What Are the Limitations of Induced Pluripotent Stem Cells?

Induced pluripotent stem cells carry a surprisingly long list of technical problems that keep them from fulfilling their therapeutic promise as quickly as early headlines suggested. The technology, which involves reprogramming adult cells back into an embryonic-like state, introduces genetic mutations, retains unwanted chemical marks from the original tissue, and produces cells that often behave more like fetal tissue than adult tissue. These are not minor wrinkles. They affect safety, reproducibility, and the practical cost of getting any iPSC-based therapy into a clinic. Understanding where the gaps are helps explain why, more than fifteen years after their invention, iPSC therapies remain rare and heavily regulated.

Genetic Damage Accumulated During Reprogramming

Turning a skin cell or blood cell back into a pluripotent state is a violent molecular event, and the genome does not come through unscathed. A study of 22 human iPSC lines made with five different reprogramming methods found that each line carried an average of about six protein-coding point mutations per exome. Most of those mutations were the kind that change a protein’s function or truncate it entirely, and they were overrepresented in genes already linked to cancer. Roughly half of these mutations pre-existed in the donor cells at very low frequencies and got amplified during reprogramming, while the rest arose fresh during or after the process.¹

Point mutations are only part of the story. Larger structural problems, including duplications and deletions spanning millions of base pairs, also show up. One comparison of integrating versus non-integrating reprogramming methods found that cell lines made with genome-integrating viruses accumulated more of these large-scale copy number changes and had larger duplications on average than lines made with non-integrating approaches.² And the genome keeps shifting even after reprogramming is finished. Over 150 days of continuous culture, researchers watched one iPSC line develop a population trisomic for an entire chromosome, which rapidly took over the culture. Both lines in that experiment picked up numerous new structural variants as they were passaged.³

This ongoing genomic instability means that quality control cannot be a one-time event. A cell line that passes genetic screening at passage five may look quite different by passage thirty, and any clinical-grade line needs monitoring at multiple time points. The enrichment of cancer-associated mutations adds a specific safety concern: cells intended for transplantation could carry the seeds of malignancy before they ever reach a patient.

Epigenetic Memory From the Donor Tissue

Even when reprogramming successfully resets a cell’s identity, chemical marks on the DNA do not always get fully erased. iPSCs tend to retain residual methylation patterns from whatever tissue they came from, a phenomenon researchers call epigenetic memory. Work comparing iPSCs derived from cord blood cells versus keratinocytes showed that the two types remained distinct in their genome-wide methylation profiles and their ability to become different cell types. Extended passaging of some clones did not fix this problem, meaning the bias persisted even after months of culture.⁴

The practical consequence is that iPSCs made from blood cells differentiate more readily into blood-related lineages, while iPSCs from other tissues resist that fate. Blood-derived iPSCs have been shown to retain methylation at loci tied to blood cell development, giving them higher colony-formation ability in blood-directed protocols compared to iPSCs from non-blood sources.⁵ This means the starting tissue is not a neutral choice. If you want to make neurons, starting from a neural-adjacent cell type may give better results than starting from a blood sample, and vice versa. But it also introduces a variable that complicates standardization, because the “best” starting material depends on the intended application.⁶

Imprinting Defects and Aberrant Methylation

Beyond leftover tissue-specific marks, reprogramming also disrupts a separate layer of gene regulation called genomic imprinting. Certain genes are supposed to be active only from the maternal or paternal copy, controlled by methylation at specific regulatory regions. iPSCs frequently show abnormal methylation at these sites. One study found consistent hypermethylation at two imprinting control regions regardless of which reprogramming method or starting cell type was used, leading to loss of parental imprinting and the abnormal activation of both copies of growth-related genes.⁷

A comprehensive analysis of imprinting in pluripotent stem cells confirmed strong correlations between methylation changes at imprinted regions and the switching of genes from their normal single-copy expression to biallelic expression, where both copies turn on.⁸ This matters because imprinted genes often regulate cell growth and metabolism. Their dysregulation is associated with developmental disorders and some cancers, so any cells derived from iPSCs with imprinting defects could behave unpredictably after transplantation.

The Teratoma Problem

Pluripotency is simultaneously the whole point of iPSCs and their most dangerous property. Any leftover undifferentiated cells in a batch of differentiated iPSC-derived tissue can form teratomas, disorganized masses containing a jumble of tissue types. Even a small number of remaining undifferentiated cells is enough to seed tumor growth after transplantation, which makes their complete removal a prerequisite for any clinical application.⁹ This risk is not theoretical; it has been flagged as a constraint that could limit potential clinical uses of iPSC-derived products if left unaddressed.¹⁰

Researchers have developed several strategies to tackle the problem. One approach uses an antibody-drug conjugate targeting CD30, a surface marker found on undifferentiated iPSCs but not on differentiated cells. Pretreating iPSC-derived insulin-producing cells with this antibody for 24 hours before transplantation completely eliminated teratoma formation in a mouse model, with zero tumors in ten treated animals compared to tumors in all eight untreated controls.¹¹ Another strategy involves engineering a “suicide gene” into the iPSCs. One version uses an inducible caspase-9 system that kills cells on demand when a specific drug is administered. This system eliminated over 94% of undifferentiated iPSCs within 24 hours in the lab and eradicated iPSC-derived tumors in mice.¹² When the same suicide gene was placed into a defined chromosomal location using gene-editing tools, concentrations as low as 0.1 nanomolar of the activating drug killed over 98% of iPSCs and their derived immune cells.¹³

These safety switches are promising, but they add manufacturing complexity, cost, and regulatory scrutiny. Every additional genetic modification to an iPSC line needs its own round of validation.

