Abstract
Proton radiotherapy may be a compelling technical option for the treatment of breast cancer due to its unique physical property known as the “Bragg peak.” This feature offers distinct advantages, promising superior dose conformity within the tumor area and reduced radiation exposure to surrounding healthy tissues, enhancing the potential for better treatment outcomes. However, proton therapy is accompanied by inherent challenges, primarily higher costs and limited accessibility when compared to well-developed photon irradiation. Thus, in clinical practice, it is important for radiation oncologists to carefully select patients before recommendation of proton therapy to ensure the transformation of dosimetric benefits into tangible clinical benefits. Yet, the optimal indications for proton therapy in breast cancer patients remain uncertain. While there is no widely recognized methodology for patient selection, numerous attempts have been made in this direction. In this review, we intended to present an inspiring summarization and discussion about the current practices and exploration on the approaches of this treatment decision-making process in terms of treatment-related side-effects, tumor control, and cost-efficiency, including the normal tissue complication probability (NTCP) model, the tumor control probability (TCP) model, genomic biomarkers, cost-effectiveness analyses (CEAs), and so on. Additionally, we conducted an evaluation of the eligibility criteria in ongoing randomized controlled trials and analyzed their reference value in patient selection. We evaluated the pros and cons of various potential patient selection approaches and proposed possible directions for further optimization and exploration. In summary, while proton therapy holds significant promise in breast cancer treatment, its integration into clinical practice calls for a thoughtful, evidence-driven strategy. By continuously refining the patient selection criteria, we can harness the full potential of proton radiotherapy while ensuring maximum benefit for breast cancer patients.
Introduction
It is estimated to be as many as 297 790 women diagnosed as breast cancer (BC) patients in the United States in 2023, accounting for 31% of female cancers. 1 Meanwhile there were 4 055 770 female BC survivors in the United States on January 1, 2022. 2 According to these statistics, BC continues to be the leading cancer in females.
While principle and standards of photon radiotherapy has been well established as an integrated part of multidisciplinary treatment of BC, proton therapy (PT) has become an interesting alternative due to its unique physical dose distribution. However, facing the high cost and low accessibility of PT, indications for protons requires careful consideration as photon therapy has already provided an effective and safe treatment for the majority of BC patients. In this review, we summarized and discussed about the current practices and prospects on the topic of patient selection for PT in BC patients in terms of organs at risk (OAR) protection, loco-regional tumor control and cost-effectiveness, basing on its physical and biological properties and available clinical outcomes.
Physical and Biological Characteristics of Proton Therapy
PT offers an advantageous dose distribution over conventional photon therapy. 3 In photon therapy, x-rays release energy the entire path as it travels through the body. The delivered dose reached maximum at depths ranging from 0.5 cm to 3.5 cm. After the x-rays exit the tumor, it continues to deposit radiation in the normal tissues beyond the target. Unlike photons, the deposited dose by a proton beam increases gradually with depth and then suddenly rises to a peak, a phenomenon known as Bragg Peak. After the Bragg peak, the radiation dose falls rapidly to zero, yielding no exit dose to the adjacent normal tissues. 4
Additionally, protons are believed to cause different degrees of radiation injury at cellular level. In fact, most treatment planning systems apply an average relative biological effectiveness (RBE) of 1.1 for PT. 5 This value, however, is subject to vary at different points along the spread-out Bragg peak (SOBP). 6 The different biological effectiveness may contribute to the induction of enhanced DNA damage and a decreased repair, leading to additional double-strand breaks (DSBs) and greater probability of genome instability.7,8
Clinical Experiences of Proton Therapy for Breast Cancer
The earliest clinical result of PT for BC was reported in 2006 by Kozak et al, 9 Over the years, the number of clinical trials about PT for BC keeps growing. Tables 1 and 2 provide a summary of 16 latest published periodic or final reports of recent studies, including 954 patients underwent PT in total.
