Management of Inflammatory Bowel Disease in the Elderly Patient: Challenges and Opportunities

Abstract

The population of older patients with inflammatory bowel disease (IBD) continues to grow, partly reflecting the aging global population in general. The debilitating effects of IBD compound age-related decrements in health and functional capacity, and make the medical management of older patients with Crohn’s disease and ulcerative colitis distinctly challenging to clinicians. Here, we review the recent literature describing the pharmacologic management of IBD in this population, with focus on the safety, tolerability, and efficacy of common treatment options, such as steroids, immunomodulators, tumor necrosis factor-α antagonists, and integrin antagonists; surgical interventions in older patients are also discussed. Few studies have systematically and prospectively evaluated the clinical challenges in the medical management of IBD in this patient population, leaving a limited evidence base to which clinicians can turn to for guidance. Treatment patterns may thus be suboptimal. For example, prolonged steroid use in the elderly was found to be common, causing significant morbidity from side effects in a particularly vulnerable population. Finally, within the context of a limited evidence base, we discuss common treatment scenarios to define the parameters within which physicians can individualize care for older patients with IBD. Overall, older IBD patients are at higher risk of adverse events and less treatment responsiveness compared with younger patients, underscoring the need for future studies to fully characterize appropriate treatment courses for this population.

INTRODUCTION

Ulcerative colitis (UC) and Crohn’s disease (CD), collectively termed inflammatory bowel diseases (IBD), are chronic diseases of the intestine with protracted courses of relapse and remission that often result in progressive bowel damage, hospitalization, and surgery.(1) The age of onset is typically 20 to 40 years, with some studies reporting a second rise in incidence rates in those aged 60 to 70 years.(2) An estimated 10% to 30% of patients with IBD worldwide are aged >60 years (Figure 1A,B).(3, 4) The incidence of IBD in this age cohort was estimated to be between 4 and 8 per 100,000 patients per year, accounting for up to 20% of all new IBD diagnoses.(3) As the global population ages (Figure 1C), the prevalence of IBD among the elderly is also expected to increase.(2)

Figure 1. Rise in the IBD over the years.

(A) Incidence rates of UC and (B) CD in Olmsted County, Minnesota between 1940 and 2000.(4) Unadjusted incidence rates for males and females were added to give the total cases per 100,000 person-years (PY). For comparison, the total age- and sex- adjusted rate between 1990 and 2000 was 8.8 per 100,000 PY for UC and 7.9 per 100,000 PY for CD. (C) General population of the United States by age group according to the United States Census Bureau.^a Projected numbers are reported in the millions.

^a2014 National population projections: Summary tables.

The diagnosis and management of IBD in elderly patients pose unique challenges. The differential diagnosis of IBD in elderly patients is complicated by comorbid conditions such as infectious colitis, segmental colitis associated with diverticular disease, nonsteroidal anti-inflammatory drug-induced intestinal injury, and ischemia, each of which can mimic the intestinal inflammation characteristic of IBD.(5) Consequently, as many as 60% of older CD patients may be initially misdiagnosed, a fourfold increase relative to younger patients.(6) Similarly, the delay in diagnosis among elderly patients may be as long as 6 years, 3 times longer than the estimated delay in younger patients.

Between 10%–15% of patients are diagnosed with IBD after the age of 60 years.(7) Late-onset IBD differs from early onset in many aspects, including location and phenotype of the disease at diagnosis, symptoms at presentation, and natural disease progression.(7) It is generally accepted that a less severe disease course occurs in late-onset UC patients despite having more severe initial attacks than early-onset patients.(7, 8) A recent meta-analysis and systematic review investigating clinical outcomes in older-onset IBD demonstrated that rates of surgical interventions were higher in older-onset than younger-onset UC at years 1 and 5 after diagnosis but similar in older-onset CD compared with younger-onset disease.(9) It is unclear whether higher rates of surgical interventions are driven by a less benign natural disease course in older-onset disease, reluctance of physicians to employ immunomodulators, aggressive treatment regimen in elderly patients or complications due to comorbid conditions.

In the absence of a definite medical cure, the current standard of care for UC and CD is long-term immunosuppressive therapy. However, age-related changes in immune function and the gut microbiome may increase susceptibility to infections(10) making chronic immunosuppressive therapy in the elderly population less than ideal. In a retrospective case-control study, IBD patients aged >50 years had greater odds of opportunistic infections relative to those aged <24 years (odds ratio [OR]: 3.0; 95% confidence interval [CI]: 1.2–7.2; P = 0.05).(11) An age-related increase in cancer incidence further complicates the risk of immunosuppression in individuals at high risk for or with a prior history of cancer. Further, comorbid illness and decline in overall health tend to complicate therapeutic planning and adjustments to IBD regimens. Physiological changes and physical limitation in elderly patients can also impact treatment choices. For example, anal sphincter incompetence and limited coordination skills may impact the patient’s ability to successfully self-administer topical aminosalicylate or corticosteroid therapy. In addition, treatments that require frequent laboratory monitoring or regular trips to infusion centers may be more challenging for older patients due to logistical considerations.(12) Therefore, treatment goals in older IBD patients may focus on immediate symptom relief and improvement of physical functioning to maintain independence—very different from the goals of younger cohorts who may prioritize prevention of long-term disease-related complications. These and other factors should be carefully considered when making treatment choices for elderly patients.

