Differences in design and their significance for interpretation

What study phases does a new active substance go through before it reaches clinical practice? What differences are there in the designs of clinical trials? Which aspects are important for interpreting results? In this CME training article, the development of clinical trial designs in inflammatory bowel disease (IBD) will be shown and it will be explained how endpoints change from clinical measurements to patient-based data and how mixed endpoints are composed of endoscopic and patient-based assessments.

What study phases does a new active substance go through before it reaches clinical practice? What are the differences in clinical trial designs? Which aspects are important for interpreting results? In this CME training article, the development of clinical trial designs in inflammatory bowel disease (IBD) is shown and it is explained how endpoints change from clinical measurements to patient-based data and how mixed endpoints are composed of endoscopic and patient-based assessments [1].

Different phases of clinical studies

There are different phases of clinical trials. The first phase – phase 1 – takes place with smaller groups of usually less than 80 participants [2]. Sometimes also with healthy volunteers, on whom the active substances are tested, especially with regard to safety and dosage. The aim is to evaluate dosage, form of administration and safety [3,4].

Once these factors have been tested, phases 2a and 2b follow. In phase 2a, the proof of concept is typically tested, phase 2b evaluates the efficacy/safety and determines the dosage and procedure for phase 3 trials [5]. In the field of inflammatory bowel disease, phase 2b is particularly important – the aim here is to understand exactly how high the dosage should be and how it should be administered in order to achieve the best efficacy.

Phase 3 usually involves between 500 and 1000 patients. The aim of this phase is to determine whether safety and efficacy are confirmed and whether they are relevant. Phase 3 trials are subject to strict statistical procedures and rigorous designs that can lead to approval. They are usually randomized and blinded to placebo [6]. It is important to have the possibility to compare: either with a placebo or another active substance. When interpreting the results, it is also important to check whether the study had a strict, predetermined statistical analysis plan [6].

The significance of the study is influenced by endpoints defined at the beginning: In IBD, clinical remission and clinical response are usually primary endpoints. Active comparators selected as controls determine the significance of the results [7,8].

Once the active substance has been approved and is on the market, observational phase 4 follows, in which significantly more patients (>100,000) can be included and observed. The endpoints are less strictly defined here, but the follow-up period is longer. In this phase, the focus is primarily on further safety aspects of the active substance, which, due to their rarity, only arise if a correspondingly large number of patients are available in the real-world setting.

Categories of clinical trial designs and their limitations

Placebo-controlled studies are not always uncontroversial, as the use of a placebo in many clinical studies appears ethically questionable. In IBD, placebos are therefore used for a short period (8-12 weeks induction) and with a small control arm (2:2:1). If the patient shows no response after this time, they can also receive the verum later. Control groups are also developed for numerous studies (umbrella/platform studies).

Then you have to find the right comparator : In general, the standard of care (SOC) or the choice of standard therapies by the treating physicians can determine the credibility of the results [8,9]. Without appropriate comparators, studies may not be able to demonstrate whether an investigational therapy is superior to prevailing practice and/or leads to bias [10].

Active comparators may not be suitable, depending on, among other things:

• where the study was conducted – in global studies, older, less effective or more toxic treatment regimens may be used [7].
• when the study was designed – the SOC at the time of study planning may no longer be the SOC at the end of the study.
• whether a Best Available Therapy (BAT) vs. SOC is used, especially if crossover is allowed; this may lead to BAT being discontinued due to patient/physician preference, especially in open-label studies.

Head-to-head studies allow a direct comparison of the efficacy and safety of two or more treatments [11]. However, there are only a few direct comparative studies(head-to-head) in which only two drugs are compared and which do not allow conclusions to be drawn about comparisons with other active substances. Other complicating factors here are differences in the study populations, the endpoints, the blinding (compared to open-label studies) and the comparator products [11,12].

Phase 3 trials are usually designed to demonstrate the superiority of a new treatment compared to SOC. They can be conducted to show that two interventions do not differ in either direction by more than a pre-specified arbitrary amount (two-sided test). It is important to design the study with sufficient statistical power – this requires a larger sample population [6,8,13–15].

