Keywords

1 Introduction

Cancer is one of the leading causes of death in the world. The International Agency for Research on Cancer (IARC) reported in 2020 that 10.0 million deaths worldwide were due to cancer and 63% of cancer deaths are reported to be from developing countries [1]. Radiation therapy (RT) is one of the well-established and effective methods for many cancer treatments. In particular, for brain tumor RT is used to treat approximately 50% of all patients and for cure or palliation it has been shown to be cost effective. However, on the down side, RT can often affect patient’s quality of life due to possible harmful late effects [2]. For instance, the radiation toxicity of normal tissue may lead to brain radiation necrosis (BRN) primarily includes inflammation and angiogenesis in which account for the breakdown of the blood–brain barrier (BBB), resulting in contrast-enhanced lesions and perilesional edema [2,3,4,5,6]. As severe toxicity in a some patients limits the doses that can be safely given to the majority, there is interest in developing a measurement method to assess individual’s tissue response during the treatment [7].

At present, the development of advanced RT techniques aim to deliver radiation to cancerous areas, while avoiding healthy tissues. These techniques include image-guided RT, intensity-modulated RT, stereotactic body RT, and proton beam therapy. Magnetic resonance (MR) guided RT combines a magnetic resonance imaging (MRI) unit with a RT unit. This allows real-time analysis of target volumes before and during irradiation, providing an instant feedback for the treatment planning. Another method in clinical practice for treatment planning is conventional photon RT which involves utilizing a computed tomography (CT) scan of the treatment position, commonly known as CT-based treatment planning. However, none of these can provide instant radiation response from the tissue.

Ionizing radiation impacts the cells by direct and indirect effects. The indirect effects are more significant interaction in RT use, when 2/3 of the oxygen radicals rise from the radiolysis of water. Radiation ionizes water molecules which generates oxygen radicals. However, technology to detect the free oxygen radicals during RT is lacking. Such technology could improve the effectiveness of the RT when measuring the concentration of oxygen radicals that correlate with the dose of irradiation.

Recently, we demonstrated that cerebral hemodynamics can be measured by functional near-infrared spectroscopy (fNIRS) during RT [8]. Furthermore, we have verified that the technology based on fiber optics can be used also combined with MR guided RT. This study further investigates the spectroscopic response to radiation at range of 650 nm to 1100 nm. Within this NIR range, the primary light-absorbing molecules in tissue are metal complex chromophores: haemoglobin, cytochrome and water, where the absorption spectra of deoxyhaemoglobin (HbR) is dominated from 650 to ~810 nm, oxyhaemoglobin (HbO) from ~810 nm to 950, and water between 950 to 1100 nm. The isobestic point wavelength at which oxy- and deoxyhaemoglobin species have the same molar absorptivity for HbO and HbR is around 810 nm and this isosbestic absorption spectra reflects total haemoglobin (HbT). In addition, cytochrome oxidase (Caa3) has a broad peak at 820–840 nm [9].

1.1 Effects of Irradiation When Measured Using Optical Techniques

Previous optics based studies on chemoradiotherapy-effects in patients with head and neck tumors indicated increased blood flow and changes in tissue oxygenation [10,11,12]. Jakubowski et al. showed that the greatest changes in hemoglobin concentrations, water and lipid content in breast tumor occurred within the first week of neoadjuvant chemo RT [13]. It seems that tissue deoxygenates in few days after the first fraction of the RT [14]. A recent human RT study by Myllylä et al. showed that there is also immediate effect in cerebral hemodynamics, particularly tissue oxygenation index (TOI) drops as a direct effect of irradiation but starts to increase again immediately after the irradiation [15]. A study by Darren et al. showed using diffuse optical spectroscopy (DOS) a correlation between oxygen saturation and erythema of skin in mice [16]. Similarly, Lee et al. showed in a mice study increased hemoglobin after irradiation. Still, most of the studies on effects of irradiation are conducted using mice models [17, 18].