Although recent literature has documented the efficacy and safety of IBD treatments in the elderly,(13–18) there are substantial gaps in our understanding of the special needs and challenges of this patient population. Most clinical trials have excluded or enrolled insufficient numbers of elderly patients, significantly limiting insights into the safety and efficacy of IBD therapy as a function of age.(3) We provide a review of the recent literature on IBD treatment options in older patients to help characterize the strengths and limitations of available evidence and the practical questions that arise for clinical decision-making. Where appropriate, we discuss studies of therapies for other autoimmune diseases (eg, rheumatoid arthritis), for which similar treatments and therapeutic algorithms are adopted.

MEDICAL MANAGEMENT IN ELDERLY IBD PATIENTS

5-Aminosalicylic Acid

Five-aminosalicylic acid drugs (eg, 5-ASA; mesalamine or mesalazine) are anti-inflammatory agents that mediate multiple but not mutually exclusive inflammation-related pathways including nuclear factor κB, cyclooxygenase/prostaglandin E2, and peroxisome proliferator-activated receptor γ signaling.(19, 20) Current national and international guidelines recommend 5-ASA for induction and maintenance of remission in patients with mild to moderately active UC.(21, 22) Meta-analysis of studies published between 1981 and 2005 demonstrated that oral 5-ASA was superior to placebo for induction and maintenance of UC remission.(23, 24) The use of 5-ASA for CD remains controversial: a meta-analysis of 6 clinical studies including 910 patients with active CD found that the 5-ASA derivative, sulfasalazine, tended to be superior to placebo at inducing remission, but there was no benefit of mesalamine over placebo. Neither compound was effective at preventing relapse in quiescent CD.(25) While most of the evidence base for 5-ASAs in IBD is for the general population, comparable efficacy of 5-ASAs was reported for both young and elderly patients. However, elimination of sulfasalazine was found to be considerably slower in the elderly, with a decreased glomerular filtration rate and lower renal clearance than in younger cohorts.(7)

The favorable safety profile and relatively benign side effects of 5-ASAs suggest that they may be preferred by elderly patients over other treatment options.(26) A retrospective observational study of elderly patients aged ≥65 years in a 20-hospital system in Pennsylvania, United States (US), revealed that 5-ASA and its derivatives were the most commonly used agents for maintenance therapy, taken by 173 of 393 (44%) patients.(27) This proportion was even higher in a French population-based registry, with 75% of the elderly IBD population exposed to ASAs.(28) Similarly, a retrospective cohort study of patients who underwent surgery for IBD at a single academic center in Wisconsin, US, stratified patients into 4 age groups and revealed that 5-ASA use was higher in the elderly cohort (aged >64 years) compared with younger cohorts.(29) While the evidence supporting the use of 5-ASA drugs for induction and maintenance of remission in UC is clear, several obstacles to optimizing therapy, particularly in the elderly, have been identified. The conventional dosing regimen for 5-ASA entails a multiple daily dosing schedule. For example, target doses of controlled-release oral formulations of 5-ASA for maintenance in UC range from 0.75 to 4.8 g/day; however, doses are available only in 250 to 500 mg capsules, requiring patients to swallow up to 16 pills every day.(19) On the other hand, topical formulations can be equally inconvenient, causing discomfort, particularly in the elderly patient in whom anal sphincter incompetence may reduce retention of enema fluid.(30) Much of the recent innovation in clinical use of 5-ASA in colitis has come from studies of novel delivery mechanisms and simplified oral dosing schedules.

The literature regarding the renal impact of 5-ASA is divisive, and it is unknown whether the risk of nephrotoxicity is related to age. A systematic review summarizing the current knowledge about the potential relationship between 5-ASA treatment and nephrotoxicity suggested that renal reactions may be idiosyncratic rather than dose-related. This conclusion was based on the low overall incidence of reported renal disease during 5-ASA treatment and the lack of clear relationship between 5-ASA dose and risk nephrotoxicity.(31)

Corticosteroids

Corticosteroids suppress multiple pro-inflammatory genes that are activated in chronic inflammatory diseases such as IBD.(32) Reduced expression of these genes downregulates lymphocyte transmigration into the gut and inhibits pro-inflammatory cytokine secretion, thereby reducing inflammation.(32, 33) Many corticosteroid formulations are available for use in IBD patients, including prednisone and budesonide, with demonstrated clinical efficacy for induction of remission as monotherapy and in combination with other IBD therapies.(34)

Side effects of short- and long-term use of corticosteroids can limit the utility of this therapy in elderly patients with IBD. Short-term effects, such as insomnia, mood instability, and delirium,(35) are most distressing for the elderly because they impair functional independence. Long-term use can lead to serious safety issues, including an increased risk of hyperglycemia, cataract formation, osteoporosis, and aseptic joint necrosis.(36, 37)

