Outstanding Challenges in IBD Treatment

In 2017, about 6.8 million people worldwide were living with inflammatory bowel disease (GBD 2017, Lancet Gastroenterol Hepatol 2020²⁶), and it incurs substantial costs in surgery, hospitalization, and lost productivity. Biological therapies have transformed IBD management since the US approved infliximab for Crohn's disease in 1998, yet a majority of patients do not reach deep remission. In real-world studies of anti-TNF drugs, about 44% of patients reached deep remission (clinical remission plus a healed bowel lining on endoscopy) at one year (Alipour et al. 2021¹⁶).

Two interrelated failures drive this treatment gap. First, clinicians lack reliable, non-invasive tools to objectively measure mucosal inflammation in real time, making treat-to-target strategies difficult to implement. Second, even when disease activity is quantified, the available therapies fail a large proportion of patients either from the outset (primary non-response) or over time (secondary loss of response). Understanding these limitations is prerequisite to designing the next generation of IBD therapeutics and monitoring strategies.

The ideal IBD biomarker would be non-invasive, cheap, rapid, and capable of accurately reflecting mucosal inflammatory activity across all disease phenotypes and locations. No current biomarker meets these criteria. The AGA's 2023 biomarker guideline reflects this directly: in patients with mild symptoms, neither a normal nor an elevated biomarker is accurate enough on its own to settle whether disease is endoscopically active, and endoscopy is still advised whenever markers and symptoms disagree³¹. The three most widely studied, C-reactive protein (CRP), fecal calprotectin (FC), and fecal lactoferrin (FL) , each carry significant limitations that restrict their clinical utility.

CRP is the most accessible serum inflammatory marker and correlates reasonably well with disease activity in Crohn's disease. It is less useful in ulcerative colitis: in studies of patients being assessed for IBD, CRP was raised in only 50–60% of those with UC, against 70–100% of those with Crohn's disease. A 2006 review by Vermeire et al.¹⁰ found CRP the best-performing conventional blood marker overall, while noting that it rises in most inflammatory diseases, not only IBD, and tracks ulcerative colitis less well than Crohn's disease.

A key practical limitation is that CRP also rises with many non-IBD processes, including infection, some cardiovascular diseases, and cancer, and baseline levels vary with body weight (Sproston & Ashworth 2018⁵⁰). In children, normal blood tests are common even at diagnosis: in one study of 135 children, ESR and CRP were normal in up to 28% of those with Crohn's disease and 42% of those with ulcerative colitis (Alper et al. 2017⁵¹).

Fecal calprotectin (a neutrophil-derived S100A8/9 heterodimer released during intestinal inflammation) is the most widely used stool test for intestinal inflammation, and the STRIDE-II treat-to-target consensus lists its normalization as a short-term treatment goal (Turner et al. 2021⁵³). A 2018 meta-analysis of 25 studies (Rokkas et al. 2018³⁰) found it performed better in UC (sensitivity 87%, specificity 77%, AUC 0.91) than in Crohn's disease (82%, 72%, AUC 0.84), and a 2023 meta-analysis focused on Crohn's disease reported sensitivity 81%, specificity 74% and AUC 0.85 (Bohra et al. 2023⁸). An earlier meta-analysis of 13 studies had reported AUCs of 0.93 in UC and 0.88 in Crohn's disease (Lin et al. 2014⁴). A meta-analysis focused on mucosal healing (Xiang et al. 2021⁵) found the best performance in UC at a cut-off of 60–75 µg/g (AUC 0.88).

However, the performance of FC may depend on disease location, which matters because isolated small bowel (ileal) Crohn's disease accounted for a quarter of a 435-patient cohort (Yang et al. 2015³⁵). A systematic review by Simon et al. (2019)⁶ found only eight studies that reported results by location: sensitivity ranged from 43% to 100% in small bowel disease and from 67% to 100% in large bowel disease, and the authors concluded that no firm conclusion can be made.

Calprotectin is also not specific to IBD: NSAIDs, proton pump inhibitors, colon polyps, diverticular disease, colorectal cancer, and gut infections can all raise it (D'Amico et al. 2021⁵⁴). There is no globally accepted cut-off, and studies have used thresholds from 48 to 400 µg/g (Rokkas et al. 2018³⁰). Stool consistency and handling affect how much calprotectin is extracted, and tests from different manufacturers are not interchangeable⁵⁴.