Treatment details about proton therapy on breast cancer patients.
aThe BED values are calculated using an α/β value of 4.0 estimated from the study of Owen et al, 26 if sufficient information is available. The boosts are not considered in the calculation for lack of relevant information.
bThe study of Naoum et al enrolled patients with and without radiotherapy, only information and results related to radiotherapy are listed.
cThe study NCT03391388 is a non-randomized three-armed trial, while its latest published outcome only analyzed the data from the proton arm.
dOnly the latest reports of studies based on overlapping populations are listed. The studies of Luo et al and Galland-Girodet et al had overlapping treatment populations with Cuaron et al 27 and Kozak et al, 9 respectively.
RBE = relative biological effectiveness; BED = biologically effective dose; SBBC = synchronous bilateral breast cancer; IBC = inflammatory breast cancer; ALND = axillary lymph node dissection; SNLB = sentinel lymph node biopsy; PSPB = passive scattering proton beam; PBS = pencil beam scanning; 3D-CPT = 3D-conformal proton therapy; IMPT = intensity modulated proton therapy; US = uniform scanning; APBI = accelerated partial breast irradiation; PT = proton therapy; NR = not reported.
Outcomes of proton therapy on breast cancer patients.
RT = radiation therapy; QoL = long-term quality of life; AE = adverse effect; SH = skin hyperpigmentation; RD = radiation dermatitis; RP = radiation pneumonitis; DFS = disease-free survival; MFS = metastasis-free survival; OS = overall survival; NR = not reported.
According to the limited trial reports which analyzed the efficacy of PT, Garda et al 12 found 2 local failures out of 11 synchronous bilateral BC patients, while the local-recurrence rate of other trials roughly ranged from 0% to 11% by the end of follow-up. Compared to outcomes of previous studies on photon therapy,28–30 the influence of PT in treatment efficacy seems inconspicuous. Galland-girodet et al also reported a similar local tumor control between proton and photon therapies in their non-randomized comparative study. 23
As to side effects of PT, records of cardiopulmonary toxicities were infrequent among those trials, except for 1 re-irradiation study, only 3 of them reported 1 case of grade ≥2 radiation pneumonitis out of 69, 42, and 18 patients, respectively.16,18,22 Meanwhile, skin toxicity is one of the most observed side-effects. Notably almost every trial listed in Table 2 reported skin or cosmetic toxicities, the rate of grade ≥ 2 skin toxicities after PT reached 100% maximally. 22 The retrospective comparative study conducted by DeCesaris et al found that PT led to significantly higher rates of grade ≥ 2 radiation dermatitis (69.2% vs 29.8%, P = .002) than photons. 19 Similarly, Galland-girodet et al 23 discovered a significantly increased risk of multiple long-term skin toxicities for PT compared to photons at a follow-up time of 7 years. Besides, given that the distal edge of SOBP of proton beams when treating BC, where the RBE peaks with higher linear energy transfer (LET),6,31 ends close to the ribs, there's a theoretical worry about rib fracture. 11 According to Table 2, 7 of the 16 clinical trials reported rib fracture cases, the rates of rib fracture ranged from 0% to 21%, despite an absence of rib fracture records in recent studies for photons.28–30 In addition, capsular contracture and subsequent reconstruction failure could be another concern for PT in BC patients with immediate reconstruction after mastectomy.10,14
Nevertheless, since the clinical trials listed vary in dose fractionation, target volume, radiation technique, patient number, and follow-up period (Table 1), it would be premature to draw rough summarization of these outcomes and compare them to previous literature for photons. The safety and efficacy of PT and the difference between protons and photons should be further verified by randomized controlled trials (RCTs) with a larger population, which remain absent. According to ClinicalTrials.gov (accessed on 20 June 2023), 32 there are 11 registered ongoing prospective trials involving PT for BC treatment (Table 3) with no published results yet, of which 5 are designed as randomized trials comparing photon and PT. The outcomes of these studies are worth expecting.
Active or recruiting prospective studies about proton therapy on breast cancer patients.
The study NCT03391388 is a non-randomized three-armed trial, while its latest published outcome only analyzed the data from the proton arm.
The study population of NCT04361240 is consist of newly enrolled trial participants of NCT02603341.
