A practical and practicable framework for implementing cardiac-sparing radiotherapy techniques in breast cancer

Advances in the treatment of breast cancer have resulted in a consistent trend toward improved outcomes worldwide. From the perspective of low- and middle-income countries (LMIC) these improved outcomes offer hope despite an increase in the incidence of breast cancer.[1] However, a substantial proportion of women treated with adjuvant radiotherapy in LMICs are at risk of developing late radiation-induced cardiac morbidity owing to the lack of appropriate radiotherapy infrastructure.[2] This risk is reflected in a higher heart mean dose (Dmean) in Asian LMICs over the last two decades and compounded by an overall higher proportion of the global burden of cardiovascular diseases.[3,4] After accounting for reporting bias, the actual heart Dmean is likely higher in routine community practice and, disconcertingly, remains unknown in those treated on Cobalt-60 (Co60) machines. Measures can be instituted to reduce the heart Dmean with increasing levels of available resources in LMICs, namely: (a) adopting hypofractionated treatment schedules; (b) avoiding internal mammary nodal irradiation in early breast cancer (EBC) when using Co60 machines; (c) excluding ribs and intercostal muscles during target delineation; (d) preferring forward-planned approaches (with breath control), and; (e) reserving inverse-planned approaches for patients with unfavorable anatomy (and/or unsuitable for breath control).[5-7] In a recent survey of more than 2000 radiation oncologists worldwide, the lowest adoption of hypofractionated treatment schedules in breast cancer was in LMICs.[8] The advantage of shorter fractionation schedules in resource constrained LMICs is obvious, yet a third of the respondents voiced concerns regarding late toxicity. The results of hypofractionated trials (most allowed Co60 treatment) should allay this concern. The anticipated effect of hypofractionated radiotherapy on cardiac function is lower than conventional fractionation, owing to a reduced heart Dmean after Equivalent Dose in 2 Gy (EQD2) conversion.[9] Dosimetric studies have demonstrated that unless the α/β ratio of the heart is lower than 1.5, almost all hypofractionated schedules have a lower EQD2 Dmean compared to conventional fractionation.[9] Acknowledging the limitations of dosimetric modeling in predicting complex cardiac events, we endorse prospective data collection on cardiac outcomes. Yet the current generation of trials in radiation oncology with cardiac-specific outcomes are designed to assess the efficacy of conventionally fractionated Proton Beam Therapy (PBT), a technology that can potentially reduce heart Dmean to near-zero, but in LMICs, this will benefit only those with financial resources.[10] Treating internal mammary nodes (IMC) to replicate the positive results of elective regional nodal irradiation (RNI) trials with Co60 machines should be reconsidered. Since a linear relationship exists between heart Dmean and the risk of major cardiac events at 10 years, the Co60 tangential pair technique (heart Dmean = 13.3 Gy) would largely offset the 1–2% (statistically non-significant) overall survival benefit, especially in EBC.[11,12] The EBC Trialists Collaborative Group’s meta-analysis of RNI reported that women included in trials during the Co60 era (1961–1978) (heart Dmean >8 Gy) experienced an increased rate of non-breast cancer mortality.[13] Consideration should instead be given to using customized heart shielding (except for lower inner quadrant primaries) or referral to a radiotherapy facility with a linear accelerator (LINAC). In facilities with computed tomography (CT)-based treatment planning, excluding the ribs and intercostal muscles (True Chest Wall, TCW) increases the distance between the whole heart contour and the target volume. This reduces the heart Dmean by 1.4 Gy when using Field-in-Field (FinF) planning technique.[14] In a systematic review on post-mastectomy chest wall recurrences (6901 patients and 340 recurrent lesions), only six lesions (1.8% of recurrent lesions; 0.1% of all patients) were located in the TCW.[15] The risk of recurrence in the TCW would be even lower in patients undergoing breast conservation surgery; therefore, its exclusion would reduce cardiac exposure in those requiring adjuvant whole breast irradiation, as recommended by the European Society for Radiotherapy and Oncology (ESTRO) target delineation guidelines.[16,17] Since the ESTRO guidelines emphasize minimizing irradiation of breast tissue beyond that included in simulator-based tangential radiotherapy planning, there is also no advantage in terms of overall plan quality and cardiac sparing when using inverse-optimized treatment planning compared to tangential radiotherapy techniques (with a few exceptions).