Impairment and Longitudinal Recovery of Older Adults Treated with Radical Cystectomy for Muscle-Invasive Bladder Cancer

Abstract

Purpose:

Treatment for muscle-invasive bladder cancer (MIBC) includes radical cystectomy (RC), a major surgery that can be associated with significant toxicity. Limited data exist related to changes in patients’ global health status and recovery following RC. We sought to use geriatric assessment (GA) to longitudinally compare health-related impairments in older and younger patients with MIBC who undergo RC.

Materials and Methods:

Older and younger patients (≥ 70 and < 70 years) with MIBC undergoing RC at an academic institution were enrolled between 2012 - 2019. Patients completed the GA pre-RC and 1, 3, and 12 months post-RC. For each GA measure, the Wilcoxon rank-sum test was used to compare score distribution between age groups at each time point. The Wilcoxon signed-rank test was used to compare distributions between time points within each age group.

Results:

80 patients (42 younger and 38 older) were enrolled. Pre-RC, 78% of patients were impaired on at least one GA measure. Both age groups had worsening physical function and nutrition at 1 month post-RC, with older patients having a greater decline in function than younger patients. Both groups recovered to baseline at 3 months post-RC and maintained this at 1 year.

Conclusions:

High rates of impairments were found across age groups in the short-term post-RC, followed by recovery to baseline.

Introduction:

Bladder cancer is the 10^th most common cancer worldwide with an estimated 549,000 new cases and 200,000 deaths in 2018.¹ It is primarily a disease of older adults with a median age at diagnosis of 73 years and nearly 75% of all diagnoses occurring in patients 65 years or older.² As the population in Europe and North American continues to age, the incidence of bladder cancer will continue to grow. With this significant burden of disease in older patients, it is crucial to understand how to optimize their care.

Geriatric assessment (GA) is a multidisciplinary evaluation of a patient’s overall health and includes measures of functional status, cognition, comorbidities, psychological status, nutrition, polypharmacy, socioeconomic support, and geriatric syndromes.^3,4 GA identifies impairments that are frequently overlooked during routine oncologic evaluation^4,5 and can predict risk for mortality or treatment-related toxicity.^6,7

For muscle-invasive bladder cancer (MIBC), standard of care is radical cystectomy (RC) with or without neoadjuvant chemotherapy (NAC), which can be associated with major morbidity. Prior studies evaluating how well older patients tolerate this treatment have been mixed, with some reporting higher perioperative morbidity and mortality compared to younger patients^8-11 while others demonstrated a benefit to aggressive treatment. ^3,12 Alternatively, patients with untreated MIBC experience frequent hospitalization due to disease progression and have a median survival of less than one year.^13,14

Patients with MIBC highly value understanding the impact of treatment on the recovery process¹⁵, yet there is a paucity of data regarding this. Given this knowledge gap, we utilized GA to prospectively evaluate impairments across the recovery period for older compared to younger patients with MIBC undergoing RC.

Materials and Methods:

Participants:

Participants had pathologically confirmed MIBC and were deemed eligible to undergo RC for definitive cancer treatment. As there is no consensus age to define “older” patients, those ≥ 70 years were considered “older” and < 70 as “younger” to reflect the median age at diagnosis of MIBC. Participants were excluded if they were unable to read or speak English or had any condition that would prohibit understanding of informed consent. All patients participated in a modern Enhanced Recovery After Surgery protocol.

Study Procedures:

Participants were enrolled at a urologic oncology office visit between April 2012 and December 2019. Participants completed the baseline GA and Functional Assessment of Cancer Therapy-Bladder (FACT-BL) questionnaires following NAC but prior to RC. Post-surgical GA and FACT-BL were repeated at 4-8 weeks (1-month assessment), 12-16 weeks (3-month assessment), and 11-14 months post-RC (1-year assessment), and all measures were included at all time points. Clinical, pathological, and surgical characteristics were abstracted from the electronic medical record. This study was approved by the Institutional Review Board at the University of North Carolina (UNC) (ClinicalTrials.gov identifier NCT01776138).

Study Measures:

Geriatric Assessment (GA)

A cancer-specific GA was validated by Hurria et al¹⁶ and has been utilized in numerous prior studies.^4,17-19 Four measures were administered by the research team: clinician-rated Karnofsky Performance Status (cKPS), Timed Up and Go (TUG), Blessed Orientation-Memory-Concentration test (BOMC), and calculation of body mass index (BMI) and percent unintentional weight loss. The remaining measures were patient-reported questionnaires completed in clinic or at home. Percent impairment was calculated as the number of measures a patient was impaired on out of total measures completed. Each measure with cut point signifying impairment is provided as supplemental material.