RT = radiation therapy; APBI = accelerated partial breast irradiation; LVEF = left ventricular ejection fraction; RVFAC = right ventricular fractional area change; PIGF = placental growth factor; GDF = growth differentiation factor; GLS = global longitudinal strain; AE = adverse event.
Potential Approaches for Patient Selection
To justify the introduction of new radiation techniques, such as PT, into clinical practice, it is essential to ensure the tangible additional benefits in normal tissue sparing, treatment efficiency or costs. For now, in the absence of results from RCTs comparing PT to photon radiotherapy, the challenge lies in exploring alternative approaches that can aid in identifying patients who will derive the maximum benefit from PT.
OAR Protection
Side effects of treatments are closely related to the quality of life for cancer survivors, while the likelihood of radiation-induced toxicities depends heavily on the exposed dose-volume of major OARs 33 . Several dosimetry comparison studies have confirmed the advantage of PT in reducing dose exposure of vital organs like the heart and lungs,34–36 which aligns with the clinical data presented in Table 2 that shows very few records about cardio-pulmonary side effects, though awaits long-term follow-up outcomes. While there is ongoing debate regarding PT's impact on skin toxicity and rib fractures, the consensus remains that the ability to spare normal tissues is a widely recognized advantage of PT. Thus, as to patient-screening for PT, the approaches focusing on the reduction of toxicities are the most explored, such as the model-based approach based on normal tissue complication probability (NTCP) models.
NTCP Models
The NTCP model is a mathematical model describing the likelihood of side effects occurring based on dose–volume data, which is first proposed in the late 1980s.37,38 The Lyman–Kutcher–Burman (LKB) model is the most well-known NTCP model with key parameters of n (the volume effect parameter), m (parameter describing the gradient of the dose–response curve), and TD50 (dose tolerance resulting in 50% complication risk) 39 that affect the calculation results.
When using the LKB NTCP model to predict the probability of complications, the key parameters can be either empirical based on published researches, 40 or population-specific in the light of the clinical follow-up data of side-effects 39 (details about NTCP model fitting methodology are provided in Appendix A) since the dose–effect relationship can be affected by factors such as patient cohort and treatment techniques. Furthermore, by introducing dose-modifying factors and latency parameters into a generalized NTCP model, 41 non-dosimetric risk factors such as age, smoking status or chronic diseases and censored time-to-event data are able to show their position in NTCP estimation (further details are provided in Appendix A).
The NTCP model can also be broadly extended as a radiation dose-based complication prediction model other than a fixed formula with predefined parameters. In this way, the dose-response relationship of OARs will be summarized from well-developed datasets with baseline information, treatment details, and follow-up outcomes. As an example, after conducting a population-based case–control study of major coronary events in 2168 women who underwent radiotherapy for BC, Darby et al 42 concluded that the relative incidence of major coronary events increases by 7.4% for each 1 Gy increase in mean heart dose (MHD) (further details are provided in Appendix A), with absolute risk without radiotherapy derived from clinical data of the population, separately for patients with or without cardiovascular risk factors.
To be noted, there are few existing NTCP models tailored for BC radiotherapy, and generally, they are obtained from photon therapy cohorts,42–44 which brings up questions about the applicability of photon NTCP models to PT. Blanchard et al verified that in patients with head and neck cancer (HNC), the accuracy of NTCP model based on photon therapy data decreased in PT, but the diversity was within the acceptable range. 45 Nevertheless, no similar study has been conducted for BC so far. There remains a need for proton-specific NTCP models to enhance the precision of model-based predictions.
The Model-Based Approach
In general, the NTCP value rises with the increase of OAR exposure. However, a certain amount of dose reduction does not always lead to a proportional decrease of NTCP value if the dose to OAR is already low enough or the Δdose area situates in the relatively flat part of the dose–response curve. 46 Therefore, the NTCP value provides a more accurate assessment of OAR sparing compared to changes in dose metrics, making it a suitable criterion for patient selection.
