Reprint of: Aging: Treating the Older Patient

THE OLDER MYELOMA PATIENT: EVALUATING AGING AND UNDERSTANDING FITNESS

Overview

Multiple myeloma (MM) is an incurable plasma cell malignancy of older adults. The median age at diagnosis is 69 years, and in the next 15 years the incidence of MM is expected to double in this age demographic [1,2]. Novel therapeutics and the routine use of autologous stem cell transplant have led to substantial improvements in overall response rates and durable remissions [3]. In contrast with younger MM patients, older patients have only modest improvements in overall survival [4–6]. MM deaths overall are highest in patients aged 75 years and greater, and early mortality is most common in those 70 years and older [3,7]. The disparity in overall survival for aging adults is multifactorial and is secondary to comorbidities, treatment strategy, toxicities, physiologic reserve, and therapy discontinuation.

Transplant “eligibility” is an active and important area of MM investigation. Autologous stem cell transplant is established in younger populations to improve survival over nontransplant therapy [8,9] and has progression-free survival advantages over delayed transplant [10]. Autologous stem cell transplant is feasible and an efficacious component of therapy for older patients with MM as well [11]. Older adults mobilize sufficient numbers of stem cells and can tolerate transplant with excellent outcomes, resulting in increased numbers of older adults undergoing autologous transplant [11]. Nearly half of autologous transplants are done in adults 60 years and greater, and this number will only increase as the population ages [12]. Referral bias for transplant still exists, likely because of historic reports depicting conflicting tolerance, response rates, and survival [13].

Transplant Eligibility

Transplant eligibility is matter of estimating a patients’ physiologic reserve for an intensive therapy [14]. Identifying and intervening on factors that contribute to vulnerability in pretransplant MM patients is imperative to balance quality of life with an efficacious therapy. One method to identify and resolve occult health factors is a geriatric assessment (GA), which is a global evaluation of the health of older adults. A GA goes beyond the disease-focused history and aims to identify unrecognized issues to intervene and prevent future complications. GA tools are established metrics to accurately assess risk of morbidity and mortality in cancer populations [15,16]. The GA consists of a multidimensional evaluation of functional status, fall history, social support, cognitive and psychological status, sensory loss, nutritional status, comorbidities, and a polypharmacy evaluation.

Table 1 depicts a set of tools often used in a cancer-specific GA. GAs have been shown to predict mortality and toxicity, independent of performance status and age [17]. Traditional metrics, such as Karnofsky performance status, are often overestimated by clinicians and are a poor indicator of treatment toxicity risk [18].

Table 1.

Geriatric Evaluation for Older AML Candidacy for Allogeneic Transplant

Domains	Common Tools	Additional Geriatric Questions
Comorbidity	HCT-CI	Remote cancer, urinary problems, visual or hearing impairment, diastolic dysfunction, prior renal impairment, osteoporosis
Function	IADL Timed up and go, grip strength, 4-m walk	Arthritis, falls, maximum ability for physical activity, exercise, balance
Social support and function	Illness-specific subscales of social support Center-specific tool	Backup caregiver, alcohol use, stairs at home, person preparing meals, power of attorney
Cognition	Mini-Mental State Exam or Montreal cognitive assessment	Prior confusion, memory impairment and duration Assess retention of information without family member interjection
Psychological	Geriatric Depression Scale Mental Health Inventory 17	Sleep problems, motivation for transplant, coping skills, preparation for setbacks, life goals (eg, specific event)
Nutritional status	Weight loss, body mass index	Last dental evaluation, dentures, use of supplements, effect of prior therapy on nutrition and weight
Polypharmacy	>5 medications	Over-the-counter medications, side effects from prior medication, clarify remote allergies

The treatment approach for MM is heterogeneous because of concerns for frailty and tolerance in older adults. Primary dose reductions and therapy discontinuations result in worse outcomes in older adults with MM [19,20]. Although it is known that functional assessments can improve on outcomes in cancer patients, feasibility, practicality, and disease specificity are barriers to implementation [21–23]. Nonetheless, a simple GA based on age, comorbidities, cognition, and physical function can predict mortality in the MM population. Frailty assessments in MM patients were predictive of death independent of treatment, cytogenetics, or stage [20].

