Do we know what to do with our nonagenarian and centenarian patients with metastatic colorectal cancer (mCRC)? Results from the South Australian mCRC registry

Introduction

In the developed world, the population is ageing as a result of increased life expectancy due to better access to high-quality health care and improved living conditions [1]. In Australia, the proportion of older people has been gradually increasing over the last few decades. In 2013, 14% of the population (3.3 million people) was aged 65 and over and 1.9% was aged ≥85 years (439,600 people). By 2053, it is estimated that 21% of the Australian population will be aged 65 and over (8.3 million people) and 4.2% aged ≥85 years (1.6 million people)[2,3].

It is well-known that age is one of the major risk factors for developing colorectal cancer (CRC). In Australia, 37% of the patients diagnosed with CRC are aged 75 and over [2,4]. Longer life expectancy increases not just the overall incidence of CRC but subsequently the number of elderly patients potentially in need of oncology treatment.

Over the last twenty years there have been advances in surgery, chemotherapeutic agents and monoclonal antibodies for the treatment of metastatic CRC (mCRC), leading to a significant improvement in the average prognosis for patients (from only 6 months of life expectancy with best supportive care (BSC) to over 2-year survival with systemic therapies in clinical trials and population registries [5]. However, elderly oncology patients are often undertreated in clinical practice, due to physicians’ concerns regarding their physical frailty and comorbidities, cognitive impairment, possible pharmacological interactions, organ function reserve, as well as issues of family/social support. Furthermore, elderly patients are generally underrepresented in trials, precluding meaningful conclusions about the optimum management for this age subgroup of patients [6,7]. Medical oncologists are now being referred patients in their nineties yet the management options for this subgroup of elderly patients are generally unknown [8,9].

This study aimed to investigate the cancer characteristics, treatments administered and outcomes for nonagenarians and centenarians diagnosed with mCRC using our state-wide population-based mCRC registry to understand the current patterns of care and outcomes in this age subgroup of patients.

Patients and methods

The SA mCRC registry is a statewide population-based database of all patients diagnosed with synchronous or metachronous mCRC since February 2006 and details have been previously reported [10]. The SA mCRC registry was analyzed to examine tumor characteristics, treatments administered and outcomes for mCRC patients aged ≥90 years. For this study, we included data collected between 2 February 2006 and 31 March 2016. Overall survival was analyzed using the Kaplan–Meier method. Statistical analysis was done using SPSS version18.0 (IBM software) and GraphPad Prism version 5 (GraphPad Software, San Diego, CA).

Results

Patients’ characteristics. One-hundred and thirty patients out of a total of 4199 (3%) were aged 90 years or older. Their age ranged from 90 to 104.8 years, with a mean age of 92.1 years. Sixty-one percent were female, 70% presented with synchronous disease. Organ involvement was as follows: 58% liver, 32% lung, 8% peritoneal and 7% bone. Primary site was right colon 46%, left colon 28%, rectum 20%, unknown 6%. Only four patients had KRAS testing – all wild-type (WT).

Treatments prescribed and patients’ outcomes. The objective of the specific anticancer intervention in our nonagenarian patients was considered to be palliative in the vast majority of the cases. Only four patients received systemic therapy beyond the age of 90 (age range 90–93). Surgery for the CRC primary was uncommon with only 44.6% having resection. Of those with synchronous disease at diagnosis, only 24% had resection of primary lesion and 3% had stoma formed for palliation. One patient (age 92) was treated with ‘curative’ intent undergoing lung resection for metastases and survived for 18.5 months. Regarding the lines of therapy delivered, these were as follows: one line of treatment in two patients, two lines of chemotherapy in one and four lines in one. Aside single agent 5FU, combination therapy (oxaliplatin/5FU ± bevacizumab) was given to two patients and cetuximab single agent in two (WT one, unknown one). For the total 130 nonagenarian patients the mOS was 3 months (95%CI 1.4–4.6 months) and the 2-year survival was 10%. All 130 patients died from CRC. For the four patients who received chemotherapy, the survival ranged from 5.3 to 13.8 months.