The model-based approach, which utilizes NTCP models, was specifically developed to select patients for PT.33,46 It has gained approval and acceptance from the Dutch Health Care Institute (Zorginstituut Nederland). As the first step of this approach, both proton and photon treatment plans are made for each patient. Comparing the NTCP value of both plans calculated through the selected NTCP model for the given complication, a ΔNTCP threshold value should be chosen—mostly based on the severity of toxicity 46 —for subsequent patient-screening. Patient may be recommended of PT only when he or she shows a NTCP-value decline reaching the threshold by the proton planning. As the second phase of this approach, a prospective observational study is required for clinical validation. 33 Notably, as a preselection tool, if the difference between the photon NTCP and the assumed NTCP with zero OAR dose is already below the selected threshold value, there is no need to proceed with a proton plan. 47
As to BC, the National Indication for Protocol Proton therapy (NIPP) for model-based selection in the Netherlands takes a threshold value of a 2% absolute lower risk on acute coronary events (ACEs) < 80 years of age as the credit for the recommendation of PT. 48 The probability of ACEs was estimated by MHD in combination with age and presence or absence of cardiovascular risk factors, basing on the theory from Darby et al. 42 The practices of model-based approach proved it clinically feasible and theoretically sound, while further observation is still required to validate its clinical significance on treatment optimization.
Second Cancer Risk Models
It has been proposed that BC patients with radiation exposure have a significantly excess risk of secondary cancers. 49 Thus, the secondary cancer risk (SCR) may be considered, to some extent, as a complication related to OAR irradiation. In line with the advantage of PT in OAR sparing, several comparative studies50–52 have concluded that protons hold an advantage over photons with reduced risk estimates for secondary carcinogenesis such as secondary lung and contralateral breast cancer in BC patients.
Yet, estimations of SCR vary from NTCP models. One of the commonly used models is presented by Schneider et al, 53 which predicts the SCR by calculating the excess absolute risk of developing a cancer in a particular organ at a particular time after radiation exposure with dose–volume data and key parameters of β (the initial slope of the dose-response curve at low dose), γe/γa (the age modifying parameters), R (the parameter characterizing the repopulation/repair-ability of the tissue), and α (a proportionality constant), of which values for different tumor sites were derived from the analysis of combined A-bomb and Hodgkin cohorts.51,53,54 However, Ntentas et al 55 estimated the SCR with concise models based on mean dose of OARs and summarized excess relative risk values from previous studies. To be noted, in Ntentas et al's study, 55 the estimation of SCRs in lung, breast and esophageal, along with the risk of radiation-related cardiovascular disease, comprised the 30-year absolute mortality risk, which was used to optimize patient selection for PT over Hodgkin lymphoma. The combination of SCR and NTCP was worthy of reference for patient screening in BC patients.
Local Tumor Control
According to the limited non-RCT results in Table 2, the overall performance of PT in tumor control is positive. However, in most BC patients indicated for radiotherapy, modern photon technique can promise an acceptable risk of cancer recurrence.28–30 Thus, it's a challenge to investigate the potential of PT in improving tumor control. Further studies applying a better defined population, such as re-irradiated patients or patients with high-risk recurrence factors in terms of surgical quality or tumor pathology, may do a contribution. Besides, for patients whose target location goes against the OAR protection, such as a tumor bed in the inferior portion of the breast or extending over two quadrants, 56 a deep-seated tumor bed, 57 a target field including the internal mammary nodes (IMN) in left-sided BC 36 and complex anatomy like pectus excavatum, 58 their dose coverage may have to be compromised to limit the irradiation exposure to OARs such as heart and lungs with photon, while the ability of normal tissue sparing beneath the beam range provides the possibility of PT to improve tumor control by delivering a sufficient dose to the targets near dose-limiting organ. In the absence of clinical trials focusing on specific populations aforementioned, alternative tools are needed to predict the performance of PT in tumor control for comparison.
TCP Models
According to Stick et al, for left-sided BC patients indicated for IMNs irradiation, proton plans were expected to reduce the risk of recurrence after 10 years by 0.9% with balanced heart-sparing and dose-coverage on IMNs 36 compared to photons. The prediction of recurrence rate in this study was based on the empirical hazard ratio for disease-free survival from the meta-analysis of previous trials.