Biomarkers of Physiologic Age

Novel aging biomarkers are being explored to provide a rapid measure of physiologic fitness [24,25]. Ideally, an aging biomarker would capture vulnerability by estimating physiologic reserve and risk for chemotherapy and/or transplant toxicity. Exploring biomarkers of aging in a cancer population is particularly challenging, because many aging biomarkers are evaluated in a community-dwelling population without a cofounding factor of cancer. Despite these challenges, many potential biomarkers of aging are being explored to evaluate the relationship among chronologic aging, frailty syndromes, and cancer. Candidate biomarkers include molecular markers, inflammatory markers, immunosenescence panels, serum and hematologic parameters, and hormones (Table 2). Each of these biomarkers has associations with aging and frailty and variable relationships with autologous hematopoietic stem cell transplant and myeloma. Inflammatory markers such as IL-6, C-reactive protein, tumor necrosis factor-α, and D-dimer have reported associations with frailty in oncology studies [26]. Inflammatory panels such as the senescence-associated secretory phenotype has a well-established relationship with aging, gerontologic conditions, and frailty [27].

Table 2.

Aging and Frailty Assessments

		Aging	Frailty	Transplant	Myeloma
Molecular markers	p16INK4A	x		x
	Leukocyte telomere length	x
	DNA methylation	x
	miRNA
Immune dysregulation	Immunosenescence	x	x		x
	SASP	x	x
Heme parameters	Anemia	x	x
Serum markers	IL-6	x	x	x	x
	CRP	x	x
	NT-proBNP		x		x
	Albumin	x	x	x
	D-dimer	x	x
	TNF
	sICAM-1		x
Clinical tools	GA metrics	x	x	x	x

SASP indicates senescence-associated secretory phenotype; CRP, C-reactive protein; NT-pro-BNP, N-terminal fragment of B-type natriuretic peptide; TNF, tumor necrosis; sICAM-1, Soluble intercellular adhesion molecule-1.

Distinct changes in the immune system are reported with aging [28,29] and are being explored with frailty [30,31]. Many of the distinct changes in T cell dysfunction found in MM patients parallels that of aging and frail populations. T cell dysfunction is commonly reported in MM where markers of immunosenescence are associated with relapsed disease [32]. Clonal expansions of T cells are common with aging and are also present with long-term survival in MM [33], although the biology of clonal expansions may differ. Anemia, inherent to MM, has been independently associated with functional disability in older adults [34]. miRNA expression profiles have been shown to correlate with age in healthy populations [35], cardiovascular disease [36,37], but not frailty in solid tumor cancer populations [38]. Other biomarkers, such as N-terminal fragment of B-type natriuretic peptide, have recently been shown as useful predictors of survival, independent of age and performance status in MM patients [39]. Markers of T cell immunosenescence and exhaustion have recently been reported to have an association with relapse after MM transplant [40].

A molecular marker of cellular senescence is peripheral blood T cell p16^INK4A (p16). p16 mRNA accumulates in aging tissues and increases with chronologic aging and with internal and external stresses [41–44]. Over the human lifespan, p16 expression levels increase more than 16-fold in peripheral blood T cells [45]. There is considerable evidence that p16^INK4A is 1 of the most robust and validated aging biomarkers. p16^INK4a is being explored for risk of chemotherapy toxicity and transplant tolerance [46,47]. External stressors can trigger p16 expression such as physical inactivity [48], tobacco use, or chemotherapy [47,49].

Aging biomarkers paired with geriatric metrics may provide a novel method to gauge risk of transplant toxicity in the MM population. The need for objective biomarkers of aging is especially important in the field of MM because of the aged population affected, heterogeneity in older adult fitness, and diverse treatment strategies that are available. Ultimately, an individualized approach to patient fitness based on geriatric metrics and aging biology can aid the clinician in decision-making for myeloma transplant.

