Radiation-induced malignancies from simultaneous integrated boost-VMAT for oropharyngeal cancer delivered with four different fractionation schemes

Abstract

Background

Late side-effects including carcinogenesis may appear following head-and-neck tumor irradiation. This study estimated the risk for developing second malignancies in heavily exposed tissues from simultaneous integrated boost (SIB) volumetric modulated arc therapy (VMAT) for oropharyngeal carcinoma.

Materials and methods

Sixteen oropharyngeal carcinoma patients needing bilateral neck irradiation were included. A three-level SIB-VMAT technique with 6-MV photons was applied. Four different treatment plans were created per patient. The first plan (VMAT_70Gy) was generated based on conventional fractionation delivering total doses to three targets of 70/63/56 Gy in 35 fractions. The VMAT_66Gy, VMAT_60Gy and VMAT_56.7Gy plans gave target doses of 66/60/54 Gy in 30 fractions, 60/56/50 Gy in 25 fractions and 56.7/52.5/46.2 Gy in 21 fractions, respectively. Differential dose-volume-histograms and a non-linear model were employed to estimate the lifetime attributable risk (LAR) for cancer induction in critical tissues adjacent to the treatment volume. The model accounted for dose fractionation and tissue-dependent proliferation ability.

Results

For 50-year-old males and females undergoing VMAT_70Gy, the average LAR for cancer induction in thyroid, salivary glands, mouth/pharynx and brain/CNS was up to 0.25%, 0.26%, 1.32% and 0.31%, respectively. The risk range for patients aged 60 years was 0.09–0.74%. These LARs were reduced up to 4.3%, 8.0% and 13.1% with VMAT_66Gy, VMAT_60Gy and VMAT_56.7Gy plans_, respectively. The three moderately hypofractionated radiotherapy techniques led to cancer risks of 0.08–1.27%.

Conclusions

The risks for radiation-induced malignancies following SIB-VMAT for oropharyngeal cancer were found to be comparable with the baseline risks for general population. Hypofractionated treatment restricted the probability for carcinogenesis.

Introduction

The cancer of oropharynx is the fourth most common head and neck malignancy with an incidence of 0.5% globally [1]. Bray et al. [1] reported 106316 new cases of oropharyngeal cancer in 2022 based on statistics derived from 185 countries. The incidence rate of this type of cancer has increased considerably in many countries worldwide over the last decades [2]. This elevated incidence was also observed in middle-aged people from 40 to 59 years [2]. The five-year survival for European males and females with oropharyngeal cancer diagnosed between 2016 and 2020 was 66.3% and 67.4%, respectively [3]. The above survival rate was increased by about 30% units through a 50-year period [3].

Radiotherapy alone or concurrent chemoradiation is employed for the definitive or postoperative management of oropharyngeal cancer [4]. Intensity-modulated radiation therapy (IMRT) or the more advanced volumetric modulated arc therapy (VMAT) are recommended for treating this malignant disease [4]. Both techniques reduce the dose to normal structures compared to conventional two- and three-dimensional radiotherapy and, subsequently, late toxicity such as xerostomia [4]. VMAT is preferable to IMRT because of the faster delivery time and the reduced number of monitor units per fraction [5, 6]. Lymph-node involvement is presented in about 70% of patients with oropharyngeal cancer [7]. Differential radiation dose delivery to multiple target areas is often required. This may be accomplished by means of the simultaneous integrated boost (SIB) which assigns different dose levels to different target areas [5, 8].

Second cancer induction constitutes a serious late side-effect attributable to head and neck cancer treatment including radiotherapy [9]. This effect is of great importance to patients with oropharyngeal malignancies because of their usual good prognosis and high expected lifespan. Theoretical second cancer risk assessments attributable to radiotherapy for head and neck carcinomas have been reported by previous studies [10–14]. The risk estimates were derived from treatments with standard daily dose fractions of 2 Gy [11, 12, 14]. Hypofractionated irradiation is an acceptable treatment option for cancer in the oropharynx often leading to improved locoregional control [15]. Several schedules for hypofractionation have been implemented in patients with this type of tumor [16–20]. The impact of the hypofrationated treatment on the late side effects is of particular concern for irradiated cancer survivors [4].

The purpose of this study was to a) estimate the risk for radiation-induced malignancies related to SIB-VMAT for oropharyngeal cancer, and, b) investigate the effect of using several hypofractionated schedules on the resultant risk of radiation carcinogenesis.