Otherwise, the tumor control probability (TCP) model was developed to simulate the tumor response in view of the effect of dose distribution to the tumor and density of clonogenic cells. 59 There is a variety of fitting methods of TCP models. Plataniotis et al 60 assumed that the cell survival after irradiations fits a Poisson distribution, thus a linear equation of two unknowns (-ln[-ln(TCP)] value and the biologically effective dose (BED) value) was deduced. Several pairs of TCP and BED values were derived from published clinical trials. By plotting -ln[-ln(TCP)] (y-axis) against BED (x-axis) values, a linear regression line corresponding to the equation with a slope α and intercept -ln(N0) (N0 is the initial number of cells before irradiation) was fitted. Prior et al 61 used the Chi-squared fitting method to estimate the fitting parameters of TCP model basing on a phenomenological sigmoidal equation, and these parameters represented the averages over the population of the involved clinical trials. However, there has been no reliable validated TCP model which is qualified for patient selection so far. After all, it remains a question that how much predicted improvement in tumor control rate should be considered clinically relevant.
Genomic Biomarkers
Biomarkers may also help identifying patients fit for protons. By analyzing the DNA repair and cell survival endpoints in multiple cell lines after irradiation, scientists have found that among the two major pathways of DNA DSB repair, cells with homologous recombination deficiencies are more sensitive to high-LET protons compared to photons,62–65 while the loss of non-homologous end-joining doesn't show the same specific sensitization. Thus, Fontana et al and Bright et al62,63 further discovered that homologous recombination inhibiting factors such as BRCA1/2 mutation and RAD51 recombinase depletion could significantly increase radiosensitivity for proton irradiation. According to another study, since the depletion of cyclin D1 seemed to have increased proton RBE in two triple-negative breast cancer cell lines, probably by blocking RAD51 recruitment to DSB sites, Choi et al assumed it as a possible classifier to predict the radio-sensitivity of triple-negative breast cancer. 66
However, when particular germline mutations or protein deficiency could be potential predictors for PT response, the resulting DNA repair deficiency may also encourage radiation-induced toxic effects.
Cost-Effectiveness
There are only 109 proton radiotherapy centers in operation worldwide for the moment. 67 PT demands high economic needs both for the investment and maintenance, thus the cost of it is two to three times higher than that of photon therapy.68–72
Cost-effectiveness analysis (CEA) is a well-known tool to evaluate if a treatment is worthy, in which a Markov model is commonly constructed to simulate different health states of the simulated patient cohorts at specific time intervals. The cost and health state utility value (ranging from 0 to 1) associated with each state can be estimated after reviewing relevant literatures. With intervention strategies investigated, the expected quality-adjusted life-years (QALYs) and costs for each strategy should be obtained from the model.73,74 The outcome can be measured by the incremental cost-effectiveness ratio (ICER) with a willingness-to-pay threshold of a certain amount of money per QALY.
As for BC, with an assumed ICER threshold of $36 000/QALY, only 1 of 16 patients could be treated with PT cost-effectively according to the CEA conducted by Austin et al 54 . While it's not very likely to decrease the absolute cost of PT, the best chance to improve its cost-effectiveness may be benefits in terms of OAR protection or tumor control. Mailhot et al noted that BC patients with no cardiac risk factor (baseline hypertension, obesity, hypercholesterolemia, history of heart disease, etc.) or with MHD below 5 Gy using photons were less cost-effective to take PT 68 . According to another study, for population with pre-existing cardiac risk factors, the cost per QALY gained of PT (the ratio of Δcost and ΔQALY between PT and conventional photon/electron radiation) decreased from €66 608 to €34 290 72 . Thus, CEA accompanied with NTCP/TCP measuring strategies is a reasonable and feasible way to make a trade-off between costs and benefits of PT.54,75 As an example, the Markov model conducted by Austin et al used TCP, NTCP and second primary cancer induction probability models to estimate the transition probability between different health states for individual patients. 54 The only patient who was found cost-effective to take PT in Austin et al's study had a comparatively great reduction on lung dose using PT compared to photon therapy (Δmean lung dose = 8.81 Gy), which decreased the risk of second pulmonary malignancy.