THE OLDER ACUTE MYELOID LEUKEMIA PATIENT: CANDIDACY AND OPTIMIZATION FOR ALLOGENEIC TRANSPLANT

Overview

The potent antileukemic effects of allogeneic hematopoietic cell transplantation must be balanced against transplantassociated morbidity and mortality, especially for older patients. Reduced-intensity allogeneic transplant achieves rates of 30% to 50% in 2-year survival for acute myeloid leukemia (AML) in patients 60 years and older. Emerging data highlight the importance of physiologic reserve beyond chronologic age alone. Multidimensional health surveys such as the GA encompassing comorbidity, function, social support, cognition, nutrition, and other domains identify unrecognized and prognostically relevant health impairments. Age-related vulnerabilities may predispose to toxicity because of inherent physiologic stress of transplant. Enrollment on prospective studies is essential to delineate the most effective risk-stratification tools to inform transplant candidacy. Interventions to attenuate transplant risks based on health vulnerabilities in older patients are proposed, although prospective studies are mandatory.

Epidemiology of AML in Older Adults

Advancing age represents the primary risk factor for AML diagnosis and disease fatality. The median age at diagnosis in Western countries is around 67 to 70 years, and older age is strongly associated with worse survival (seer.cancer.gov) [50]. Adverse biologic factors inherent to the disease, competing health conditions, and less treatment together interfere to limit long-term success. The biology of older age as a risk for AML and inferior outcomes remain an area of interest, although age-related clonal myeloid hematopoiesis appears to predispose to leukemia and possibly chemotherapy resistance clones [51].

A large series of AML patients enrolled on Cancer and Leukemia Group B trials illustrates the influence of age on outcomes. Patients 60 years and older had less frequent favorable risk disease using the European LeukemiaNet criteria and more frequent adverse cytogenetics (31% adverse disease for older versus 22% in patients under 60 years). For the favorable risk group, only 24% of older patients achieved 3-year disease-free survival compared with 55% in younger patients. Nontransplant protocol–based treatment rarely accomplished leukemia control for intermediate and adverse risk disease: rates of 3-year disease-free survival were 11% and 6%, respectively [52]. The large United Kingdom National Cancer Research Institute AML16 trial of intensive therapy showed similar results, with rates of 5-year survival of around 14% to 15% [53].

Older Age and Allogeneic Transplant for AML

The historical barrier of older age to allogeneic hematopoietic cell transplant has slowly been lifted over the past 2 decades. The Center for International Blood and Marrow Transplant Research (CIBMTR) showed 22% of allogeneic transplant recipients were 60 years or older in 2007 to 2013 among common malignant conditions [54]. Likewise, US allografts for patients 70 years and older have risen 10-fold over the past decade, with AML as the leading indication (Figure 1) [55]. Few older AML patients receive allogeneic transplant; a population study of AML diagnosed between 2000 and 2007 showed only 4% of AML patients 65 to 74 years underwent allografts [56]. Swedish data described a marked uptick in transplant use for those 61 to 70 years in recent years yet no transplants for those 70 years and older for AML [57].

Figure 1.

Allogeneic transplant malignant indications and change over time in the United States reported to the Center for International Blood and Marrow Transplantation Research.

Transplant outcomes for older AML patients

In a review of studies of adults 60 years and older undergoing allogeneic transplant for AML among 749 patients, 13 studies reported rates of overall survival at 1, 2, and 3 years of 58%, 45%, and 38%, respectively [58]. Devine et al. [59] published a multi-institutional prospective study of a low-intensity fludarabine-busulfan-antithymocyte globulin regimen using matched related and unrelated donors for AML in first remission. Patients 60 to 74 years of age achieved a favorable 2-year survival rate of 48% and a low nonrelapse mortality (NRM) rate of 14%. Registry data for AML and myelodysplastic syndrome patients 70 years and older using all donor sources and conditioning regimens showed 2-year survival rate of 38% and 2-year NRM rate of 30% [55]. Haploidentical and umbilical cord graft studies in older AML patients have been described [60–62].

Advancing age still confers greater risks of transplant-related mortality for AML in first remission [63]. However, limiting the comparison to HLA matched grafts after reduced-intensity conditioning showed no major influence of age alone [64,65]. A report by Versluis et al. [66] derived from prospective AML studies of patients 60 years and older revealed reduced-intensity transplant from matched related or unrelated donors improved 5-year overall survival rates at 35% versus 26% for other forms of consolidation therapy (hazard ratio [HR] = .71, P = .017).

