Let's talk about your next study New study? Request a quote →
Loading...

Cell Therapy Bioanalytical Services

Molecular and immunology readouts for CAR-T, TCR-T, and NK cell therapy programs, from ex vivo mechanism through clinical biomarker studies

A cell therapy is a living product, which changes the questions a program has to answer. How many engineered cells are present at each timepoint, what state those cells are in, whether they can still kill, whether the construct they carry is intact and expressed, and what the recipient’s immune system has done to them are separate measurements, and the readouts that look interchangeable are not. A transgene copy number establishes that the construct is present but not that a viable cell is doing anything with it, a specific lysis value establishes that killing occurred but not which cells did it or whether they can do it again, and a copy number counts molecules whether or not the construct inside them is complete. The picture a program needs comes from reading them together, and which of them a program runs, on which matrix and at which timepoint, is settled before samples are collected rather than reconstructed afterward.

Accelevir runs the molecular and the cellular readouts of cell therapy bioanalysis in one laboratory. A program can therefore measure the construct and the cells that carry it from a single sample, and ask whether an integrated construct is complete rather than only whether it is present. The same design, controls, and team carry from preclinical studies into clinical studies, without changing laboratories. Assays supporting clinical studies are analytically qualified under our quality management system, with qualification documentation available to support regulatory filings.

How to Measure a Cell Therapy Product and the Response to It

A cell therapy program measures three things: the engineered product, what the product does, and what the host does to it. Construct expression and integrity, phenotype, cytotoxic function, expansion, persistence, and the immune response against the construct fall within those three, and few programs need all of them at every timepoint.

These measurements are also a regulatory expectation. FDA’s guidance on the development of chimeric antigen receptor T cell products notes that these are living drugs capable of proliferation, so conventional pharmacokinetic models do not apply, and recommends a sampling schedule with sufficient timepoints during the expansion phase. The guidance notes that persistence may be monitored by measuring levels of transgene and CAR expression, and recommends developing assays to detect humoral or cellular immune responses against the engineered cells, or both. Which readouts a program needs depends on the construct, the delivery method, and the stage, and our scientists settle that with you before samples are collected.

The Engineered Product: Copy Number, Integrity, Expression, and Phenotype

Copy number, construct integrity, expression, and composition describe the engineered product itself, and copy number is where most programs begin. Integrity asks whether what was counted is complete, expression whether the construct reaches the surface, and phenotype what kind of cells carry it. All four can be measured on the product before dosing, and copy number, expression, and phenotype can also be followed in the recipient afterward, with integrity and sequence available there for lentiviral constructs.

How many copies of the construct are present?

Is the construct intact?

Is the construct expressed, and in which cells?

What is the product composed of?

01

Accelevir measures vector copy number by digital PCR against a host reference gene, reported per cell equivalent. This is the standard measurement and the most readily available one, and it carries furthest, since it describes the product before dosing and the engineered cells that persist after it. One qualification of that single design covers the matrices a study uses.

02

Accelevir measures construct integrity by interrogating a single construct molecule at several positions at once, which separates complete constructs from truncated or rearranged ones. A copy number assay amplifies one target site, so a product carrying truncated or rearranged integrants can return an unremarkable copy number. Reading several positions at once is the design principle behind the HIV-1 IPDA®, which we co-developed with the Siliciano laboratory at Johns Hopkins University. That assay reads a single proviral genome at several conserved sites at once, so an intact provirus can be distinguished from a defective one rather than simply counted.

03

Accelevir measures receptor expression by flow cytometry and transgene transcript by reverse-transcription digital PCR (RT-dPCR). Flow cytometry reports expression per cell with subset identity attached, and depends on a detection reagent for your construct, which you either supply or tell us to source. RT-dPCR reports expression independently of any such reagent, which matters early, before one specific to your binding domain has been made, and it is the molecular readout for an RNA-delivered construct.

