Human systemic exposure to a [14C]-para-phenylenediamine-containing oxidative hair dye and correlation with in vitro percutaneous absorption in human or pig skin

https://doi.org/10.1016/j.fct.2004.02.020Get rights and content

Abstract

We investigated the absorption of a commercial [14C]-PPD-containing oxidative dark-shade hair dye in human volunteers as well as in vitro using human or pig ear skin. The hair of eight male volunteers was cut to a standard length, dyed, washed, dried, clipped and collected. Hair, washing water, materials used in the study and a 24-h scalp wash were collected for determination of radioactivity. Blood, urine and faeces were analysed up to 120 h after hair dyeing. An identical [14C]-PPD-containing hair dye formulation was applied in vitro for 0.5 h to human and pig ear skin, and radioactivity was determined in skin compartments after 24 h. In humans, the recovery rate was 95.7 ± 1.5% of the applied radioactivity. Washing water, cut hair, gloves, paper towels, caps or scalp wash contained a total of 95.16 ± 1.46% of the applied [14C]. Absorbed radioactivity amounted to 0.50 ± 0.24% in the urine and 0.04 ± 0.04% in the faeces, corresponding to a mean of 7.0 ± 3.4 mg [14C]-PPD-equivalents absorbed. Within 24 h after application, most of the radioactivity was eliminated. The Cmax of [14C]-PPD-equivalents in the plasma was 0.087 μgeq/ml, the Tmax was approximately 2 h, and the mean the AUC0–12 h was 0.67 μgeq h/ml. In vitro tests in human or pig skin found total absorbed amounts of 2.4 ± 1.6% (10.6 ± 6.7 μgeq/cm2) or 3.4 ± 1.7% (14.6 ± 6.9 μgeq/cm2), respectively. Percentage-based in vitro results were considerably higher than corresponding in vivo data, whereas, in units of μg/cm2, they corresponded to a total absorbed amount of 7.40 or 10.22 mgeq for human or pig skin, respectively. All results suggested that hair dyeing with oxidative hair dyes produces minimal systemic exposure that is unlikely to pose a risk to human health.

Introduction

Presence, magnitude and nature of human systemic exposure are pivotal parameters of the safety evaluation of substances that are applied to human skin or come into contact with the surface of the human body. Although contact allergy to hair dyes has been a major safety issue, the prevalence of hair dye-induced allergy appears to have decreased in recent years against an increasing use of hair dyes in the industrialised world. For example, the North American Contact Dermatitis Group reported a decreasing prevalence for allergenic reactions to para-phenylenediamine (PPD), i.e. from rank 3 in 1984 to rank 10 in 1996 (Maouad et al., 1999). This is consistent with European observations, which described PPD as the 5th most frequent allergen during 1985–1990, and as the 15th most frequent allergen during 1991–1996 (Goossens and Merckx, 1997).

Given their chemical class of oxidative hair colour ingredients (aromatic amines) and the large size of the potentially exposed population (more than 50% of the female population of industrialised countries) the safety of hair dyes has been extensively studied. Toxicological properties of hair dyes and their ingredients have been subject of numerous investigations including their acute, subchronic, chronic, genetic and reproductive toxicities or carcinogenicity (see Corbett et al., 1999; Nohynek et al., 2004). In contrast, little is known about the magnitude and nature of the actual systemic exposure of the human organism to hair dyes. Thus a pivotal parameter of the risk assessment of hair colours remains poorly understood. Today, the estimation of the human systemic exposure to hair dyes relies principally on in vitro models, i.e. percutaneous absorption/penetration studies in human or animal skin. Although the test methods have been somewhat standardised (SCCNFP, 1999; OECD, 2000a, OECD, 2000b; Diembeck et al., 1999), their results are subject to considerable variation depending on the skin type (human or animal), origin of the skin (body region) and test protocols. In addition, the question whether presence of a substance in skin compartments, such as the stratum corneum or the epidermis suggests potential bioavailability is unresolved (Dressler, 1999; Schaefer and Redelmeier, 1996). Recent results comparing percutaneous absorption of a [14C]-labelled ultraviolet filter in vitro and in vivo in humans suggested that in vitro models may substantially over-estimate systemic exposure and the subsequent risk to human health (Benech-Kieffer et al., 2003). Overall, in the absence of validation studies, in vitro percutaneous absorption data may be regarded as an insufficiently validated endpoint for human systemic exposure. Therefore, it is of pivotal importance that the uncertainties of in vitro models are addressed by in vivo validation studies in man or suitable in vivo models that resemble the human organism and his skin. Some in vivo studies on the percutaneous absorption of oxidative hair dyes in humans or non-human primates were reported in the 1980s. One study measured the systemic absorption in man of oxidative hair dye precursors such as para-phenylenediamine (PPD) by analysis of human urinary metabolites, such as N-mono- or N,N-di-acetylated PPD (Goetz et al., 1988). Other studies applied [14C]-labelled oxidative hair colours to the hair of human volunteers or monkeys and estimated systemic absorption by determination of urinary excretion of [14C]-labelled compounds (Maibach and Wolfram, 1981; Wolfram and Maibach, 1985). Although all results suggested that use of oxidative hair dyes produced little systemic exposure in man (less than 1.0% of the radioactivity applied to the hair excreted in the urine), none of the investigations applied a mass balance approach or attempted to quantify the systemic exposure via determination of blood levels. In the absence of an adequate mass balance it cannot be ascertained whether a part of a topically applied compound has been lost, missed in the analysis or retained in a compartment the organism. In addition, no studies have been conducted that compared human absorption of hair dyes in vivo with that of in vitro models using identical formulations and exposure conditions.