Materials and methods

Patients

This retrospective study included sixteen patients (10 males) who were subjected to external-beam radiotherapy for oropharyngeal carcinoma in our department. The disease stage of the participants was III–IVa. Chemotherapy was given concurrently to radiation therapy. Chemotherapy protocols and dosing schedules were individualized at the discretion of the medical oncology team, involving either cisplatin or carboplatine — paclitaxel. Bilateral neck irradiation was delivered in all participants. Moreover, a three-dose level prescription was applied for the selected patients. The median age of the participants was 57.1 ± 7.3 years with a range of 46–70 years. This planning study received an institutional review board approval.

Planning CT and contouring

A treatment planning computed tomography (CT) was performed on a Revolution GSI scanner (General Electric Medical Systems, WI, USA) before radiotherapy. All patients were immobilized in a supine treatment position using thermoplastic masks. Contiguous 3-mm-thick CT scans were acquired and transferred to a Monaco workstation (Monaco, version 5.11.03, Elekta AB, Sweden) for contouring and treatment planning. A radiation oncologist with more than 20 years of experience in the head and neck cancer treatment was responsible for the manual delineation of all structures of interest. The primary gross tumor volume (GTV) was delineated using CT, positron emission tomography (PET)-CT and clinical examination to ensure involving areas. Nodes were considered involved if they were larger than 1 cm, had necrosis or showed high fluorodeoxyglucose (FDG) uptake on PET [maximal standardized uptake value (SUV_max) > 3.0]. The GTV included gross tumor, with clinical target volume (CTV) expansions 5 mm around the GTV accounting for subclinical disease spread. The planning target volume (PTV) was defined with a uniform expansion of 5 mm from CTV. The PTV1 was the high-risk target receiving the highest therapeutic dose and it encompassed the primary tumor and grossly involved lymph nodes. The PTV2 included high-risk subclinical regions adjacent to the tumor and involved lymph nodes. An intermediate therapeutic dose was administered to PTV2. The PTV3 covered low-risk elective nodal regions at risk of microscopic disease and received the lowest radiation dose. Contours of the following normal organs were generated on the CT scans: left and right parotids, left and right submandibular glands, spinal cord, brain stem, oral cavity, pharynx, left and right inner ears, mandible, left and right optic nerves, optic chiasm and esophagus.

VMAT planning

The VMAT planning of the patients with cancer in the oropharynx involved double arcs in clockwise and counter-clockwise directions. A full gantry rotation of 360° was used in both arcs. A SIB technique with three prescription dose levels to the PTV1, PTV2 and PTV3 was applied. Four different plans were created for each study participant on the basis of four different fractionation schemes which were previously reported [16–19]. The first scheme (VMAT_70Gy) corresponded to the conventional fractionation delivering a total dose of 70 Gy to the PTV1 at 2.0 Gy/fraction for 35 days. The PTV2 and PTV3 doses were 63 and 56 Gy. The three other schemes, denoted as VMAT_66Gy, VMAT_60Gy and VMAT_56.7Gy, involved moderately hypofractionated treatment. The VMAT_66Gy involved 30 fractions and gave to PTV1, PTV2 and PTV3 doses of 66, 60 and 54 Gy. The corresponding target doses with VMAT_60Gy were 60/56/50 Gy in 25 fractions whereas those of VMAT_56.7Gy were 56.7/52.5/46.2 Gy delivered in 21 daily fractions. The fraction dose to the high-risk target from VMAT_66Gy, VMAT_60Gy and VMAT_56.7Gy was 2.2, 2.4 and 2.7 Gy, respectively. The 95% of the volume of each PTV had to receive at least 95% of the prescribed tumor dose. Sixty-four VMAT plans in total were generated by a senior medical physicist.

Organ equivalent dose calculations

Second cancer risk assessments were made with the mechanistic non-linear model introduced by Schneider et al. [21]. The above model, based on the organ equivalent dose (OED) concept, has been broadly applied to assess the risk for sarcoma or carcinoma induction after radiotherapy of malignant or benign conditions [12, 22–25]. Model-based second cancer risk assessments can be obtained for many normal organs and tissues of the human body which receive an inhomogeneous dose distribution [21]. This model provides data for estimating the second cancer risk to salivary glands, thyroid gland, mouth/pharynx and brain/central nervous system (CNS). The dose to the above regions of interest was inhomogeneously distributed in the generated VMAT plans. Parts of the above regions were always exposed to high therapeutic doses of more than 2.5 Gy. The salivary glands consisted of the contours of both parotids and both submandibular glands. The brain/CNS was taken as a single structure including the brain, brainstem and spinal cord. The union of mouth and pharynx was taken as one critical structure for cancer risk estimation.