The overuse of corticosteroids among elderly IBD patients is of particular concern because their risk of reduced renal and hepatic clearance leads to increased exposure and makes these patients more susceptible to steroid toxicity.(38) Infections in particular are a significant risk among elderly patients with IBD.(11, 15) In an analysis of 300 IBD patients across a range of ages, corticosteroid use was associated with significantly increased odds for opportunistic infections (OR: 3.4; 95% CI: 1.8–6.2).(11) Specifically among 3552 patients with elderly-onset IBD, the rate of serious infections was significantly higher in patients exposed to oral corticosteroids at any time during the preceding 6-month period compared with unexposed patients (adjusted rate ratio: 2.3; 95% CI: 1.8–2.9).(15) Steroid use also confers an increased risk of osteoporosis among elderly IBD patients. In a study of adult IBD patients, multivariate analysis showed that cumulative steroid dose and age (≥55 years) were significant, independent risk factors for low bone mineral density in the femoral neck.(37) The prevalence of malabsorption and calcium and vitamin D deficiency in the elderly emphasizes the importance of early and regular bone densitometry assessments, and use of bisphosphonates should be considered along with vitamin D and calcium supplementation.(39) Other concerns with chronic steroid use are increased risks of hyperglycemia and diabetes, both already common conditions in the elderly.(40, 41) In a case control study of CD patients aged >50 years, the relative risk (RR) for developing hyperglycemia after steroid therapy demonstrated a numerical but not statistically significant increase relative to younger patients (RR: 1.53; 95% CI: 0.54–4.32).(40) And, in a retrospective population-based cohort study of patients aged >65 years, oral corticosteroid users were significantly more likely to develop diabetes mellitus than the control group receiving a proton pump inhibitor (adjusted RR: 2.31; 95% CI: 2.11–2.54).(41) A recent study of elderly CD patients showed that those who received continued steroid therapy reported worsened anxiety, sleep, and fatigue compared with patients who did not use corticosteroids.(42) Moreover, elderly patients who received corticosteroid monotherapy versus tumor necrosis factor (TNF)-α antagonist or immunomodulator regimens showed significantly more depressive symptoms and anxiety.(42)

Current treatment guidelines for the general IBD population thus recommend tapering of systemic steroids after clinical remission is achieved. With significant variability in steroid tapering protocols,(22, 34, 43) tapering should be individualized and based on the rate and completeness of treatment response, and duration of exposure to steroids.(43)

Steroid dependency, defined as the inability to discontinue or reduce dosage of systemic steroids below 10 mg/day, is prevalent, affecting up to 40% of patients with moderately to severely active IBD.(44) In these patients, early initiation of corticosteroid-sparing therapies, including thiopurines, methotrexate, and biologics, is recommended to reduce steroid exposure and represent a measure of high-quality IBD care in several guidelines.(44–47)

Despite these evidence-based guidelines, older IBD patients, in contrast to younger IBD patients, are more often treated with maintenance corticosteroid therapy and are less frequently initiated on such steroid-sparing strategies (Figure 2).(27) In a retrospective observational study, Juneja et al followed 393 IBD patients aged ≥65 years and found that corticosteroid use nearly doubled, from 36% between 1991 and 2000 to 64% between 2001 and 2010.(27) Further, elderly-onset patients (those diagnosed after the age of 65) were on higher average prednisone doses than those diagnosed at age <65 years (8 mg/day and 2.2 mg/day, respectively; P < 0.005).(27) Corticosteroid use for >6 months was identified in 32% of IBD patients aged ≥65 years.(27) The extended use of corticosteroids in older patient populations is further demonstrated by the inadequate implementation of steroid-tapering protocols.(14, 22, 43) In a retrospective cohort study by Johnson et al, which included a nationally representative population of IBD patients aged ≥65 years (N = 8502), nearly 20% of those treated with a TNF-α antagonist received corticosteroids for >3 months and 10% received a maintenance course lasting ≥6 months. Further, 43% of older patients treated concomitantly with a TNF-α antagonist and corticosteroids had multiple steroid courses spanning >3 months.(14) Despite the evidence-based data demonstrating high rates of adverse outcomes with steroids, familiarity, ‘perceived safety’, and the perceived concern about other immunomodulators or biologic agents contribute to the widespread use of corticosteroids in the elderly.(27)

Figure 2. Utilization of different therapeutic agents in patients aged 65 years or older.(27).

In a retrospective observational study, electronic medical records from a 20-hospital setting were reviewed from January 1, 1991 through December 31, 2010. Patterns for IBD medication in a total of 383 IBD patients aged ≥65 years were summarized.

Altogether, findings from these studies confirm and extend the evidence for substantial steroid-related toxicity in elderly patients with IBD. Budesonide, a newer corticosteroid, has some advantages in safety relative to others. Evidence suggests that topical budesonide does not accumulate in the body and compared with other corticosteroids, reaches more central parts of the colon and shows less suppression of plasma cortisol, without diminished efficacy.(48) A study comparing oral budesonide and prednisone showed similar efficacy rates in active and extensive UC.(49) A Cochrane review of budesonide in CD suggested that it may have lower short-term efficacy than other corticosteroids but fewer adverse events. Because of its better tolerability profile compared with conventional corticosteroids, budesonide may be a first-line steroid in elderly patients with disease extent and severity amenable to its action. However, cost and insurance coverage in some regions including the US may limit access, and in clinical practice, low-dose prednisone is the only alternative.(27) The frequent use of prednisone and associated risks highlight the need to monitor the duration of corticosteroid use and implement steroid-sparing strategies particularly in the elderly population.