Provisional Recommendations on Candidacy

GA and risk categorization

No well-designed studies have characterized the morbidity and risks of transplant in older adults of hematopoietic cell transplant versus nonhematopoietic cell transplant approaches. Decisions to pursue allogeneic hematopoietic cell transplant demand individualization and attention to disease status, donor type, regimen, center experience, and patient goals. A GA has been the standard tool within geriatric oncology to better characterize health in older patients. The GA in Table 1 was adopted from the Cancer and Aging Research Group and supplemented by additional medical history that may be relevant to transplant [67,68]. The extensive resources committed for transplant and appreciable risks may justify this low-risk low-cost evaluation. Table 3 illustrates a schema used at our institution that divides patients into anticipated transplant-associated mortality risk into fit (ie, consider similar to younger patients), vulnerable (higher risk but transplant risks may be manageable, review optimization strategies), and unfit (avoid transplant because of excessive risks unless on trial or exceptional circumstances). This schema has not been tested and should not be considered a guideline. The available data and/or rationale for such criteria are summarized below.

Table 3.

Provisional Framework for Allogeneic Hematopoietic Cell Transplant Eligibility Criteria for AML Patients 60 Years and Older with Adequate Disease Control

Factor	Fit (All of the Below)	Vulnerable (Any of the Below)	Usually Exclude or Trial (Any of the Below)
Age, yr	<66	66–75	>75
Comorbidity	HCT-CI 0–2	HCT-CI 3–5	HCT-CI 6+ or severe organ dysfunction
KPS, %	90–100	70–80	<70
Function
IADL*	Normal	Any limitation	Limitations and HCT-CI 4+
Performance†	No impairment	Impaired	Severely limited
Cognitive	No impairment	Mild cognitive impairment, prior confusion	Moderate cognitive impairment
Other factors	Normal	Weight loss, single or limited caregiver, falls, albumin < 3.5 g/dL, limited motivation	Weight loss, single or limited caregiver, falls, albumin < 3.5 g/dL, limited motivation + others

KPS indicates Karnofsky performance score.

^*IADL used here omits the category of ability to do laundry.

^†Performance was measured by timed up and go, 4-m walk, and grip strength; 6-minute walk used for prehabilitation baseline.

Age

The preponderance of data supporting allografting for older AML patients has been in patients 65 years or less and sparse for 70 years and above [55,69]. Ideally, patients above 75 years of age would undergo transplant on clinical trials specifically addressing age-related issues. Chronologic age alone between 60 and 75 years will not usually dictate transplant candidacy, and complementary evaluation may aide decision-making. Patients younger than 60 years may also benefit from a comprehensive health evaluation.

Comorbidity

The hematopoietic cell transplantation–specific comorbidity index (HCT-CI) summates various comorbid conditions extracted from history and objective testing [70]. In adults 60 years and older undergoing nonablative HLA matched allografts, Sorror et al. [71] showed a borderline significant increase in NRM for intermediate comorbidity of HCT-CI scores of 1 to 2 (HR, 1.69; 95% confidence interval [CI], .91 to 3.14) and 3 or more (HR, 2.0; 95% CI, 1.12 to 3.6) relative to no comorbidity and inferior survival for those with the higher comorbidity. In a registry validation, rates of 1-year allogeneic transplant NRM were 17%, 21%, and 26% for HCT-CI scores of 0, 1 to 2, and 3, respectively (P < .001). NRM did not differ for HCT-CI scores of 1 to 2 relative to 0 after reduced-intensity regimens. High comorbid burden (ie, HCT-CI score of 5 or more) exacted a profound effect on NRM (HR, 1.77; 95% CI, 1.50 to 2.10). Similar effect sizes for NRM and survival have been reported by degree of comorbid burden [72]. Comorbidity and age are additive to the prognostic effects on NRM and survival [71,73,74]. Thus, older patients entering transplant with HCT-CI scores of 3 to 5 are vulnerable, and at our center we simply avoid transplant for patients with HCT-CI scores of 6 unless exceptional circumstances or trial. Following the scoring guidelines by Sorror [75] will maximize accuracy in tabulating comorbidity and likely prognostic discrimination.

GA and Functional Status

Functional status

Low physician rated Karnofsky performance status (eg, <60% to 70%) often excludes patients from allogeneic transplant consideration and among older transplanted patients. A Karnofsky performance status of 80% or less may have worse outcomes relative to 90% to 100% [76].