Fecal lactoferrin (FL), a glycoprotein released by activated neutrophils, performed about as well as FC in Crohn's disease: pooled sensitivity 75%, specificity 80%, AUC 0.81, though from only four studies (Bohra et al. 2023⁸). It has not been widely adopted in routine practice, for reasons that are not clear⁸. Fecal S100A12 distinguished IBD from other conditions in children (pooled sensitivity 95%, specificity 97%, AUC 0.99; Witarto et al. 2023⁹), but that review assessed diagnosis rather than disease activity and pooled only seven small studies.

Doctors often look at several blood tests together, but no blood marker has been shown to be ideal or superior to current diagnostic tools (Vermeire et al. 2006¹⁰), and none can replace endoscopy for judging mucosal healing, the key target in modern treat-to-target protocols. A 17-gene blood test was developed to predict how aggressive IBD will become rather than to measure current inflammation (Biasci, Lee et al. 2019³²), but in the PROFILE trial of 386 patients with newly diagnosed Crohn's disease it did not help choose treatment (Noor et al. 2024⁵²).

Modern IBD therapy spans several mechanistic classes: anti-TNF agents (infliximab, adalimumab, certolizumab pegol, golimumab), anti-integrin agents (vedolizumab), IL-12/23 and IL-23 blockers (ustekinumab, risankizumab, mirikizumab, guselkumab), JAK inhibitors (tofacitinib, upadacitinib, and filgotinib, which is not approved in the US), and S1P receptor modulators (ozanimod, etrasimod). Despite this mechanistic breadth, a consistent pattern emerges across trials: induction remission rates are modest, maintenance remission decays over time, and net remission across all enrolled patients is substantially lower than headline trial data suggest.

Critically, patients who have already used an anti-TNF tend to respond less well to their next advanced therapy. In a meta-analysis, patients whose first anti-TNF never worked (primary non-response) were 24% less likely to reach remission on a second-line biologic than patients who had stopped for side effects (Singh et al. 2018³). In the PANTS cohort, a low drug level at week 14 was the only factor independently associated with primary non-response, which is part of why drug-level monitoring matters²⁹.

A reanalysis by Kayal et al. (2023)¹¹ recalculated pivotal Crohn's disease trials as net remission: the share of everyone enrolled at induction who was in remission at the end of maintenance, instead of counting only induction responders who were re-randomized. Among anti-TNF-naive patients, net remission at about one year was 16.7% for infliximab, 28.5% for adalimumab, 34.8% for ustekinumab, and 18.7% for vedolizumab, all below 40%.

How much an earlier failure matters depends on the next drug. In the same meta-analysis, patients whose first anti-TNF never worked responded less well to ustekinumab, but not to vedolizumab (Singh et al. 2018³). The reasons are still being studied.

In UC, the 2024 AGA evidence synthesis (Ananthakrishnan et al. 2024¹²) compared advanced therapies by network meta-analysis, separately in patients starting their first advanced therapy and in those who had already used one. In previously exposed patients, upadacitinib ranked highest for inducing remission, followed by tofacitinib and ustekinumab. Absolute rates in the pivotal trials were modest: in the two upadacitinib induction trials, 26% and 34% of patients were in clinical remission at week 8, against 4% to 5% on placebo (Danese et al. 2022⁴⁷).

A fundamental reason for variable biomarker performance and inconsistent treatment response is that neither "Crohn's disease" nor "ulcerative colitis" is a single, homogeneous entity. Both represent broad clinical syndromes with heterogeneous underlying immunobiology, genetic architecture, microbiome composition, and disease behavior. Verstockt et al. (2022)¹³ argued in Gastroenterology that current classification tools are suboptimal and should be complemented by molecular subclassification. The Montreal classification used today records age at diagnosis (A1–A3), location (L1–L3, with an L4 modifier for upper GI disease), behavior (B1–B3), and a perianal modifier (Satsangi et al. 2006⁵⁵).

