Cell Biology and Toxicology

, Volume 24, Issue 4, pp 315–319

Correlation of visual in vitro cytotoxicity ratings of biomaterials with quantitative in vitro cell viability measurements

Authors

    • Biochemical Sciences and Engineering, Central Research and DevelopmentDuPont Experimental Station
    • DuPont Experimental Station
  • Ann B. Yetter
    • Biochemical Sciences and Engineering, Central Research and DevelopmentDuPont Experimental Station
Article

DOI: 10.1007/s10565-007-9040-z

Cite this article as:
Bhatia, S.K. & Yetter, A.B. Cell Biol Toxicol (2008) 24: 315. doi:10.1007/s10565-007-9040-z

Abstract

Medical devices and implanted biomaterials are often assessed for biological reactivity using visual scores of cell–material interactions. In such testing, biomaterials are assigned cytotoxicity ratings based on visual evidence of morphological cellular changes, including cell lysis, rounding, spreading, and proliferation. For example, ISO 10993 cytotoxicity testing of medical devices allows the use of a visual grading scale. The present study compared visual in vitro cytotoxicity ratings to quantitative in vitro cytotoxicity measurements for biomaterials to determine the level of correlation between visual scoring and a quantitative cell viability assay. Biomaterials representing a spectrum of biological reactivity levels were evaluated, including organo-tin polyvinylchloride (PVC; a known cytotoxic material), ultra-high molecular weight polyethylene (a known non-cytotoxic material), and implantable tissue adhesives. Each material was incubated in direct contact with mouse 3T3 fibroblast cell cultures for 24 h. Visual scores were assigned to the materials using a 5-point rating scale; the scorer was blinded to the material identities. Quantitative measurements of cell viability were performed using a 3-(4,5-dimethylthiozol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) colorimetric assay; again, the assay operator was blinded to material identities. The investigation revealed a high degree of correlation between visual cytotoxicity ratings and quantitative cell viability measurements; a Pearson’s correlation gave a correlation coefficient of 0.90 between the visual cytotoxicity score and the percent viable cells. An equation relating the visual cytotoxicity score and the percent viable cells was derived. The results of this study are significant for the design and interpretation of in vitro cytotoxicity studies of novel biomaterials.

Keywords

3T3 cell lineAdhesivesBiomaterialsCytotoxicityFibroblasts

Abbreviations

MTT

3-(4,5-dimethylthiozol-2-yl)-2,5-diphenyltetrazolium bromide

Introduction

To possess clinical value, medical devices and implanted biomaterials must be non-toxic, eliciting no adverse response from the application site or surrounding tissues. There is great necessity for reliable in vitro cytotoxicity assays, so that medical biomaterials can be correctly evaluated before preclinical or clinical studies. An accurate and precise in vitro cytotoxicity assay can reduce the number of animal studies needed to develop a new medical device. At the same time, it is desirable that in vitro cytotoxicity evaluations be sufficiently rapid to allow the screening of large numbers of potential biomaterial candidates. For this reason, visual scoring systems are often used to report cytotoxic effects of medical biomaterials. In such systems, biomaterials are assigned cytotoxicity ratings based on visual evidence of morphological cellular changes, including cell lysis, rounding, spreading, and proliferation. A visual grading method requires only microscopic inspection of cell–material interactions, in contrast to quantitative methods for cytotoxicity measurement, which require time-consuming colorimetric assays and spectrophotometer equipment. While visual cytotoxicity scoring provides advantages in terms of speed and convenience, it is critical to ensure that accuracy is not compromised.

Important disparities exist in regulatory recommendations for cytotoxicity testing of new medical devices. The International Organization for Standardization has set forth ISO-10993-5 guidelines for “Biological Evaluation of Medical Devices” that allow for the use of visual grading systems in cytotoxicity determinations (ISO 10993-5 1999). The US Food and Drug Administration guidelines for medical device evaluation are largely in concordance with ISO-10993 requirements (FDA G95-1 1995) so that medical device cytotoxicity may be evaluated according to ISO-10993-5 recommendations, with a visual scoring method. In contrast, the Japanese Ministry for Health and Welfare applies a more stringent standard for medical device approval and specifies that medical device cytotoxicity be evaluated using a quantitative assessment of surviving cells (MHW notification no. 99 1995). Thus, there is disagreement between regulatory bodies regarding the appropriateness of visual scoring methods for cytotoxicity assessment of new medical devices.