Non-inferiority ( NI) studies, on the other hand, are designed to show that a novel treatment is no less effective than a certain amount of the standard intervention (one-sided test) . The tested interventions are considered non-inferior if the difference between the treatments is within a predefined equivalence range. The sample sizes required to demonstrate NI (and therefore the cost of the study) are often taken into account when choosing an NI design, as they can be controlled by setting the NI margin to specific values. NI studies should generally be reserved for assessing whether a treatment is less expensive, less toxic or less invasive than SOC (Fig. 1).

The limitations of NI studies include the impossibility of interpreting the results as proof of efficacy, as they cannot distinguish between an effective and a less effective or ineffective treatment. In addition, bias can occur if there is little adherence in the control group (in open-label studies) so that the study basically has no comparison group for the experimental arm. Given the limitations of NI studies, the efficacy results in superiority studies are therefore generally easier to interpret.

Generalizability of study populations to “real-world patients”

Once a study has been completed, it is important to ensure that the active substance is also suitable for the population to be treated. A study includes a specific group of subjects that rarely corresponds to the general population: study participants are usually younger and have fewer comorbidities than heterogeneous, real-world patients, resulting in a higher incidence of adverse events (AEs) in the target populations after approval [10]. With the growing awareness of the need to adapt study results to real patients, the assessment of generalizability, the so-called generalizability assessment, is increasing. Generalizability assessments can be differentiated according to when the assessment took place [17]:

A priori: eligibility-related generalizability (i.e., eligibility criteria are mainly used before the start of the study), which assesses how representative the study population is of the target population.

A posteriori: sample-driven generalizability (i.e. after the study), which assesses how representative the study sample (enrolled patients) is of the target population.

However, few studies use the a priori method; most studies are designed in such a way that concerns about generalizability are not apparent until after the study has ended, missing the opportunity to design the study so that it is applicable to the target population.

The importance of generalizability is highlighted by the increasing appreciation of real-world evidence (RWE) to support the use, benefits and risks of a new therapy [18,19]: RWE, which relates to post-market safety, is increasingly being considered in regulatory decisions and in the design of clinical trials and observational studies.

RWE complements randomized controlled trials (RCTs), which are carefully controlled experiments designed to test specific hypotheses about the efficacy and safety of new drugs that do not necessarily reflect current clinical practice. So if there is a drug or class of drugs with a large amount of data from clinical trials, RWE can contribute to knowledge about patient safety and the comparative effectiveness of drugs.

Development of clinical endpoints over time

Over time, the clinical endpoints have also changed: on the one hand due to the development of new active ingredients and mechanisms that lead to a better cure of diseases, but also with regard to a stronger focus on the patient. The so-called Patient Reported Outcomes (PROs) take this into account. Nowadays, patients can use tablets to enter clinical scores themselves and contribute their assessments.

Many studies continue to work with Clinical Reported Outcomes such as the CDAI or MAYO score, but more and more are also analyzing stool frequency and abdominal pain in Crohn’s disease or stool frequency and blood in ulcerative colitis, which fit better with the clinical or inflammatory biomarkers (e.g. CRP, fCal). Observational Reported Outcomes (usually assessed by nurses) and endoscopic progression (as a surrogate endpoint, not sufficient on its own) are also included in the overall picture [1]. The most important PROs in the area of IBD such as stool frequency, abdominal pain or urgency are shown in Table 1.

Finally, there are aspects that are primarily important for the patient, but also play a role for the practitioner in the long term, such as quality of life. This can be assessed using various general scores, but also a specific score, the 32-item Inflammatory Bowel Disease Questionnaire (IBDQ-32) [20]. In addition, there are also factors such as economic or financial aspects for which there are scores – these are many questionnaires for the patient, but these questionnaires are relevant in order to obtain a more meaningful overall picture.

Composite endpoints and hierarchical testing

Limitations relate primarily to older scores such as CDAI and MAYO (Table 2), which are often somewhat subjective and do not allow a comprehensive analysis of the endpoints. This affects Crohn’s disease even more than ulcerative colitis, as the symptoms of CU are easier to assess.

Composite endpoints, a combination of objective inflammatory markers with qualified PROs, are used in an attempt to bring together the multitude of different aspects. The aim is to develop and ultimately launch therapies that are both effective and safe.