Immunomodulators

For elderly patients who are resistant to aminosalicylates or show corticosteroid dependence (approximately 30% of the elderly population(40)), immunomodulator therapy with thiopurines or methotrexate may be indicated to maintain IBD remission.(22, 43) Thiopurines (eg, azathioprine, 6-mercaptopurine) are immunosuppressive by deactivating pro-inflammatory T lymphocytes,(50) whereas methotrexate is a folate antagonist with immunosuppressive properties and established efficacy in several inflammatory disorders including CD.(51)

Guidelines recommend the use of immunomodulators primarily for their steroid-sparing effects, rather than as monotherapy to induce remission.(22, 43) A review of nearly 40 randomized controlled trials in CD found that neither azathioprine/6-mercaptopurine nor methotrexate improved the odds of achieving remission over placebo.(52) However, a significant difference over placebo was observed for both azathioprine/6-mercaptopurine and methotrexate for the maintenance of remission.(52) The use of methotrexate in UC has not been well-established.(22, 43)

Despite data supporting their efficacy and the risk of corticosteroid therapy in the elderly, immunomodulators are used less frequently in older patients. A retrospective observational study of 393 patients aged ≥65 years in a 20-hospital system revealed that merely 6% of patients received maintenance therapy with either 6-mercaptopurine or azathioprine and 1% received methotrexate therapy, compared with 32% receiving maintenance prednisone (Figure 2).(27)

Although recommended for their steroid-sparing properties, immunomodulators cause systemic immunosuppression and are associated with significant safety issues of their own. A small prospective single-center study of azathioprine in elderly IBD patients (n = 25) concluded that azathioprine was efficacious but less well tolerated in the older population relative to other age groups with IBD, with 48% of 25 patients aged ≥75 years intolerant within 3 months versus 36% of 353 patients overall.(53) The increased risk of opportunistic infections with azathioprine and 6-mercaptopurine (OR: 3.1; 95% CI: 1.7–5.5) use may be compounded with the overall increased risk in patients aged >50 years.(11)

Several studies have reported that immunomodulator therapy is associated with increased risk of malignancy.(54–57) In a prospective observational cohort study of 19,486 IBD patients, the multivariate-adjusted hazard ratio (HR) of lymphoproliferative disorder (LPD) between thiopurine-treated and thiopurine-naive patients (N = 18,486) was 5.28 (P = 0.0007).(54) Thiopurine-treated patients aged >65 years had a yearly LPD incidence rate of 5.41 per 1000 patient-years (PY) compared with 2.58 in patients aged 50 to 65 years. Markedly, there was an approximately 15-fold difference in yearly LPD incidence rate between patients aged >65 years and those aged <50 years (5.41 and 0.37 per 1000 PY, respectively). Older age and longer disease duration are independently associated with increased risk of incident LPDs.(54) Older age (>65 years) was also predictive of lymphoma in a 10-year retrospective cohort study of UC patients from the Veterans Affairs health care system (N = 36,891; HR: 2.6; P = 0.02).(55) The age-, sex-, and race-adjusted HR of developing lymphoma while being treated with thiopurines was 4.2 (P < 0.0001) relative to patients who had not been exposed to thiopurines. The incidence rate of nonmelanoma skin cancer in a retrospective study spanning 11 years was also higher with thiopurine exposure, and highest among patients aged >65 years (12.2 per 1000 PY vs 4.8 per 1000 PY without exposure).(57) Similarly, Peyrin-Biroulet et al found that patients aged >65 years were at the highest risk of developing nonmelanoma skin cancer with thiopurine use (4.04 per 1000 PY vs 0.84 per 1000 PY without thiopurine use vs 0.66 per 1000 PY with thiopurine use in patients <50 years old).(56)

Biologics

Biologic therapies for IBD include TNF-α, integrin, and interleukin antagonists.(58) TNF-α is a pro-inflammatory cytokine with a key role in the signaling cascade that results in pathologic, chronic intestinal inflammation.(59) Elevated levels of TNF-α have been observed in some patients with UC and CD,(60) and TNF-α antagonists (ie, infliximab,(61, 62) adalimumab,(63, 64) certolizumab pegol,(65, 66) and golimumab(67, 68)) have demonstrated efficacy as induction and maintenance therapy for UC and/or CD.

The integrins are cell-surface proteins that mediate adhesion of circulating lymphocytes to cell adhesion molecules on endothelia, permitting extravasation to inflamed tissues.(69) Integrin antagonist therapies include the humanized monoclonal antibodies, natalizumab (indicated for CD and multiple sclerosis)(70) and vedolizumab (indicated for UC(71) and CD(72)). Natalizumab interacts with the α₄ subunit of the α₄β₁ and α₄β₇ integrins, whereas vedolizumab selectively targets the α₄β₇ integrin. Binding to the α₄β₇ integrin prevents its interaction with mucosal addressin cell adhesion molecule-1 (MAdCAM-1) on microvascular endothelia, inhibiting migration of inflammatory cells to the gastrointestinal tract.(69) By also binding α₄β₁ integrin, natalizumab blocks its interaction with vascular cell adhesion molecule-1 (VCAM-1), thereby preventing migration of leukocytes across the blood-brain barrier into the central nervous system.(70) This latter property of natalizumab may account for its efficacy in multiple sclerosis, whereas the absence of binding to α₄β₁ integrin allows vedolizumab a gut-selective mechanism of action.

