown in Fig. S3.

Optimizations of MS/MS parameters

Mass spectrometric conditions were systematically optimized to achieve high sensitivity and selectivity for the trace-level detection of LOX-associated NDSRIs. The optimization strategy was guided by the structural characteristics of the analytes, particularly the presence of tertiary amine and nitroso functional groups, which favor protonation under positive electrospray ionization (ESI +) conditions. Initial ionization experiments were conducted using both positive and negative ESI modes. Positive ion mode was selected due to the formation of stable, abundant protonated molecular ions [M + H] ⁺ for both N-NMP and N-NDLOX. In contrast, negative ion mode yielded significantly lower signal intensity and poor reproducibility. This observation is consistent with the analytes' basic nature and their tendency to protonate in acidic environments. Representative quantifier and qualifier MRM chromatograms obtained for N-NMP and N-NDLOX are presented in Fig. S4, demonstrating the selectivity of the selected transitions used for analyte confirmation and quantification.

To ensure high selectivity and regulatory compliance for nitrosamine analysis, MRM transitions were carefully optimized for each analyte. Precursor ions were identified by direct infusion, followed by systematic optimization of cone voltage and collision energy to generate stable, abundant product ions. The selection of quantifier and qualifier transitions was based not only on signal intensity but also on fragmentation specificity, ensuring reliable confirmation of analyte identity in accordance with regulatory expectations for trace-level impurity analysis. Fragmentation behavior was evaluated to identify structurally informative product ions, particularly those associated with cleavage of the nitroso functional group and piperazine moiety. This approach enhanced method selectivity by minimizing the risk of false positives arising from structurally related matrix components. The optimized MRM transitions and corresponding parameters are summarized in Table 4.

Table 4.

Finalized tuning MRM mass spectrometry conditions for LOX-NDSRIs in ESI^+ve mode

Compound	Chemical formula	Precursor ion (m/z)	Product ion (m/z)	Spectral window (min)	Dwell time (ms)	Collision energy (V)	Cone Voltage (V)
N-NMP	C5H11N3O	130.033	100.017 (Quantifier ion)	3.61–3.95	25	16	14
			57.958 (Qualifier ion)		25	14	14
N-NDLOX	C17H17FN4O5	377.223	347.126 (Quantifier ion)	8.70–8.95	25	14	4
			259.084 (Qualifier ion)		25	18	4

Source-dependent parameters, including capillary voltage, desolvation temperature, and gas flow rates, were systematically optimized to maximize ionization efficiency and signal stability. A desolvation temperature of 450 °C and a gas flow rate of 850 L/h were found to efficiently desolvate droplets, thereby reducing ion suppression and improving signal-to-noise ratio. The optimized ion source conditions ensured a consistent response across replicate injections. Dwell time was optimized to balance sensitivity and chromatographic peak definition. A dwell time of 25 ms provided an adequate number of data points for each peak while maintaining high signal intensity, ensuring accurate quantification at trace concentrations. To further enhance method robustness and reduce matrix-related interference, a time-segmented flow diversion strategy was implemented. The LC effluent was directed to waste during non-relevant elution windows and introduced into the mass spectrometer only during analyte elution. This approach minimized contamination of the ion source, improved long-term instrument stability, and enhanced overall method sensitivity. The selected MRM transitions were supported by characteristic fragmentation behavior of the target analytes. The proposed MS/MS fragmentation pathways for N-NMP and N-NDLOX are illustrated in Fig. S5 and provide structural justification for the selected quantifier and qualifier ions.

Carryover was evaluated and mitigated by optimizing needle wash conditions. A strong wash solvent consisting of acetonitrile and water (80:20, v/v) effectively eliminated residual analyte signal between injections, ensuring reliable quantification at low concentration levels. The MS/MS optimization was driven by a combination of structural considerations, ionization efficiency, and regulatory requirements for selectivity and sensitivity, resulting in a robust, highly selective detection system suitable for trace-level NDSRI analysis.

Robustness study

According to ICH Q2(R2) guidelines, robustness is a critical aspect of method development that ensures reliable, consistent performance across varying experimental conditions. The robustness of the method was evaluated by examining its performance under slight variations in chromatographic and MS conditions, in accordance with the guidelines outlined in ICH Q2(R2) Appendix I. Parameters such as flow rate, column temperature, DT, CE, and DGF were deliberately altered. Even under these conditions, the method showed consistent performance, with variation in LOX-NDSRI content of < 10%, confirming its robustness (Table S1).