[16] Inverse-optimized treatment planning has improved patients’ quality of life by decreasing toxicity at several sites. But a vast body of dosimetric literature in left breast cancer has led to the conclusion that, unlike other anatomical sites, attempting to optimize coverage by adding more fields (Inverse intensity-modulated radiation therapy [IMRT]), control points [Volumetric Modulated Arc Therapy (VMAT)], and helical treatment delivery (Tomotherapy) results in an exit dose toward the heart which is challenging to constrain. A systematic review demonstrated that tangential radiotherapy (either two-dimensional [2D], three-dimensional conformal radiation therapy [3DCRT], or FinF; all without breathing control) consistently led to lower heart Dmean.[3] Attempts to improve upon the tangential radiotherapy planning technique are reflected in the reported time trends between 2011 and 2013, demonstrating higher heart Dmean compared to 2003–2010 (17 planning regimens in 2003; 115 planning regimens in 2013). Further “evolution” of these techniques focused on replicating the unique advantage of standard tangential radiotherapy by restricting fields, control points, and helical delivery in a tangential orientation only (Tangential IMRT, Tangential VMAT, and Tomo Direct, respectively). However, the increased planning complexity of these techniques effectively precluded widespread adoption and limited their application to patients with challenging anatomy (pectus excavatum/carinatum, large breast, wide separation) or those unable to perform cardiac-sparing maneuvers (breath control or alternative positioning measures). Therefore, institutions in LMICs could adopt a standardized approach to guide the selection of planning techniques by balancing dosimetry, complexity, existing infrastructure, and expertise.[17] Only treatment delivery with voluntary deep inspiration breath-hold (DIBH) and PBT have consistently better dosimetry than tangential radiotherapy planning.[3] A systematic review comparing dosimetry between DIBH and non-DIBH treatment plans demonstrated less cardiac exposure with comparable target coverage.[18] All included studies (except one) utilized the combination of DIBH with tangential radiotherapy planning, and there were no differences between included DIBH platforms. While PBT is dosimetrically superior, justification for widespread adoption is challenged by the fact that for most patients, DIBH with tangential radiotherapy techniques is adequate. In a dosimetric analysis of 179 patients preceding the initiation of the Danish Breast Cancer Group (DBCG) Proton trial, only 22% of plans using DIBH and tangential radiotherapy exceeded heart Dmean ≥4 Gy and/or lung V17 Gy/20 Gy ≥37% (versus 54% of plans using inverse-optimized techniques).[19] The design of this trial highlights the utility of a model-based approach for selecting patients likely to benefit from PBT. The modern understanding of radiation-induced cardiac morbidity as a “no threshold dose” phenomenon is a paradigm shift from the previous understanding of “inconsequential low dose.”[11] Despite ingenious planning complexity, the tangential FinF planning technique remains the solid benchmark for comparing all planning techniques.[3] The heart Dmean is currently the most validated surrogate for exposure to critical substructures (left ventricle, left anterior descending coronary artery), and a pragmatic, stepwise approach toward its systematic reduction needs to be implemented, with increasing plan complexity that mirrors the improvement in infrastructure [Figure 1].[11]Figure 1: Strategies to reduce cardiac exposure are arranged in the suggested order (from below upward) in which they should be adopted depending on the available infrastructure in LMICs. Upward and downward pointing arrows in parentheses next to the proportion of facilities denote increasing and decreasing acquisition trends in LMIC, respectively. (The reported proportions are provided as a range, as information was incomplete on 291 facilities in DIRAC, as of Jun 01, 2022). Co60 = Cobalt 60 machines, FinF = Field-in-Field, IMC = Internal Mammary nodal Chain, LINAC = Linear Accelerator, LMIC = Low- and Middle-Income Countries, PBT = Proton Beam Therapy, RT = RadiotherapyThere is a need for the global radiation oncology community to realign our emphasis on the appropriate utilization of existing infrastructure rather than prioritizing the acquisition of technologies that provide incremental gains at a steep cost and reduced throughput.[2] The “cancer moonshot” we strive for may be on the horizon; in the meantime, investing a fraction of those resources in providing access to cost-effective, proven radiotherapy techniques in LMICs will save more lives by reducing long-term morbidity. Financial support and sponsorship Nil. Conflicts of interest There are no conflicts of interest.