Ustekinumab, a monoclonal antibody to the p40 subunit of interleukin-12 and interleukin-23 previously used for the treatment of psoriasis and psoriatic arthritis, was recently approved for the treatment of moderately to severely active CD.(73) Due to the limited information related to its use of treating CD in the elderly, ustekinumab will not be discussed in this section.

TNF-α antagonists

Although TNF-α antagonists have been available for over a decade, data for use in the elderly and comparisons between agents in this population are sparse. Whether a cause or consequence of the limited evidence base, in practice, TNF-α antagonists are used less frequently in older IBD patients relative to other immunosuppressives (Figure 2).(27) Several studies have indicated that fewer than 4% of IBD patients aged >65 years received TNF-α antagonist therapy compared with approximately 11% of IBD patients in the general population (mean age: 50.8 years).(14, 27, 74) In one such study, Johnson et al reported that among 8502 Medicare beneficiaries (aged 65 to 104 years) initiating TNF-α antagonist therapy, 67% were ≤75 years, 33% were aged >75 years, and just 4% were ≥85 years old; no patient >90 years of age received TNF-α antagonist therapy.(14)

One possible reason for the low use of TNF-α antagonist treatment in the older population may be the reduced efficacy reported by some studies. For instance, in an observational retrospective study by Lobaton et al, clinical response rates at week 10 were 68% in 66 patients who initiated TNF-α antagonist treatment at age ≥65 and 80% in the younger cohort (n = 112; P < 0.001); whereas differences were not significant after ≥6 months of therapy (79% and 83%, respectively (P = 0.639)).(17) In contrast, other studies have reported similar TNF-α antagonist efficacy in both older and younger patient populations. In a study of 95 IBD patients older than 65 years who were treated with infliximab or adalimumab between 2000 and 2009, 59% of UC patients and 65% of CD patients achieved clinical remission whereas remission rates in a control group of 190 UC and CD patients aged ≤65 years were 57% and 59%, respectively.(75)

Among those who initiate therapy with a TNF-α antagonist, older patients are less likely to remain on treatment long-term. In a single-center retrospective study, discontinuation rates at 12 months were higher among patients aged >60 years than in the control population who initiated therapy for IBD before the age of 60 (25% vs 7%).(76) The only significant predictor of early treatment discontinuation among older patients was the use of combination therapy: concomitant use of azathioprine was associated with a two-fold increase in treatment cessation compared with TNF-α antagonist monotherapy,(76) a reflection, perhaps, of the increased risk of infection with combination therapy in this population.

Safety concerns with TNF-α antagonist therapy in the general IBD population may become heightened when patients are elderly. In a single-center retrospective study, the risk for any serious adverse event with TNF-α antagonist therapy was nearly fivefold higher when patients were ≥65 years of age (n = 66; RR: 4.7; P < 0.001).(17) As with other immunosuppressive therapies, TNF-α antagonist-specific safety issues among elderly IBD patients include risk of infection, malignancy, and death. In a study of 3079 IBD patients receiving TNF-α antagonist therapy, higher rates of severe infection (11%) and mortality (10%) were observed in the 95 patients older than 65 years compared with younger patients (2.6% and 1%, respectively) or age-matched controls on nonbiologic therapy (0.5% and 2%, respectively).(75) Toruner and colleagues reported that infliximab treatment significantly increased the odds of opportunistic infection (OR: 4.4; 95% CI: 1.2–17.1) across a range of ages, with a higher magnitude in the elderly.(11) In contrast, in a cohort study of Medicare beneficiaries with rheumatoid arthritis aged ≥65 years, there was no increase in serious bacterial infections among initiators of TNF-α antagonist therapy (RR: 1.0; 95% CI: 0.6–1.7) or disease-modifying anti-rheumatic drugs versus methotrexate initiators.(77)

Several case reports,(78, 79) a small case series,(80) and analyses of administrative data (81, 82) have quantified the risk of lymphoma after TNF-α antagonist therapy in IBD patients. Others have reported cases of hepatosplenic T-cell lymphoma with combined TNF-α antagonist and thiopurine exposure, although primarily in adolescent or young adult males.(83) However, there does not appear to be an association between TNF-α antagonist use and a risk for all cancers.(84, 85) Since many TNF-α antagonist users received concomitant immunomodulator therapy or have a history of recent use of such therapies, the relative contributions of each drug class to lymphoma or other malignancy risk have been difficult to ascertain.