Analytical method validation study

The developed LC–TQ–MS/MS method was validated in accordance with ICH guidelines to establish its suitability for the trace-level quantification of LOX-associated NDSRIs. Given the stringent regulatory requirements for nitrosamine impurities, particular emphasis was placed on sensitivity, selectivity, and reproducibility at low concentration levels [40–45].

Sensitivity and system suitability test (SST)

System suitability was evaluated before validation experiments. SST solution containing 10 ng mL⁻¹ of each N-NMP and N-NDLOX was injected in six replicates. Precision was assessed using peak area and retention time, with coefficients of variation (CVs) kept below 5%. Sensitivity was evaluated using signal-to-noise (S/N) ratios, with a minimum acceptance criterion of S/N ≥ 10. The results demonstrated CVs of 1.39% for N-NMP and 1.12% for the N-NDLOX, with corresponding S/N values of 84 and 61, confirming adequate system performance and sensitivity.

Specificity

The method's specificity was evaluated by analyzing blank, placebo, and spiked samples to assess potential interferences at the analyte retention times. No significant interfering peaks were observed, confirming the method's ability to selectively quantify N-NMP and N-NDLOX in the presence of LOX and excipient matrices. The use of LC–MS/MS in MRM mode provided an additional level of selectivity, as each analyte was identified based on both retention time and characteristic transition ions. The consistency of quantifier-to-qualifier ion ratios further confirmed analyte identity, ensuring compliance with regulatory expectations for trace-level nitrosamine analysis. Representative chromatograms demonstrating method specificity are shown in Fig. 5

Detection limit (DL) and quantitation limit (QL)

The DL and QL for N-NMP and the N-NDLOX were determined based on signal-to-noise ratios of approximately 3:1 and 10:1, respectively, using standard solutions of known concentrations. The DL and QL were established at 0.2 and 0.6 ng/mL for N-NMP and 0.3 and 1.0 ng/mL for N-NDLOX. Repeatability at the QL level was evaluated by six replicate injections, yielding CVs of 2.17% for N-NMP and 5.25% for the N-NDLOX (Table 5). The achieved LOQ values are significantly below the interim control limits derived from CPCA, confirming method suitability for regulatory implementation.

Table 5.

Summary of method validation result

Method validation parameter	Sample and procedure	Target criteria	Results
			N-NMP	N-NDLOX
Specificity	Diluent, placebo, standard solution, and sample solutions	No interference with peaks of interest Q-to-q deviation should be less than ≤ 20%	No interference 2.64	No interference 2.18
Precision	Six replicate injections of LOX-NDSRIs at specification level spiked to LOX sample	The % CV of the 6 recoveries is ≤ 20%	5.31	3.75
Intermediate Precision	The same samples prepared for precision were used for analysis on different day with the same instrument	% CV of the 6 recoveries is ≤ 20%; The % CV of the 12 recoveries is ≤ 25%	5.56 5.43	4.12 3.90
Detection limit (DL)		Signal-to-noise ratio (s/n) ≥ 3	5.6	4.41
Quantitation limit (QL)		Signal-to-noise ratio (s/n) ≥ 10	14.9	11.9
QL precision		The % CV of the 6 replications is ≤ 20	2.17	5.25
Linearity	Range (ng/mL) Slope Intercept Pearson coefficient (> 0.99)	Pearson coefficient > 0.99	0.6–6 91,753 1338 0.9984	1–30 1479.7 1165.1 0.9980
Recovery	For LOX drug substance: The level at QL mean ± SD The level at 100% mean ± SD The level at 150% mean ± SD For LOX tablet formulations: The level at QL mean ± SD The level at 100% mean ± SD The level at 150% mean ± SD	The average recovery at each level is within 80%–120%	94.13 ± 1.08 96.55 ± 2.51 99.69 ± 2.36 93.09 ± 1.05 97.45 ± 1.89 98.89 ± 1.35	95.38 ± 1.05 99.63 ± 0.13 99.13 ± 1.51 94.27 ± 1.10 99.50 ± 0.25 98.51 ± 1.48
Solution stability		Peak area (0 h to 24 h, % difference with initial < 10 for both nitrosamine impurities)	1.02	1.65

Method precision and intermediate precision

Method precision was assessed by analyzing six replicate preparations of LOX drug substance and 500 mg strength of LOX tablets (10 mg/mL), spiked with both NDSRIs at the interim control limit (0.5 and 2 ppm for N-NMP and N-NDLOX). Intermediate precision was evaluated by reanalyzing the same samples on a different day using the same instrument. The %CV values for both repeatability and intermediate precision were below 10%, demonstrating excellent method precision (Table 5).

