Updates from VARSITY: histologic remission

Professor Jean-Frédéric Colombel (Icahn School of Medicine at Mount Sinai, New York, USA) opened the symposium with an overview of how treatment targets in UC have evolved from improving symptoms.

Written by Maria Dalby

Dr Edward Loftus (Rochester, USA) gives us an update on the histologic remission sub-analysis from the VARSITY study

Professor Jean-Frédéric Colombel (Icahn School of Medicine at Mount Sinai, New York, USA) opened the symposium with an overview of how treatment targets in UC have evolved from improving symptoms and inducing symptomatic remission to aiming for endoscopic remission and the current STRIDE treatment targets which combine clinical remission, defined as resolution of rectal bleeding and normalisation of bowel habits, with endoscopic remission defined as a Mayo endoscopic subscore of ≤1 (1). Histology targets were discussed in the development of STRIDE, but were not included because of insufficient evidence and a lack of a validated histological index. This was also echoed in regulatory guidelines at the time: in 2016 FDA stated that histology was not required to assess mucosal healing, and in 2018 EMA recommended that histological evaluation be included as a secondary endpoint in clinical trials. This attitude is now changing; ECCO and the American College of Gastroenterologists have both issued statements calling for a standardised and validated scoring system for assessing histologic remission.

A major stumbling block in the evolution of histologic healing as a treatment target has been the lack of consistent terminology regarding mucosal healing – until now, many studies have equated mucosal healing with endoscopic remission only, but Professor Colombel stressed that the correct definition of mucosal healing should include both endoscopic and histologic outcomes. When defined as simply a Mayo endoscopic subscore of ≤1, achieving mucosal healing has been shown to significantly reduce the need for colectomy and improve the rate of long-term clinical and steroid-free remission. Failure to achieve endoscopic remission has been associated with increased risk of relapse and hospitalisation.

However, recent data has shown that endoscopic remission is not completely correlated with histologic remission, and even patients with a Mayo endoscoic subscore of 0 may have histologic evidence of active inflammation. Likewise, only around one third of patients who achieve endoscopic remission will have complete resolution of clinical symptoms. In contrast, there is now retrospective data to show that histologic remission is associated with freedom from symptoms and reduced risk of relapse and colorectal cancer, and is also a better predictor of corticosteroid use and hospitalisation than endoscopic remission, suggesting that histologic remission may be more predictive of favourable long-term outcomes in UC.

Far from being a new concept in UC, the relationship between clinical presentation and the histological appearance of biopsies was considered as far back as the 1950s in the works by Truelone and colleagues (2). What is new is the need for robust methodology and validated indices to allow cross-trial comparisons and evaluation. In the next talk of the symposium, Professor Laurent Peyrin-Biroulet (University Hospital of Nancy, France) discussed the scoring systems that are available for assessing histologic inflammation in UC. The most widely used index is the Geboes grading scale, which was published in 2000. This scale was developed empirically and has not been validated, and its complexity makes it unsuitable for use in clinical practice.

Another index that is simple enough to allow it to be used in clinical practice is the Rutgeerts index, but this is also not validated. More recently, two indices have been developed based on the Geboes grading scale and undergone full validation: the Robarts histopathology index (RHI) and the Nancy histopathology index (NHI). Both rely on grading the level of neutrophil infiltration as a marker of inflammation activity. While the RHI is still relatively cumbersome to perform, the NHI has been designed to be easy to use both in clinical trials and routine practice. The Geboes, RHI and NHI indices were used to assess disease activity in the recent TOUCHSTONE phase-2, placebo-controlled study of sphingosine-1-phosphate (S1P) receptor agonist ozanimod, and were shown to be almost identical in terms of reliability and responsiveness in discriminating between treatment and placebo.

In the final presentation of the symposium, Professor Silvio Danese (Humanitas University, Milan, Italy) summarised the recent VARSITY study and discussed the insights it provides with respect to histologic outcomes. Reports in the medical literature of histologic improvement and remission biologics and small molecules in UC have until now been highly variable, because of differences in scoring indices, patient populations etc. VARSITY is the first study to directly compare the outcomes achieved with vedolizumab versus adalimumab, including their effect on histologic activity. VARSITY was designed as a phase-3b, randomised, double-blind, double-dummy, multi-centre, active-controlled study and included a total of 769 patients with moderate to severe UC (3). The study met its primary endpoint in that a significantly higher proportion of patients in the vedolizumab arm were in clinical remission at Week 52 compared with in the adalimumab arm (31.3% v 22.5%; Δ=8.8%; 95% 2.5, 15.0; p=0.006). Vedolizumab was also superior to adalimumab with respect to the rate of endoscopic remission at Week 52 (39.7% v 27.7%; Δ=11.9%; 95% CI 5.3, 18.5; p<0.001). The difference in clinical response between the two treatment arms was evident from Week 6 and was sustained at each assessment until Week 52. Biopsies were performed at Weeks 2, 14 and 52 and assessed using the Geboes grading scale.

This analysis showed that significantly more patients in the vedolizumab arm had minimal histologic activity, defined as a Geboes score of less than 3.2, compared with the adalimumab arm at Week 14 (21.1% v 12.7%; Δ=8.4%; 95% CI 3.2, 13.6) and that the number of patients achieving minimal histologic activity increased progressively in the vedolizumab arm until Week 52 (33.4% v 13.7%; Δ=19.6%; 95% CI 13.8, 25.5). Using the RHI index and a score of 5 or less to define minimal histologic activity resulted in similar significant differences in favour of vedolizumab at Weeks 14 and 52. The difference remained evident when applying the much stricter definitions of Geboes <2 and RHI <3, but with smaller proportions of patients achieving each score. Professor Danese stressed that while these results provide important insights in terms what can be achieved with respect to histologic outcomes with vedolizumab versus adalimumab, incorporating histologic healing as a treatment target in routine clinical practice will require evaluation in a large prospective clinical trial over a prolonged period of time.

  1.  Peyrin-Biroulet L, Sandborn W, Sands BE, Reinisch W, Bemelman W, Bryant R V., et al. Selecting Therapeutic Targets in Inflammatory Bowel Disease (STRIDE): Determining Therapeutic Goals for Treat-to-Target. Am J Gastroenterol. 2015;
  2. Truelove SC, Richards WID. Biopsy Studies in Ulcerative Colitis. Br Med J. 1956;
  3. Sands BE, Peyrin-Biroulet L, Loftus E V., Danese S, Colombel J-F, Törüner M, et al. Vedolizumab versus Adalimumab for Moderate-to-Severe Ulcerative Colitis. N Engl J Med. 2019;