In this paper, we measured absorption effects of irradiation in human skin at near infrared (NIR) light range of 650 nm to 1100 nm (in vivo) and in chicken thigh skin (ex vivo) to compare in vivo vs. ex vivo tissue effects of irradiation. Based on light propagation studies in tissue, when using fiber source-detector distance of 10 mm placed on human forehead, the spectrometer measurement volume does not reach the brain cortex [19] however, it covers few millimeters depth in human skin, including stratum corneum (0.02 mm), epidermis (0.07 mm to 0.13 mm), dermis (1 mm) and subcutaneous fat (1.2 mm) [20, 21]. Within the NIR region, we can generally consider four substances to dominate the absorption of light in skin tissue at ~2 mm depth: hemoglobin (oxy and deoxy), and water (H2O) and melanin in the range 620–720 nm [20] and their irradiation effects to NIR absorbance are discussed in this study.

2 Methodology

The spectroscopic measurements included patient (in vivo) measurements and store bought fresh chicken sample (ex vivo) measurements. All measurements were performed in Oulu University Hospital (Oulu, Finland) under the approval of the ethical committee of the North Ostrobothnia’s Hospital district (Finland) (No. 237/2018). Irradiation was performed with the Oulu University Hospital’s linear accelerators (Clinac iX and True beam, Varian medical systems). Figure 1 illustrates the measurement setup.

Fig. 1.
figure 1

Absorbance measurement setup used in the study. Optical fibres are guided to the RT chamber and attached at 1 cm source-detector distance to centre of the irradiated area (yellow) in the measured sample. (Color figure online)

Full size image

For absorbance measurement, QE pro-ocean optics spectrometer system was used with a tungsten-halogen broadband light source (HL-2000-HP, Ocean Optics) having emission spectral range of 350 nm to 2400 nm. The optical fibers for both light source and spectrometer detector had a refractive index of approximately 1.58 and an internal transmittance of 0.999 in the range of 600 to 1100 nm, fabricated by Schott and then customized for RT environment usage. Fibre tip with a diameter of 2.5 mm and a length of 10 mm was attached perpendicular to the patient’s skin and sample using 3D printed fibre probe holder. Both fibers had the same dimensions. Patients were measured in the treatment position under the medical linear accelerator and the fiber probes (tips touching skin) were attached to the face mask, which are typically used in RT (Fig. 2). Chicken leg sample was treated similarly. Prior to each data collection background and reference spectra measurements were performed using the standard procedures provided by the spectrometer manufacturer.

The study included 20 patients with total fractions of 80 measurements. Measured patients were undergoing whole brain RT (WBRT) for 20 Gy delivered in small doses called fractions in 4 Gy. Delivery of radiation dose implemented using forward-intensity modulated RT (FIMRT) and open field techniques. All the treatments were implemented with the dose rate of 600 monitor units (MU)/min. During the treatment session, the patient is in a supine position while two opposed radiation fields are targeted in the brain bilaterally, in 2 field setups of 1st irradiation of 25 s and 2nd irradiation of 24 s. For the tissue response comparison, chicken samples with skin were irradiated similarly, with same amount of treatment dose and setup as patient’s measurement were implemented.

Fig. 2.
figure 2

Patient measurement setup for spectroscopic measurement of forehead skin. Both optical fibers were attached to the face mask at source-detector distance of 10 mm. Patient provided consent for use of the image.

Full size image

3 Results

Figure 3 and 4 shows absorption effects of irradiation in the spectral range of 650 nm to 1100 nm when measured from patient’s (on the left) and corresponding effects of irradiation measured from chicken sample (on the right). Both figures include four spectra responses representing average response before radiation, during irradiations, and after irradiation. In patient measurement, on the left, can be seen that absorbance increases as function of the irradiation. In chicken measurement there are no obvious changes in absorbance spectra caused by the irradiation.

Fig. 3.
figure 3

Radiation effects on patient’s skin (left) vs chicken skin (right) measured in 650–1100 nm spectral range of absorbance. Measurements were done during 1st irradiation continuously before, during and after irradiation. Spectra includes average of 5 s before radiation, average of 20 s of irradiation in 5 s period and average of 5 s after radiation.