Patient-reported and/or performance-based function is an essential core element of a multidimensional GA. We reported a high frequency of patient-reported functional impairments for patients 50 years and older before allografting because 40% had at least 1 limitation in a core functionally tool, the Lawton Instrumental Activities of Daily Living (IADL), encompassing the ability to manage medications, finances, meals, grocery shop, telephone, transportation and driving, and housekeeping/chores [77]. Other reports from Houston and San Francisco have likewise demonstrated a high frequency of functional vulnerabilities by GA testing before transplant [78,79]. In a Chicago cohort (N = 203) the presence of any IADL limitation independently imparted high risks of NRM and inferior survival but not relapse rates. Adverse effects were amplified in patients 60 and older (HR, 3.3) relative to patients 50 to 59 (HR, 1.9), although CIs overlapped. Patient function through a self-report quality of life instrument also conveyed worse 1-year survival rates in a multi-institutional study by Wood et al. [80] of all ages supporting the premise of patient-reported function for prognostication.

A slow 4-m walk speed also predicted higher mortality in the Chicago series. In studies of adult patients of all ages, cardiopulmonary fitness tools of 6-minute walk distance or cycle ergometry may risk-stratify patients [81,82]. Whether patient-reported function, performance-based function, or both will be optimal is unknown. Other studies in geriatrics and geriatric oncology support the IADL to stratify for outcome using this convenient survey tool [67,83].

Other Health Domains

Emotional health

The Houston and Chicago series showed surveys of emotional health uncovered a higher frequency of emotional health impairments than by the HCT-CI alone [78]. The Mental Health Inventory-5 demonstrated 36% of patients in the Houston study expressed anxiety or depression. Using the Short-Form 36 quality of life instrument mental composite summary subscale, slightly over half of the patients in the Chicago GA indicated mental health at least 1 standard deviation below population normals [74]. Emotional health and motivation are clearly relevant factors, although the attendant risks from these disturbances require additional study.

Cognition

The Houston series on GA described a 16% prevalence of mild cognitive impairment before transplant [78]. In 27 unselected patients 60 years and older before allografting, we found a 10% rate of cognitive impairment by the Mini-Mental Status Exam (A. Artz and M.Q. Lacy, personal communication, 2015). A diagnosis of cognitive impairment should not be based on screening tests alone. One approach to establish cognitive impairment is to perform neuropsychological testing and/or consult geriatrics. The consequences of allografting in cognitive impairment are unknown, although Klepin et al. [84] reported pretreatment cognitive impairment negatively affected early survival after receipt of AML induction therapy in older adults. Therefore, one must exercise caution in pursuing allogeneic transplant if cognitive deficits are uncovered.

Composite models and biomarker

Data-driven models should be developed to maximally guide candidacy. Multidimensional risk scores using comorbidity, function, and age, if not other markers as used in geriatric oncology and noncancer older populations [67,85], will likely emerge. For example, we found a combination of high comorbidity (ie, HCT-CI scores of 3 or more) and functional impairment by IADL produced a simple 3-level scoring system with appreciable discrimination: 2-year survival rates exceeded 60% for no risk factors compared with 29% and 0% for those with 1 or 2 risk factors in patients 60 years and older, respectively. Biomarkers such as elevated C-reactive protein and depressed albumin add prognostication discrimination to comorbidity and age [72,86]. Hypoalbuminemia (<3.5 g/dL) can be applied as a readily available biomarker in light of the established and independent risks of NRM.

Regimen Intensity

The decision of regimen and graft sources is beyond the scope of this review, yet risk stratification must account for the intended regimen and graft source. Unfortunately, a low-intensity regimen may not necessarily improve outcomes for older adults because of heightened risks of relapse. The preliminary data of a benefit in relapse-free survival of myeloablative chemotherapy for AML and myelodysplastic syndrome in patients up to age 65 years with low comorbidity suggest one may consider an ablative regimen for older fit AML patients [87].

Transplant Optimization

Rationale

Multidisciplinary optimization strategies to reduce transplant morbidity are attractive when considering the substantial risks associated with specific vulnerabilities and rapid transplant growth in patients 70 years and older. The promise has not been prospectively tested.