To this end we investigated the exposure of the human organism after hair dyeing with a [14C]-labelled, commercial dark-shade hair dye (containing a maximal concentration of PPD) under controlled conditions and using a stringent mass balance approach. In order to permit comparison of the in vivo results with those of in vitro models, we performed parallel in vitro investigations with human and pig skin with the same oxidative hair dye formulation under conditions that resembled those of the in vivo study as closely as technically feasible. Our aim was (a) to elucidate the actual human systemic exposure to an oxidative hair dye under use conditions and (b) to assess the relevance of currently used in vitro models.

Section snippets

Chemicals

For the human and in vitro studies [14C]-phenyl-labelled para-phenylenediamine di-hydrochloride (PPD) was prepared by Amersham Biosciences, Cardiff, UK, and had a specific activity of 2.11 GBq/mmol or 57 mCi/mmol. The radiochemical purity determined by HPLC was 98.2%. Unlabeled PPD was obtained from DuPont, France. Dulbecco modified phosphate buffer saline solution (w/o Ca2+, Mg2+, Instamed 9.55 g/l) was obtained from Biochrom, Chapniers, France and sodium lauryl sulfate from Sigma, Saint

Study in humans

The individual and mean applied doses and recovery data per study subject are summarised in Table 1. The total recovery rate was 95.7 ± 1.5% (range: 93.0–97.1%) of the applied radioactivity.

Non-absorbed radioactivity

The results of the analyses of radioactivity in the hair wash water, cut-off hair, scalp wash, gloves, towels and caps are summarised in Table 2.

Discussion

The recovery rate of the human study (95.7 ± 1.5% of the applied radioactivity) suggests an adequate mass balance. Although some (<4.3%) of the applied radioactivity remained unaccounted for, this is not surprising, taking into account that the dyed hair was not completely shaved off, but clipped, leaving about 1–2 mm of hair on the scalp. Residual hair retained an estimated 1–1.5% of the applied radioactivity. In addition, residual radioactivity in the skin or hair follicles was not measured in

Acknowledgements

We wish to express our gratitude for Catherine Noe for her dedication and excellent technical support.

References (25)

  • S.J. Gaudet et al.

    Identification and characterization of two isoenzymic forms of arylamine N-acetyltransferase in Syrian hamster skin

    J. Invest. Dermatol.

    (1993)
  • N. Goetz et al.

    Percutaneous absorption of p-phenylene-diamine during an actual hair dyeing process

    Int. J. Cosm. Sci.

    (1988)
  • Cited by (72)

    • In vivo quantitative molecular absorption of glycerol in human skin using coherent anti-Stokes Raman scattering (CARS) and two-photon auto-fluorescence

      2019, Journal of Controlled Release
      Citation Excerpt :

      Addressing this complex question would require an in vivo experimental framework capable of providing a detailed anatomical skin description together with a 3D active molecule concentration view over the time course of the experiment. This is in sharp contrast with the information provided by conventional skin bioavailability approaches such as ex vivo human skin mounted on Franz cell diffusion set-up together with validated analytical methods (such as radio-labeled compounds [5] or liquid chromatography-tandem mass spectroscopy [6]) or destructive sequential adhesive tape stripping extraction to determine active molecule concentration profile at specific time point [7]. Recently many efforts have been made to develop skin imaging modalities that use inelastic scattered light by the vibrational motions of chemical bonds [8].

    • Human systemic exposure to [<sup>14</sup>C]-paraphenylenediamine-containing oxidative hair dyes: Absorption, kinetics, metabolism, excretion and safety assessment

      2015, Food and Chemical Toxicology
      Citation Excerpt :

      In addition, although the Hueber-Becker et al. (2004) study included a mass balance of applied radioactivity and also attempted to measure plasma kinetics of [14C]-PPD-equivalents, validated analytical methods of sufficient sensitivity were not available to determine whether the radioactivity present in plasma represented metabolites or whether the parent compound PPD was also present. In a subsequent investigation of urine samples collected in the study by Hueber-Becker et al. (2004), it was found that human urine contained mainly two PPD metabolites, i.e. N-mono- and N,N′-diacetyl-PPD (Nohynek et al., 2004b). The same metabolites were also detected after in vitro incubation of human epidermis or human hepatocytes with PPD (Nohynek et al., 2005).

    • Quantitative 3D molecular cutaneous absorption in human skin using label free nonlinear microscopy

      2015, Journal of Controlled Release
      Citation Excerpt :

      This is far from conventional skin bioavailability approaches used to quantify the penetration of active compounds into the skin following topical application. Classical approach uses ex vivo human skin mounted on Franz cell diffusion set-up [3] and any appropriate validated analytical methods to quantify chemicals (in SC, epidermis, dermis and subcutaneous receptor fluid) such as liquid scintillation counting with radio-labeled compounds [4] or liquid chromatography–tandem mass spectrometry (LC–MS/MS) [5]. Another method, which can be used in vivo, involves sequential adhesive tape stripping of SC and chromatographic analyses of compounds of interest after strip extraction to determine concentration profile across the skin outermost layers at specific time points [6].

    View all citing articles on Scopus

    Some of the data were presented at the EUROTOX 2003, i.e. Meuling, W., Hueber-Becker, F., Roza, L., Benech-Kieffer, F., Leclaire, J., Nohynek, G.J., 2003. Percutaneous absorption of para-phenylenediamine (PPD) after application of a [14C]-PPD-containing hair dye to human volunteers. Toxicology Letters 144 (Suppl. 1), 160.

    View full text