Radiopharmaceutical discovery naturally rewards peak performance. Researchers look for compounds with high target affinity, favorable biodistribution, strong tumor-to-background ratios and radiolabeling methods capable of producing high radiochemical yields. Those characteristics determine which candidates look most promising in preclinical studies and which molecules eventually advance toward clinical development.
But there is another test that usually comes later, and it can be just as unforgiving: can the radiopharmaceutical actually be manufactured reliably?
A recent position paper in EJNMMI Radiopharmacy and Chemistry, From Peak Performance to Practical Robustness in Radiopharmaceutical Design, argues that radiopharmaceutical development has historically placed too much emphasis on achieving the highest possible radiochemical yield under optimized laboratory conditions and not enough on whether that chemistry remains reliable when conditions inevitably vary. The authors contend that robustness should be considered from the earliest stages of molecular and radiochemical design rather than treated as an engineering problem to solve after a drug candidate has already been selected.
A radiolabeling reaction can perform exceptionally well when every parameter is tightly controlled. Temperature, reaction time, precursor concentration, pH, trace-metal contamination, solvent composition and starting activity can all influence whether a synthesis succeeds, and an academic laboratory can often optimize those variables until a reaction produces an impressive yield.
Commercial manufacturing does not operate inside a single ideal experiment. A process must perform reliably across multiple batches, potentially on different synthesis modules, with different reagent lots, at different activity levels and eventually at different manufacturing sites. Small changes that have little consequence in conventional pharmaceutical production can become significant in radiochemistry because the product is being manufactured under time pressure while the radionuclide is continuously decaying.
The new position paper argues that this creates a distinction between peak radiochemical performance and radiochemical robustness. A reaction that achieves a very high yield only within a narrow operating window may ultimately be less valuable than one that produces a slightly lower yield but remains consistent when process parameters vary within realistic manufacturing ranges.
That distinction could become increasingly important as radiopharmaceutical development moves from small academic programs into multicenter trials and commercial-scale production.
Automated synthesis modules have become essential to modern radiopharmaceutical manufacturing because they reduce radiation exposure to personnel, improve reproducibility and create standardized workflows that can be validated for clinical use. But automation also removes some of the flexibility available to an experienced radiochemist working manually at the bench.
A chemist can sometimes observe a reaction, adjust conditions or compensate for small deviations during manual synthesis. An automated system executes a predefined sequence. Fluid transfers occur through fixed tubing and cartridges, reaction vessels have particular heating characteristics, and cassette design can constrain volumes, mixing and purification strategies.
Chemistry that depends on extremely precise conditions can therefore become difficult to transfer between systems. According to the EJNMMI position paper, radiolabeling reactions may be particularly vulnerable to kinetic limitations, trace-metal interference and radiolysis at clinically relevant activity levels. A synthesis that works consistently at small research-scale activities may behave differently when significantly more radioactivity is introduced for clinical or commercial production.
That is not a theoretical concern. A 2026 head-to-head study of [F-18]PSMA-1007 production on two automated synthesis platforms found that both systems produced material meeting pharmacopoeial quality requirements, but radiochemical yields differed substantially. FASTlab2 produced an average decay-corrected yield of 43.7%, compared with 28.5% using the NEPTIS Perform platform, despite both manufacturing the same radiopharmaceutical. Read the automated synthesis comparison.
The drug molecule did not change. The manufacturing environment did.
Increasing production activity introduces another challenge. Radiopharmaceutical developers eventually need to manufacture batches large enough to support clinical trials, multiple patients or centralized distribution, but radioactive concentration itself can alter the stability of a formulation.
Radiolysis becomes particularly important at high activity levels. Radiation emitted by the radionuclide can generate reactive chemical species capable of degrading the radiopharmaceutical being manufactured. A synthesis that appears stable at research scale can therefore become less predictable when production moves toward commercial batch sizes.