Integrin antagonists

Evidence for the safety and efficacy of integrin antagonists in older patients is limited. Despite its efficacy, natalizumab use for CD has been limited because of serious safety concerns. Progressive multifocal leukoencephalopathy (PML), a rare and often fatal viral infection, has been reported in patients with CD and multiple sclerosis treated with natalizumab.(86) There have been no reports of PML with vedolizumab exposure to date, likely because of its gut-selective mechanism of action.

Prospective randomized trials of vedolizumab enrolled limited numbers of older IBD patients. A descriptive retrospective analysis of pivotal studies in UC and CD demonstrated that the safety profiles of vedolizumab and placebo were similar in patients aged <35 years, 35 to 55 years, and ≥55 years.(18) Vedolizumab-treated UC and CD patients in the oldest subgroup (n = 147) had fewest serious infections (<1%) and adverse events leading to hospitalization (11%) compared with patients in younger subgroups. Early experience in clinical practice (N = 66 IBD patients aged ≥18 years) has shown that older age may correlate with a higher likelihood of vedolizumab response.(87) The mean age of patients who responded to vedolizumab, defined as having abnormal C-reactive protein or elevated physician global assessment score, were significantly older (42.5 vs 25.5 years; P = 0.01) at week 6 and at week 14 (43.6 vs 20.9 years; P = 0.02).

Vedolizumab has no known systemic immunosuppressive effects and reported rates of serious infection are low.(88) There have been no head-to-head studies comparing the relative safety of vedolizumab with TNF-α antagonists or other immunosuppressive agents. Indirect comparisons through meta-analyses have revealed no significant differences in safety profiles between the different biologics.(89, 90) As further safety data accumulate, such targeted therapy with low likelihood of systemic immunosuppression may prove to be an attractive option in older patients.

SURGERY IN ELDERLY IBD PATIENTS

Surgery remains an important treatment option in patients for whom medical management has failed.(91) Approximately 25% of all intestinal surgeries for IBD are performed in patients >55 years old.(92) Total proctocolectomy with ileal pouch-anal anastomosis (IPAA; commonly described as the “J pouch”) is currently the standard of care in the surgical management of UC that is refractory to conservative management.(91) Historically, IPAAs were not widely performed in elderly IBD patients, largely owing to concerns about incontinence. The American Society of Colon and Rectal Surgeons recommends careful attention to the patient’s mental status, anal sphincter function, and underlying comorbidities, as well as patient education on the associated risks and potential functional difficulties when considering IPAA surgery.(93)

As for medical therapy, there may be a higher frequency of complications in the immediate post-operative period in older patients. In an analysis of administrative data on patients who underwent IPAA for UC, postoperative complication rates were similar among those aged ≤60 and >60 years, and hospital stays were slightly longer among older versus younger patients (8.2 days vs 7.4 days; P = 0.035).(94) Pellino et al reported that while elderly UC patients had more comorbidities than their younger counterparts, there were no significant between-group differences in bowel control after IPAA.(95) Further, in a study of patients undergoing elective surgery (ie, restorative proctocolectomy), postoperative complication rates were comparable between the younger and older age cohorts.(96) In contrast, in a retrospective database study of patients who received IPAA for UC, multivariate analysis revealed that the only significant factor predisposing patients to fecal leakage was older age at the time of surgery (OR: 1.07; 95% CI: 1.02–1.12; P = 0.005).(97) In a more recent analysis, 30-day postoperative mortality rates were significantly higher among CD patients aged ≥65 versus those aged <65 years (4.2% vs 0.3%; P < 0.001).(98) Among UC patients, corresponding rates were 6.1% and 0.7%, respectively (P < 0.001).(98)

Patient age, type of admission, presence and number of comorbid conditions, including congestive heart failure, thromboembolic disease, and liver and renal diseases are associated with increased postoperative mortality.(92) In one small study of elderly UC patients who underwent emergency surgery during a 26-year period, 27% (8/30) of patients aged ≥60 years died within 30 postoperative days. By contrast, <1% of elderly patients (1/114) died after elective surgery for UC.(99)

In general, elderly patients are hospitalized more often than younger patients. Among IBD-related hospital admissions in the US in 2004, 25% of patients were older than 65 years.(100) Further, data from the Nationwide Inpatient Sample indicate that elderly patients experience worse outcomes after hospitalization, with 15.7% of IBD patients aged 65 to 84 years and 35% of patients aged over 85 years requiring discharge to a nursing home or rehabilitation facility compared with less than 1% in individuals aged under 45 years. Moreover, home healthcare after hospital discharge was required by 12.6% of patients aged 65 to 84 years versus 4.7% of those aged 18 to 45 years.(100) According to a recent meta-analysis, immunomodulator and biologic therapies were significantly less likely to be used in older-onset IBD patients and the rate of surgery was higher in older-onset compared to early-onset UC patients. The use of biologic therapies in older-onset CD and UC patients could potentially reduce the likelihood of surgery.(9)

CLINICAL CHALLENGES IN THE ELDERLY

Pharmacologic Management of IBD in the Setting of Comorbidities

A particular challenge in the management of elderly IBD patients is selection and continuation of therapy in the setting of comorbidities and polypharmacy.(101) For example, poor lung function in patients with comorbid chronic obstructive pulmonary disease may make an IBD patient more susceptible to developing pneumonia or have more severe outcomes as a consequence. Other conditions become more prevalent in the elderly population, including cardiovascular disease and diabetes, presenting distinct challenges for clinicians who manage patients with comorbid IBD.(101) For example, patients with diabetes mellitus and ulcerative colitis share a host of complications including peripheral neuropathy, reduced bone density, infection, and thromboembolism that challenge both diagnosis and treatment.(102)