04

Accelevir measures product phenotype by multiparameter flow cytometry, and we report lineage, memory and differentiation distribution, and activation and exhaustion markers. Products are heterogeneous as dispensed for dosing, and a composition measured beforehand is what a later shift during the study can be read against. Copy number, integrity, expression, and phenotype are study endpoints, and whether a therapy works is a clinical endpoint, which no molecular or cellular readout answers.

That principle carries to a lentiviral therapeutic construct because deletions, rearrangements, and other structural defects arise during reverse transcription and integration, both of which a lentiviral vector completes using HIV-1 enzymes packaged into the particle. Deletion and rearrangement are the classes that carry over most directly, and the assays we built to detect them in HIV-1 apply to a lentiviral construct. How often such defects arise in an integrated lentiviral product has not been established in the literature, which is a reason to measure integrity rather than assume it. The conserved regulatory elements a lentiviral vector inherits, among them the LTRs, the packaging signal, and RRE, are regions we already target, in designs refined across many HIV-1 studies. We adjust the multiplex design for element spacing, which differs between a lentiviral construct and HIV-1. What the lineage does not supply is the payload: the transgene is the point of the product, so the assay has to confirm it is present, and we design that target for the construct in front of us.

Sequencing answers the same question at higher resolution, because reading a single integrated lentiviral genome end to end confirms the sequence of the construct in the cells that carry it rather than inferring it from the regions an assay targets. Targeted sequencing on DNA or RNA, by long or short read, resolves a defined region in more depth, and the more elements a construct carries, whether a safety switch, an armoring cytokine, or several transgene elements, the more that resolution matters. These are methods we built for HIV-1 and run routinely, so they reach a lentiviral construct as a design adapted to it rather than one built from scratch. What carries from HIV-1 is specific to lentiviral vectors. Recombination and deletion during reverse transcription are documented for gammaretroviral vectors as well, but their frequency in an integrated clinical product has not been characterized, so a gammaretroviral integrity assay is designed from the construct rather than from that inheritance. Multiplex digital PCR design applies to any integrating vector.

What the Engineered Cells Do

Function is measured by putting the product in front of a target cell expressing your antigen. A single co-culture then supports killing, the identity and output of the effectors that produced it, the number of responding cells, and their capacity to divide again, with antigen-negative targets and non-transduced effectors from the same donor as controls.

Do the cells kill, how potently, and how fast?

Which effector cells respond, and what do they produce?

How many target-responding cells are present?

Can responding cells expand?

01

Accelevir measures cytotoxic function, the potency readout for an engineered product, in a flow-based immune cell-mediated killing assay. We report it as specific lysis across a range of effector-to-target ratios and co-culture durations. We label target cells with a tracker dye that distinguishes them from effectors in the acquired sample, and we score death by viability dye uptake and apoptosis markers within the labeled population. Off-target cells can carry a second tracker dye in the same well, so on-target and bystander killing are read together rather than inferred from separate experiments. For engineered products, we can normalize the effector-to-target ratio to construct-positive cells rather than to total cells, and we report the basis with the result, because the two conventions are not interchangeable.

02

Intracellular cytokine staining, run on the effectors from that same co-culture, measures cytokines and cytotoxic effector proteins retained after a secretion block. We assign each profile to CD4+ or CD8+ cells and their memory subsets, and where a detection reagent is available we gate construct-positive cells in the same panel, so function is attributed to the engineered fraction rather than to the well.

03

ELISpot counts cells secreting a given cytokine in response to the target, reported as spot-forming units per million input cells, and FluoroSpot resolves which cells secreted more than one. We run these against target alone, against unmodified mononuclear cells, and against both together, so antigen-driven responses are separated from background activity.

04

T cell proliferation by dye dilution measures division following antigen-specific stimulation, reported by subset and activation state, a different property from secretion during a short stimulation. We can run it on the product’s own engineered cells engaging target-expressing cells, and on recipient cells responding to the construct.