Standard Platform

Team approach

Our center has attempted to leverage vulnerabilities detected by GA by assembling an interdisciplinary team to develop targeted interventions as outlined in Table 4. The team consists of a dietician, social worker, geriatric oncologist, transplant physician (often separate from the primary transplant physician), transplant nurse, transplant advanced practice provider, physical therapist, and a data or clinical coordinator. Most team members exist in transplant programs. Other team members to consider include a pharmacist, psychologist, and patient advocate. Cultivating relationships with medical subspecialists for prevalent comorbid conditions such as cardiac and pulmonary dysfunction promotes a mindset of optimal monitoring and management and away from clearance alone.

Table 4.

Standard Optimization Platform for Older Allogeneic Transplant Recipients Used at University of Chicago

Health Area	Standard Optimization Recommendations
GA	Interventions individualized based on GA-detected limitations in Table 2. Multidisciplinary team meeting and written optimization plan before transplant.
Comorbid conditions	Cultivate preferred subspecialists and request post-transplant follow-up. Referral to geriatrics or geriatric oncology for geriatric syndromes (eg, falls, incontinence).
Function	Structured prehabilitation based on baseline activity and limitations. Home safety assessment (fall risk, stairs, proximity of bathroom).
Social support	Pretransplant family meeting, caregiver present during cytopenia and “on-call” if issues. Request secondary caregivers. Enlist caregivers in optimization plan.
Cognition	Delirium awareness and precautions, medication avoidance, encourage presence of caregivers, add written material to verbal instructions, call patient to reinforce education.
Emotional health	Offer cognitive therapy and support group to avoid additional medications unless required
Nutrition	Develop nutritional plan before transplant included preferred supplements if used.
Medications	Reduce unnecessary medications. Re-evaluate days 30 to 100 post-transplant.
Other	Infection mitigation plan for grandchildren, evaluate remote drug allergies (eg, penicillin allergy in childhood), more frequent follow-up, smooth survivorship transition from transplant center.

Specific interventions

Randomized transplant studies of exercise during and after transplant have shown only modest benefits, and data were primarily derived in younger adults at lower risk of functional disability [88]. Structured exercise may be preferred over low intensity education based on the reduction in disability in nontransplant older adults [89]. Starting before transplant (prehabilitation) will likely maximize yield, although extended time for prehabilitation in AML patients may not be feasible. The conditioning regimen of transplant may still be an opportunity to gain strength and endurance and can be combined with prehabilitation. Medication avoidance can follow the updated Beers criteria and requires diligence [90].

Augmented social support can facilitate patient monitoring, compliance, and optimization across domains. Older adults often rely on older spouses or siblings who may have their own health issues. A family meeting of all potential caregivers may combat patient reluctance to enlist children or other social support. We routinely request a backup care-giver in case the primary caregiver becomes ill and inquire about the health of the primary caregiver. During anticipated cytopenia and toxicities from conditioning therapy (when such a regimen is used), a fortified caregiver presence (ie, 24/7 available) is a consideration.

GA-guided interventions

Based on the GA, specific vulnerabilities often emerge for which individualized therapy can be directed often by nonphysician team members (eg, dietician, physical therapist, social worker). For example, dentures may prompt regular dietician input and modified nutritional recommendations. Subtle or mild cognitive impairment justifies greater attention to medication avoidance, focus on sleep hygiene, and consideration of serial monitoring for delirium (including caregiver education) when hospitalized and/or geriatrics consultation. Weakness in a specific arm may guide placement of central venous catheters, modified exercises, initiation of an assist device (eg, cane, walker), and special instructions on rising from a bed or chair.

Timing

Time to transplant may be 1 of the most modifiable factors and a key reason to characterize fitness for transplant. The high rates of relapse and short disease-free survival in older adults justify moving quickly to allogeneic transplant for appropriate candidates. A common clinical dilemma for the vulnerable older AML patient is the tendency toward a sluggish pursuit of transplant because of uncertainty about the harms and benefits of transplant. Here, the decision-making process for transplant may be as important as the criteria used [91,92].

Future Directions

Transplant studies must ensure adequate representation of older adults and compare results with nontransplanted treatment strategies. Large prospective studies incorporating multidimensional tools such as GAs are required to better inform transplant candidacy. Finally, optimization strategies should be tested to reduce transplant-associated morbidity and perhaps extend transplant to older and more vulnerable AML patients.