Some comorbidities may suppress immune function; therefore, caution should be exercised when treating these patients with systemic immunosuppressive agents. In a study of 2,766 patients receiving immunomodulator therapy, 19% of patients with underlying diabetes developed an infection within 1 year compared with only 8% of those without diabetes, with this effect remaining significant on multivariable analysis.(16) Medications used for the treatment of IBD may in turn worsen a preexisting comorbidity. For example, hypertension, hyperglycemia, and congestive heart failure may all be worsened by steroid use.(101) Steroid use can also precipitate psychotic manifestations.(101) TNF-α antagonists are contraindicated in patients with class III-IV New York Heart Association heart failure.(103)

Given the association between age and cancer, many older IBD patients may have had prior malignancy, and are thus at higher risk for cancer recurrence or new primary tumors. Although active untreated cancer remains a contraindication to immunosuppression, emerging data suggest that thiopurine or TNF-α antagonists may be safe in patients with prior malignancy.(85, 104, 105) In a retrospective analysis of patients with previous cancer, the time to a new cancer was similar whether patients received TNF-α antagonist monotherapy, TNF-α antagonist and methotrexate/thiopurine combination therapy, or methotrexate or thiopurine monotherapy (P = 0.22).(85) Similarly, in another observational cohort study, while a history of cancer increased the risk of developing a new or recurrent cancer, treatment with immunomodulators did not influence that risk.(104) A meta-analysis of 16 studies of 11,702 patients with IBD, rheumatoid arthritis, or psoriasis and prior diagnosis of cancer found rates of cancer recurrence to be 33.8 per 1000 PY with TNF-α antagonist monotherapy, 36.2 per 1000 PY with immunomodulatory monotherapy, and 37.5 per 1000 PY without immunosuppressive therapy.(105) Although more long-term experience is needed, current evidence does not support an increase in malignancy risk with integrin antagonist use.(88) However, cancer-specific risks exist, such as between thiopurines and nonmelanoma skin cancer,(56) making them a less suitable option for those with a past history of this cancer.

Ultimately, minimizing unnecessary pharmacologic exposure that might exacerbate a comorbid condition is essential in all patients. The presence of comorbid conditions necessitates close collaboration among treatment providers for coordinated decision-making and optimal outcomes.

Combination Therapy

The relative risks and benefits of combination therapy in an older population are unclear. In randomized controlled trials and meta-analyses in the general population, combination immunomodulator therapy with infliximab is associated with higher trough levels, reduced rates of anti-drug antibodies, and greater clinical and endoscopic response.(106) In addition, risks with such therapy appear low except for a small increase in risk of lymphoma, primarily associated with thiopurine use.(55) However, the use of combination therapy in an older cohort was associated with greater rates of treatment discontinuation contrasting with results obtained in younger cohorts, whereby combination therapy resulted in superior outcomes or had no effect on treatment durability.(76) These observations are substantiated by the Markov modeling by Scott and colleagues, who found that for patients older than 65 years, the benefits of combination therapy may be outweighed by the associated risk of lymphoma, especially with long-term exposure.(107) However, further research is needed to accurately inform practice and decisions need to be individualized based on age, comorbidity, and likelihood of aggressive course of IBD.

Treatment Adherence

Adherence is a complex and multifactorial issue, influenced by several factors including the physician–patient relationship, lack of insight into illness, perception of lack of benefit, financial burden, and forgetfulness.(108) In elderly patients, complex dosing regimens and polypharmacy add to the rate of nonadherence.(3) Medication nonadherence in IBD has been associated with disease flares, increased healthcare costs, anxiety, and moodiness.(109, 110)

Polypharmacy is common in the IBD elderly population, with 22% of patients taking more than 5 different medications and 94.7% taking more than 3 medications. A mean number of 7 agents per patient were reported to be taken regularly by elderly IBD patients.(27) A meta-analysis of studies published between 1986 and 2000 examined the relationship between number of daily doses and rate of adherence. Analysis of 76 studies from several disease areas suggested that less frequent oral dosing is related to higher adherence. In fact, significant differences in the rate of adherence were demonstrated between once daily and three-times daily or four-times daily dosing.(111)

In addition to polypharmacy, the complex regimens often required to achieve clinical response or remission in IBD increase the likelihood of nonadherence and medication-related errors.(12) Thus, forgetting to take medication is reported to be one of the most common reason for nonadherence across a variety of therapeutic classes.(3) Elderly patients with underlying cognitive decline or with limited functional ability may be especially prone to nonadherence with regimens involving multiple pills taken several times a day or with treatment administered outside the home. Visiting nurse programs and other social support mechanisms could play an important role in ensuring adherence and optimal outcomes in this cohort.(100)

Insurance-Related Obstacles

The benefits of using highly potent agents earlier in IBD treatment, especially in CD, are becoming more widely accepted. However, this strategy has yet to be formally assessed in elderly patients. The lack of clinical data demonstrating the effectiveness of biologics and immunomodulators in the elderly and the cost of these drugs constitute a big obstacle to their use in this patient cohort. Recently, the center for Medicare and Medicaid Services adopted IBD-specific quality measures targeting drug therapy optimization and calling for the use of steroid-sparing therapies.(112) Despite these efforts, coverage by public insurance (eg, Medicare) is limited or not available for certain drugs such as budesonide, extended release 5-ASA, and injectable drugs, creating a barrier to treatment access for the elderly.

