Modelling physiological and biochemical aspects of scalp cooling

Chemotherapy induced hair loss is a feared side effect of cancer treatment. Scalp cooling during the administration of cytotoxic drugs can reduce this hair loss. Cooling can be achieved by means of a cap, that is pre–cooled in a freezer or that exchanges coolant with a reservoir. The current hypothesis for the hair preservative effect of scalp cooling is that cooling of the scalp skin reduces blood flow (perfusion) and chemical reaction rates. Reduced perfusion leads to less cytotoxic drugs available for uptake, while the reduced temperature decreases uptake of and damage by chemotherapy. Altogether, less damage is done to the hair cells, and the hair is preserved. However, the effect of scalp cooling varies strongly. A systematic evaluation of the current hypothesis is necessary for a better understanding of the various important parameters of scalp cooling. In our study, we wanted to quantify the contribution of the putative mechanisms by which scalp cooling prevents hair loss, and with this investigate possible options for improving effectiveness of current day scalp cooling protocols. A computational model has been developed based on the current hypothesis of the mechanisms of scalp cooling. The full computational model consists of two sub–models that describe heat transfer in the human head and transport of doxorubicin (a specific chemotherapy agent) in the human body. Experiments have validated and improved the different computational models. The heat transfer model uses the Pennes’ equation to describe heat transfer in the human head. Parameter studies with the heat transfer model show that key parameters that determine the actual skin temperature during scalp cooling are the size of both the sub–cutaneous fat–layer and the hair–layer. We measured the reduction in perfusion due to cooling of the head, for different subjects and for different sites. A scalp cooling system was used to slowly cool nine subjects for 90 minutes. Afterwards, subjects were re-warmed for 60 minutes to investigate hysteresis effects. Skin temperature and perfusion were monitored by thermocouples and laser Doppler perfusion probes, respectively. Results show that intra–individual variability is small compared to inter–individual variability. During cooling, perfusion was gradually reduced to a lowest level of 20% § 10% at a skin temperature decrease of 20oC. A (physiologically based) pharmacokinetic model was created to describe doxorubicin distribution in the human body. It consists of eight compartments representing individual organs. Transport, clearance and metabolism between these compartments is described using mass balance ordinary differential equations. Parameter studies show that key parameters in the model are body mass, cardiac output, and both blood flow to and volume of the different tissues. The effect of local drug concentration and local tissue temperature on hair cell damage was investigated using in vitro experiments on keratinocytes. Cells were exposed for 4 hours to a wide range of doxorubicin concentrations. During exposure, cells were kept at different temperatures. Cell viability was determined after 3 days using a modified MTT viability test. Control samples were used to establish a concentration-viability curve. Results show that cell survival is significantly higher in cooled cells (T <22oC) than in non–cooled cells (T = 37oC), but no significant differences are visible between T = 10oC and T = 22oC. A preliminary study showed no significant differences in survival between T = 26oC and T <22oC. Information on variability was used to develop a population based computational model for scalp cooling. Results of the in vitro cell experiments were fitted to an equation for viability, with temperature and concentration as independent variables. Simulations were compared with studies from literature on the success of scalp cooling. With this, a critical viability could be established, Scrit = 0:88, defined as the viability value above which hair loss is prevented. Current day scalp cooling protocols were evaluated and an optimized scalp cooling protocol was developed. This protocol may be used for doxorubicin doses up to 60–70 mg, and aims to lower skin temperature to 17–18oC. To this end, a cap temperature of Tcap = ¡10oC, infusion time of 2 hours and a cooling time equal to infusion time plus one hour is used. Furthermore, a haircut is advised to decrease the thermal resistance between head and cap. Our scalp cooling model shows that with this protocol, effectiveness of scalp cooling was highest.