A recent example comes from cGMP-Compliant High-Yield Automated Production of F-18 AlF-FAPI-74, where researchers optimized an automated production process capable of producing an average 36 GBq of F-18 AlF-FAPI-74, with batches reaching 42.5 GBq. The validated process achieved an average decay-corrected radiochemical yield of approximately 50%, radiochemical purity above 95% and stability for at least 10 hours after synthesis.
The significance is not simply that researchers produced FAPI-74 successfully. Earlier automated approaches had reported lower yields and batch activities that the authors considered insufficient for centralized production and distribution. The newer process was deliberately engineered around automation, quality control, radioprotection and higher-activity manufacturing.
That is the difference between demonstrating that a molecule can be radiolabeled and demonstrating that the molecule can potentially become a product.
This leads to a potentially important shift in radiopharmaceutical development strategy. Candidate selection has traditionally been dominated by biological performance: which molecule binds best, clears fastest from healthy tissue, produces the highest tumor uptake or delivers the most favorable dosimetry.
Those criteria will remain essential, but the authors of the new position paper argue that radiochemical robustness should become a co-equal criterion during initial probe selection. If two candidate molecules demonstrate similar biological performance but one can be labeled reliably across a broad range of conditions while the other requires an extremely narrow operating window, the more robust compound may ultimately have a much greater chance of reaching patients.
That approach resembles Quality by Design principles already used throughout pharmaceutical manufacturing. Instead of defining one perfect synthesis condition and attempting to reproduce it forever, developers characterize a broader design space within which the process consistently produces material meeting predefined quality standards.
For radiopharmaceuticals, that could mean deliberately testing variations in temperature, precursor concentration, activity, reagent composition and other critical process parameters early in development. The objective would not simply be to discover the single condition that produces the highest radiochemical yield, but to understand how much variation the chemistry can tolerate before product quality begins to deteriorate.
A synthesis method can be highly successful at the institution where it was invented and still become difficult to reproduce elsewhere. That becomes a serious problem when a radiopharmaceutical enters a multicenter clinical trial and multiple manufacturing sites need to produce comparable doses for patients.
Different centers may use different automated modules, cyclotrons, hot cells, analytical equipment and local workflows. Even when protocols are standardized, small operational differences are inevitable. A fragile synthesis can therefore create batch failures, production delays and site-to-site variability that have little to do with the biological performance of the drug itself.
Those failures have commercial consequences. Every failed batch consumes expensive isotope, precursor, synthesis-module time and personnel resources while potentially causing a patient treatment or imaging appointment to be delayed or canceled. For short-lived radionuclides, there may be no opportunity simply to remake the batch later in the day.
Robust chemistry therefore becomes part of patient access. A radiopharmaceutical that works beautifully but cannot be produced consistently across the network needed to distribute it has limited clinical value, regardless of how impressive its preclinical data may be.
Radiopharma is increasingly becoming a manufacturing industry as much as a drug-development industry. Companies are investing heavily in cyclotrons, isotope production, hot cells, automated synthesis systems, fill-finish capacity, quality-control laboratories and distribution networks because successful radiopharmaceuticals have to move from chemistry bench to patient within extraordinarily constrained timelines.
That means manufacturability may eventually influence competition between products in ways that are difficult to see during early development. Two radiopharmaceuticals targeting the same biology could demonstrate similar clinical performance, while one proves significantly easier to manufacture, produces fewer failed batches, supports larger batch sizes or can be transferred more readily across production sites.
In that situation, the operationally robust molecule could become the better commercial drug even if another candidate originally produced more impressive chemistry data under optimized laboratory conditions.
The radiopharmaceutical industry has spent years learning how to optimize molecules for targets, biodistribution and dosimetry. The next stage of maturation may require developers to ask another question much earlier in the process: can we reliably make this drug everywhere it eventually needs to be made?
Because in radiopharma, a molecule that works only under perfect laboratory conditions may not really be a successful drug candidate at all.