From each SIB-VMAT plan, differential dose volume histograms (DVHs) of the thyroid, salivary glands, brain/CNS and mouth/pharynx were extracted. These DVHs were used to find the OED as follows:

$$OED=\frac{1}{V}\sum _i{V}_{D_i}\frac{\text{exp}(-a_i^{\prime }{D}_i)}{a_i^{\prime }R}\left[1-2R+R^2\hspace{0.17em}\text{exp}(a_i^{\prime }{D}_i)-{(1-R)}^2\text{exp\hspace{0.17em}}\left(-\frac{R}{1-R}{a}_i^{\prime }{D}_i\right)\right]$$ (1)

where V is the total volume of the structure under examination as derived from CT scans, V_Di is the volume of the structure absorbing a D_i dose and R is the parameter showing the tissue-dependent repopulation ability. The a′_i, known as cell-kill parameter, was computed with the following formula:

$$a_i^{\prime }=a+\beta \frac{1}{n}{D}_i$$ (2)

where n is the number of dose fractions of the entire treatment course and α, β are parameters obtained by the linear-quadratic model. The parameter β was determined by assuming an α/β ratio equal to 3 for all structures of interest in accordance with previously reported data [21]. For the VMAT_70Gy, VMAT_66Gy, VMAT_60Gy and VMAT_56.7Gy, the average OED, denoted as OED_av, for thyroid, salivary glands, mouth/pharynx and brain/CNS was found.

Lifetime cancer risk assessments

The tissue-specific OED_av from each VMAT-SIB plan category was then employed for estimating the EAR for developing a radiation-induced second malignancy to thyroid, salivary glands, mouth/pharynx and brain/CNS as follows:

$$EAR={\beta }_{EAR}\hspace{0.17em}OE{D}_{av}\hspace{0.17em}\text{exp}\left[{\gamma }_e(ag{e}_e-30)+{\gamma }_a\hspace{0.17em}ln\left(\frac{ag{e}_a}{70}\right)\right]$$ (3)

where β_EAR is the slope of the linear-no-threshold model, age_a is the age of the patient during radiotherapy, age_a is the attained age of the patient and y_e, y_a, are the age-modifying factors. The β_EAR introduced in equation (3) was transferred to a population in Western countries as previously described [21]. The tissue-dependent model parameters for OED and EAR calculations were taken from the literature [12, 21] and they are summarized in Table 1.

Table 1.

Model parameters for the organ equivalent dose and cancer risk estimations

Structure	βEAR (105 PY Gy)−1	γe	γa	a(Gy−1)	β	R
Thyroid	0.40	−0.046	0.60	0.087	0.029	0.23
Mouth/pharynx	0.73	−0.024	2.38	0.043	0.014	0.97
Salivary glands	0.73	−0.024	2.38	0.087	0.029	0.23
Brain/CNS	0.70	−0.024	2.38	0.018	0.006	0.93

CNS — central nervous system

The LAR for second cancer induction was estimated by assuming a patient of 50 and 60 years old during radiotherapy. A latency of 5 years for developing solid tumors was considered for the above assessments [26] and it was added to the above age_e. The EARs were estimated for all ages up to a final attained age of 80 years and they were summed to give the LAR as follows:

$$LAR=\sum _{ag{e}_e+5}^{80}EAR(ag{e}_e,ag{e}_a)P(ag{e}_e,ag{e}_a)$$ (4)

where P(age_e, age_a) is the surviving probability derived from the United States life tables [27]. Different LARs were calculated for 50- and 60-year-old male and female patients by using the corresponding surviving probabilities. A specially designed software tool facilitated the OED and cancer risk calculations [28].

Results

The ratio of each OED from the VMAT_66Gy, VMAT_60Gy and VMAT_56.7Gy plans to the OED from VMAT_70Gy shows the relative patient-specific second cancer risk from hypofractionated treatments in respect to conventionally fractionated radiotherapy for oropharyngeal cancer. These ratios for the four structures of interest are illustrated in Figure 1. The risk for developing secondary malignancies to thyroid, mouth/pharynx, salivary glands and brain/CNS due to VMAT_66Gy was decreased on average by 4.3 ± 0.7%, 3.8 ± 0.5%, 3.1 ± 0.8% and 3.0 ± 1.6% in respect to VMAT_70Gy, respectively. For VMAT_60Gy, the corresponding cancer risk reductions were 8.0 ± 0.6%, 7.9 ± 0.5%, 5.9 ± 1.1% and 8.0 ± 1.5% whereas those associated with VMAT_56.7Gy were 12.5 ± 0.6%, 12.6 ± 0.8%, 9.1 ± 1.2% and 13.1 ± 1.7%.