Linearity and calibration curve performance

Linearity of the method was established across concentration ranges of 0.6–7.5 ng/mL for N-NMP and 1–30 ng/mL for N-NDLOX, corresponding to 0.06–0.6 and 0.1–3 ppm, respectively, in a 10 mg/mL LOX matrix. Calibration curves were constructed at seven concentration levels using a 1/x weighting factor. Linear regression analysis yielded the equations Y = 91753X + 1338 for N-NMP and Y = 1479.7X + 1165.1 for the N-NDLOX, with excellent correlation coefficients. Back-calculated concentrations at each level showed biases within ± 10%, confirming calibration accuracy. A strong linear relationship between peak area and analyte concentration was observed across the tested range (Table 5).

Matrix effects

Matrix effects were evaluated by comparing the slopes of calibration curves prepared in solvent and in the LOX tablet formulation matrix. A slope ratio within the range of 0.85–1.15 was considered indicative of negligible matrix interference. The obtained slope ratio of 0.94 demonstrated minimal signal suppression or enhancement, confirming that the matrix had no significant impact on analyte ionization or quantification and that the method is robust and reliable. The slope-comparison approach was selected because it provides a direct assessment of the matrix's influence on the analyte response under the final analytical conditions. The observed slope ratio of 0.94 indicated negligible ion suppression or enhancement and demonstrated that the sample matrix did not significantly affect quantitative performance. Although matrix factor determination and post-column infusion experiments may provide additional mechanistic information regarding ionization effects, the results obtained from the matrix-matched calibration study were considered adequate for the intended purpose of method validation and routine regulatory application.

Accuracy and recovery

Method accuracy was evaluated using a standard addition technique by spiking LOX drug substance and tablet formulation samples at three concentration levels: QL, 100%, and 150% of the specified control limit. The spike levels for N-NMP were 0.06, 0.5, and 0.6 ppm, while those for N-NDLOX were 0.1, 2.0, and 3.0 ppm. Each concentration level was analyzed in triplicate. The mean recoveries for N-NMP ranged from 94.13–99.69% in the LOX drug substance and 93.09–98.89% in the tablet formulation. For N-NDLOX, the recoveries ranged from 95.38–99.63% and 94.27–99.50% in the drug substance and tablet formulation, respectively, demonstrating satisfactory method accuracy (Table 5).

Solution stability

Solution stability of N-NMP and N-NDLOX was evaluated by spiking LOX drug substance and formulated samples with 1.0 ppm LOX and storing them under ambient conditions. Samples were analysed at intervals from 0 to 24 h. No significant changes in peak area (< 10%) were observed compared with freshly prepared solutions, demonstrating that both nitrosamine impurities are stable in solution for at least 24 h under the tested conditions.

Real-time analysis of commercial LOX product

Each detected nitrosamine impurity in the drug substance or drug product must be quantified relative to its respective AI. When multiple nitrosamines are present, their combined exposure must not exceed 100% of the permitted limit. To demonstrate the method's applicability, three marketed LOX products in 500 mg tablet formulations were analyzed using the validated LC–TQ–MS/MS method. Triplicate analyses revealed no detectable LOX-related NDSRIs, confirming compliance with regulatory requirements.

The analytical performance of the developed method was critically evaluated in the context of reported LC–MS-based nitrosamine determination strategies. Most existing methods are primarily designed for low-molecular-weight, volatile nitrosamines and often rely on extensive sample preparation or derivatization to achieve acceptable sensitivity. In contrast, the present method enables direct quantification of structurally complex NDSRIs without derivatization, while achieving limits of quantification in the low ng/mL range. Furthermore, unlike conventional methods that prioritize sensitivity alone, the current approach integrates multi-metric sustainability assessment, thereby providing a more comprehensive analytical framework. This distinction highlights the proposed method's ability to address both regulatory sensitivity requirements and emerging sustainability expectations, which are rarely considered simultaneously in reported methodologies.