The objective of the present study is to compare visual in vitro cytotoxicity ratings to quantitative in vitro cytotoxicity measurements for biomaterials to determine the level of correlation between visual scoring and a quantitative cell viability assay. Biomaterials representing a spectrum of biological reactivity levels were evaluated, including organo-tin polyvinylchloride (PVC; a known cytotoxic material), ultra-high molecular weight polyethylene (a known non-cytotoxic material), and implantable tissue adhesives. Polyethylene and organo-tin PVC were chosen for the dynamic cytotoxicity assay because these two materials typically represent the extremes of non-toxicity and toxicity, respectively, in ISO 10993-5 standardized static cytotoxicity tests of medical devices (AAMI 2003). Tissue adhesives were chosen for the study as representative biomaterials intended for use in the abdominal space. A total of 33 material assessments were performed. The mouse fibroblast cell line NIH 3T3 was chosen for the cytotoxicity study, as fibroblasts are the main cellular component of dense connective tissues and are the typical cell line used for cytotoxicity studies of biomaterials. Biomaterial samples were placed in direct contact with 3T3 fibroblasts to more closely mimic the physiological situation (Ratner et al. 1996). Each material was incubated in direct contact with mouse 3T3 fibroblast cell cultures for 24 h. Visual scores were assigned to the materials using a 5-point rating scale; the scorer was blinded to the material identities. Quantitative measurements of cell viability were performed using a 3-(4,5-dimethylthiozol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) colorimetric assay; again, the assay operator was blinded to material identities. The results of this study are significant for the selection and interpretation of in vitro cytotoxicity studies for medical biomaterials.

Materials and methods

Biomaterial samples

Organo-tin PVC strips were obtained from Smiths Medical (Kent, UK). Polyethylene samples were obtained from the US Pharmacopeia (Rockville, MD). Polysaccharide-based tissue adhesives composed of dextran aldehyde and multi-arm polyethylene glycol (PEG) amine were obtained from Dr. George Kodokian at DuPont; the synthesis of these adhesives has been previously described (Kodokian and Arthur 2006). The commercial cyanoacrylate-based tissue adhesive Dermabond™ was obtained from Johnson & Johnson/Ethicon (Somerville, NJ).

Cell culture and reagents

The mouse fibroblast NIH 3T3 cell line was obtained from American Type Culture Collection (ATCC). Cells were grown at 37°C in a 5% (v/v) CO2 incubator in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% (v/v) fetal bovine serum (Invitrogen; Carlsbad, CA).

Cytotoxicity assay with visual scoring

Biomaterial samples were randomized, and the assay operator was blinded to material identities. A 30-mg sample of biomaterial was placed into an empty well of a six-well culture plate. Mouse 3T3 fibroblast cells were then seeded into the well at a density of 50,000 cells per well. The plate was incubated at 37°C for 24 h. After the incubation, fibroblast cell growth and morphology were visualized directly, using a Nikon microscope with a 10× objective, equipped with a Nikon 35-mm camera. A visual cytotoxicity score was assigned according to a 5-point rating scale, ranging from 1 = cytotoxic to 5 = non-cytotoxic, based on observable characteristics of cell spreading and cell lysis (Fig. 1). A score of 5 is assigned for a confluent monolayer of well-defined cells exhibiting cell-to-cell contact; cell morphology and cell density are not altered by the presence of a biomaterial, and discrete intracytoplasmic granules are observed. No cell lysis is observed, indicating a non-cytotoxic reaction. A score of 4 is assigned when occasional lysed cells are present; not more than 20% of the cells appear to be round, loosely attached, and without cytoplasmic granules. A score of 3 is assigned when cell lysis becomes more prevalent, but no more than 50% of the cells are round and devoid of intracytoplasmic granules. A score of 2 is assigned when the majority of cells are affected, but not more than 70% of the cells are rounded or lysed. A score of 1 is assigned when destruction and lysis of cells is nearly complete; considerable open areas between cells indicate that extensive cell lysis has occurred, indicating a cytotoxic reaction.
https://static-content.springer.com/image/art%3A10.1007%2Fs10565-007-9040-z/MediaObjects/10565_2007_9040_Fig1_HTML.gif
Fig. 1

Scheme for assigning visual cytotoxicity scores to biomaterials

Cytotoxicity assay with quantitative cell viability measurement

Biomaterial samples were randomized, and the assay operator was blinded to material identities. A 30-mg sample of biomaterial was placed into an empty well of a six-well culture plate. Mouse 3T3 fibroblast cells were then seeded into the well at a density of 50,000 cells per well. The plate was incubated at 37°C for 24 h. After the incubation, the cell viability was determined using a MTT colorimetric assay kit from ATCC; the protocol for the MTT assay has been extensively described (Sgouras and Duncan 1990).

