Breast cancer is considered one of the most prevalent cancers in adulthood with nearly a third of the newly diagnosed cases arising in patients older than 65 years.[1] Compared with their younger counterparts, older patients tend to develop more hormonal/estrogen-positive (ER+) tumors because of chronic exposure to estrogen stimuli.[2,3] Such favorable tumor biology allows better outcomes as ER-positive (ER+) tumors grow and spread slower relative to ER-negative tumors. Despite that, for patients (65–70 years or older) receiving lumpectomy for early-stage (T1N0) and ER+ breast cancer, the optimal adjuvant treatment is debatable.[4] The projected life expectancy along with competing comorbidities, polypharmacy, and the behavior of those small size ER+ tumors present a challenge for the treating oncologist.[5]
In a recent national survey of 174 oncology providers from different specialties,[5] performance status influenced the treatment recommendations for this specific population, where systemic and radiation therapy (RT) were recommended more for patients with better performance status. This shows that in daily practice, biological age and performance status may play a bigger role in treatment recommendations regardless of clinical trial results.[6] This is not surprising as the International Society of Geriatric Oncology (SIOG) suggests that these patients are a heterogeneous population in regard to functional status, cognitive abilities, nutritional and psychological distress, social support, or engagement.[7–9] Moreover, the clinical trial end points focus on overall survival (OS) and disease-free-survival (DFS), which can be less meaningful for older patients where quality of life (QOL) and treatment tolerance matter more.[9] Such challenges leave the older patients vulnerable to either overtreatment or undertreatment. Overtreatment can stimulate a cascade of diminished QOL and financial burdens while undertreatment and de-escalation can lead to worse tumor outcomes.[8,10]
Many trials examined the omission of conventional radiation (3–6 weeks) for low-risk breast cancers (Table 1). Notably, CALGB-9343[11] and PRIME-II[15,17] confirmed that the addition of conventional RT to antiestrogens, such as tamoxifen (TAM), improved the 10-year local control rates (LC) from 90% to 98% and from 90.2% to 99.1%, respectively, without any improvement in (OS) or (DFS). Interestingly, in the CALGB-9343 study, the 5-year results revealed that LC was 96% with TAM only and dropped to 90% at 10 years, whereas TAM plus RT had 99% and 98% at 5 and 10 years, respectively. This reflects that with omission of RT, the risk of local failure increases 1% per-year for these patients. Other trials (Table 1) included younger patients making CALGB and PRIME II the seminal trials for patients 65 years and older. Hence, the rhetorical practical question becomes who would want 6 weeks of daily RT for zero OS gain, while a simple daily pill for 5 years can provide all the benefits of systemic tumor control? Moreover, 90% LC rates at 10 years with omission of RT—for a 70-year-old—outweigh the burden of daily RT and its potential cardiopulmonary side effects. Consequently, recommending only aromatase inhibitors (AI)—which has been shown to be superior to TAM for this population[18,19]—and omitting 5–6 weeks of daily RT seemed a convenient option for many caregivers and patients in 1994 (when CALGB started). In fact, the current National Comprehensive Cancer Network (NCCN) guidelines endorse hormonal therapy alone for 70-year-old patients with early hormonal-sensitive tumors on the basis of CALGB and PRIME-II results.[20]
As we build evidence over the years and improve RT techniques, many patients (older than 65 years) with T1N0 luminal-A today are being denied improvement in local control through modern RT. Furthermore, these trials did not stratify by performance status and enrolled patients between 1994–1999, when the 6-week RT burden was completely different than today. At that time, the image guidance techniques, the accelerated partial breast irradiation (APBI), breath-hold technique, and ultrahypofractionation; all these concepts were not yet developed. Although all phase III trials (Table 1) suggested better LC with RT addition to hormonal treatment, the interpretation was always that RT could be omitted. Today, after more than two decades and significant advances in RT, we still apply CALGB-9343 and PRIME-II results, and more patients are offered RT omission.[21,22]
Moreover, the current clinical trials in the pipeline are designed to use advanced biological markers to also omit RT on the basis of the CALGB-9343 and PRIME-II concepts. For example, IDEA trial (ClinicalTrials.gov identifier: NCT02400190) is examining omission of RT for postmenopausal women (50–69 years) with low oncotype score; LUMINA trial[23] examines omission of RT for women age 55 years and older with early-stage luminal-A and low KI-67, and PRECISION (ClinicalTrials.gov identifier: NCT02653755) is another trial studying omitting RT for younger women (50–75 years) using PAM-50. Although those trials are single arm with no comparator, they reflect that the mindset is to omit RT regardless the RT evolution and AI tolerance and compliance. The only trial that has a comparator group is the DEBRA trial (ClinicalTrials.gov identifier: NCT04852887) where women age 50–70 years with T1N0 ER+ and oncotype score < 18 are randomly assigned to RT or no RT and everyone receives hormonal therapy.
