Electron Beam Terminal Sterilization of a Thermolabile Retatrutide Injectable in a Multidose Cartridge

1. Introduction

Terminal sterilization provides a strong sterility assurance concept because the drug product is sterilized in its final container. Regulatory guidance therefore expects terminal processing whenever it is feasible, while any alternative must be justified through product and process evidence. For peptide injectables, the preferred method cannot be selected from modality labels alone: heat sensitivity, radiolysis, oxidation, aggregation, container closure integrity, dose penetration, and microbiological resistance must be considered together.

Retatrutide is a synthetic peptide agonist under clinical development and is not an approved active ingredient for the cartridge formulation evaluated here. The development candidate comprises retatrutide in an aqueous tromethamine mannitol phenol vehicle. Aqueous peptide products are especially vulnerable to ionizing radiation because water radiolysis generates reactive species that can induce oxidation, fragmentation, crosslinking, or aggregation. Evidence from other peptides, including insulin, indicates that acceptable recovery after irradiation cannot be assumed from low process temperature or from chromatographic main peak area alone.

This Original Article reframes formulation development around the terminal sterilization decision and addresses modality selection, electron beam process definition, microbiological assurance modelling, and analytical control. All batch, dose, sterility, and chromatographic results are synthetic. They were created to test a decision framework and must not be represented as measurements.

2. Materials and Methods

2.1. Formulation and Container System

The model product was a 3 mL multidose cartridge containing the candidate composition shown in Table 1. The formulation is a development hypothesis and not a clinical, commercial, or validated product. The primary package was represented as a glass cartridge with elastomeric closure components. Material grades, siliconization, extractables and leachables, headspace oxygen, and closure dimensions require product specific definition before irradiation studies.

2.2. Selection of the Sterilization Strategy

Five strategies were evaluated against peptide stability, applicability to a sealed liquid cartridge, residual risk, scalability, and regulatory burden. The decision was framed as selection of a confirmatory development pathway, not approval of a manufacturing process.

Table 2. Product specific comparison of sterilization strategies.

Method	Thermal burden	Final container fit	Dominant product risk	Decision
Moist heat	High	Excellent when compatible	High risk of peptide degradation and aggregation	Not selected for the aqueous peptide
Dry heat	Very high	Not suitable for aqueous filled cartridges	Product and closure incompatibility	Not selected
Ethylene oxide	Low	Poor fit for a sealed aqueous injectable	Residuals, diffusion, aeration, material effects	Not selected
Sterile filtration and aseptic filling	Low	Not terminal	Strong baseline option but relies on aseptic control	Fallback and comparator
Electron beam	Low average thermal load	Potential terminal application	Aqueous radiolysis and dose nonuniformity	Selected for experimental feasibility testing

Table 2. Product specific comparison of sterilization strategies.

Method	Thermal burden	Final container fit	Dominant product risk	Decision
Moist heat	High	Excellent when compatible	High risk of peptide degradation and aggregation	Not selected for the aqueous peptide
Dry heat	Very high	Not suitable for aqueous filled cartridges	Product and closure incompatibility	Not selected
Ethylene oxide	Low	Poor fit for a sealed aqueous injectable	Residuals, diffusion, aeration, material effects	Not selected
Sterile filtration and aseptic filling	Low	Not terminal	Strong baseline option but relies on aseptic control	Fallback and comparator
Electron beam	Low average thermal load	Potential terminal application	Aqueous radiolysis and dose nonuniformity	Selected for experimental feasibility testing

2.3. Electron Accelerator Configuration and Dose Concept

The simulated process used a UELR-10-15-S-1 electron accelerator, unit No. 128170401, with the supplied machine settings listed in Table 3. Nominal beam power was calculated as 9.5 MeV multiplied by 1.19 mA, equivalent to approximately 11.3 kW. This value describes electrical beam output and is not an absorbed dose.