These readouts run with native and engineered effectors, and with both lineages. NK cell killing does not depend on antigen presentation and is governed by the balance of activating and inhibitory signals, so reduced-MHC targets can be used to assess missing-self recognition, and we can measure antibody-dependent killing through CD16 where that mechanism is part of the design.

A single co-culture reports one encounter, so we can build a repeated-stimulation design into a study, exposing the same effectors to fresh targets across successive rounds and reading killing, expansion, and activation and exhaustion markers at each round. A product that kills well on first exposure and loses function by the third is then distinguishable from one that sustains it, although the two look identical in a single-timepoint lysis value.

A co-culture removes trafficking, the microenvironment, and host immune pressure, so what it measures is capacity rather than behavior in the recipient. Cellular kinetics after dosing and, in nonclinical work, distribution by tissue describe what happens once those pressures are back.

Cellular Kinetics After Dosing

Expansion and persistence are two measurements rather than one, and they can disagree. Accelevir quantifies the transgene in blood, bone marrow, or tissue by digital PCR against a host reference gene, reported per cell equivalent. We include the reference gene as a matter of course for cell therapy work, because a result per cell equivalent is more accurate than one normalized to input DNA mass. Flow cytometry enumerates construct-positive cells and reports what state they are in, and the two, run from the same cryopreserved aliquot, describe the same cells. Copy number without viable cells points to persisting genomic material rather than a persisting product, while a flow count at background with copy number persisting usually points to cells that carry the construct without expressing it, through silencing or loss of expression. The other explanation is a sensitivity difference between the readouts, which acquiring more events on the flow panel narrows, and we run the two together and report them against each other for that reason.

We bank timepoints as they are collected, then thaw and run them in one batch rather than assay by assay. A cryopreserved sample is only as informative as the processing behind it, so sample handling is part of the measurement rather than preparation for it. Accelevir isolates and cryopreserves PBMCs and harmonizes isolation across collection sites. We qualify sites, vendors, and operators on post-thaw viability, recovery, and phenotype, measured on the vials those sites produce, which is what makes banked material from several clinical sites comparable.

What the Host Does to the Product

The recipient’s immune system can respond to the construct itself. We read that response with the same assays pointed at a different stimulus, an overlapping peptide pool spanning the construct in place of a target cell. This is immune monitoring of the recipient rather than characterization of the product. The pool covers the non-self binding domain in particular, and restimulating participant mononuclear cells with it reads a cellular response against the construct by ELISpot or intracellular cytokine staining, in the format we already run for antigen-specific responses in infectious disease work. Peptide pools are commercially available or client-provided, and we can qualify custom panels. The same recipient response can also be read by proliferation.

Antibody against the construct is measured for every approved product and has incidence data behind it, while T cell responses against construct-derived epitopes are measured inconsistently and usually as an exploratory endpoint. The cellular side has been argued to carry the clearer mechanistic link to loss of persistence, on a much smaller evidence base, and its weight has been argued to rise wherever the recipient’s immune system is intact: without lymphodepletion, on repeat dosing, and for allogeneic and in vivo products.

We can develop anti-construct binding antibody assays, often called anti-CAR antibody assays, for your program, in a plate-based format against a recombinant extracellular domain or binding protein, or in a cell-based format in which we incubate construct-positive cells with serum and detect bound IgG. The tiered structure follows our anti-drug antibody work: screening against a statistically defined cut point, confirmation of specificity, and titer.

A neutralizing readout answers whether those antibodies block function rather than simply bind, and we can build it on the killing assay, repeating the co-culture in the presence of study serum against matched pre-dose and control serum and reporting inhibition of specific lysis. Because cytolysis is the mechanism for many of these products, measuring blocked cytolysis measures that mechanism rather than a proxy for it. We settle two constraints at scoping. The assay needs a defined effector lot and an established cut point, so that a fall in lysis is attributable to the serum rather than to the effectors. For an autologous product, the effectors in the assay are not the participant’s own cells, so the result describes neutralization of a construct-matched effector rather than of that participant’s own product.