Results

Thirty-three biomaterial samples were assessed for cytotoxicity using both a visual rating scale and a quantitative MTT colorimetric assay. The results are summarized in Table 1. The results indicate a high degree of correlation between the visual cytotoxicity rating and the quantitative cell viability measurement (Fig. 2). When the data are plotted to show the relationship between the visual cytotoxicity score and the MTT colorimetric technique for quantifying cell viability, a linear relationship is suggested. The visual rating and quantitative cell viability are related by the following equation:
$${\text{\% }}\;{\text{cell}}\;{\text{viability = 18}}{\text{.8}}\left( {{\text{visual}}\;{\text{score}}} \right)$$
https://static-content.springer.com/image/art%3A10.1007%2Fs10565-007-9040-z/MediaObjects/10565_2007_9040_Fig2_HTML.gif
Fig. 2

Relationship between quantitative cell viability and visual cytotoxicity ratings for medical biomaterials. Fibroblast cells were incubated in direct contact with material samples for 24 h at 37°C. Visual cytotoxicity scores were assigned according to a 5-point rating scale. Quantitative cell viability measurements were performed using an MTT colorimetric assay. Linear correlation was performed with Pearson’s linear correlation, with the equation of the line and r value shown

Table 1

Visual cytotoxicity ratings and quantitative cell viability data for 33 medical biomaterial samples

Visual cytotoxicity rating

% Cell viability

1

3

1

3

1

3

1

27

1

30

1

43

2

25

2

26

2

25

2

45

2

49

2

54

2

50

2

53

2

50

3

32

3

32

3

31

4

76

4

77

4

76

4

79

4

85

4

79

4

83

4

82

4

83

5

100

5

95

5

84

5

100

5

94

5

84

A Pearson’s correlation gives a correlation coefficient of 0.90 between the visual cytotoxicity score and the percent viable cells.

Discussion

The above results indicate that a linear relationship exists between visual in vitro cytotoxicity ratings and quantitative in vitro cell viability measurements for medical biomaterials. When mouse 3T3 fibroblasts are incubated in direct contact with biomaterial samples, visual cytotoxicity scores are predictive of the percentage of viable cells. This finding is significant because it suggests that visual scoring is a valid method for rapidly estimating the cytotoxicity of new materials to clinically relevant cell lines. Visual inspection can be performed more quickly than a colorimetric assay and avoids the necessity for additional equipment, so the visual rating method may be particularly advantageous for screening a large number of biomaterial samples. It is important to note, however, that the medical biomaterial samples given a visual rating of 3 (in the middle of the rating scale) all demonstrated quantitative cell viabilities well below 50%, suggesting that special caution should be used in interpreting visual ratings in the middle of the rating scale. The visual rating system used in the present study suggests that biomaterials should raise serious concern for cytotoxicity when the visual rating is 3 or lower.

It is critical that in vitro assays for cytotoxicity accurately assess the potential of biomaterials to damage cells; a lack of stringency during in vitro testing can lead to unnecessary and harmful animal studies and disastrous results during in vivo preclinical testing and clinical use. Indeed, cyanoacrylate-based adhesives have been widely used in medical and surgical procedures, yet concerns have recently been raised regarding the hazards of cyanoacrylate use to both patients and healthcare workers (Leggat et al. 2007). Cyanoacrylate adhesives have been shown to cause inflammation and tissue necrosis in vivo (Toriumi et al. 1991; Kim et al. 1995). In another instance, a commercial albumin/glutaraldehyde tissue adhesive was shown to cause significant damage to lung and liver tissue, and investigators have recommended that its use be restricted to emergency procedures (Furst and Banerjee 2005). Direct contact testing of novel biomaterials with in vitro cell cultures may provide an early indicator of potential problems in vivo. We therefore recommend that any subjective in vitro cytotoxicity assessment that utilizes visual rating scales be validated against a quantitative in vitro cell viability assay before routine use of the visual inspection scheme. The visual rating system must be validated for the specific cell lines and biomaterials being examined. Such an assessment increases the likelihood of regulatory approval for novel biomaterials, and more importantly, it represents responsible product stewardship.

Acknowledgements

This study was performed as part of a DuPont Six Sigma project. We are grateful to Deana DiCosimo, Harvey Gold, and Lourdes Puig from the DuPont Six Sigma organization for their guidance and suggestions.

Copyright information

© Springer Science+Business Media B.V. 2007