The Digital Divide Between Drug Substance and Drug Product Manufacturing

Key Takeaways
Automation within a unit operation does not guarantee reliable data continuity between drug substance, drug product, and quality systems.
Upstream, purification, and fill–finish operations have different process analytical technology priorities and data-context requirements.
Named manufacturer case studies report reductions in documentation and review burdens, but do not establish an industry-wide DS–DP return on investment.
At the DS–DP interface, material readiness, analytical results, disposition status, and process histories must support coordinated decisions.
Integration investments should be prioritized by operational friction, quality criticality, implementation feasibility, and measurable outcomes.

From Automated Production to Connected Manufacturing

A biologic drug substance batch may complete production in a facility equipped with automated bioreactors, sophisticated purification systems, and extensive process monitoring. However, before that material can move into drug product manufacturing, analytical results must be evaluated, material status confirmed, manufacturing schedules coordinated, and the records needed for subsequent operations made available. When these activities depend on separate systems and manual communication, an otherwise highly automated process can encounter delays at the boundaries between operations.

This disconnect illustrates a challenge in pharmaceutical digitalization. Equipment controls do not necessarily translate into continuous information flows across production stages. BioPhorum identifies inadequate integration among manufacturing execution systems (MES), enterprise resource planning (ERP), laboratory information management systems (LIMS), and quality management systems (QMS) as a contributor to manual data entry, deviations, longer release times, and costs.¹

The distinction is particularly important in drug substance (DS) and drug product (DP) manufacturing, where upstream bioprocessing, downstream purification, formulation, and fill–finish operations generate different types of data and rely on different control strategies. Even when individual production systems are highly automated, information may remain difficult to exchange across manufacturing stages. Data captured during DS production, for example, may provide valuable context for subsequent DP operations, but realizing that value requires systems that can preserve the information, connect it to the relevant materials and batches, and make it accessible when decisions are made.

Different Processes, Different Digital Requirements

Biologics manufacturing comprises operations whose control objectives change as material progresses toward the finished product. Upstream processing tracks cell culture conditions and process trajectories. Downstream purification adds chromatography, filtration, and separation measurements, along with data on material characteristics and purification performance.

These differences shape process analytical technology (PAT) priorities. In an industry assessment published in Biotechnology and Bioengineering, experts from Merck, Bristol Myers Squibb, GSK, Pfizer, Boehringer Ingelheim, and BioPhorum surveyed technical and business value across biologics operations.² They identified on-line liquid chromatography, in-line Raman spectroscopy, and gas analysis as high-value approaches for production bioreactors. For Protein A purification and polishing, they highlighted variable path-length ultraviolet–visible measurements, on-line chromatography, multiangle light scattering, and automated sampling.

Downstream measurements address product-related characteristics, including aggregates and charge variants, and process-related impurities, including host cell proteins and residual nucleic acids. No single analytical method can comprehensively measure every critical quality attribute, making complementary approaches important.³

Preserving relationships across successive operations adds another requirement. Investigating whether a shift in cell culture conditions affected purification performance may demand correlation between bioreactor time-series records and later chromatography or filtration measurements, with the material and processing context intact.

DP production brings its own monitoring requirements. Formulation, sterile filtration, filling, and lyophilization may generate pressure, temperature, and chamber-condition records that must be tied to the correct lot and evaluated with quality results. In 2020, a collaboration among Exputec, the Vienna University of Technology, and Takeda studied a multivariate monitoring workflow covering 58 formulation, fill, and finish manufacturing lots.⁴ The researchers encountered equipment data that were not directly associated with manufacturing lots. They aligned signals, identified process phases, and extracted features into a shared statistical structure. The analysis distinguished patterns among lots and flagged atypical process behavior without establishing a broadly validated abnormal-batch detection system. The study demonstrates additional analytical value from contextualizing measurements already being collected.

Where Manufacturing Information Loses Continuity

Information exchange becomes more complicated when data must pass from equipment controls and process historians into manufacturing, laboratory, and quality platforms. The relevant records must be associated with the correct material, batch, process step, and decision, not simply transmitted as files.

BioPhorum’s 2024 MES integration paper identifies a lack of standardized processes as an obstacle to scalability and describes manual entry, deviations, and longer release times associated with poorly integrated systems.¹

Certificates of analysis (CoAs) illustrate the challenge beyond one facility. BioPhorum characterizes digital CoA exchange as generally immature and identifies 11 use cases spanning raw materials through finished products. Sponsors and contract development and manufacturing organizations (CDMOs) have developed local exchange standards, reinforcing the need for consistent data models.⁵

The European Medicines Agency (EMA) emphasizes the retention of original electronic records, metadata, and audit trails so that manufacturing and analytical activities can be reconstructed. Its guidance also calls for review of relevant electronic data before decisions such as batch release.⁶ Effective connectivity must therefore preserve provenance and reviewability alongside availability.