Eligibility Criteria of Ongoing Trials
The eligibility criteria for the two largest ongoing RCTs of PT for BC are listed in Table 4, which are the RADCOMP 76 trial and the DBCG 77 trial. Since the patients enrolled for RCTs will be randomly assigned to receive either proton or photon radiation therapy, these criteria are defined broadly to maximize the generalizability of results and have limited reference value in patient selection for PT.
Eligible criteria of ongoing RCTs.
The DBCG study, 77 for instance, only enrolls individuals with a minimum MHD of 4 Gy and/or a volume of ipsilateral lung receiving 20 Gy (V20) of at least 37% when using photon radiotherapy. Although cardiorespiratory doses were taken into account, this threshold was decided mainly in consideration of the annual number of patients adjuvant loco-regional IMN radiotherapy in Denmark, the patients’ willingness to participate in clinical trials and the capacity limitation of proton centers, which would allow approximately 22% of all patients requiring loco-regional IMN radiotherapy to enter the randomization process. 78 Meanwhile, the inclusion criteria for the UK PARABLE 79 trial require a 2% or higher estimated absolute lifetime risk of radiation-induced cardiac toxicity. This estimation is based on factors such as patients’ age, cardiac risk profile, and MHD, which is similar to the Dutch model-based proton selection strategy and emphasizes on the benefits of normal tissue sparing more specifically. The significance of these criteria in improving the treatment appropriateness of PT remains to be validated through the results of these trials.
Discussion
Further Development of NTCP Model-Based Approach in BC
Currently, the NIPP for model-based selection in the Netherlands holds the only clinical practice of NTCP model-based patient-selection for PT in BC patients.
48
However, the involved NTCP model developed by Darby et al focuses solely on ACEs, highlighting a limitation wherein each NTCP model can assess only a single specific side-effect. Taking this into consideration, some of the current practical applications of model-based approach may use multiple NTCP models to synthetically analyze the main side-effects of radiation therapy for specific diseases. As an example, to implement the NIPP for model-based selection of HNC patients, three NTCP models corresponding to severe xerostomia, grade ≥ 2 dysphagia and tube feeding dependence were selected
47
. Patients were qualified for PT with at least one of the ΔNTCP or
On the other hand, the NTCP model-based approach may lead to wastage of medical resources and delays in treatment as it requires both photon and proton treatment planning for each patient. Given that special thoracic anatomy like pectus excavatum is believed to cause dosimetric differences, 58 it makes sense to explore the impact of baseline anatomic parameters, such as the distance from chest wall to heart or the thoracic aspect ratio, to OAR exposure, target dose coverage or the differences in NTCP value between proton and photon plans. Hopefully, a quicker patient-screening basing on pre-irradiation images may be developed for further selecting procedure to improve decision-making efficiency. Besides, the development of artificial intelligence including automated treatment planning techniques is also expected to expedite the whole patient-screening procedure. As a good example, Kierkels et al 80 developed a dose mimicking and reducing algorithm to automatically optimize robust proton plans from a photon reference dose and integrated it into an automated NTCP model-based patient selection framework for HNC patients.
Concerns About the Increased Risk of Certain side Effects
Regarding the current clinical outcomes of PT for BC, concerns have arisen that the advantage of cardiopulmonary protection may be partially counterbalanced by an elevated risk of skin toxicity or rib fracture.
However, the performance of PT on skin toxicity is believed to be impacted by different techniques used in beam delivery. As an early-stage technique, passive scattering beam necessitates the delivery of a nearly full prescription dose to the skin. Whereas, with scanning technique, in particular with pencil beam scanning utilized in intensity-modulated proton therapy, a better optimization of skin dose can be provided by applying dose constraints and adjusting the density or energy of beams.81–83 According to a recently published recent systematic review, 84 the overall incidence of severe dermatitis in the published clinical results of PT for BC is significantly higher with scattering PT than with scanning PT in both partial breast (P = .02) and whole breast/chest wall (P < .001) irradiation. Moreover, the rate of severe dermatitis after scanning PT to the whole breast/chest wall (6%) was comparable to the latest hypofractionated photon therapy 85 .