One can argue that we are overselecting patients to solely omit RT without clear reasons. Is it the side effects of radiation? Since RT was found beneficial, we should rather work on improving its tolerance and convenience as treatment, not studying its omission. Was hormonal therapy found to be more tolerable and more efficacious when compared head-to-head with RT?
For elderly patients, AI has been shown to be more effective than TAM. AI comes with a heavy cost of cardiovascular side effects, accelerated osteoporosis, and musculoskeletal weakness leading to poor compliance.[24,25] Besides, the competing comorbidities, geriatric psychology, other AI side effects and polypharmacy result in poor QOL.[26] The adherence rates in the literature for the past two decades are surprisingly puzzling, with 5-year adherence rates ranging from < 20% up to > 75%, reflecting the dynamically changing nature of such problem governed by QOL, polypharmacy, and side effects tolerance.[27–29] Moreover, a report from Michigan Radiation Oncology Quality Consortium studying 8,711 patients showed that patients 70 years and older significantly reported less toxicity, less fatigue, and less breast pain compared with their younger counterparts.[30] This finding was the same for both conventionally fractionated and hypofractionated RT. In contrast, an international survey encompassing 2,353 women showed that 91% of patients receiving hormonal therapy reported musculoskeletal side effects and physical changes.[31] This led to discontinuation of therapy among 33% of the respondents, with more discontinuation rates in the United States. Interestingly enough, AI administration was associated with higher odds of reporting cardiovascular, musculoskeletal side effects and concerns about medication cost, compared with TAM-only users. When such side effects and concerns were raised by the patients, 31.5% felt dismissed or minimized by the provider. This alludes that RT could be well tolerated in this patient population compared with AI. In fact, another study surveying 130 patients with a median age of 74 years showed that 66% of patients had significant concerns regarding AI versus 39% expressing concerns from RT. Moreover, 57% of patients in that study would rather receive RT alone.[32]
Therefore, initial geriatric assessment and discussions regarding both approaches with their toxicities will allow patients to set the expectations about their QOL. Geriatric assessment, despite being recommended by NCCN and SIOG, remains not yet well implemented in daily oncology practice.[33,34] This multidisciplinary diagnostic process helps detecting medical, psychosocial, and functional problems which subsequently guide the QOL expectations and potentially avoid undertreatment or overtreatment for this population. Nonetheless, older patients may have greater attention to the information given to them and are more willing to participate in the decisions about their care which warrants presenting all the pieces of information.[34,35] The timing and methodologies for conducting such assessments have been well described in previous articles.[36] A crucial part about recommending AI alone is a clear discussion about adherence to treatment and side effects monitoring. Exploring factors as social support, functional status, and ability to thrive through such geriatric tools can help directing AI candidacy. Other than the side effects we mentioned earlier, some published data showed that AI side effects solely can lead to significant distress and mental health deterioration in such population[37] and hence AI discontinuation. In fact, AI discontinuation was associated with increased local failure rates without any compromise of DFS/OS.[38,39]
Recently, the FAST-FORWARD trial started the era of ultra-hypofractionation. Excitingly, 26 Gy in five consecutive daily fractions for T1-T2/N0 tumors for any age group has similar local control rates compared with 3-week hypofractionated regimen.[40,41] Furthermore, the patient-reported outcomes showed no significant difference between 26 and 40 Gy in all outcomes, with breast hardness approaching nonsignificance (odds ratio, 1.2; 95% CI, 1.0 to 1.5; P = .048).[40] Of note, this trial included all age groups, with approximately 49% between age 60 and 79 years; nonetheless, the effectiveness and safety of such dose/regimen can be applied to this population.