Carolina Frailty Index (CFI)

The CFI is based on the deficit accumulation model of frailty and is predictive of all-cause mortality in older adults with cancer.²⁰ It includes 36 items derived from the GA, providing a summary measure of a patient’s degree of impairment. The CFI is calculated as the number of deficits divided by the total number of items, giving a score between 0 and 1, and was calculated for each patient at each time point. Patients were categorized as robust (0-0.2), pre-frail (0.21-0.35), or frail (>0.35).

FACT-BL

The FACT-BL is a validated bladder cancer-specific measure of quality of life (QOL).^21,22 It includes five subscales: physical well-being, social/family well-being, emotional well-being, functional well-being, and bladder cancer-specific concerns. Items are rated on a 5-point scale with higher scores corresponding to better quality of life. The subscales are summed to provide a total FACT-BL score (range 0 – 156). A change in FACT-BL score of ≥ 10 is considered clinically meaningful.²²

Statistical Analysis:

Descriptive statistics were used to characterize demographics and clinical characteristics. The proportion of patients impaired on each measure by age group was compared using Fisher’s exact test. The Wilcoxon rank-sum test was used to compare the distribution of scores between age groups and the Wilcoxon signed-rank test was used to compare distributions within age groups between time points, including only patients that completed the measure at both times.

All patients completed the pre-RC assessment; however, not all patients completed every post-RC assessment, or they did so earlier or later than planned per protocol. Therefore, the distribution of time from surgery to completion of each assessment was evaluated, and patients in the 10^th percentile or 90^th percentile were excluded from analysis for that time point. Sensitivity analyses were performed to evaluate the potential influence of year of surgery, receipt of NAC, and choice of age cut point on results. All statistical analyses were performed using STATA v16.0.

Results:

Participant characteristics:

Of the 84 patients enrolled, 4 did not undergo RC and were excluded from this analysis. Most patients were white (84%) and male (78%) (Table 1). Median age was 62 years for younger patients (range 41-69; n = 42) and 75 for older patients (range 70-83; n = 38). There was no difference in clinical stage or frequency of mixed histology between groups (p=0.94; p=0.33, respectively). Younger patients were significantly more likely to receive NAC than older patients (83% vs 63%, p=0.05).

Table 1.

Clinical, pathologic, and surgical characteristics of included patients.

	Entire cohort (n = 80)	< 70 (n = 42)	≥ 70 (n = 38)	p value
Median age, years	69 (62 - 75)	62 (56 - 66)	75 (73 - 79)	< 0.001
Male	62 (78)	30 (71)	32 (84)	0.19
White	67 (84)	33 (79)	34 (90)	0.23
Clinical stage				0.94
Stage 2	46 (58)	25 (60)	21 (55)
Stage 3a	26 (33)	13 (31)	13 (34)
Stage 3b	5 (6)	2 (5)	3 (8)
Stage 4a	2 (2)	1 (2)	1 (3)
Unknown	1 (1)	1 (2)	0
Received NAC	59 (74)	35 (83)	24 (63)	0.05
NAC regimen				0.69
Gem/cis	39 (66)	24 (69)	15 (63)
ddMVAC	7 (12)	5 (14)	2 (8)
Gem/carbo	3 (5)	2 (6)	1 (4)
Etoposide/cis	2 (3)	1 (3)	1 (4)
Clinical trial	8 (14)	3 (8)	5 (21)
Histology				0.33
Pure urothelial	46 (58)	24 (57)	22 (58)
Pure spindle cell	1 (1)	1 (2)	0
Pure squamous	1 (1)	1 (2)	0
Pure small cell	1 (1)	1 (3)	0
Mixed	31 (39)	15 (36)	16 (42)
Pathologic T stage				0.40
pT0	15 (19)	11 (26)	4 (10)
pTis	10 (13)	5 (12)	5 (13)
pTa	2 (2)	1 (2)	1 (3)
pT1	3 (4)	3 (7)	0
pT2	8 (10)	4 (10)	4 (11)
pT3a	21 (26)	9 (22)	12 (32)
pT3b	2 (2)	0	2 (5)
pT4a	12 (15)	6 (14)	6 (16)
pT4b	7 (9)	3 (7)	4 (10)
Pathologic N stage				0.35
pN0	54 (68)	30 (71)	24 (63)
pN1	10 (13)	4 (10)	6 (16)
pN2	9 (11)	5 (12)	4 (11)
pN3	2 (2)	2 (5)	0
pNxa	5 (6)	1 (2)	4 (10)
Surgical approach				0.66
Open	34 (42)	19 (45)	15 (39)
Robotic	46 (58)	23 (55)	23 (61)
Any post-operative complication	62 (78)	32 (76)	30 (79)	0.80
Grade 3+ post-operative complication	23 (29)	14 (33)	9 (24)	0.46
Median hospital LOS, days	5 (4-6)	4 (4-5)	5 (4-6)	0.03
Post-operative care				0.02
Home without HH	31 (39)	22 (53)	9 (24)
Home with HH	46 (57)	19 (45)	27 (71)
SNF/LTACH	3 (4)	1 (2)	2 (5)
30-day readmission	21 (26)	11 (26)	10 (26)	1.0