Cells That Never Fully Grow Up

One of the most frustrating limitations for researchers trying to model adult diseases is that iPSC-derived cells tend to stall at an immature developmental stage. The problem is best documented in heart cells. iPSC-derived cardiomyocytes beat rhythmically in a dish, which looks impressive, but their structure, gene expression, metabolism, and contractile properties resemble fetal heart cells rather than adult ones.¹⁴ They rely heavily on sugar-based metabolism instead of the fatty acid oxidation that adult heart cells use, their internal structures are disorganized, and their electrical properties differ from what you would record in an adult heart. This immaturity limits their usefulness both for modeling adult heart diseases and for drug screening, where the cells need to respond the way adult tissue would.¹⁵

The maturation problem is not unique to heart cells. iPSC-derived neurons, liver cells, and pancreatic cells all tend toward embryonic or fetal phenotypes. Various strategies exist to push them toward maturity, including prolonged culture, mechanical stimulation, electrical pacing, and co-culture with other cell types. Progress has been made, but no protocol reliably produces cells that are indistinguishable from their adult counterparts. For conditions that only manifest in adult tissue, like age-related heart failure or late-onset neurodegeneration, this gap between fetal-like and adult cell behavior is a real obstacle.

Immunogenicity Is Not Zero

One of the original selling points of iPSCs was that patient-specific cells would be immune-matched and therefore tolerated by the patient’s own immune system without suppression. That assumption has turned out to be more complicated than expected. Studies testing iPSC immunogenicity have produced conflicting results, and there are biological reasons to suspect that reprogrammed cells might express proteins in unusual patterns that the immune system could recognize as foreign.¹⁶

An alternative strategy to patient-specific lines is to bank iPSCs from donors who are homozygous for common immune-matching markers, so that a relatively small number of lines could serve a large portion of the population. But even when donor and recipient are matched on the major immune markers, rejection can still occur because immune responses are influenced by other factors beyond those main markers.¹⁷ The result is that some degree of immune suppression may still be needed for many iPSC-based transplants, which undercuts one of the technology’s main advantages over organ transplantation or embryonic stem cell therapy.

Line-to-Line Variability

iPSC research has a reproducibility problem rooted in the cells themselves. Different iPSC lines, even ones derived from the same donor, can differ substantially in their differentiation ability, gene expression, protein levels, and morphology. These differences arise from the donor’s genetic background, the specific reprogramming event, and drift during culture.¹⁸

The laborious and expensive nature of iPSC work means that many studies use very few cell lines, often just two or three. With so much inherent variability between clones and between donors, small sample sizes make it hard to distinguish a genuine experimental effect from random line-to-line noise. A systematic look at the proteome of iPSC-derived astrocytes confirmed that variability exists at every level: between donors, between different clones from the same donor, and even between supposedly identical differentiations of the same clone.¹⁹ For disease modeling and drug screening, where you need to confidently attribute differences to the disease or the drug rather than to the cells, this variability is a serious confounder. Best-practice guidelines now recommend using multiple clones from multiple donors, but that multiplies the already high cost and labor involved.

Reprogramming Methods and Their Trade-Offs

The original iPSC recipe used retroviruses to deliver the reprogramming genes, and this method carries well-known hazards: the viral DNA inserts permanently into the genome, can land inside or near important genes, and the inserted sequences can reactivate later or trigger immune responses.²⁰ The risk of insertional mutagenesis was recognized early as a major limitation,²¹ and it drove the development of non-integrating alternatives such as episomal plasmids, Sendai virus, mRNA delivery, and small-molecule cocktails.

Non-integrating methods avoid permanent genome insertion, but they come with their own downsides. They tend to be less efficient, meaning fewer cells successfully reprogram, and the resulting iPSCs still pick up point mutations and chromosomal changes during the process. The field has converged on non-integrating approaches as the safer default for clinical-grade work, yet no method completely eliminates the genetic and epigenetic problems described earlier. Each approach involves a trade-off between efficiency, safety, and the quality of the resulting cells.