Comprehensive sustainability and performance assessment

The sustainability profile of the developed LC–MS/MS method was evaluated using an integrated, multi-metric framework that captured not only environmental impact but also analytical performance and real-world applicability. In contrast to approaches that equate greenness solely with solvent substitution or instrument simplification, the present work adopts a function-oriented perspective, in which sustainability is understood as achieving regulatory-level sensitivity with minimal procedural burden. The combined outputs of AGREE, GAPI, RAPI, and EPPI are presented in Fig. 6, providing a unified visualization of this balance [46–50].

Fig. 6.

Integrated Multi-Metric Evaluation of Analytical Sustainability: a AGREE, b GAPI, c RAPI, and d EPPI Assessment of the Developed LC–MS/MS Method

From a design perspective, the method incorporates several intrinsically green elements that are often overlooked in high-sensitivity LC–MS/MS workflows. These include the absence of derivatization, elimination of complex extraction or clean-up procedures, and the use of a single-step dilution–filtration approach for sample preparation. The analytical workflow is therefore streamlined, reducing chemical handling, minimizing potential error propagation, and limiting resource consumption associated with multi-step processing. In addition, the method enables simultaneous determination of multiple analytes within a single run, thereby improving analytical efficiency per sample and reducing cumulative solvent and energy usage across large analytical batches. These features collectively contribute to what may be described as functional greenness, in which sustainability is achieved through procedural simplicity and analytical efficiency rather than through isolated material substitutions.

The AGREE assessment yielded a score of 0.62, reflecting a moderate but realistic level of greenness. This score is primarily influenced by the use of organic solvents and the energy requirements of LC–MS/MS instrumentation, both of which are intrinsic to achieving trace-level detection of nitrosamine impurities. Importantly, the assignment of individual AGREE criteria was based on experimentally verified conditions. The classification of the method as an off-line analysis arises from the necessity of sample preparation before injection. At the same time, favorable scores in derivatization and procedural steps reflect the inherent simplicity of the method design. Solvent-related penalties were retained despite the widespread analytical acceptance of acetonitrile and methanol, ensuring that the score reflects environmental reality rather than perceived acceptability. Waste generation was calculated per run, providing a more accurate representation of process intensity. Thus, the AGREE score captures a balanced scenario in which environmentally favorable design elements coexist with unavoidable instrumental constraints.

The GAPI assessment provides further insight by resolving the analytical workflow into discrete stages. The resulting pictogram demonstrates that the environmental impact is not uniformly distributed but is primarily localized in solvent use and instrumentation. In contrast, sample preparation, which often represents a major source of environmental burden in analytical chemistry, is characterized by minimal intervention, absence of auxiliary reagents, and reduced operational complexity. The classification of the method within the simple procedural domain reflects the absence of labor-intensive or resource-intensive steps such as extraction, derivatization, or purification. Similarly, although solvent use is not green, it is limited in volume per analysis, and waste generation remains within a controlled range. This distribution highlights a critical aspect of the method: environmental impact is driven by analytical necessity rather than inefficient design.

While AGREE and GAPI focus on environmental aspects, the RAPI evaluation confirms that these considerations have not compromised analytical performance. The method achieved an RAPI score of 75–78, indicative of high analytical reliability. This performance is supported by experimentally validated parameters, including low limits of quantification, excellent linearity, acceptable precision, and high recovery. The evaluation was conducted conservatively, with reproducibility not assumed in the absence of inter-laboratory validation and trueness derived from recovery studies rather than external reference materials. The resulting score, therefore, reflects demonstrated performance rather than inferred capability. The high selectivity afforded by LC–MS/MS in MRM mode further reinforces the method’s suitability for trace-level impurity determination in complex matrices.

The EPPI assessment was conducted according to the previously reported framework. EPPI provides a comprehensive evaluation of analytical methods by integrating two equally weighted sub-indices: the Environmental Impact Index (EI) and the Performance and Practicality Index (PPI), each contributing 50% to the final EPPI score. The EI index evaluates the environmental sustainability of the analytical procedure based on four key components, namely sample preparation, instrumentation, reagent consumption, and waste generation. The PPI index assesses analytical performance and practical applicability through predefined criteria including method type, validation status, sensitivity, reagent availability, instrument accessibility, maintenance requirements, throughput, and operational feasibility. The final EPPI score was calculated by combining the EI and PPI scores according to the published scoring algorithm. This balanced approach enables simultaneous assessment of environmental impact, analytical performance, and practical applicability, consistent with the principles of White Analytical Chemistry. Detailed scoring matrices and criterion-specific weighting procedures have been previously described in the original EPPI publication and are therefore not reproduced here.