Defining Treatment Goals in the Elderly

Treatment goals for IBD have been shifting over recent years away from symptom control toward objectively-measured endpoints.(113) Mucosal healing and the normalization of disease biomarkers, such as fecal calprotectin in UC, are two such objective targets that are increasingly being incorporated as either primary or exploratory endpoints in clinical trials and in practice.

There is growing evidence supporting mucosal healing as a driver for sustained remission and steroid-free remission, as well as for reduction in rates of surgery and hospitalization. Rutgeerts and colleagues demonstrated in a postoperative setting that CD patients with endoscopic evidence of active disease were more likely to have recurrence of symptoms within the following 3 years than those with healed mucosa.(114) In fact, even in asymptomatic patients, over half have endoscopic evidence of active inflammation and structural bowel damage—eg, strictures and/or fistula—10 years after diagnosis.(115, 116) A retrospective analysis of 495 patients (354 with UC and 141 with CD) after 1 year of conventional treatment found that mucosal healing predicted less surgery in patients with UC.(117) In patients with CD, mucosal healing was associated with fewer complications and reduced steroid treatment.(117) Notably, no patient in this analysis had received biologic treatment and 50% had achieved mucosal healing. Together, these and other studies support the importance of mucosal healing for long-term treatment success. However, the potential benefits of preventing long-term complications and improving long-term outcomes in older patients should be weighed against the risks of therapy escalation. Older patients without risk of aggressive disease may benefit from symptom improvement alone to attain functional independence.

Utilizing biomarkers such as fecal calprotectin can provide noninvasive and cost-effective disease evaluations for the elderly population.(113) Fecal calprotectin levels of <50 μg/g are considered normal and have been reported to predict mucosal healing in patients with IBD.(118) However, the degree to which change in fecal calprotectin levels correlate with treatment response and remission is still under active investigation. In a recent review, Bressler and colleagues suggested initiating strategies to control inflammation when calprotectin levels are >250 μg/g; confirming possible inflammation with further testing when calprotectin levels are >100 μg/g to 250 μg/g, and continuing therapy when calprotectin levels are <100 μg/g.(119)

Colon Cancer Screening

It is well established that IBD can predispose a patient to colorectal cancer. A 2001 review of 116 clinical trials estimated an overall prevalence of colorectal cancer in any UC patient of 3.7% (95% CI: 3.2–4.2).(120) Further, the cumulative probability of cancer increased with disease duration from 2% at 10 years to 18% at 30 years.

Several factors should be considered when deciding on the frequency of colon cancer screening in IBD patients, including age of symptom onset, duration and extent of disease, and family history. Because of a significant risk of colorectal cancer after 8 years of IBD, the American Gastroenterological Association guidelines recommend examinations every 1 to 3 years beginning 8 years after IBD symptom onset or within 2 years in those with extensive or left-sided colitis.(121) After two consecutive negative readings, screenings can be extended out to 3 years, but are still relatively frequent in comparison to healthy adult screenings. In healthy adults, the American College of Gastroenterology recommends a colonoscopy every 10 years starting at age 50 years.(122) Optimizing the frequency of screening for colorectal neoplasia should depend on treatment and additional risk factors for malignancy, including previous malignancy. Additional risks of performing colonoscopy in elderly IBD patients, including cardiopulmonary complications and perforation, should be weighed against the potential benefits of timely diagnosis of neoplasia.(123)

CONCLUSIONS

Management of IBD in the older patient poses unique challenges. Analysis of real-world practice suggests frequent suboptimal management of such patients with overuse of corticosteroids and underuse of steroid-sparing regimens including immunomodulators and biologics, stemming in part from concerns of therapy-related adverse events, in particular infection. As evidence accumulates about the relative safety and efficacy of each therapy in an older IBD patient, early initiation of effective therapies may reduce steroid exposure and related side effects in the elderly. Further data are also needed on whether gut-selective immunosuppression may be a safer option in older patients, while achieving the same level of efficacy as systemic immunosuppression. Additional unique challenges in older patients are recognizing the impact of comorbidity on outcomes with treatment of IBD and conversely the impact of IBD therapy on comorbidity. Evolving trends in the IBD treatment landscape should always be translated to the older patient before they are utilized in this population. Management of IBD in the older patient requires personalization, setting of individual therapeutic goals (which should also include preservation of physical status and maintenance of functional independence), and weighing risks versus benefits of combination immunosuppression. Real-world studies, shared clinical experience, and prospective studies with well-defined endpoints and patient populations can help structure evidence-based decision-making and improve long-term clinical outcomes.