This article is more than 5 years old. Information may no longer be current.
The ‘puzzle’ of noncompliance: Skipped radiation and lives lost
Chu and colleagues report that as many as 35% of high-risk patients with breast cancer in the United States who have N2 to N3 locally advanced disease do not receive a potentially curative locoregional radiotherapy.
Is this a problem and, if so, to what extent?
History of treatment
Joseph Ragaz
Radiation with early hormones was the primary treatment of choice for breast cancer in the 1960s and 1970s; however, it was gradually replaced by chemotherapy by the early and mid-1980s. As chemotherapy was considered curative and radiotherapy too toxic, radiation was discontinued in parallel with the increasing use of chemotherapy.
In the early 1980s, our group revisited the interest for chemotherapy combined with a more modern and improved-quality radiation. Emerging evidence based on preclinical data indicated that chemoradiation sensitization would substantially improve breast cancer outcomes compared with either modality alone. This concept, it was felt, would be particularly important for patients with positive lymph nodes.
Thus, we re-evaluated the addition of radiation to chemotherapy in a prospectively randomized trial, started in the mid-1980s. Our 15-year follow-up results — published in 1997 in The New England Journal of Medicine — demonstrated a 30% reduction in breast cancer mortality by adding radiation to adjuvant chemotherapy.
Soon after we began our trial, the Danish Breast Cancer Cooperative Group evaluated the same concepts. Their 10-year results, published neck-to-neck with our data in the same issue of NEJM, eventually involved a much larger patient sample and provided an independent confirmation of radiation benefit. Their data demonstrated approximately a 20% to 30% reduction in breast cancer mortality, according to subsets.
Subsequently, the Oxford Overview group conducted several meta-analyses of their Early Breast Cancer Trialists’ Collaborative Group (EBCTCG) and confirmed a significant reduction of both locoregional and systemic breast cancer recurrences — and of overall mortality — with radiation. This benefit was shown for both the post-mastectomy type of radiation therapy — with the chest wall and locoregional lymph node fields — as well as for the more restricted BREAST–ON field, covering breast and part of ipsilateral axilla.
A proof of the principle was thus established: Chemotherapy alone cannot handle locoregionally disseminated breast cancer, and the addition of radiation therapy to chemotherapy will significantly reduce breast cancer mortality. It also has been established that the higher the risk as identified at diagnosis (ie, as judged from the nodal status), the higher the absolute radiation therapy benefit. As a result of these initiatives, soon after the 1997 NEJM publications, adjuvant post-mastectomy radiation combined with chemotherapy became a guideline recommendation internationally for all patients with breast cancer who have positive lymph nodes.
Surprising results
In their current publication, Chu and colleagues utilized a large U.S.-based National Cancer Data Base of about 2.7 million patients diagnosed with breast cancer and treated between 1998 and 2011. A subset of 56,990 patients had N2/N3 node-positive disease, a classification that represents extremely high-risk disease. For these patients, the routine use of both radiation therapy and chemotherapy is clearly indicated, as it is evidence-based.
In a detailed multivariate analysis that assessed socioeconomic status, race, income, type of the institution, etc, researchers determined that up to 35% of patients with N2/N3 disease did not receive radiation. This is a surprisingly high rate of noncompliance, the reasons for which the authors describe as a “puzzle.” The results also showed that patients who did not receive radiation therapy also were more likely to forgo chemotherapy.
A majority of the patients evaluated were urban non-Hispanic white (80.9%) and represented a medium-to-higher socioeconomic strata compared with the rest of U.S. population. This analysis was restricted to N2/N3 patients, and noncompliance rates among those with N1 disease — a much more voluminous breast cancer subset — have not been evaluated.
Consequences of noncompliance
There are several questions arising from this review. What do these figures mean for the population health overall, and specifically for the ultimate outcome — the annual breast cancer-specific and overall mortality rates? Also, how much should we care about this 35% or even higher noncompliance rate with a potentially curative therapy within this very high-risk breast cancer subset?
Like any curative therapy, its absence and non-delivery, in simple terms, means higher mortality. Higher mortality is defined here as “avoidable deaths” in proportion to noncompliance, as previously described. How many “avoidable deaths” can one project annually due to radiation therapy noncompliance, taking the proportion of N2/3 cases in the population and the noncompliance data of Chu and colleagues?
Untreated, N2/N3 patients with breast cancer typically face at least a 60% to 70% recurrence rate and an equivalent rate of breast cancer mortality. Assuming a major improvement with chemo-hormonal therapy (ie, without radiation), conservatively, one would expect a much improved outcome but still close to 50% mortality of these N2/N3 patients treated with chemotherapy and hormone therapy alone without radiation. These estimates translate to 500 of 1,000 new N2/N3 patients still dying of their disease if not irradiated.
If we take these 1,000 patients and now give them additional radiation and assume a 30% reduction of mortality with radiation as established by the earlier publications (RR = 0.7), we will see 150 additional lives saved. However, if only 650 patients instead of all 1,000 patients are irradiated, then instead of 150 saved lives, only 100 lives are saved and 50 would die unnecessarily. Prorated for the U.S. annual estimates of N2/N3 cases, these figures indicate a loss of at least 300 N2/N3 patients with breast cancer each year due to radiation noncompliance, and, overall, a stunning 3,500 avoidable deaths during the 1998 to 2011 time span studied by Chu and colleagues.
If we consider a similar noncompliance rates in all N1 patients diagnosed annually, the number of potentially avoidable deaths could be substantially higher. Also, if 35% of this higher socioeconomic breast cancer population did not receive radiation, the noncompliance figures may be much higher in the lower socioeconomic sectors of the U.S. population, which would encompass a higher proportion of non-Caucasians with lower insurance affordability and other factors typical for noncompliance.
Take-home message
The study by Chu and colleagues indicates that omission of curative therapies, for whatever reasons, may have serious consequences.
Our estimates indicate that these policies may adversely impact overall outcomes of high-risk patients with breast cancer.
The issue of quantitating the exact nature of noncompliance, albeit complex, indicates that identification of reasons for noncompliance — and staging major educational and logistical steps for overcoming the noncompliance — would be an extremely cost-effective strategy.
References:
Chu QD, et al. J Am Coll Surg. 2015;doi:10.1016/j.jamcollsurg.2014.12.045.
Early Breast Cancer Trialists’ Collaborative Group. Lancet. 2000;355:1757-1770.
Early Breast Cancer Trialists’ Collaborative Group. Lancet. 2005;366:2087-2106.
Overgaard M, et al. N Engl J Med. 1997;337:949-955.
Ragaz J, et al. Cancer Res. 2009;doi:10.1158/0008-5472.SABCS-09-2063.
Ragaz J, et al. N Engl J Med. 1997;337:956-962.
For more information:
Joseph Ragaz, MD, FRCP, is a medical oncologist and clinical professor at the University of British Columbia. He can be reached at Suite 218-809 West 41st Ave., Vancouver, BC Canada V5Z 2N6; email: joseph.ragaz@ubc.ca.
Disclosure: Ragaz reports no relevant financial disclosures.