We develop these assays per construct, and the development effort is part of the proposal. The reason to run them is not only safety: when a product stops persisting, a response against the construct is one of the few measurements that offers an explanation, and reading it alongside the kinetics data is what turns a disappearance into a finding.

Accelevir measures cytokines in two compartments, running single-analyte ELISA in our laboratory and sending broader multiplex panels to a specialist laboratory. Which analytes are worth measuring follows the question rather than a standard list, and the compartment matters as much as the panel. Co-culture supernatant reports what the product released and is attributable to the cells in that well, while serum integrates the whole host response, including the inflammatory signature associated with cytokine release syndrome. Blocking secretion is what makes intracellular staining possible, so it reports what a cell retained at one timepoint, while the supernatant reports what the well released across the culture.

Delivery: Lentiviral, Gammaretroviral, and RNA

The delivery method determines which molecular readouts exist at all, and therefore what a sampling schedule can be built on.

01

Integrating vectors

Lentiviral and gammaretroviral constructs integrate, so copy number and persistence can be followed for as long as the cells remain. Where each is used has diverged. Gammaretroviral vectors have been displaced by self-inactivating lentiviral vectors for hematopoietic stem cell targets, while remaining in use for ex vivo transduction of activated T cells, including in approved CAR-T products, and they play no part in in vivo delivery. For lentiviral constructs, integrity and sequence are available as well. Clonal tracking by integration site analysis, which resolves whether a persisting pool is polyclonal or dominated by a few clones, is coming soon here. Genotoxicity and long-term follow-up are a separate program of work on a regulatory schedule, scoped apart from the mechanism and biomarker readouts here.

02

RNA delivery

An RNA-delivered construct does not integrate and expression is transient, so transcript is the molecular readout and there is no integrated DNA to count. The sampling schedule follows expression decay rather than expansion and persistence, and the question becomes how long expression lasts and what happens on redosing. With a short expression window, depletion of the target compartment and its reconstitution over time, measured by flow cytometry, often carries more of the pharmacodynamic question than the molecular readouts do.

03

In vivo delivered programs

Where the construct is delivered to the recipient rather than to cells in culture, the engineered cell is made inside the body. There is no product to characterize before dosing, so transduction level, expression, and composition become measurements in the participant across a time course rather than attributes of a manufactured product. Which cells received the construct also becomes a clinical question rather than a nonclinical one, because the distribution of the vector or payload is itself the open question.

Two delivery routes are in early clinical study, receptor-targeted lentiviral vectors and lipid nanoparticles carrying messenger RNA, and each is read by the molecular readout of the delivery in use: integrated DNA for a vector, transcript for an RNA payload.

Two readouts in our laboratory resolve which cells received the construct. Flow cytometry with a construct detection reagent alongside lineage markers reads transgene expression by population directly, with no separation step. Immunomagnetic isolation of target and non-target fractions, followed by digital PCR on each, gives a molecular measure per population, including the fractions a program most wants to see clear. A bulk measurement on mononuclear cells cannot answer that, because it averages across the populations the question is about. Biodistribution by digital PCR extends that question to tissue in the nonclinical work behind the program. Where a program doses more than once, the response to the delivery vehicle becomes a question of its own.

Discuss your delivery method with a scientist

What Each Cell Therapy Program Asks of the Measurement

What differs between product types is which question carries the decision, and which readouts exist at all given how the construct is delivered.

Program What the readouts have to answer
CAR-T, autologous
Construct expression, integrity, and product phenotype; killing against antigen-positive and antigen-negative targets; expansion and persistence after dosing; the state of the cells that persist; the host response to the construct
CAR-T, allogeneic
The same, and how long a product expected to be cleared actually persists
TCR-T
The same as autologous, with MHC restriction determining what a valid target is, and multimer staining available where the epitope and allele are defined
CAR-NK and NK cell therapies
Killing that does not require antigen presentation, including against reduced-MHC targets, antibody-dependent killing through CD16, and persistence of a product expected to be transient
Engineered TIL and other engineered immune effectors
Product characterization, antigen-specific function against autologous or matched targets, and persistence in blood and in tumor tissue
CAR-macrophage
Construct expression and phenotype in a myeloid effector, target engagement, and the cytokine program that accompanies it
In vivo delivered CAR programs
Which immune populations were transduced and which were not, whether the construct is intact and expressed, where the vehicle and payload went, and how long expression persisted
Autoimmune indications
The same product readouts, with depth and duration of target cell depletion as the pharmacodynamic endpoint, and reconstitution of that compartment over time