To integrate environmental and operational considerations, EPPI was applied as a composite metric combining the EI and the PPI. The EI score of 75.8% reflects moderate environmental impact, primarily associated with solvent hazard profiles and energy consumption. However, this is counterbalanced by a PPI score of 87.0%, which captures the strong practical advantages of the method. These include minimal sample preparation, absence of derivatization, high throughput, and multianalyte capability. The resulting EPPI score of 81.4% demonstrates that the method achieves a favorable balance between environmental responsibility and analytical efficiency, consistent with the principles of White Analytical Chemistry.

The deliberate selection of AGREE, GAPI, RAPI, and EPPI reflects the need to overcome the limitations of single-metric evaluation. AGREE provides a standardized quantitative measure of greenness but does not address process-level contributions, unlike GAPI. RAPI introduces a necessary performance dimension, ensuring that sustainability does not come at the expense of analytical quality. EPPI integrates these perspectives, enabling a holistic interpretation of the method’s overall value. This layered assessment strategy allows the method to be evaluated not as an idealized green system, but as a realistic, performance-driven analytical platform in which sustainability is achieved through optimization rather than simplification.

The developed LC–MS/MS method cannot be classified as purely green in the conventional sense; however, it represents a sustainability-balanced approach that controls environmental impact without compromising analytical performance or practicality. The integrated assessment presented in Fig. 6 demonstrates that the method aligns with contemporary expectations of analytical science, particularly in regulatory contexts where sensitivity and reliability are paramount. This work, therefore, contributes to a more nuanced understanding of analytical sustainability, emphasizing that meaningful greenness arises not from isolated parameter optimization but from the rational integration of performance, practicality, and environmental considerations.

Comparative performance and sustainability analysis

The analytical performance of the developed LC–TQ–MS/MS method was critically evaluated against reported methodologies for nitrosamine determination. Existing analytical approaches predominantly focus on low-molecular-weight, volatile nitrosamines and often require extensive sample preparation, derivatization, or enrichment strategies to achieve acceptable sensitivity. Such approaches are not directly applicable to structurally complex NDSRIs, which exhibit distinct physicochemical properties and pose analytical challenges.

In contrast, the present method enables the direct quantification of LOX-associated NDSRIs without the need for derivatization or complex extraction procedures. The achieved limits of quantification in the low ng/ml range demonstrate sensitivity comparable to, or exceeding, that of reported LC–MS-based methods, while maintaining a simplified analytical workflow. This balance between sensitivity and procedural efficiency represents a key advancement over conventional nitrosamine analysis strategies.

Furthermore, most reported methods emphasize analytical sensitivity as the primary performance criterion, with limited consideration of environmental impact or operational practicality. The combined use of AGREE, GAPI, RAPI, and EPPI tools enables simultaneous evaluation of environmental impact, analytical performance, and method practicality, which is rarely achieved in reported methodologies.

The sustainability profile of the developed method indicates a moderate environmental impact, primarily associated with solvent use and energy-intensive LC–MS/MS instrumentation. However, these limitations are effectively balanced by reduced sample preparation complexity, absence of derivatization, multianalyte capability, and high throughput. This results in a method that aligns with the principles of White Analytical Chemistry, where sustainability is achieved through optimization of the overall analytical system rather than isolated parameter modification.

Importantly, integrating (Q)-SAR-based mutagenicity prediction, CPCA, and analytical quantification provides a unified, risk-informed framework that extends beyond conventional method development. This approach aligns toxicological relevance with analytical sensitivity, ensuring that method performance is directly linked to regulatory requirements for impurity control.

Overall, the proposed method distinguishes itself from existing analytical strategies by combining high sensitivity, simplified workflow, and integrated sustainability evaluation within a single framework. This comprehensive approach not only addresses current regulatory expectations but also provides a scalable model for developing future analytical methods targeting structurally complex nitrosamine impurities.