Molecular studies are beginning to find subgroups that respond differently to treatment. In Crohn's disease, a module of immune and stromal cells in inflamed tissue was linked to failure of anti-TNF therapy (Martin et al. 2019⁵⁶). Across IBD, a pattern of deep ulceration with neutrophil and fibroblast signatures was more common in patients who did not respond to several therapies (Friedrich et al. 2021⁵⁷). Patients with high oncostatin M (OSM) expression in mucosal biopsies show markedly inferior response to anti-TNF therapy (West et al. 2017, Nat Med¹⁴), yet no OSM pathway inhibitor is approved and OSM testing is not part of any routine clinical workup.

No routine test yet predicts which drug will work for a given patient, so current practice remains empirical sequence therapy: patients cycle through agents until one works. Drugs tend to work less often after earlier biologics have failed. In separate ustekinumab trials in Crohn's disease, 34% of patients whose anti-TNF had failed responded by week 6, against 52% to 56% of patients whose conventional therapy had failed (Feagan et al. 2016³⁸).

Several promising strategies are in development or early clinical adoption. For biomarker improvement, point-of-care FC testing now enables near-real-time results in outpatient settings, reducing the lag between sample collection and treatment adjustment.

For treatment optimization, the evidence for proactive therapeutic drug monitoring (checking drug levels on a schedule, not only when problems arise) is mixed. NOR-DRUM B found a benefit during infliximab maintenance³³, but a 2022 meta-analysis of randomized trials in IBD found no clear benefit from routine proactive TDM (Nguyen et al. 2022⁶⁰), and a 2024 BMJ guideline made a weak recommendation for it during infliximab maintenance only (Kawano-Dourado et al. 2024⁶¹). Combination biological therapy targets complementary pathways at once and may overcome the ceiling of single-target blockade. In the phase 2 VEGA trial in ulcerative colitis, 83% of patients on guselkumab plus golimumab had a clinical response at week 12, against 61% on golimumab and 75% on guselkumab alone, and the authors called for larger trials (Feagan et al. 2023⁶²). In the larger DUET trials of a combined guselkumab and golimumab product, reported in 2026, clinical remission at week 48 was 41.0% on the combination against 34.0% on guselkumab and 11.5% on golimumab in ulcerative colitis, and 50.8% against 42.5% and 25.4% in Crohn's disease; phase 3 trials are starting (Johnson & Johnson 2026⁶³).

Ultimately, resolving IBD's treatment challenges likely requires moving from reactive, symptom-driven management to a model of precision IBD care: molecular endotyping at diagnosis, biomarker-informed monitoring using disease-location-appropriate tests, proactive TDM, and timely escalation before irreversible structural damage accrues.

IBD is most often diagnosed between the ages of 15 and 35 (Crohn's & Colitis Foundation⁶⁴), precisely the years during which education is completed, careers are established, and long-term earning trajectories are set. A Danish register study of 3,398 patients diagnosed before age 25 (Rasmussen et al. 2025²⁰) found low labor market participation slightly more often in IBD than in comparators (16.4% against 14.4%), but income was not reduced. The risk was concentrated: only patients with severe disease carried an elevated risk (RR 1.46), and among those who also had a mental health condition, 46% had low labor market participation (RR 5.03).

A 2024 systematic review and meta-analysis by Youssef et al. (J Crohn's Colitis)¹⁸, pooling 134 studies, quantified the scale of professional impairment. Only 65.6% of IBD patients were employed, and about one in three reported having lost a job because of their disease, although only a few studies asked. Across all IBD patients, the pooled estimate was 23.9 working days lost per year, with 4.1 hours per week of missed work time. Even while present at work, a further 3.9 hours per week were lost to presenteeism (reduced on-the-job effectiveness). The pooled overall productivity loss across all IBD patients was 39.4%.

Long-term outcomes are similarly stark. The Norwegian IBSEN cohort, followed for 10 years after IBD diagnosis, found an overall work disability rate of 18.8%, with a relative risk of 1.8 in UC and 2.0 in CD compared with the general population (Høivik et al. 2013¹⁹). The relative risk of a disability pension was highest in patients under 40, and patients over 60 had no increased risk. Steroid treatment at one year predicted work disability in both diseases; in UC, raised inflammatory markers at diagnosis, early colectomy, and more than two relapses in the first year did too. Disease activity also matters in the pooled data: patients with active or moderate-to-severe disease had more absenteeism, presenteeism, and overall work impairment¹⁸.