Discussion

There is very little data in the literature analyzing the oncological treatment of nonagenarian mCRC patients. An ‘elderly population’ in clinical trials is generally defined as the subgroup of patients aged 65 years of age and older [11]. The definition of ‘old’ has evolved over time as medical care has improved and people now live longer and fitter. With this, the number of fit among very old patients presenting with cancer will increase and therefore an understanding of patterns of care, and where available, outcomes of interventions will be important to guide decision-making. For colorectal cancer, there are few reports on specifically the nonagenarian and older patients and these primarily focus on surgery and come from institutional experience [12–14]. Moreover, the number of recruited patients with mCRC aged ≥90 years in randomized trials is understandable scarce, and therefore the information about how these patients do on any sort of palliative treatment is almost inexistent [6].

Our data are derived from a large, population-based registry where there is no selection bias provides an insight into the frequency of this very old subgroup and currently suggests the vast majority do not receive systemic therapy and not surprisingly, have a very poor prognosis with a mOS of 3 months, which is lower than the general mOS for mCRC treated with BSC only [15]. The cause is likely to be multifactorial. One explanation is that this patient population may behave differently than younger elderly subpopulations (i.e., septuagenarians) and this may have a molecular basis [16,17]. Furthermore, tolerance of full doses of chemotherapy may not be possible [18,19]. Increased rate of comorbidities will also be important [20]. The location of the primary CRC tumors may also be relevant given the different prognosis between left and right primary site, with right side having poorer survival and previous reports have suggested that the frequency of right-sided colonic tumors does increase with patient age as confirmed in our analysis [21].

Only four (3%) of our nonagenarian patients received active systemic treatment. The current standard of care for unresectable mCRC includes fluoropyrimidine-based chemotherapy with or without the combination of a monoclonal antibody (cetuximab and panitumumab or bevacizumab, depending on tumoral RAS status) and the mOS for mCRC in clinical trials is greater than 24 months [22]. If 70 years or older is used as the cut off for ‘elderly’, in clinical trials the chemotherapy-associated survival outcomes are not statistically different compared with younger patients [23–25]. Furthermore, data from our registry supports equivalent survival for patients over and under 75 years with active interventions [26]. However, these data do not guide us for those very old patients. Several studies have pointed out that old age is a strong discriminating factor in determining whether elderly mCRC patients are offered chemotherapy and therefore the number of very old receiving chemotherapy are often very small as seen in our analysis [26]. Our results show very poor mOS (range 5.25–13.75) for the four patients who received systemic therapy. Similar survival was reported by Yu et al. [27] who studied the feasibility and tolerability of chemotherapy and/or radiation in CRC patients aged 80 years or older. Over a 10 year period, 29 patients were identified for the study although all the stages for CRC were included and some of these patients were treated with curative intent. The median age for that cohort was 82 years (range: 80–93) and their reported mOS for patients with stage IV disease was 6 months (95%CI: 5–33). Reddy et al. [28] worked in a similar retrospective review in this setting, including also octogenarians and nonagenarians with mCRC on chemotherapy or chemo-radiation therapy. Thirty patients in their 80’s and 90’s were treated with palliative intent in that study and, in spite of the starting attenuated dose of chemotherapy and the high rate of treatment toxicity, the mOS achieved among these patients was 20.6 months (95% confidence interval, 11.1–26.4 months). This likely reflects the number of patients aged less than 90 years of age. These authors suggested that selected octogenarians and nonagenarians with mCRC could benefit from chemotherapy, with similar outcomes seen when compared with their younger counterparts. The difficulty with all of these analyses is that patients aged between 80 and 90 years were also included. Rivoirard et al. [29] have attempted to assess a nonagenarian population from two institutions, although they have included multiple cancer types, including hematological malignancies, and only included 12 patients. Only three patients with mCRC diagnosis were included in the study. Given the very heterogeneous cancer population included in the study overall, the diverse chemotherapy treatments and unequal number of chemotherapy lines is very difficult to draw any clear conclusion.

There are limitations of our analysis. We do not have data on co-morbidities to understand potentially why so few patients received systemic therapy. We do not collect data on whether these decisions were driven by the patients and their relatives or the clinicians themselves.

Conclusions

This analysis gives us some insight into the outcomes for the very old and is likely the largest series focused on patients aged 90 years and over. Female sex and right-sided primary cancers are more frequent. Systemic therapy is rare and the survival is poor. This may well be appropriate based on frailty, co-morbidity and patient expectations. However, as the number of patients reaching 90 years will increase over time further research to understand whether active therapy is possible or warranted in this age group should be a considered. Furthermore, it opens the interesting question of resource allocation and whether society can offer all therapy options without age cut offs.

Disclosure statement

No potential conflict of interest was reported by the authors.