• 2019-10
  • 2019-11
  • 2020-03
  • 2020-07
  • 2020-08
  • br tients Although initial studies questioned the benefit of


    tients [20, 21, 25]. Although initial studies questioned the benefit of
    Table 2
    Causes of death in older and younger patients.
    Older group
    Younger group p
    Table 3
    Causes of death in older patients.
    All deceased older patients
    Chemo group
    Table 4
    Demographic Information in older patient cohort (n = 369).
    Congestive Heart Failure 17 4.6
    Coronary Artery Disease 96 26.0
    Abbreviations: ECOG, Eastern Cooperative Oncology Group; BMI, Body Mass Index; COPD, chronic obstructive pulmonary disease
    oxaliplatin in patients aged above 70 years [2], there are now excellent large data sets that show the benefits of doublet chemotherapy are sim-ilar in patients at least up to 80 years of age [21, 26–28]. Furthermore, evidence suggests that older patients do not experience significantly in-creased toxicities on standard 5-flurouracil based chemotherapy [19, 20, 29, 30].
    Our study is consistent with existing literature that older patients are far less likely to receive chemotherapy compared to younger pa-tients. In our cohort, 90% of patients aged 50–69 received chemother-apy, compared to only 60% in the age group over 70. This improved over historical numbers where the rate of chemotherapy use in older patients has been around 50% [13, 21, 31]. In our previous work, we ex-amined reasons for withholding chemotherapy in older patients with colon cancer. Although patient preference was the most common rea-son stated by treating clinicians, age alone, comorbidities, and small per-ceived benefit accounted for N50% of the reasons given for withholding chemotherapy [9].
    One assumption that our data challenges is that older patients will not benefit from chemotherapy because they 3-Deazaneplanocin A will die from other causes. In our study, potential “overtreatment” only occurred in a small proportion of older patients. We found that 11% of older patients treated with chemotherapy ultimately died from non-colon cancer causes. This implies that their competing risk of death was higher than that of cancer, suggesting that they may not have directly benefited from che-motherapy. Thus, incorporating a more complete geriatric assessment that includes an evaluation of functional status and comorbidities may help spare some of these patients from the toxicities of chemotherapy, especially when their chance of clinical benefit is low [10, 18, 32]. In our chart review, performance status, even with a simple ECOG, was often not documented, an observation that is consistent with existing literature [33]. Survivorship care in older patients with colon cancer should also be explored further. Existing data suggest that patients with a cancer diagnosis may receive less aggressive cardiovascular interventions and preventative care compared to their counterparts [34–36]. These disparities may in fact increase the risk of non-colon cancer deaths, regardless of the treatment patients receive for their cancer.
    More concerning is the issue of under-treatment. Although older pa-tients were more likely to die from non-colon cancer causes, nearly 40% of older patients in our cohort who were not offered chemotherapy ul-timately died of colon cancer. Given the robust benefit of adjuvant che-motherapy in patients with Stage 3 disease, we feel that some of these deaths may be potentially preventable with the judicious use of chemo-therapy in patients with expected life expectancy of 5–10 years or more. 
    Our data indicate that there is clearly a need for enhancing education of both patients and providers about potential survival benefits of adju-vant chemotherapy.
    In this study of patients aged 50 and over, we found a relatively high proportion of patients diagnosed with or dying of a second malignancy. Our database is population based and diagnosis dates of malignancies were captured with high fidelity. This was true in both the younger and older cohorts. While the precise mechanisms for high rates of a sec-ond malignancy is unclear, it is likely due in part to common lifestyle and hereditary risk factors. Until the pathophysiology is clarified, clini-cians should continue to counsel their patients about routine cancer screening, and address lifestyle changes even in patients for whom che-motherapy is not being offered. This further underscores the need for survivorship plans in older patients with cancer that extend beyond sur-veillance for recurrence.
    Our study has important limitations. We did not seek to examine the benefit of chemotherapy in this cohort. Thus, there were more colon cancer deaths in the chemotherapy group since this likely represented a group with higher risk of recurrence. Our analysis of predictors of death was limited both by small sample size and incomplete data on the chart review. Despite manual chart review data about comorbidities and performance status was limited given that standard documentation of these was not mandatory in this population based retrospective re-view. This question requires further study and may be best addressed by prospective studies which include a standardized geriatric assess-ment. Further study is also warranted to see if existing predictors of life expectancy [15, 17] can be applied and validated in the oncology population. Development or adaptation of these tools will be key for cli-nicians moving forward, and may help select patients who warrant a comprehensive geriatric assessment as part of decision-making for ad-juvant therapy. Our data did not allow us to examine the risks and ben-efits of chemotherapy in frail older patients, a population where decision-making remains extremely challenging. Finally, our data did not include analysis of compliance or completion of chemotherapy, which has been addressed in other papers [37, 38]. Furthermore, pub-lished results of randomized control trials about optimal duration of chemotherapy are still pending, which will be particularly pertinent to the older population who may fear risk of long term toxicity such as neuropathy [39].