The professional burden also shows up in quieter ways. Many people with IBD work shorter days because of symptoms or find it hard to plan their careers¹⁸, and cost studies usually cannot count lost promotions or training, so their totals likely run low²¹. A scoping review of eighteen studies (Cheng et al. 2022²²) grouped the difficulties into career planning, work performance, managing physical and cognitive symptoms, social effects at work, and the physical work environment and commute. It grouped the supports that help into five categories: flexibility, changes to the physical workspace, social support, self-management, and extended health benefits. The review notes that this research has mostly been descriptive, and that future studies still need to test which supports actually work.

A Swedish nationwide registry study by Khalili et al. (2020)²¹ found that yearly societal costs (health care, sick leave, and disability pensions) were consistently 2 to 3 times higher in IBD patients than in matched general population controls, and more than 3 times higher in the first year after diagnosis. Controlling the disease is therefore an economic goal as well as a clinical one: active or severe IBD is linked to more lost work and disability (Youssef et al. 2024¹⁸; Høivik et al. 2013¹⁹; Rasmussen et al. 2025²⁰).

IBD imposes a three-layered financial burden on patients: direct out-of-pocket healthcare costs, indirect costs from lost productivity and disability, and the long-term wealth penalty from reduced lifetime earnings. Across all three layers, IBD patients fare substantially worse than the general population. In US claims data, IBD costs began rising after 2013 while costs for matched controls did not (Park et al. 2020²³).

The most comprehensive US cost analysis, the Crohn's & Colitis Foundation's IBD Cost Initiative (Park et al. 2020²³), compared per-patient costs against matched non-IBD controls over 52,782 patients. IBD care cost $22,987 per member per year in paid claims against $6,956 for matched controls, more than a threefold difference, and out-of-pocket costs ran $2,213 against $979. The first year after diagnosis was the most expensive, averaging $26,555. A separate 2016 analysis put the combined direct and indirect national burden at $14.6–31.6 billion for 2014 (Mehta 2016³⁶), though that estimate rests on pharmacoeconomic data from 2004–2008 and predates most current biologic pricing.

A landmark analysis of lifetime costs by Lichtenstein et al. (2020, Clin Gastroenterol Hepatol)²⁴ modeled lifetime costs for Crohn's disease at $622,056 in total, of which $416,352 is the incremental cost over matched controls averaged across all ages at diagnosis. Outpatient care is the single largest component ($273,056), ahead of inpatient care ($164,298) and pharmacy ($163,722). Earlier diagnosis costs more: the incremental lifetime burden runs $707,711 for a diagnosis before age 12 against $177,614 at 70 or older, which is the economic shape of a disease that arrives young.

Out-of-pocket costs vary enormously by insurance status. In the claims analysis above, patients with IBD reported $2,213 a year out of pocket against $979 for matched controls (Park et al. 2020²³). Biologic list prices run into the tens of thousands of dollars a year, but published figures date quickly and vary by drug, dose, payer and rebate, so no specific list price is quoted here. What matters for a patient is that out-of-pocket exposure to the same drug differs enormously by coverage, and that this can lead to cost-related non-adherence, drug discontinuation, and preventable flares. For commercially insured patients with manufacturer copay assistance, costs can be near zero, but federal law bars manufacturer copay coupons in Medicare, Medicaid, and other federal health programs (HHS Office of Inspector General 2014⁶⁵), which disproportionately affects lower-income and older IBD patients.

The indirect cost burden is equally significant and frequently underestimated. A US productivity analysis by Ding et al. (2022)²⁵ measured indirect costs from lost work against disease activity rather than against a control group. In Crohn's disease these ran from $7,169 per patient per year in remission to $29,524 in moderately-to-severely active disease, and in ulcerative colitis from $4,348 to $24,283. These costs are largely absorbed by patients themselves and their employers rather than by healthcare systems. The Youssef meta-analysis put indirect costs from overall work impairment at €5,131 per patient per year, based on the few studies that reported costs (Youssef et al. 2024¹⁸). The US IBD population is about 3.1 million adults (Dahlhamer et al. 2016²⁸), but no published national total for indirect costs is cited here, because the audit of this page could not find one.

The financial burden also falls unevenly. Patients who need biologics or steroids have the highest costs in US claims data (Park et al. 2020²³). This can create a reinforcing cycle, in which suboptimal treatment efficacy drives higher costs while cost-related barriers to care impede access to the treatments most likely to induce remission.