A model target window of 15.0 to 18.5 kGy was used solely for feasibility calculations. The dose uniformity ratio was defined as maximum absorbed dose divided by minimum absorbed dose. Real minimum and maximum doses must be established using calibrated dosimeters placed at mapped product locations in a fully specified load. Product density, cartridge orientation, scan overlap, conveyor loading, irradiation from one or more sides, and electron energy losses must be recorded. The machine settings in Table 3 cannot be converted into a defensible product dose without those data.

2.4. Simulated Microbiological Qualification

Three virtual batches were assigned low presterilization bioburden distributions and batch specific minimum doses. The probability of a nonsterile unit was estimated from an exponential survivor approximation using sampled initial contamination, absorbed dose, and D10 values. For each batch, 200000 Monte Carlo iterations used a median D10 of 1.30 kGy with bounded uncertainty from 0.80 to 2.00 kGy. The study criterion was a model probability of a nonsterile unit not greater than 10 to the power of minus 6. Simulated verification dose tests contained 100 units per batch, and simulated final sterility tests contained 20 units per batch. These organism counts and test outcomes are generated values; they do not constitute ISO dose substantiation, pharmacopoeial sterility testing, or validation evidence.

2.5. RP HPLC Method for Retatrutide and Related Substances

A stability indicating RP HPLC diode array detection method was defined as the primary chromatographic control. A wide pore C18 column, 150 by 4.6 mm, 3.5 micrometre particles, and 300 angstrom pores was modelled at 30 °C. Mobile phase A was water with 0.1 percent trifluoroacetic acid and mobile phase B was acetonitrile with 0.1 percent trifluoroacetic acid. The gradient increased from 32 to 62 percent B over 20 min at 0.8 mL/min. Detection was at 214 nm, injection volume was 10 microlitres, and nominal sample concentration was 1.0 mg/mL. Assay was calculated against an external retatrutide reference standard. Related substances were calculated by corrected area normalization using experimentally established relative response factors when available.

The model system suitability criteria were retention time precision not greater than 1.0 percent relative standard deviation, peak area precision not greater than 2.0 percent, tailing factor not greater than 2.0, theoretical plates not less than 10000, and resolution from the nearest specified impurity not less than 1.5. Selectivity must be challenged with heat, oxidation, light, pH, and irradiation stressed samples. The proposed method can support assay and chromatographic purity but cannot by itself establish complete peptide identity, primary sequence preservation, aggregation state, or biological activity.

2.6. Simulated Analytical Validation and Decision Criteria

Validation characteristics were modelled in alignment with ICH Q2(R2). The simulated range was 0.5 to 1.5 mg/mL for assay and 0.05 to 3.0 percent for related substances. Provisional post irradiation decision criteria were retained assay of at least 95.0 percent of label claim, total related substances not greater than 3.0 percent, and no new individual impurity greater than 1.0 percent. These thresholds are study rules rather than release or shelf life specifications.

Table 4. Simulated RP HPLC validation and system suitability outputs.

Characteristic	Simulated result	Model criterion
Assay linearity	R squared 0.9998	R squared at least 0.999
Accuracy	99.2 to 100.6 percent	98.0 to 102.0 percent
Repeatability	0.46 percent RSD	Not greater than 2.0 percent
Intermediate precision	0.78 percent RSD	Not greater than 3.0 percent
Impurity quantitation limit	0.05 percent	Not greater than 0.10 percent
System suitability peak area	0.42 percent RSD	Not greater than 2.0 percent
Tailing factor	1.08	Not greater than 2.0
Theoretical plates	28400	Not less than 10000
Nearest peak resolution	2.35	Not less than 1.5

Table 4. Simulated RP HPLC validation and system suitability outputs.

