Skip to main content
EU Science Hub

These Cell Transformation Assays (CTAs) aim to predict the carcinogenic potential of chemicals.

The methods are designed to evaluate the effects of chemicals on the growth of specific cell types and their potential progression through a transformation process from being normal cells to fully malignant cells.

We coordinated a series of studies that aimed to standardise and subsequently evaluate three CTA protocols in terms of transferability and reproducibility.

Two methods focussed on the Syrian hamster embryo (SHE) cells (at pH 6.7 and pH 7.0) and one method used a mouse fibroblast cell line called BALB/c 3T3.

The results of the validation studies and full reports are availableon TSAR, the Tracking System for Alternative methods towards Regulatory acceptance.

Read more about the SHE CTA at pH 6.7, SHE CTA at pH 7.0 and in vitro BALB/c 3T3 CTA on TSAR


Carcinogenesis refers to a complex multistage process in which normal cells transform into cancer cells.

The progressive changes happen at a number of cellular, genetic and epigenetic levels that result in a cell undergoing uncontrolled cell division, the formation of a malignant mass and eventually a tumour. While the process of carcinogenesis has been the focus of much research, complete mechanisms underlying each step in the process are still not fully understood.

Cell transformation is a multistage process which models various stages of in vivo carcinogenesis:

  • a block in cellular differentiation visualised as morphological changes in the cells
  • the acquisition of immortality - unlimited lifespan, an aneuploid karyotype and genetic instability
  • the acquisition of tumourogenicity - visualised through foci formation, anchorage independent growth in semi-solid agar and autocrine factor production
  • full malignancy when cells are injected into a suitable host.

The CTA has the potential to detect both genotoxic and some non-genotoxic carcinogens and through the use of two-stage protocols it has the potential to distinguish chemicals that are likely tumour initiators or promoters.

Three Cell Transformation Assays

In all three of the methods that were assessed, cells are exposed to test chemicals for a specific length of time and are then visually scored according to specific parameters relating to phenotypes and growth patterns.

This is achieved using light microscopy, where either the number of transformed colonies (for SHE) and/or the number of foci (for BALB/c 3T3) formed is scored.

Animal testing replacement

For the purposes of regulatory assessment of the carcinogenic potential of chemicals, the only accepted approach is the rodent-based life-time cancer bioassays.

These assays, depending on experimental design, can involve hundreds of animals (usually rats), they take a considerable amount of time to complete (minimum 3 years) and cost a lot of money (millions is not unusual).

While they are required for regulatory purposes, their human relevance has been questioned. Moreover the economic perspective and the perspective of the 3Rs all call for the development of alternative new approaches to assess carcinogenicity.

Due to the complexity of events that make up carcinogenicity, it is unlikely that a single in vitro method will be able to provide enough information for an unequivocal assessment of carcinogenic potential.

Therefore, although the CTAs appear to provide relevant information about the transforming potential of chemicals, they should not be used as stand-alone methods. They should always be used in combination with other complementary information sources as part of a weight of evidence approach or an integrated testing strategy.

Depending on the extent of other information available, it is conceivable that information on the transforming potential of chemicals may be sufficient for decision making and may thus allow waiving of the rodent bioassay.

Validation study

In vitro CTAs have been in use (widely) for almost 40 years and have found application in academia and various industrial settings as a method to assess potential carcinogenicity of chemicals.

Despite the widespread use of the method, they have not found application in a regulatory setting, and that is despite the possibility for considerable reductions in animal usage should they be applied (see above).

Regulators have consistently criticised the approach for lack of standardisation of protocols and its lack of objective criteria for scoring in the assay (it is visually scored, meaning there is a risk of variation due to human operator differences).

To remedy this, a number of initiatives were undertaken to look at how possible it would be to validate the approaches so the methods could be considered for regulatory purposes.

A workshop organised by EURL ECVAM in 1998 concluded that CTA tests were promising but still required further development, standardisation and validation before they could be used for regulatory purposes.

In parallel, the OECD initiated a review of the performance of CTAs and largely came to the same conclusions. However, in this case they specifically identified the SHE and BALB/c 3T3 CTAs as being developed sufficiently and proposed they be developed into an OECD Test Guideline.

On that basis EURL ECVAM was asked to coordinate a study that would standardize and subsequently evaluate three CTA protocols in terms of transferability and reproducibility. The study aimed to provide prospective data on reliability that would complement the review data of the OECD. In that sense it was designed to provide limited assessments of predictive capacity, which was largely provided by the substantial OECD review.

The study was completed in 2010 and was subsequently peer reviewed by ESAC, which issued an opinion in early 2011. It was subsequently provided to the OECD for further consideration.

Validation study outcomes

As a result of the study, well described standardized protocols are now available for both the SHE and BALB/c 3T3 methods. The two SHE protocols appeared to be transferable, at least to experienced laboratories, and reproducible. The BALB/c 3T3 method likely requires further attention from the perspective of reproducibility and data analysis.

In all three cases, the production of photo catalogues for each variant of the method helped identify typical transformed colonies and foci in each assay and this likely led to more consistent scoring during both training and the evaluations of the assay - notably addressing the concerns of regulators relating to objectivity of scoring.

EURL ECVAM recommendations

On the basis of this study and the ESAC opinion, the following recommendations were made:

  • The reproducibility and transferability of the two SHE CTA methods suggests that an OECD test guideline should be developed. The fact that the protocols and results were similar between the two methods suggests that one OECD test guideline should be developed.
  • Any use of the three CTA protocols (and for that matter any other CTA methods) should include the requirement of appropriate training of scorers when it comes to assessing outcomes (i.e. the number of transformed colonies and/or the number of foci formed).
  • The BALB/c 3T3 protocol assessment suffered from certain (limited, but still significant) deficiencies. This means that recommendations we can make are limited in terms of reliability. Nevertheless, we concluded that refinement of the method in terms of further validation testing is likely to yield positive results.
  • The BALB/c 3T3 protocol also involves a new statistical approach which should be refined to be more widely applicable. Currently, it uses advanced statistical approaches that should be more refined and more widely useable.
  • Despite this outcome for the BALB/c 3T3 protocol, we are convinced that it is much more applicable from a 3Rs perspective. It involves the use of an immortalised cell line rather than primary cells from embryos from pregnant female rats, as is the case for SHE cells.
  • Even though the SHE approach still involves the use of animals in the production of the cells required for the assay, it still represents a considerable reduction in numbers of animals required for the full assessment of carcinogenic potential of chemicals (as assessed via the two-year long animal based life time bioassays).
  • The performance characteristics of the SHE CTA methods should be further evaluated according to pre-existing data and complemented by any prospective data that might be generated in the meantime.
  • The use of CTAs has the potential to partially replace or partially reduce the use of animals when used in a weight of evidence approach for the identification of chemicals that may have carcinogenic properties.