Samples and Species

Human, mouse, and NHP samples

Accelevir works with human, mouse, and non-human primate (NHP) samples, most often rhesus or cynomolgus macaque, with assay availability varying by species. We partner with the vivarium of your choice for animal studies and test the samples here. A method does not cross unchanged from a mouse study into a clinical trial, because the matrix, the species reagents, and the qualification all change. What carries across is the design, the controls, and the team, which keeps a preclinical dataset and a clinical one readable against each other and shortens the step between them.

Sample types

We accept the product itself, whole blood, bone marrow aspirate, fresh or cryopreserved mononuclear cells, serum and plasma, dissociated tumor and lymphoid tissue, and co-culture supernatant. All samples are processed in our BSL-2+ laboratory, which adds enhanced biosafety practices to BSL-2 containment for work with bloodborne pathogens and viral vectors.

Tissue, where the question is anatomical

Blood samples the circulating pool, and bone marrow, tumor, and lymphoid tissue can carry a different picture. Where a product has to act in tissue, the tissue is where depletion and persistence have to be read, so we can process tissue alongside blood and read the two against each other.

We can scope total RNA sequencing and mRNA sequencing into an integrated biomarker program, for transcriptional profiling of the product or of the cells that persist.

How a Cell Therapy Program Runs With Us

01

Scoping

02

Assay development and qualification

03

Testing

04

Reporting

Scoping

An Accelevir scientist reviews the construct, its delivery method, the target antigen, and the decision the data has to support. We then return a proposed readout panel, the sample requirements for each readout, and an itemized quote.

Assay development and qualification

Most programs need molecular, cellular, and immunology readouts together, so what we build is a testing plan rather than an assay. The plan covers the construct-specific designs, typically the integrity assay, any anti-construct antibody work, and the detection strategy for flow, alongside the standing assays it draws on. Each construct-specific design is built once and qualified once, with testing in the matrices the study will use forming part of that qualification, before study samples are tested and with acceptance criteria fixed in the method. Preclinical and discovery-stage work can run as research-use-only when the sponsor does not need qualification documentation.

Testing

We process material on receipt or thaw it from cryopreserved stocks, check viability and recovery, and run it with the agreed conditions and controls.

Reporting

You receive results in each assay's units with the controls behind them, the gating strategy and gated data files where flow cytometry is used, QC documentation, and tables linking the molecular and cellular results by subject and timepoint. Statistical analysis is optional.

Our Laboratory and Scientific Team

Accelevir Diagnostics, LLC is a bioanalytical contract research organization (CRO) in Baltimore, Maryland, built around measurements where the signal of interest is rare and the background is large. An engineered cell population after dosing is one of those, and so is an integrated construct that may or may not be complete.

Accelevir has supported more than 100 clinical studies, dozens of them interventional clinical trials, and serves seven program areas: Cell Therapy, Gene Therapy, Gene & RNA Editing, RNA Therapeutics, Vaccines, Antibody & Bispecific Therapeutics, and Antivirals.

We support clinical trials under the comprehensive quality management system of our CAP-accredited, CLIA-certified laboratory, a system CAP inspects as part of our accreditation. Cell therapy bioanalysis runs within it under documented standard operating procedures, and we host sponsor audits. We coordinate sample shipments from clinical sites and in-life partners, and we receive and track every sample under our quality management system. A chain-of-custody record follows each sample from receipt, and quality assurance reviews every report before it reaches you. Data transfer follows specifications built to your requirements, with the format and schedule agreed before the first samples arrive.