In some studies, the various endpoints are subsequently tested hierarchically and checked for type I errors by defining the so-called primary endpoint in the study protocol in advance [6,21,22]. This must ultimately be positive or at least show a difference so that the study is positive. Further analyses, the so-called secondary endpoints, will only take place if the primary endpoint is positive or is more relevant. This creates a hierarchy of different analyses, subgroups or endpoints, which can only be interpreted if the aspect previously defined in the sequence was positive (Fig. 2). If an endpoint in the hierarchy does not reach statistical significance (level-limiting factor), no further statistics are performed and all endpoints lower in the hierarchy are not considered statistically significant/positive results.

There are various aspects of clinical trial design that can ultimately complicate analysis and interpretation and may determine whether a study actually has the relevance it would like to have. These include a crossover of subjects between the control and experimental groups or the occurrence of bias (Table 3).

The combination of a biological agent and an immunosuppressant has proven to be effective and safe in the treatment of chronic inflammatory bowel disease. A new emerging therapeutic concept is Advanced Combination Treatment (ACT [23]), which involves the combination of at least two biologic agents or one biologic agent and a small molecule drug with different mechanisms of action to achieve optimal disease control in certain patients [24]. The idea behind this is that simultaneously influencing several pathogenic signaling pathways can provide an additive or, in the best case, even synergistic benefit, which could offer an option especially for patients in whom disease control could not yet be achieved with monotherapy alone or who have concomitant immune-mediated inflammatory diseases. For example, the phase 2a VEGA study showed that combined induction therapy with guselkumab and golimumab was more effective in patients with ulcerative colitis than monotherapy with one of the two substances, without any increased safety concerns [25].

Head-to-head studies

Historically, head-to-head studies in IBD initially began with the analysis of steroids. At the end of the 1970s, sulfasalazine vs. prednisone vs. azathioprine vs. placebo were compared in Crohn’s disease. In 2010, SONIC [26] was the first head-to-head study in IBD that included biologics therapy (infliximab + azathioprine vs. infliximab vs. azathioprine in Crohn’s disease) and demonstrated that anti-TNF in combination with thiopurines produced a significant improvement. This was followed in 2017 by the first head-to-head study comparing a biologics originator with a biosimilar in patients with IBD (infliximab in MC) [27]. In such a case, it is ultimately not necessary to show much more than that the drugs have the same effect, as they are to be used for the same indication.

Nowadays, head-to-head comparisons are sometimes already included in the phase 3 trials, in that the pivotal trial already has a comparative arm. Two examples of this are the SEAVUE study [28], in which adalimumab is compared with ustekinumab, and the VARSITY study [29], which compares adalimumab with vedolizumab.

In head-to-head studies, several aspects must be considered and taken into account, including subjective influences of investigators or so-called censoring, in which observations of patients are removed, e.g. if they have a dropout at follow-up (Overview 1).

Take-Home-Messages

• Endpoints in chronic inflammatory bowel disease change rapidly.
• Composite endpoints are gaining in importance: combination of objective measurements and PROs (PROs are more subjective, factors such as inflammation parameters and endoscopy are more objective).
• PROs and flexible modular endpoints could enable studies and approvals for IBD therapies that target other aspects of IBD beyond inflammation.
• Clinical head-to-head studies are important for deciding between different treatment alternatives, but should be interpreted with caution.

Literature:

1. Thabane L, et al: A tutorial on sensitivity analyses in clinical trials: the what, why, when and how. BMC Med Res Method 2013; 13: 92.
2. Juillerat P: Clinical Trial Designs – How to distinguish differences and limitations? IBDmatters – CME lecture; https://medizinonline.com/ibd-matters-clinical-trial-designs/
3. FDA: www.fda.gov/patients/drug-development-process/step-3-clinical-research; last accessed: 15.04.2024.
4. NCI: Phases of Clinical Trials. www.cancer.gov/about-cancer/treatment/clinical-trials/what-are-trials/phases; last accessed: 15.04.2024.
5. Le Tourneau C, et al: J Natl Cancer Inst 2009; 101(10): 708-720.
6. Yuan J, et al: J Biopharm Stat 2016; 26(5): 912-923.
7. Ananthakrishnan R, et al: Crit Rev Oncol Hematol 2013; 88: 144-153.
8. Prasad VK: Adv Drug Develop 2019; 17(3): 150-152.
9. ICH: www.fda.gov/media/71349/download; last accessed: 15.04.2024.
10. FDA: www.fda.gov/media/130326/download; last accessed: 15.04.2024.
11. Kim MS, Prasad V.: Cancer 2020; 126(19): 4270-4272.
12. Kim H, et al: Br J Clin Pharmacol 2013; 77(1): 116-121.
13. Karanicolas PJ, et al: Can J Surg 2010; 53(5): 345-348.
14. ICH: www.fda.gov/media/71336/download; last accessed: 15.04.2024.
15. Head SJ, et al: Eur Heart J 2012; 33: 1318-1324.
16. Haslam A, et al: J Natl Compr Canc Netw 2020; 18(2): 161-167.
17. Pouillon, et al: Head-to-head trials in inflammatory bowel disease: past, present and future. Nature Reviews Gastroenterology & Hepatology 2020; 17: 365-376.
18. He Z, et al: Clin Transl Sci 2020; 0: 1-10.
19. FDA: www.fda.gov/science-research/science-and-research-special-topics/real-world-evidence; last accessed: 15.04.2024.
20. Sherman, et al: N Engl J Med 2016; 375: 2293-2297.
21. Yarlas A, et al: The Inflammatory Bowel Disease Questionnaire in Randomized Controlled Trials of Treatment for Ulcerative Colitis: Systematic Review and Meta-Analysis. J Patient Cent Res Rev 2020; 7(2): 189-205; doi: 10.17294/2330-0698.1722.
22. Korn EL, et al: In: Niederhuber JE, et al, eds: Abeloff’s Clinical Oncology, 6th ed. 2020: 296-307.
23. FDA: www.fda.gov/media/102657/download; last accessed: 15.04.2024.
24. Solitano V, et al: Advanced Combination Treatment With Biologic Agents and Novel Small Molecule Drugs for Inflammatory Bowel Disease. Gastroenterol Hepatol 2023; 19(5): 251-263.
25. Danese S, et al: The future of drug development for inflammatory bowel disease: the need to ACT (advanced combination treatment). Gut 2022; 71(12): 2380-2387.
26. Feagan BG, et al, VEGA Study Group: Guselkumab plus golimumab combination therapy versus guselkumab or golimumab monotherapy in patients with ulcerative colitis (VEGA): a randomized, double-blind, controlled, phase 2, proof-of-concept trial. Lancet Gastroenterol Hepatol 2023; 8(4): 307-320.
27. Colombel JF, et al: Infliximab, Azathioprine, or Combination Therapy for Crohn’s Disease. N Engl J Med 2010; 362: 1383-1395; doi: 10.1056/NEJMoa0904492.
28. Duk Ye B, et al: Efficacy and safety of biosimilar CT-P13 compared with originator infliximab in patients with active Crohn’s disease: an international, randomized, double-blind, phase 3 non-inferiority study. Lancet 2019; 393: 1699-1707.
29. Sands BE, et al: Ustekinumab versus adalimumab for induction and maintenance therapy in biologic-naive patients with moderately to severely active Crohn’s disease: a multicentre, randomized, double-blind, parallel-group, phase 3b trial. Lancet 2022; 399(10342): 2200-2211;
    doi: 10.1016/S0140-6736(22)00688-2.
30. Sands BE, et al: Vedolizumab versus Adalimumab for Moderate-to-Severe Ulcerative Colitis. N Engl J Med 2019; 381: 1215-1226; doi: 10.1056/NEJMoa1905725.
31. Zhang J, et al: Systematic bias between blinded independent central review and local assessment: literature review and analyses of 76 phase III randomized controlled trials in 45 688 patients with advanced solid tumour. BMJ Open 2018; 8: e017240;
    doi: 10.1136/bmjopen-2017-017240.
32. FDA: www.fda.gov/media/71195/download; last accessed: 15.04.2024.

FAMILY PHYSICIAN PRACTICE 2024; 19(7): 4-9