Characteristic	Simulated result	Model criterion
Assay linearity	R squared 0.9998	R squared at least 0.999
Accuracy	99.2 to 100.6 percent	98.0 to 102.0 percent
Repeatability	0.46 percent RSD	Not greater than 2.0 percent
Intermediate precision	0.78 percent RSD	Not greater than 3.0 percent
Impurity quantitation limit	0.05 percent	Not greater than 0.10 percent
System suitability peak area	0.42 percent RSD	Not greater than 2.0 percent
Tailing factor	1.08	Not greater than 2.0
Theoretical plates	28400	Not less than 10000
Nearest peak resolution	2.35	Not less than 1.5

3. Results

3.1. Simulated Dose Mapping and Sterility Assurance

Table 5. Simulated microbiological and dose outputs for three virtual batches.

Batch	Bioburden CFU per cartridge	Dmin kGy	Dmax kGy	DUR	Positive verification tests	Positive sterility tests	Median model PNS	Upper 95 percent PNS	Iterations meeting PNS criterion percent
EB-R01	1.6	15.2	18.1	1.19	0/100	0/20	3.2e-12	2.1e-09	100.00
EB-R02	2.0	15.4	18.4	1.19	0/100	0/20	2.8e-12	2.0e-09	100.00
EB-R03	1.8	15.1	18.3	1.21	0/100	0/20	4.3e-12	2.6e-09	100.00

Table 5. Simulated microbiological and dose outputs for three virtual batches.

Batch	Bioburden CFU per cartridge	Dmin kGy	Dmax kGy	DUR	Positive verification tests	Positive sterility tests	Median model PNS	Upper 95 percent PNS	Iterations meeting PNS criterion percent
EB-R01	1.6	15.2	18.1	1.19	0/100	0/20	3.2e-12	2.1e-09	100.00
EB-R02	2.0	15.4	18.4	1.19	0/100	0/20	2.8e-12	2.0e-09	100.00
EB-R03	1.8	15.1	18.3	1.21	0/100	0/20	4.3e-12	2.6e-09	100.00

The simulated dose uniformity ratio ranged from 1.19 to 1.21. No positive units were generated in the model verification dose or sterility test datasets. Monte Carlo estimates met the predefined probability criterion in all batches. These favorable outputs reflect model assumptions, including low bioburden and the selected D10 distribution; zero simulated positives do not prove sterility or validate the dose window.

3.2. Simulated Post Irradiation RP HPLC Performance.

Table 6. Simulated RP HPLC results before and after electron beam treatment.

Batch	Assay pre percent	Assay post percent	Main peak pre area percent	Main peak post area percent	Total RS pre percent	Total RS post percent	Largest new impurity percent	Model decision
EB-R01	100.1	97.6	99.34	97.82	0.66	2.18	0.54	Pass
EB-R02	99.8	97.2	99.28	97.61	0.72	2.39	0.65	Pass
EB-R03	100.3	97.8	99.41	97.90	0.59	2.10	0.50	Pass

Table 6. Simulated RP HPLC results before and after electron beam treatment.

Batch	Assay pre percent	Assay post percent	Main peak pre area percent	Main peak post area percent	Total RS pre percent	Total RS post percent	Largest new impurity percent	Model decision
EB-R01	100.1	97.6	99.34	97.82	0.66	2.18	0.54	Pass
EB-R02	99.8	97.2	99.28	97.61	0.72	2.39	0.65	Pass
EB-R03	100.3	97.8	99.41	97.90	0.59	2.10	0.50	Pass

Across the three virtual batches, post irradiation assay was 97.53 ± 0.31 percent and total related substances were 2.22 ± 0.15 percent. Each batch passed the provisional model criteria. The decrease in assay and increase in related substances were intentionally included to represent a measurable radiation burden rather than a no effect assumption.

Figure 1. Simulated absorbed dose and RP HPLC assay for three virtual batches. Values are generated outputs and are not experimental measurements.

Figure 1. Simulated absorbed dose and RP HPLC assay for three virtual batches. Values are generated outputs and are not experimental measurements.