Figure 1.

Patient-specific second cancer risk to: A. Thyroid gland; B. Mouth/pharynx; C. Salivary glands; D. Brain/central nervous system (CNS) from hypofractionated volumetric modulated arc therapy (VMAT): VMAT_66Gy, VMAT6_0Gy and VMAT_56.7Gy relative to the conventionally fractionated VMAT_70Gy for oropharyngeal carcinoma

The average OED associated with VMAT for oropharyngeal cancer delivered with the different fractionation schemes are presented in Table 2. These OED values varied from 2.6 to 14.2 Gy by the structure under examination and the applied fractionation schedule. The age-specific average lifetime second cancer risks are presented in Figure 2. The LAR for radiation-induced second malignancies to thyroid, mouth/pharynx, salivary glands and brain/CNS in 50-year-old females undergoing VMAT_70Gy was 0.25%, 1.32%, 0.26%, and 0.31%, respectively. The corresponding risks for males irradiated at the above age were 0.23%, 1.21%, 0.24% and 0.28%. The probabilities of carcinogenesis for both male and female patients aged 60 years at the time of conventionally fractionated radiotherapy were 0.09% to 0.74%. The range of the second cancer risk estimates related to moderately hypofractionated VMAT_66Gy, VMAT_60Gy and VMAT_56.7Gy in male and female patients at both examined ages was 0.08–1.27%.

Table 2.

Average organ equivalent dose (OED) ± one standard deviation (SD) related to volumetric modulated arc therapy (VMAT): VMAT_70Gy, VMAT_66Gy, VMAT_60Gy and VMAT_56.7Gy for oropharyngeal cancer

Structure	Average OED ± SD [Gy]
	VMAT70Gy	VMAT66Gy	VMAT60Gy	VMAT56.7Gy
Thyroid	7.23 ± 0.09	6.92 ± 0.09	6.65 ± 0.10	6.33 ± 0.10
Mouth/pharynx	14.15 ± 0.18	13.61 ± 0.21	13.03 ± 0.20	12.36 ± 0.24
Salivary glands	2.83 ± 0.21	2.75 ± 0.19	2.67 ± 0.20	2.58 ± 0.20
Brain/CNS	3.44 ± 1.04	3.33 ± 0.99	3.16 ± 0.95	2.99 ± 0.90

CNS — central nervous system

Figure 2.

Average lifetime attributable risk (LAR) for second cancer induction to: A. Thyroid gland; B. Mouth/pharynx; C. Salivary glands; D. Brain/central nervous system (CNS) in 50- and 60-year-old male (denoted as M) and female (denoted as F) patients undergoing volumetric modulated arc therapy (VMAT): VMAT_70Gy, VMAT6_6Gy, VMAT_60Gy and VMAT_56.7Gy for oropharyngeal cancer

Discussion

Hypofractionated radiotherapy may be considered for head and neck cancer patients because it leads to similar outcomes as conventional fractionation and it simultaneously shortens the duration of the treatment [29]. The impact of the applied fractionation on the probability of carcinogenesis in patients irradiated for a head and neck cancer has not been investigated. This study provided second cancer risk assessments from radiotherapy for oropharyngeal carcinoma delivered with a commonly applied fractionation schedule involving 70 Gy at 2.0 Gy per fraction [4, 30]. Risks of carcinogenesis were also estimated with three different hypofractionation regimens. All patients were planned with the three-level VMAT-SIB technique. The use of VMAT_66Gy giving a total PTV1 dose of 66 Gy in 30 fractions slightly decreased the probability for radiation-induced second malignancies up to 4.3% compared with the risk from standard fractionated VMAT_70Gy. More pronounced reductions up to 8.0% were found with VMAT_60Gy involving 25 fractions giving 60 Gy to the gross disease. This reduction reached 13.1% when VMAT delivered 56.7 Gy to the PTV1 in only 21 fractions.

The average cancer risk was estimated to four different heavily irradiated structures surrounding the target for typical 50- and 60-year-old male and female patients treated for oropharyngeal carcinoma. The highest probability of carcinogenesis was always observed for the mouth/pharynx. This probability from VMAT_70Gy reached 1.21% and 1.32% for 50-year-old males and females, respectively. The risk for patients aged 60 years was up to 0.74%. The risks for cancer induction to thyroid, salivary glands and brain/CNS due to VMAT_70Gy were 0.09–0.31%. The second cancer risks from treatment delivered with the three hypofractionation schemes were 0.08–1.27%.