Age and hospital length of stay reported as median (Interquartile range), all others as n (%). NAC regimen percentages reported as number of patients receiving each regimen out of total number of patients receiving NAC. Abbreviations: Carbo, carboplatin; cis, cisplatin; ddMVAC, dose-dense methotrexate, vinblastine, adriamycin, cisplatin; gem, gemcitabine; HH, home health; LOS, length of stay; LTACH, long-term acute care hospital; NAC, neoadjuvant chemotherapy; SNF, skilled nursing facility.

^aPatients that went for cystectomy but were found to have unresectable disease, and so no lymph nodes removed.

There was no difference in surgical approach (open versus robotic, p=0.66), all-grade post-operative complications (p=0.80), grade 3+ post-operative complications (p=0.46), or 30-day hospital readmission (p=1.0) between groups, but older patients did have a significantly longer median hospital length of stay (5 days vs. 4 days, p=0.03) and were more likely to be discharged home with home health (71% vs. 45%, p=0.02) than younger patients.

Baseline

At baseline, 78% of patients were impaired on at least one GA measure. Older patients had better scores for both tangible (i.e. someone to prepare meals or take them to the doctor; 100 vs 95, p=0.007) and emotional support (i.e. someone to confide in or provide advice; 100 vs 90, p <0.001) compared to younger patients (Table 2). There were no differences between groups on any physical function measure or in the proportion of patients impaired by domain (Table 2; Figure 1). CFI and the proportion of patients categorized as robust, pre-frail, or frail, was not different between groups (Figure 2). FACT-BL scores were not different between groups (Supplemental Figure 1).

Table 2.

Median scores by measure for each age group across time points.

	Normal Range	Baseline			1 month			3 months			1 year
		< 70 n = 42	≥ 70 n = 38	p	< 70 n = 27	≥ 70 n = 20	p	< 70 n = 21	≥ 70 n = 20	p	< 70 n = 18	≥ 70 n = 13	p
Up and go	< 14	9.5	11	0.13	12	12a	0.96	9	10.2	0.02	8.8	10.1	0.73
IADL	≥ 14	14	14	0.57	14a	12a	0.10	14	14b	0.13	14	14	0.20
Physical health	≥ 14	18	16	0.18	14a	9a	0.05	18b	14b	0.06	16c	13	0.09
Clinician KPS	≥ 80	90	90	0.81	80	70a	0.02	90b	90b	0.85	90	90	0.63
Patient KPS	≥ 80	90	90	0.51	80	60a	0.03	90b	95b	0.45	90	85	0.07
Falls	0	0	0	0.06	0	0	0.40	0	0	0.11	0	0	0.89
BOMC	≥ 11	4	4	0.67	2a	2a	0.28	0b	2	0.02	3	2	0.92
Medications	< 9	4	7.5	0.008	2	6	0.04	3	6	0.18	3	5.5	0.09
Comorbidities	< 4	1	2	0.16	1	1	0.59	1	2	0.06	1	3	0.03
Depression	< 12	6.5	6.5	0.82	6.5	7	0.84	3	6b	0.38	5	5.5	0.66
Anxiety	< 6	5	3	0.2	4	3	0.28	3	2	0.39	4	2	0.15
BMI	18.5 – 24.9	27.2	28	0.48	26a	27a	0.38	26.6	26	0.96	26c	27	0.61
Percent weight loss	<10%	0	−2	0.25	−5a	−7a	0.68	−2	−4	0.67	+0.3	−0.3	0.92
Social activities	>8	9	10	0.33	8a	7a	0.91	11	12b	0.88	11	11.5	0.86
Tangible support	>50	95	100	0.007	95	100	0.007	95	100	0.12	97.5	100	0.38
Emotional support	>50	90	100	0.0005	93	100	0.06	98.7	100	0.46	93.7	100	0.13
Percent impairment	N/A	12	14	0.83	23a	34a	0.44	7b	12b	0.66	13	13	0.81
CFI	< 0.2	0.1	0.12	0.35	0.19a	0.25a	0.22	0.1b	0.15b	0.22	0.07	0.16	0.17

Important clinically relevant differences are in bold font. Reported p value is for comparison between age groups at each time point. Abbreviations: BMI = body mass index; BOMC = Blessed Orientation Memory Concentration test; CFI = Carolina frailty index; IADL = instrumental activities of daily living; KPS = Karnofsky performance status; N/A = no established normal range.