CRISPR and the p53 Bottleneck

Gene editing with CRISPR-Cas9 is often paired with iPSC technology to correct disease-causing mutations or to create disease models. But applying CRISPR to iPSCs introduces a specific complication. When Cas9 cuts both strands of DNA in a human pluripotent stem cell, the damage triggers the p53 pathway, a tumor-suppressor response that kills most of the edited cells. One study found that while indel efficiency exceeded 80%, the DNA breaks were so toxic that most cells died, severely reducing the yield of precisely edited lines.²²

The concern goes deeper than low efficiency. Because p53 selects against cells with DNA damage, the surviving edited cells may be enriched for those that have lost or mutated their p53 gene. In other words, the very cells that survive CRISPR editing could be the ones with a broken tumor suppressor, which is exactly the wrong starting material for a therapy you plan to put into a human body. Newer approaches like base editing, which chemically modifies individual DNA letters without cutting both strands, generate far less p53 activation and may avoid this selection bias.²³ But base editing can only make certain types of changes, so it is not a universal replacement for traditional CRISPR cutting.

Manufacturing Cost and Regulatory Barriers

Even when the biology cooperates, getting iPSC-derived products to patients faces steep logistical hurdles. Manufacturing under good manufacturing practice (GMP) conditions demands reproducibility at every step, extensive quality control, and documentation that meets regulatory requirements. These demands drive costs high enough to prevent many research groups from even initiating clinical trials.²⁴ Reliable scale-up under GMP conditions remains a major technical hurdle, and fundamental questions about whether transplanted cells will survive long-term, integrate into the target tissue, and avoid immune rejection are still unresolved.²⁵

Intellectual property also plays a role. The core iPSC patents and the reagents needed for manufacturing have been covered by overlapping licenses that pile on costs. Annual licensing fees for reagents and core intellectual property in master cell bank manufacturing have been reported to exceed $600,000.²⁶ Some of the foundational patents are beginning to expire, which may eventually lower the barrier, but for now the combination of biology, regulation, and licensing makes iPSC therapies among the most expensive cell products to develop.

Mitochondrial and Metabolic Quirks

Reprogramming resets a cell’s nucleus, but the mitochondria, the cell’s energy-producing compartments, carry their own DNA and their own problems. iPSCs derived from patients with mitochondrial mutations showed reduced maximum respiratory capacity compared to controls, along with lower spare respiratory capacity, meaning the cells had less metabolic headroom under stress.²⁷ These bioenergetic deficits also correlated with altered differentiation behavior. The takeaway is that reprogramming does not fix pre-existing mitochondrial disease, and mitochondrial mutations can propagate into any cell type derived from those iPSCs, potentially confounding both disease models and therapeutic applications.

The Gap Between Dish and Body

Most iPSC-derived cells are grown on flat plastic surfaces, which is radically different from the three-dimensional environment of living tissue. Growing cells in 3D organoid formats can overcome many limitations of flat culture and allows researchers to study cells in a more architecturally realistic setting. But generating and studying organoids from iPSCs, particularly cardiac organoids containing multiple cardiovascular cell types, remains a challenge.²⁸ Organoids tend to lack blood vessels, are limited in size by nutrient diffusion, and do not fully recapitulate the mechanical forces that shape tissues in the body. So while they represent a meaningful step up from flat culture, they are still a rough approximation of what happens in a living organ.

Why iPSCs Outside Humans and Mice Remain Difficult

Nearly all iPSC research has been done in human and mouse cells, and extending the technology to other species has proven stubbornly difficult. Non-model species lack the annotated genomes, validated reagents, and species-specific culture conditions that make iPSC work routine in humans and mice. On top of this infrastructure deficit, genuine biological differences in how pluripotency networks are wired across species mean that protocols optimized for human cells often simply do not work in, say, livestock or endangered wildlife.²⁹ For conservation biology and veterinary medicine, this limits the potential applications that iPSC technology could offer, from preserving genetic diversity in endangered populations to developing cell therapies for animals.

How iPSCs Compare to Embryonic Stem Cells

iPSCs were initially celebrated as a functional equivalent of embryonic stem cells (ESCs) without the ethical controversy of destroying embryos. The two cell types are broadly similar, but systematic comparisons have made clear that they are not identical. They differ in gene expression patterns, epigenetic profiles, and differentiation tendencies.³⁰ Some of these differences trace back to the epigenetic memory and reprogramming artifacts already discussed. Others may reflect fundamental aspects of how reprogramming works compared to natural embryonic development. For researchers choosing between the two, the decision often depends on the application: iPSCs offer patient-specificity and avoid embryo destruction, while ESCs remain the benchmark for pluripotency and are sometimes preferred for basic research where genetic and epigenetic consistency matters more.

Ethical Terrain That iPSCs Did Not Eliminate

iPSCs sidestepped the central ethical issue of embryonic stem cell research by not requiring embryo destruction. But they introduced new ethical questions of their own. Because iPSCs can theoretically be coaxed into forming gametes, sperm and egg cells, they raise the possibility of creating human embryos from reprogrammed cells. This prospect has prompted discussion about what norms should govern such embryos, including their quality, reliability, and the circumstances under which creating them would be acceptable.³¹ The revised guidelines from the International Society for Stem Cell Research have addressed some of these concerns, but the regulatory landscape remains uneven across countries, and the ethical conversation continues to evolve as the technology’s capabilities expand.