Accelevir is scientist-run. One team works alongside yours from protocol design to final report, so a design can adapt as the biology develops, and you have direct access to the scientists running your samples.

About Accelevir

The construct and the cells that carry it, measured together

Copy number, integrity, transcript, and cellular readouts run on material from one collection, so a program is not splitting limited sample across vendors or reconciling datasets that were never meant to be read together.

Immunology and molecular readouts from one study design

Killing, phenotype, intracellular function, and secreted cytokines are designed alongside the molecular endpoints rather than after them, and our immunology team includes scientists with cell-based immunotherapy development experience.

Preclinical to clinical without changing laboratories

Moving a method from a preclinical vendor to a clinical laboratory can mean requalifying it mid-program. We partner with the vivarium of your choice for mouse and NHP studies, then carry the same design, controls, and team into analytically qualified testing for clinical studies here.

Talk with our scientific team

FAQ

Frequently Asked Questions

Do you have an established assay for our construct?

Accelevir runs established singleplex digital PCR assays for the elements constructs share, such as the packaging signal, the LTRs, RRE, and WPRE. None of them targets your transgene or separates a complete construct from a truncated one. An integrity assay, an anti-construct antibody assay, and a flow detection strategy all depend on your construct’s architecture, so Accelevir builds each for the program that needs it. Killing, phenotyping, intracellular cytokine staining, ELISpot, FluoroSpot, proliferation, copy number by digital PCR, and biodistribution are established methods in our laboratory, and what changes per program is the target, the panel, and the analytes.

Why would we measure construct integrity as well as copy number?

Because the two readouts can disagree, and the disagreement changes what the copy number means. Accelevir reports integrity alongside copy number so that a count can be read as a count of complete constructs rather than of molecules carrying one target site. For a complex construct, one carrying several transgene elements, a safety switch, or an armoring cytokine, that is the difference between a dose that is understood and one that is only measured.

Can you work with samples from a study that has already finished?

Accelevir can often work from banked material, within the limits the material itself sets. We measure viability, recovery, and available volume before any sample is committed, so what a banked set can support is settled before it is spent.

How much material does each readout need?

Accelevir states the material requirement for each readout at scoping, because the totals depend on the study design. Cell input follows from the conditions, controls, and replicates in each cellular assay, and from the number of timepoints, while molecular input follows from the expected copy range and the matrix. Our scientists prioritize the readouts your next decision depends on when material is limited.

Should persistence be measured by digital PCR or by flow cytometry?

Accelevir recommends both, run from the same cryopreserved aliquot, because they answer different questions and can disagree in interpretable ways. Digital PCR quantifies the transgene, while flow cytometry establishes that viable cells are present and reports their state. Where the transgene persists and the flow signal does not, the usual reading is that cells carry the construct without expressing it. An Accelevir scientist will say which of the two carries your endpoint.

What is the qualification status of these assays?

The assays Accelevir runs to support clinical studies are analytically qualified under our quality management system, with qualification documentation available to support regulatory filings. That statement does not cover multiplex cytokine panels, which we send to a specialist laboratory. We do not offer product release testing.

What turnaround should we expect?

Turnaround depends on the number of samples, how many readouts the design includes, and whether Accelevir first has to develop and qualify a method for your construct. We confirm a timeline in writing when we scope the study.

Match the Readouts to Your Program's Question

Bring us the construct and how it is delivered, the target antigen and the cells that express it, and the species and stage. Tell us the samples you have or plan to collect, the timepoints they span, and the decision your next milestone turns on. Our scientists will recommend a readout panel matched to that question, state the sample requirements for each, and return an itemized proposal matched to your study design.

Accelevir Diagnostics, LLC is a CAP-accredited, CLIA-certified laboratory. Services are provided for research use in support of clinical and preclinical studies and are not intended for diagnostic, therapeutic, prognostic, or patient-management use.