^aComparison of scores within age group at baseline compared to 1 month significant at p < 0.05

^bComparison of scores within age group at 1 month compared to 3 months significant at p < 0.05

^cComparison of scores within age group at 3 months compared to 1 year significant at p < 0.05

Figure 1.

Percentage of patients impaired on each domain across time points for (A) patients under 70 years old and (B) patients age 70 and older. There were no significant differences between age groups in percentage of patients impaired on each domain at each time point.

Figure 2.

Percentage of patients categorized as robust, pre-frail, or frail at each time point for younger and older patients. Distribution of frailty categorization was significantly different between age groups at 3 months (p = 0.03).

* p < 0.05.

1 Month Post-RC

Percent impairment increased for both groups at 1 month compared to baseline (younger: 23% vs 12%, p=0.02; older: 34% vs 14%, p=0.003), primarily in the physical function and nutrition domains (Figure 1). Older patients had lower clinician- and patient-rated KPS (pKPS) and a greater decline from baseline in both KPS scores (−20 vs 0 for both) than younger patients. CFI was worse compared to baseline for both groups (younger: 0.19 vs 0.1, p <0.001, older: 0.25 vs 0.12, p <0.001), while FACT-BL scores were only worse for older patients (younger: 119 vs 120, p=0.97; older: 109 vs 123, p=0.01).

There was no significant difference in CFI by post-operative care type (home without home health = 0.19, home with home health = 0.24, p=0.40), presence of any grade or grade 3+ post-operative complication (no complications = 0.20, any grade complication = 0.21, p=0.99; no grade 3+ complication = 0.24, grade 3+ complication = 0.19, p=0.21), or 30-day hospital readmission (no readmission = 0.24, readmission = 0.19; p=0.70).

3 Months Post-RC

Percent impairment decreased at 3 months compared to 1 month (younger: 7% vs 23%, p=0.005; older: 12% vs 34%, p=0.02), and was not different compared to baseline for either group (younger: p=0.25; older: p=0.96). Compared to 1 month, both groups improved on physical health, cKPS, pKPS, and social activities. Older patients also improved on IADLs and depression, while younger patients improved on BOMC. However, a greater proportion of older patients remained impaired on physical health compared to younger patients (53% vs 16%, p=0.04). Both groups had persistently lower BMI at 3 months but there were no other differences compared to baseline.

CFI improved at 3 months compared to 1 month (younger: 0.1 vs 0.19, p <0.001; older: 0.15 vs 0.25, p=0.01), with no difference compared to baseline. A greater proportion of older patients remained pre-frail or frail at 3 months compared to younger patients (p=0.03). FACT-BL scores improved compared to 1 month (younger: 134 vs. 119, p=0.04; older: 127 vs 109, p=0.01). At 3 months, 45% of older and 31% of younger patients had an improvement in FACT-BL score of ≥10 compared to baseline, while 15% of older and 13% of younger patients had a decrease of ≥10.

1 Year Post-RC

Percent impairment was not different between groups (younger: 13%, older: 13%, p=0.81) and there were no statistically significant differences between groups on any measure at 1 year. Older patients had no differences in scores on any measure at 1 year compared to 3 months or to baseline. Younger patients had lower physical health scores and BMI at 1 year compared to 3 months, but had improvement in BOMC, social activity, and emotional support at 1 year compared to baseline. CFI and frailty categorization were not different between groups. FACT-BL scores were not different compared to baseline or 3 months for either group.

Sensitivity analyses

There were no significant differences in the proportion of patients receiving NAC (p=0.34), patients ≥ 70 years (p=0.32), or median CFI at each time point by year of surgery (baseline p=0.08; 1 month p=0.30; 3 month p=0.93, 1 year p=0.58). The pattern of results was not different using 65 as an age cut point rather than 70, with no significant difference in CFI between groups at any time. There was no significant difference in CFI at baseline or 1 month by receipt of NAC (Baseline: NAC = 0.1, no NAC = 0.13, p=0.39; 1 month: NAC = 0.21, no NAC = 0.22, p=0.85).

