**4. Experimental models**

#### **4.1 Animal models**

Due to the ethical problems in obtaining scalp biopsies from chemotherapy patients, little is known about the mechanisms of CIA in humans. In the last decade, important information

The basic principle of chemotherapy is to impair the mitotic and metabolic process of cancer cells. Unfortunately, certain normal cells and tissues with rapid metabolic and mitotic rates such as the hair follicles are also affected by the chemotherapy. Up to 90% of hair follicles undergo anagen, an active growth phase, at a given time. The rapid hair growth as well as the high blood flow rate around the hair bulb leading to the accumulation of drugs is a key predisposing factor for rapid and extensive alopecia (Batchelor, 2001). In humans, CIA usually begins approximately 2 to 4 weeks and is complete at 1 to 2 months after the initiation of chemotherapy (Batchelor, 2001). Hair might be easily depilated as early as 1 to 2 weeks after the treatment due to the weakening and breakage of hair shaft. The hair would fall out upon combing and in the bedding area. The degree of CIA depends on the type of chemotherapy, dosage regimen and route of administration. Almost all chemotherapies cause alopecia but with varying degrees of severity and frequency (Apisanthanarax and

> Amscarine Carmusine Carboplatin Gemcitabine Interleukin-2 Mercaptopurine Mitomycin Procarbazine Vinorelbine

A high-dose intravenous chemotherapy is commonly associated with more rapid and extensive alopecia. By contrast, oral therapy at lower doses on a weekly schedule tends to cause less alopecia even though the total dose may be large (Wilkes, 1996). Combination therapy consisting of two or more chemotherapeutic agents normally causes a higher incidence and more severe CIA compared to single agent therapy. Long-term chemotherapy

CIA is usually reversible with the hair regrowth generally occurring 3 to 6 months after the end of treatment. However, in most cases the new hair is grey or differs in color, representing the distortion of pigmentation process. Moreover, the new hair typically exhibits some changes in hair structure and texture, e.g. coarser, slow growth, and reduced density (Wang et al., 2006; Trueb, 2009). Permanent alopecia has been reported but rarely

Due to the ethical problems in obtaining scalp biopsies from chemotherapy patients, little is known about the mechanisms of CIA in humans. In the last decade, important information

Busulfan Chlorambucil Epirubicin Hydroxyurea Melphalan Methotrexate Mitoxantrone Teniposide

**3.2 Pathophysiology** 

Bleomycin Cytarabine Dacarbazine Docetaxel Etoposide Idarubicin Interferon-α Mechlorethamine

Paclitaxel Topotecan Vincristine

occurs (Betcheler, 2001).

**4.1 Animal models** 

**4. Experimental models** 

Duvic, 2003) as summarized in Table 1.

More common or severe Less common or severe

Cyclophosphamide

Cisplatin Dactinomycin Doxorubicin Fluorouracil Ifosfamide Irinotecan Nitroureas Thiotepa Vinblastine Vindesine

Table 1. Chemotherapeutic agents associated with alopecia.

may also result in the loss of pubic, axillary and facial hair.

about the CIA comes mostly from animal models. The commonly used animal models are neonatal rats and adult mice. However, there are some differences in human and rat/mouse hair growth pattern. In humans, the hair growth cycle occurs in a mosaic or asynchronous pattern, where the growth cycle of individual hair follicles is independent of neighbouring hair follicles. In contrast, rodent hair growth occurs in a wave pattern, beginning from the head and moving towards the tail. A group of hair follicles at a specific area are usually in the same stage of hair cycle. In general, only 10% of hair follicles in adult mice or rats are in the anagen phase as compared to 90% in adult humans. Some animals including guinea pigs and Angora rabbits exhibit a mosaic hair growth pattern but are not commonly used due to their insensitivity to CIA.

#### **4.1.1 Neonatal rat model**

The early model for CIA was established in newborn rats (Hussein et al., 1990; Hussein, 1993). Seven to eight-day old Sprague Dawley rats exhibit spontaneous anagen hair growth for about a week. In this model, administration of chemotherapeutic agents such as cytosine arabinoside, doxorubicin, cyclophosphamide, and etoposide induces alopecia one week after the treatment with the severity of CIA depending on the agents (Hussein et al., 1990; Hessein, 1991; Jimenez and Yunis, 1992).

The major advantage of neonatal rat model is the rapid and easily noticeable CIA due to progressive hair loss from the head and throughout the body in about 2 days. Several drawbacks and limitations of this animal model have been reported. For examples, the level of growth factors and cytokines and the hair follicle structure in neonatal rats differ substantially from those in mature animals, thus altering the response of hair follicles to treatment agents. Also, the lack of hair pigmentation in Sprague Dawley rats, which have a white fur, limits the study of drug effects on melanocytes. Indeed, some observations in newborn rats appear irrelevant to humans. For instance, the protective effect of topical application of 1,25-dihydroxyvitamin D3 on alopecia induced by cyclophosphamide was observed in neonatal rats but not in humans (Jimenez and Yunis, 1992; Hidalgo et al., 1999).