3.3. What HPLC Can and Cannot Confirm

The model RP HPLC method confirmed only that the simulated intact retatrutide peak remained quantitatively predominant and that the generated impurity burden remained within the predefined study rules. In a real study, coeluting oxidation products or sequence variants may remain undetected, and area percent may diverge from peptide mass recovery. A defensible integrity conclusion therefore requires mass balance and orthogonal evidence. At minimum, LC MS or LC MS MS identity, peptide mapping, size exclusion chromatography for high molecular weight species, subvisible particle testing, and a receptor based potency assay should accompany RP HPLC.

4. Discussion

4.1. Rationale for a Nonthermal Terminal Method

The selection exercise rejected heat based terminal methods because the aqueous peptide and phenol containing formulation has no demonstrated thermal tolerance at sterilizing conditions. Electron beam irradiation offers short exposure and low average heat input, but these advantages do not eliminate chemical risk. The dominant product specific hazard is indirect radiation chemistry in water, not bulk temperature. Published insulin work showed severe loss in unprotected aqueous solution at 10 kGy and improved recovery only after formulation protection and cryogenic irradiation. Accordingly, the favorable retatrutide simulation should be treated as a testable hypothesis rather than evidence that the selected dose window is safe.

4.2. Dose Setting and Process Validation

The model window of 15.0 to 18.5 kGy was chosen to create a narrow feasibility scenario below the conventional 25 kGy reference point. It does not establish a sterilization dose. An experimental programme must first characterize bioburden number and resistance, select an applicable substantiation approach, conduct verification dose experiments, and map minimum and maximum absorbed dose throughout the worst case load. Routine processing would require calibrated dosimetry, process interruption studies, load configuration control, and investigation limits. For a drug product, the strategy must also satisfy the applicable pharmacopoeial and regional regulatory expectations; medical device dose methods should not be imported uncritically.

4.3. Product Quality and Container Closure Risk

A phase appropriate irradiation matrix should compare 0, 5, 10, 15, 18.5, and 25 kGy at ambient, refrigerated, and frozen conditions, with air and nitrogen headspace where technically feasible. Immediate and time dependent testing should include assay, related substances, peptide mapping, molecular mass, aggregation, particles, appearance, pH, osmolality, phenol content, biological potency, endotoxin, sterility, preservative effectiveness, and container closure integrity. Glass, elastomer, lubricant, and adhesive changes should be assessed through functionality and extractables and leachables studies. Dose dependent changes should be linked to toxicological qualification of new impurities.

4.4. Decision Gate for Development

Electron beam development should advance only if the experimentally measured minimum dose achieves the microbiological target while the experimentally measured maximum dose remains within a validated product quality envelope. A predefined stop rule is recommended: if retained assay falls below 95 percent, a new impurity exceeds its qualified limit, potency falls outside the development criterion, aggregation increases materially, or container closure integrity fails, the terminal irradiation branch should be discontinued or reformulated. Sterile filtration followed by aseptic filling remains the scientifically credible comparator and fallback strategy.

4.5. Limitations

This study has five principal limitations. First, every dose, microbiological, and chromatographic result is simulated. Second, no organism distribution or product specific D10 value was measured. Third, accelerator settings were not linked to a defined load geometry or physical dosimetry. Fourth, the RP HPLC method was not experimentally validated and cannot alone establish full peptide integrity. Fifth, the formulation and container system are development candidates. The work therefore cannot support a sterility claim, a terminal sterilization validation, a release decision, a shelf life, or administration to humans.

5. Conclusions

This in silico Original Article identifies electron beam irradiation as a plausible candidate for terminal sterilization of a thermolabile retatrutide cartridge when moist heat is not compatible. Three explicitly simulated batches met the model dose, sterility assurance, assay, and related substance criteria. The result is a development decision framework, not experimental confirmation. Product dose must be measured by calibrated dosimetry, microbiological efficacy must be substantiated with real bioburden and verification studies, and peptide integrity must be demonstrated by validated RP HPLC plus orthogonal structural and biological methods. Until those requirements are met, sterile filtration and aseptic filling remain the default control strategy.