Connecting Manufacturing and Quality Delivers Measurable Benefits

Published manufacturer implementations suggest that integration can improve efficiency even when the physical production process remains substantially unchanged. In a GSK pilot involving pharmaceutical batch documentation and existing automation software, AspenTech reports a 95% reduction in order preparation time and a reduction of more than 50% in record review time.⁷ The finding illustrates how digitalizing manufacturing information can improve its accessibility and reduce the manual effort required to manage production records. Similar challenges arise when information must move between manufacturing stages, particularly when different systems capture, organize, and retain process data.

At Cipla’s Patalganga facility, electronic batch recording integrated MES with ERP, supervisory control and data acquisition (SCADA), building management, in-process quality control, and warehouse management systems. Körber, the relevant technology provider, reported approximately 75% fewer manual data entries and faster review-by-exception workflows.⁸

Körber also reported 60% faster batch record reviews and 25–40% less production documentation time in a Ferring Pharmaceuticals implementation.⁹ At IDT Biologika, a Körber case study describes reduced manual effort and simplified customer reviews of completion certificates instead of full paper batch records.¹⁰

These are vendor-reported outcomes at individual implementations, not industry benchmarks. They document distinct benefits in data entry, record preparation, and review, but a shorter batch-record review does not necessarily mean an equivalent reduction in overall batch-release time.

The DS–DP Interface: Connecting Materials, Testing, and Manufacturing Decisions

The DS-to-DP transition requires continuity in three categories: material and production status, analytical and quality decisions, and process history. DP teams need to know when incoming material will be ready; quality personnel need results and disposition status; and prior processing records may be needed during investigations, technology transfer, or evaluation of unexpected characteristics.

Consider a DS batch awaiting completion of testing while a DP site prepares for filling. Planning depends on anticipated availability, while quality personnel must determine whether the applicable requirements for use have been satisfied. If the records sit in separate systems, personnel may need to assemble them manually. A connected workflow could expose authorized users to the current material status, relevant results, and approval record without displacing required quality decisions.

Samsung Biologics describes a time-sensitive transfer in which DS and DP activities operated under one quality system and conditional release allowed DP filling to begin while some DS testing continued.¹¹ The manufacturer states that specified basic tests, including endotoxin and bioburden, were performed and that overlapping operations saved weeks; the program delivered 40 batches and 10 million vials on schedule.

The company attributes this outcome to co-located operations, quality coordination, and client collaboration, not a specific digital platform. The example nevertheless highlights information dependencies that connected systems could support: material readiness, testing progress, permitted use, and production scheduling.

Across different sites and CDMOs, software and record structures may vary. Structured exchange of CoAs and related quality information could help receiving sites access relevant records without rebuilding material histories manually. Direct quantitative evidence of the manufacturing benefit of such end-to-end DS–DP integration, however, remains limited.⁵

From Connected Data to Advanced Process Understanding

Advanced analytics across operations require an architecture that integrates diverse instruments, established equipment, and modeling systems. In the International Journal of Pharmaceutics, researchers from the University of Delaware, Rutgers University, and Purdue University demonstrated a cloud-connected data management framework at two academic continuous DP manufacturing pilot plants.¹² The framework integrated PAT tools, legacy equipment, and simulation models while considering cybersecurity.

The results show a practical architecture for connecting heterogeneous data sources and preparing for modeling and digital twins. They do not demonstrate commercial biologics DS–DP integration or a quantified manufacturing improvement. Predictive applications require suitable data sets, defined process questions, and evaluation of model performance beyond connectivity alone.

The FDA’s PAT framework encourages innovative process monitoring and control within appropriate manufacturing and quality safeguards.¹³ As models gain influence over operational decisions, governance must address access, validation, changes to analytical methods, cybersecurity, and the traceability of underlying records.

Prioritizing Integration Where It Creates Operational Value

The first integration investment should address a specific manufacturing constraint rather than the appeal of a new platform. Teams can identify where operators duplicate data entry, reconcile records, wait for results, or lack the status needed to make a timely decision, then determine the quality risks and production consequences.

For facilities burdened by documentation, MES-to-equipment and quality-review integration may offer the clearest starting point. For multi-site networks, material availability, testing status, or production readiness may matter more. Useful measures include record review time, manual transfers, exception frequency, schedule disruptions, and the interval from DS disposition to DP readiness.

Information criticality and integration feasibility should temper that prioritization. Product-disposition interfaces demand rigorous permissions and data integrity, while legacy systems may require mapping or revised workflows. A project should be evaluated against its original operational objective. Fewer manual entries may save documentation effort without adding capacity; faster visibility into a laboratory result may aid scheduling even when release remains limited by required testing. Baseline measurements help distinguish intermediate workflow gains from changes to the complete manufacturing timeline.

Building Continuity Across the Manufacturing Life Cycle

Electronic batch recording, integrated manufacturing systems, and advanced process monitoring show that existing manufacturing data can be put to more effective use. Extending those capabilities across DS and DP operations could improve material transfers, testing coordination, planning, and quality decisions, although direct end-to-end performance evidence remains comparatively thin.

As manufacturers seek higher efficiency across drug substance and drug product production, the ability to coordinate material availability, analytical testing, quality decisions, and manufacturing schedules may become as important as the sophistication of the equipment performing each operation.