The radiotherapy-induced second cancer risks need to be seen in conjunction with the natural frequency of these adverse effects for unexposed persons. The Surveillance, Epidemiology, and End Results (SEER) cancer statistics review reported the age- and gender-specific lifetime baseline risk (LBR) for developing thyroid, brain/CNS and mouth/pharynx carcinomas in the US population [31]. For 50- and 60-year-old people of both sexes, the LBR for the presence of brain/CNS cancer and thyroid cancer was 0.37–0.55% and 0.38–1.05%, respectively [31]. These values are increased compared to the respective risks of 0.08–0.31% related to VMAT for oropahryngeal cancer with all fractionation schemes. The radiotherapy-induced risk for the development of mouth/pharynx carcinoma in males aged 50- and 60-years was estimated to be 1.21% and 0.68%, respectively. The corresponding LBRs were higher and equal to 1.62% and 1.37% [31]. For the mouth/pharynx cancer, the above risk relation was not observed in females irradiated for oropharyngeal carcinoma. The risk for the appearance of this malignancy following radiotherapy was 0.74–1.31% depending upon the female patient’s age while the LBR was lower and equal to 0.57–0.66%. The LAR assessments were found to be in the same order of magnitude as the LBR corresponding to unexposed people. Regarding the salivary gland cancer, Mousavi et al. [32] recently reported that the cumulative risk for the presence of this malignant disease by age 70 is 0.12% worldwide. More elevated model-based LAR estimates reaching 0.26% were found after the administration of VMAT in patients with oropharyngeal cancer.

The lack of data about target doses and PTV definition or the expression of the cancer risks with quantities other than the tissue-specific lifetime risk from the entire treatment course makes the direct comparison of our results inconsistent with those previously published [10, 14]. Two reports used different models than that employed in our work [11, 13]. In the phantom study of Rehman et al. [13], the average organ dose was found to assess the cancer risk in critical organs exposed to high therapeutic doses. The use of this dosimetric quantity for cancer risk estimates with the linear approach may be applied in the low dose region but not recommended for high doses exceeding 2.5 Gy where cell sterilization effects exist [33, 34]. A linear-exponential model was also employed by Jain et al. [11] for estimating the probability of carcinogenesis in patients treated for cancer in the oropharynx. This model relies on the risk fall-off at high doses and ignores the tissue-repair effects in fractionated treatment. The cancer risk estimates presented here were derived from a model accounting for tumor dose fractionation and cell proliferation. Johnson et al. [12] combined this model with VMAT planning data of five patients receiving 70 Gy in 35 fractions for oropharyngeal cancer. They found a cancer risk to mouth/pharynx of 1.06%, which is similar to that presented here. The cancer risk to brainstem was estimated to be 0.48%. This value is higher than the maximum probability of carcinogenesis of 0.31% for a structure encompassing the brain, brainstem and spinal cord. Beal et al. [35] studied 3025 patients with salivary gland tumors. The salivary gland tumors of eighteen patients were considered as second cancers following head and neck irradiation mostly for Hodgkin’s disease. The median age during the initial radiotherapy course was only 22 years, whereas our risks were estimated for older patients of 50 and 60 years old irradiated for cancer of the oropharynx.

The results of this study are limited by the inaccuracies of the parameters of the non-linear model employed for estimating the second cancer risks. The model parameters were determined based on data taken from A-bomb survivors and Hodgkin’s patients subjected to radiotherapy [21]. These patients having hematologic malignancies were followed for a mean period of 8 years and received lower tumor doses than those given to the participants of this work with oropharyngeal cancer. It is well known that daily imaging with cone beam CT is needed for the accurate setup of head and neck cancer patients in the linear accelerator prior to treatment [36, 37]. The cancer risks presented here solely refer to the administration of radiation therapy. Further work is required to find the alteration of the radiotherapy-induced cancer risk by the imaging procedures involving ionizing radiation.

Conclusions

This study estimated the risk for developing thyroid, brain/CNS, salivary glands and mouth/pharynx second malignancies following VMAT-SIB for oropharyngeal cancer delivered with four different fractionation schemes. The use of the standard fractionation schedule involving 70 Gy to the high-risk volume in 35 fractions led to second cancer risks of 0.09–1.32% depending upon the tissue-at-risk, patient’s age and gender. Hypofractionated treatments given to gross disease doses of 56.7 to 66 Gy in daily fractions of 2.2–2.7 Gy led to smaller risks of 0.08–1.27%. The more pronounced risk reductions of 9.1–13.1% were observed with VMAT_56.7Gy. The risk for radiation-induced malignancies in the surrounding heavily irradiated sites were comparable with the natural frequency of these effects for general population.

Data availability statement

Data may available upon request to the corresponding author.