Conclusion

A highly sensitive and selective LC–TQ–MS/MS method was successfully developed and validated for the trace-level quantification of levofloxacin-associated NDSRIs, namely N-NMP and N-NDLOX, in both bulk drug substance and pharmaceutical formulations. The method demonstrated excellent analytical performance, including low limits of quantification at the nanogram-per-milliliter level, high accuracy, and robust precision, confirming its suitability for routine quality control and regulatory compliance. Beyond conventional analytical method development, this study establishes a unified framework integrating toxicological risk assessment, and multi-metric sustainability evaluation. The incorporation of (Q)-SAR prediction and CPCA enables scientifically justified control strategies for NDSRIs, addressing a critical gap in current regulatory practice for structurally complex nitrosamine impurities. This integrated approach ensures that analytical sensitivity is aligned with toxicological relevance, thereby enhancing the reliability of impurity risk management. The sustainability assessment performed using AGREE, GAPI, RAPI, and EPPI tools revealed that the developed method achieves a balanced analytical profile, characterized by moderate environmental impact and high practical applicability. While the use of organic solvents and LC–MS/MS instrumentation imposes inherent environmental constraints, these are offset by streamlined sample preparation, absence of derivatization, multianalyte capability, and high analytical efficiency. The resulting EPPI score further confirms that the method aligns with the principles of White Analytical Chemistry, where environmental considerations are integrated with performance and operational feasibility rather than treated in isolation. Importantly, this work demonstrates that sustainability in analytical chemistry should not be interpreted as the elimination of instrumental or solvent-intensive techniques, but rather as the rational optimization of analytical systems to achieve maximal efficiency with controlled environmental impact. The proposed methodology, therefore, represents a shift from conventional greenness-focused evaluation to a more comprehensive, performance-driven sustainability paradigm. Despite the promising analytical performance and comprehensive toxicological and sustainability assessment demonstrated in this study, certain limitations should be acknowledged. The developed framework was evaluated using two levofloxacin-associated NDSRIs, namely N-NMP and N-NDLOX, and therefore its applicability to other NDSRIs requires further investigation. In addition, the method was validated within a single laboratory setting, and inter-laboratory studies were not conducted. Future studies involving a broader range of pharmaceutical NDSRIs and multi-laboratory evaluations would further strengthen the regulatory applicability and generalizability of the proposed approach. Overall, the developed LC–MS/MS method provides a regulatory-relevant, scalable, and sustainability-aware analytical strategy for controlling NDSRIs in pharmaceutical systems. The integrated framework presented in this study can be extended to other drug substances and impurity classes, providing a robust template for future analytical method development in response to evolving regulatory and environmental expectations.

Abbreviations

NDSRIs

Nitrosamine drug substance–related impurities

API

Active pharmaceutical ingredient

FDA

Food and Drug Administration

EMA

European Medicines Agency

LOX

Levofloxacin

N-NDLOX

N-nitroso desmethyl levofloxacin

N-NMP

N-nitroso methyl piperazine

DLOX

Desmethyl levofloxacin

NMP

N-methyl piperazine

GAC

Green analytical chemistry

CPCA

Carcinogenic potency categorization

AGREE

Analytical GREEnness

GAPI

Green analytical procedure index

RAPI

Red analytical performance index

EPPI

Environmental performance and practicality index

MRM

Multiple reaction monitoring

NMR

Nuclear magnetic resonance

MDD

Maximum daily dose

ATP

Analytical target profile

SST

System suitability test

CVs

Coefficients of variation

S/N

Signal-to-noise

DL

Detection limit

QL

Quantitation limit

EI

Environmental impact index

PPI

Performance and practicality index

(Q)-SAR

Quantitative structure–activity relationship

Author contributions

Srinivas Nakka, Vishnu Murthy Marisetti, Surendra Babu Manabolu Surya, and Hemanth Kumar Chanduluru* contributed to this work. Srinivas Nakka: Writing—original draft, Resources, Investigation, Formal analysis. Vishnu Murthy Marisetti: Software, Methodology, Formal analysis, Data curation. Surendra Babu Manabolu Surya: Review & editing, Validation, Resources, Investigation, Formal analysis, Conceptualization. Hemanth Kumar Chanduluru: Writing—original draft, Software, Formal analysis, Data curation, Greenness Assessments.

Funding

Open access funding provided by SRM Institute of Science and Technology for SRMIST – Medical & Health Sciences.

Data availability

Data Availability Statement All data generated or analyzed during this study are included in this published article and its Supplementary Information files. Additional datasets generated and/or analyzed during the current study are available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

Ethics approval was not required for this study, as it did not involve human participants or animals.

Consent for publication

Not applicable.

Competing of interests

The authors declare no competing interests

Declaration of generative AI and AI-assisted technologies in the writing process

During the preparation of this work, the author(s) used ChatGPT to check for grammatical corrections. After using this tool/service, the author(s) reviewed and edited the content as needed and take full responsibility for the content of the publication.