#### **4.1.2 Adult mouse model**

The adult black C57BL/6 mouse model for CIA was first developed in 1994 (Paus et al., 1994). In this mouse strain, the skin melanocytes are confined to hair follicles and the stage of hair growth is indicated by the skin color, i.e., pink during the telogen phase and black during the anagen phase. To mimic human hair scalp, depilation is performed to induce the mouse hair follicles at telogen phase to enter anagen phase, which is normally achieved in about 9 days. At around 16 days after the depilation, morphological signs of catagen are detectable. At day 20 after the depilation, all hair follicles are in the telogen phase. The CIA model was used to study the effect of cyclophosphamide (120-150 mg/kg, ip) on day 9 after the depilation (anagen phase). Cyclophosphamide was found to induce premature catagen development, dystrophic follicles, and complete alopecia in 6 days. In the past decade, progress in the understanding of hair follicle damage and pathogenesis of CIA has been obtained largely by using this model. On a cellular level, cyclophosphamide induces massive apoptosis of keratinocytes and melanocytes, although the precise mechanism of induction is largely unknown (Hendrix et al., 2005).

Chemotherapy-Induced Alopecia 59

Scalp torniques are the application of bands around the head to occlude the superficial blood flow to scalp, thus reducing the amount of drugs delivered to the hair follicles. The torniques range from 10 mmHg above systolic pressure to 300 mmHg around the scalp. These torniques are applied 5-10 minutes prior to or at the time of chemotherapy until up to 30 minutes after the drug administration (Cline, 1984). Although reports described mild to moderate prevention of CIA induced by vincristine, cyclophosphamide, and doxorubicin, this technique is no longer recommended due to patient discomfort (Wang et al., 2006).

Scalp cooling or hypothermia is the application of cold to the scalp using a device (cap) that is pre-cooled in a freezer or exchanges coolant with reservoir. A period of cooling lasts from 5 minutes prior to chemotherapy until an hour or more after the drug administration. Many studies have shown that the efficacy of scalp cooling can range from 0-90% (for review, *see* Grevelman and Breed, 2005). A recent study reported that scalp cooling helps reduce major CIA in patients receiving doxorubicin (60 mg/m2), docetaxel (80 mg/m2), or combination of 5-fluorouracil (600 mg/m2), epirubicin (60 mg/m2), and cyclophosphamide (600 mg/m2) for 6 to 9 cycles (Auvinen et al., 2010). The current hypotheses of the protective effect are: (i) cooling reduces blood flow to hair follicles by vasoconstriction, resulting in a decrease in the amount of drugs available for uptake; and (ii) cooling decreases cellular metabolism and drug uptake. Scalp cooling to 20°C was shown to reduce blood flow to 20% of normal flow; however, further decrease in the temperature (<18°C) did not result in further decrease in scalp blood flow (Janssen et al., 2007). Recent *in vitro* studies indicate the significant role of temperature on keratinocyte cell viability upon doxorubicin chemotherapy; however, there is no difference in cell survival between 10°C and 22°C (Janssen et al., 2008). Based on these findings, it appears that there is an optimal temperature for scalp cooling (~20°C), and that increasing the cooling will only result in patient discomfort. Other factors affecting the effectiveness of this method include drug regimen, application and duration of cooling, and

Scalp cooling is practically ineffective if the chemotherapeutic agent is administered as a continuous infusion over a prolonged period. Additionally, scalp cooling increases the risk of scalp metastasis, and is therefore contraindicated in patients with hematological malignancies and cutaneous T-cell lymphoma (Dean et al., 1979; Apisanthanarax and Duvic,

Stress protein response is one of the cellular protective mechanisms against various adverse conditions. Enhanced expression of stress proteins such as Hsp90, Hsp70, and Hsp25 has been observed in response to certain physical and chemical stresses, which has been linked to increased stress tolerance. Previous *in vitro* studies have shown that heat treatment and overexpression of stress response proteins, e.g., Hsp70 and Hsp27, could protect against the cytotoxic effects of anticancer drugs such as doxorubicin, cyclophosphamide, etoposide, and taxol (Kampinga, 1995; Jaattela et al., 1998; Kwak et al., 1998; Ito et al., 1999; Xia et al., 1999), leading to the investigation of the protective effect of stress protein activation on CIA in neonatal rats (Jimenez et al., 2008). In this study, heat was locally applied to the skin with a copper cylinder through which heated water was circulated. Conducting gel (Vaseline) was

**5.1 Physical prevention 5.1.1 Scalp torniques** 

**5.1.2 Scalp cooling** 

the cool conductivity (Betcheler, 2001).

2003).

**5.1.3 Heat treatment** 

#### **4.2 Culture models 4.2.1 Organ culture model**

Although animal models have provided valuable information regarding the pathogenesis of CIA, the animal physiological and pathological conditions do not necessarily mimic human conditions. The first human organ-culture hair follicle model for CIA was developed in 2007 (Bodo et al., 2007). In this model, mature anagen (anagen VI) human hair follicles are microdissected intact from the occipital scalp of healthy adults. The isolated hair follicles are maintained in William's E medium containing L-glutamine, hydrocortisone, and insulin. 4- Hydroperoxycyclophosphamide (4-HC), a key cyclophosphamide metabolite, was used to verify key *in vivo* characteristics of CIA since clinical and animal data of cyclophosphamideinduced CIA are well established. 4-HC was shown to induce melanin clumping and incontinence, reduce keratinocyte proliferation, induce keratinocyte and dermal papilla cell apoptosis, and induce premature catagen, resembling *in vivo* hair follicle dystrophy. Comparison of the test results in adult mouse model and organ culture model in response to cyclophosphamide is shown in Table 2. The human organ-culture hair follicle system could be used to study the pathogenesis and potential treatment of CIA, i.e., to identify key molecular targets and inhibitors of CIA.


Table 2. Comparison of adult mouse model and organ culture model.