Still, does not these 5 days radiation treatment potentially cause cardiopulmonary side-effects? The answer is majority of those elderly patients with T1N0 luminal-A are favorable candidates for APBI providing safe local control as whole breast irradiation with negligible cardiopulmonary side effects.[42] In addition, the 10-year updated results of the Florence trial using APBI, 30 GY in five fractions (daily and every other day RT were allowed) confirmed the safety of such short APBI courses, and it had both excellent patient- and physician-reported cosmetic outcomes.[43] Nowadays, what was once considered overwhelming treatment in terms of transportation, side effects, cost, and daily setup for a patient older than 65 years can be accomplished in 5 days only. No doubt that this sounds more appealing to patients than daily AI for 5 years.
However, the question becomes is it safe to omit hormonal therapy for this population? This was partially answered in the NSABP-21[13] trial where patients of different age groups with a hormonal-sensitive invasive tumor ≤ 1 cm had no difference in DFS between TAM only and RT only after lumpectomy. However, this trial included younger patients where 50% were younger than 60 years, and they lived longer and had higher incidences of contralateral breast cancers without TAM. This is different in postmenopausal patients, as a recent SEER analysis of more than 13,000 patients found that RT alone without AI for postmenopausal women was not associated with increased risks of secondary breast cancer.[44] Moreover, BASO II trial[14] showed that for patients younger than 70 years (mean age: 57 years) with early-stage grade 1 invasive tumors, the DFS/OS rates were similar between RT-alone and AI-alone after lumpectomy. Those results were replicable in elderly patients as well. In a series from Memorial-Sloan-Kettering, 888 women 65 years and older with T1, N0, ER+ tumors, 5-year local failure rates were similar between AI-alone and RT-alone and no significant difference in DFS or OS.[45] Moreover, in a Danish Cohort, elderly patients with T1aN0 (< 10 mm), hormonal-sensitive tumors, the survival without hormonal therapy was similar to age-matched cancer-free women in general population.[46] An NCDB analysis encompassing more than 30,000 patients showed no interaction between hormonal treatment and age.[47]
On the basis of the above, we think that Meattini et al did it right. In their ongoing trial EUROPA (NCT04134598), investigators are comparing exclusive AI versus APBI only for women older than 70 years in a phase III trial enrolling 926 patients. Another phase-II trial (CAMERAN, NCT054722782) aims to recruit 90 patients for the same random assignment with a primary end point of QOL. Nevertheless, a similar trial possibly including younger age (younger than 70 years old) should be considered in the United States. In the enhanced precision medicine era, using oncotype scores and next-generation sequencing, we should be able to stratify younger patients (between 55 and 70 years) with early-stage luminal-A to study RT versus hormonal therapy instead of studying RT omission only as discussed earlier. Although we await the results of EUROPA trial and with many ongoing trials using expensive genetic panels to selectively omit RT, the easily available geriatric assessment tools can be implemented in daily practice to identify the poor candidates for AI therapy. Collectively, with the studies we discussed earlier, omission of hormonal therapy for elderly population is not a dangerous approach.
To conclude, we do think it is time to revisit the omission of adjuvant RT for elderly women 65 years and older with T1N0 luminal-A breast cancer and consider testing omitting hormonal treatment. Meanwhile, APBI/UK-FAST-FORWARD regimen in five consecutive daily fractions or Florence trial every other day regimen after lumpectomy for this population appeals a reasonable approach as the primary treatment, and AI could be optional for patients who would tolerate it.
J Clin Oncol. 2023;41(13):2331-2336. © 2023 American Society of Clinical Oncology