Full size image
Fig. 4.
figure 4

Standard deviation of measured spectra from patient on the left and chicken sample on the right.

Full size image

Figure 4 shows the change in absorbance due to irradiation. In the patient’s measurement, as can be seen in the range 650 nm–800 nm absorbance starts to increase most strongly while between 800 nm–950 nm increase is slightly lower. In the range 950 nm–1100 nm where water is dominating the absorbance increase is lowest.

Fig. 5.
figure 5

Change of absorbance due to irradiation in patient RT. The absorbance increases due to irradiation, strongly from 700 nm to 800 nm and less in range of 800 nm to 1100 nm. On the right is shown dose dependent increase of absorbance at spectral range from 650 nm to 1100 nm.

Full size image
Fig. 6.
figure 6

Change of absorbance after irradiation in 2 s steps patients (left) and chicken (right). As can be seen after irradiation the absorbance starts to decrease back to its normal initial level, whereas no changes in absorbance can be seen in chicken sample.

Full size image

4 Discussion

The WBRT is palliative treatment, which is given to patients with identifiable brain metastases. In this study, we measured NIR absorbance effects in forehead skin during the RT of brain tumor patients. Moreover, for a comparison study, chicken skin using the same measurement method was performed. Irradiation caused an increase in tissue absorbance in the whole measured spectral range of 650 nm–1100 nm in all patient measurements and the absorbance linearly correlates with the given dose (Fig. 5). In general, the spectral range is dominantly affected by HbR and HbO approximately at range of 650 nm–950, lipid at around 940 and water at around 980 nm. Significant changes from 650 nm to 800 nm indicate strong hemodynamic changes due to irradiation, which is interestingly visible only in patient’s measurements (Fig. 5) and (Fig. 4). This indicates that oxygenation of blood in patients is strongly affected by irradiation. After irradiation the spectral response relaxes with time (Fig. 6).

Also, The LINACs used in radiation oncology that produce radiation in pulsed microseconds-long bursts, generated by the accelerator waveguide, and effects of Cherenkov radiation could be possible. Figure 5 (left) from patient’s response shows similar to the radio dense tissue under 1 mm depth, effects of the Cherenkov emission spectrum at the surface [22], however it was not visible in chicken skin response.

The lack of absorbance effects in chicken sample (ex vivo) measurement may be due to the absence of blood circulation and/or due to the fact that it did not contain much blood. Therefore, obvious conclusion is that the absorbance effects of irradiation are mostly related to dynamical changes in tissue fluids particularly blood flow. This is supported also by our previous study [23] showing the temporal effects in cerebral haemodynamics during irradiations of WBRT measured by fNIRS.

Interestingly, irradiation causes changes also in the range between 950 nm to 1100 nm that is dominated by water absorbance, however, in chicken sample measurements there were in this spectral range neither visible effects of irradiation detected. The water absorbance effects in the patient measurements may be also related to blood circulation and particularly to the water bound with blood.

Prior research investigating tissue oxygenation and blood flow during RT indicates an association with increased blood flow and tissue oxygenation [10,11,12]. The efficacy of RT is known to be dependent on tumor oxygen status, as it enhances the free radicals which is responsible for DNA damage and cell death. Hence, the interaction of radiation in oxygenated tissue is more effective than the interaction in hypoxic cells. Usually tissue deoxygenates in few days after the first fraction of the radiation therapy [14]. However, there are studies reveal instances where some well-oxygenated tumors did not respond, while certain hypoxic tumors exhibited positive responses, possibly due to the dynamic changes during treatment in tumor oxygen status induced by radiation. Therefore, continuous monitoring of individual tumor hemodynamic status during therapy may provide valuable predictive information for treatment outcomes [11, 24]. In future, we aim to identify dose dependent absorbance response at selected NIR ranges during patient RT. The possibility to detect immediate tissue NIR absorbance changes during the RT can lead to different strategies in radiation dose planning and potentially improve the outcome of the RT when used in individual patient treatment planning.