Discussion:

In our study of adults with MIBC treated with RC, patients experienced a high rate of impairment on GA measures at all time points regardless of their age. In both older and younger patients, we observed a pattern of worsening functional status, increased frailty, and decreased QOL at 1 month post-RC, followed by recovery to baseline at 3 months and maintenance of recovery at 1 year. However, older patients had a greater decrease in physical function and QOL at 1 month and were more likely to remain pre-frail or frail at 3 months compared to younger patients. Understanding the impact of treatment on overall health and function is important to patients with MIBC,¹⁵ and our study provides preliminary evidence that might inform patient-provider communication regarding expectations of the recovery process.

Deciding upon the best course of MIBC treatment can be complex. Treatment with curative intent has the potential for significant toxicity, but also the potential for significant benefit in terms of longer survival and reduced cancer-related symptoms. A thorough assessment of each patient is required to determine the most appropriate strategy. Our study demonstrates the utility of GA in providing a comprehensive assessment of MIBC patients, as it reveals important impairments that may negatively impact treatment tolerance yet otherwise be overlooked. Our study also demonstrates that these impairments are not exclusive to older patients, underscoring the concept that functional age does not equate with chronological age. Similarly, a study using GA in women with breast cancer found substantial deficits in patients of all ages, including reduced physical function in over 60% of patients under 50 years old.²³ These data suggest that GA can be beneficial in the evaluation of all patients, not just those considered “geriatric.”

While performing GA more widely is an important step in enhancing cancer care, we must also move beyond assessments and toward interventions. Our patients had high rates of GA-identified impairments for which intervention is possible. Few studies have attempted to use GA to guide interventions in the oncology setting. However, these studies have relied on the primary oncologist to execute interventions and have suffered from low implementation rates.^19,24,25 Better success has been demonstrated in surgery. In one study, patients undergoing vascular surgery were randomized to pre-operative GA with impairment optimization or standard of care.²⁶ Patients in the intervention arm had a 40% reduction in hospital length of stay, decreased post-operative delirium, and fewer cardiac complications than controls. A similar strategy of GA-guided intervention could be applied to patients with MIBC planning to undergo cystectomy, incorporating this into existing Enhanced Recovery After Surgery protocols.

While our study is the first description of longitudinal changes in impairments among cystectomy patients, it does have important limitations. First, only patients who underwent RC were included. These patients may have better baseline health than the overall bladder cancer population, some of whom may not be surgical candidates. Second, our study population was primarily white males. Although this is representative of the demographics of patients with MIBC, our results may not be generalizable to all patients, particularly members of minority groups. Third, the initial assessment was completed post-NAC, and therefore may not represent a true “baseline” for patients who received NAC. This was unlikely to significantly impact our results as the median time from NAC completion to baseline assessment was 36 days, allowing time for patients to recover from chemotherapy, and there was no difference in CFI at baseline or 1 month by receipt of NAC. Not all patients completed their assessment at every time point, however post-RC assessment completion did not differ by baseline frailty or age group. Finally, we were unable to apply statistical methods to control for possible confounders due to our small sample size. However, the groups were very similar in terms of key pathologic and perioperative characteristics. Despite these limitations, our study is the largest evaluation of GA among patients with MIBC and is the first to use repeated assessments to evaluate changes over the course of treatment.

Future studies should investigate similar repeated GAs among MIBC patients who undergo tri-modality therapy or receive systemic treatment for metastatic disease, and how these outcomes compare to those undergoing RC. Understanding impairments among these patients would further inform patient-provider conversations regarding treatment options in the context of functional outcomes.

Conclusions:

We described longitudinal changes in impairments of older compared to younger patients undergoing cystectomy for MIBC using GA. High rates of GA-identified impairments were noted across age groups with a pattern of worsening physical function in the short term post-operatively followed by recovery to baseline. Further work is needed to evaluate impairments in patients receiving other treatment modalities and to develop programs to successfully intervene on these impairments.

Abbreviations:

BOMC

Blessed Orientation-Memory-Concentration

BMI

body mass index

CFI

Carolina Frailty Index

cKPS

clinician-rated Karnofsky Performance Status

FACT-BL

Functional Assessment of Cancer Therapy-Bladder

GA

geriatric assessment

IADLs

instrumental activities of daily living

MHI

Mental Health Index

MOS

Medical Outcomes Study

MIBC

muscle-invasive bladder cancer

NAC

neoadjuvant chemotherapy

OARS

Older Americans Resources and Services

pKPS

patient-rated Karnofsky Performance Status

QOL

quality of life

RC

radical cystectomy

TUG

Timed Up and Go