Concerning the risk of rib fractures, as previously mentioned, it is attributed to the RBE peak at the distal edge of SOBP. To optimize the rib dose, the RADCOMP Breast Cancer Atlas recommends a well-defined posterior border of target volume for PT as “up to but not including ribs.” Meanwhile, the currently applied average RBE of 1.1 is doubted in plan evaluation since the average RBE value at the distal end of SOBP is reported as approximately 1.35. 5 As Liu et al noted, 86 the biological dose of ribs calculated with a uniform RBE of 1.1 led to a 13% underestimation in D0.5cc when compared to the biological dose calculated with a dose-averaged LET-based model in proton plans for BC. Thus, a better understanding of RBE values and an exploration of variable RBE models in the evaluation of PT plans for BC are called for further optimization of rib and even other normal tissue doses.
In summary, it is anticipated that advancements in technology, treatment standards, and radiotherapy plan evaluation systems have the potential to mitigate these risks. For now, evaluating
Besides, regarding the elevated risk of capsular contracture in reconstructed breasts caused by PT, 14 given that the implants are non-biological and replaceable, relevant concerns are more inclined towards the costs and damages associated with subsequent revision surgeries. In our view, this risk for post-reconstruction patients has to be informed before radiotherapy decision, whereas, it should not override treatment recommendations based on other clinical benefits of PT.
Financing Considerations
The cost-effectiveness mentioned earlier actually reflects the balance between expenses and returns rather than a decrease in absolute costs. From a practical point of view, patients’ affordability may be a peripheral factor worth considering when making treatment recommendations.
To alleviate the financial burden on patients, exploring the efficacy of hypofractionated PT 10 may help cutting the expenditure by reducing the number of treatment sessions. However, from the perspective of personal interests, insurance support holds more significance than individual payment capacity. In the study from Mendenhall et al, 89 116 out of 131 insured BC patients (89%) received insurance approval for PT, of which most cases required a medical review, comparison plan or peer-to-peer discussion. In general, insurers tend to approve PT for malignancies characterized by poor outcomes with photons or high risks of adverse effects. In other words, they prioritize cases with a certain benefit from PT, or one could say, an appreciated cost-effectiveness. Therefore, from this perspective, the capability of clinicians goes no further than presenting compelling evidence through various patient selection strategies to convince the insurers into insurance approval or to advocate for the expansion of insurance coverage. For individuals without health insurance, these evidences might serve as valuable references for assistance from charitable foundations.
Conclusion
PT is a highly promising technology in modern radiotherapy which shows potential to substantially decrease normal tissue irradiation and improve tumor control in high-risk BC patients. However, due to its high cost and limited accessibility, PT should better be offered to highly-selected patients to maximize its clinical benefits. Currently, the most practical way of patient-selection for BC is the NTCP model-based approach with clinical practice and national approval in the Netherlands. Ongoing clinical trials are expected to provide valuable clinical experience that will help directing treatment decisions in the future.
Supplemental Material
sj-docx-1-tct-10.1177_15330338241234788 - Supplemental material for Proton Therapy in Breast Cancer: A Review of Potential Approaches for Patient Selection
Supplemental material, sj-docx-1-tct-10.1177_15330338241234788 for Proton Therapy in Breast Cancer: A Review of Potential Approaches for Patient Selection by Xiao-Yu Wu, Mei Chen, Lu Cao, Min Li and Jia-Yi Chen in Technology in Cancer Research & Treatment
Footnotes
Abbreviations
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Ethics Statement for Animal and Human Studies
Not applicable.
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: National Key Research and Development Program of China (grant number 2022YFC2404602), Shanghai Hospital Development Center Foundation (grant number SHDC12023108), Scientific and Technological Innovation Action Plan of Shanghai Science and Technology Committee (grant number 22Y31900103), Clinical Research of Shanghai Municipal Health Commission (grant number 20224Y0025), Beijing Science and Technology Innovation Medical Development Foundation (grant number KC2021-JX-0170-9), and National Natural Science Foundation of China (grant numbers 82373514, 82373202, 81972963).
Supplemental Material
Supplemental material for this article is available online.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
