Volume 1 - Issue 2, June 2017

The Long March to Personalized Medicine in IBD Treatment Back

Large unmet needs in IBD treatment

Inflammatory bowel diseases (IBD) are chronic diseases. If given no treatment, patients with Crohn’s disease (CD) can develop complications such as structural damage of the intestinal tract and infections within two years. Severe complications may require surgery or hospitalization. However, if anti-TNF therapy is initiated soon after diagnosis, the risk of complications can be dramatically attenuated 1.

Many biologics have demonstrated therapeutic effects in IBD treatment. Clinical trials have shown that infliximab, certolizumab pegol, adalimumab and vedolizumab resulted in significantly more patients with ulcerative colitis (UC) reaching clinical remission2, 3. All four drugs plus ustekinumab have also been proven to be effective in treating patients with CD2, 3. However, these biologic agents showed no benefit in more than 50% of IBD patients. Moreover, previous studies have shown that anti-TNF therapies lose their efficacy after the first year. Though some patients reached clinical remission after treatment with infliximab, adalimumab, or certolizumab pegol, remission was maintained in less than half of these patients after one year of therapy4-6. Thus, it can be seen that there are still large unmet needs in IBD treatment.

Precision medicine may provide better treatment for patients with IBD

Despite low rates of remission, biologic agents can still provide benefit for IBD patients. Patients with endoscopic scores of 0 at week 8 had a much higher chance of reaching steroid-free remission at week 30 when treated with infliximab (46% vs. 6.5% in patients with endoscopic scores of 3)7. One-year mucosal healing has been shown to be a predictor for resection in patients with UC and CD8. Patients with better mucosal healing rates after one year of treatment were found to be at lower risk for both colectomy and resection.

Biologics deliver mucosal healing at different speeds. Vedolizumab is equally as strong as infliximab and delivers similar rates of mucosal healing7, 9. However, most CD patients given these agents showed no benefit until after week 40, and healing effects were also slow for UC patients9.

Precision medicine is based on individual characteristics such as disease progression, treatment response, response to changes in therapy, and long-term outcomes10. As such, precision medicine can provide a personalized approach to patient management. As patients on IBD treatments show varying outcomes, researchers have begun to look for specific predictors of patient’s response.

Practical predictors for personalized medicine have yet to be found

The thiopurine methyltransferase (TPMT) genotype is currently the most accurate at predicting adverse events from azathioprine. Low and intermediate TPMT activity have been shown to significantly increase the risk of adverse events such as myelosuppression in patients receiving thiopurines for IBD (79% for moderately low activity vs. 35% for normal activity, P < 0.001)11. It should be noted that most predictors found in clinical trials showed poor results at the individual level or were not applicable in clinical patient care settings.

Polymorphisms on specific genes, matrix-based prediction tools, and micro-RNA have all been found to be significantly associated with anti-TNF treatment response in IBD patients12-14. However, due to overly small study cohorts and small differences in odds ratios, these predictors produced poor results in individualized medicine. The only potential predictor of anti-TNF response was prior IBD surgery13, 15, which resulted in worsening inflammation.

In the 7-year certolizumab pegol study, low hematocrit was shown to be associated with deceased probability of maintaining remission. Despite this discovery, only about 20% patients still remained in remission at the end of the study16, 17. Hematocrit can predict patient response, but this finding alone is not enough to trigger advances in personalized medicine. Another long-term study found that consistently low albumin levels were indicative of subsequent bowel stenosis, perforation, or fistula18, 19. Low serum concentrations of golimumab, certolizumab, and vedolizumab have all been found to be correlated with lower rates of patient response20, 21.

One thing we can predict in IBD treatment is adverse effects from drugs. For example, on-study narcotic use and on-study corticosteroid use were found to be more likely to cause serious infections in patients with prior anti-TNF failure who were given vedolizumab22. However, no study has found a good predictor for IBD treatment response that can be applied clinically to precision medicine.

Patients should be evaluated early to optimize therapy outcomes

In terms of ex post predictions, patients who responded well in the early stages of anti-TNF treatment tended to maintain remission during long-term therapy23-25. Patients should be assessed early on when receiving biologic treatment to determine if the biologic agent should be continued or if another agent should be chosen for the next step.

Vedolizumab can serve as alternative therapy in patients with previous anti-TNF failure

Vedolizumab has been proven to be effective in treating CD and UC, even in patients with prior anti-TNF failure.

GEMINI II studied the use of vedolizumab in induction and maintenance therapy in patients with moderate to severe CD26. GEMINI III investigated the effects of vedolizumab in population with larger proportion of prior anti-TNF failure27. Patient outcomes at week 6 and week 10 showed significantly different rates of clinical remission as compared to the placebo group, regardless of whether patients had previous anti-TNF failure (see Fig. 1)26, 27 .

Figure 1. Patient outcomes at week 10 of GEMINI III

In anti-TNF-naïve patients, response at week 6 was the same as response at week 10. However, response at week 10 was superior to response at week 6 in anti-TNF-experienced patients, suggesting that these patients need more time to respond to vedolizumab27. Therefore, evaluations of anti-TNF-experienced patients conducted at week 6 may not be entirely accurate.

In the long-term study, vedolizumab treatment showed continued efficacy beyond 52 weeks. If patients responded to vedolizumab in the first year, this effect was continued regardless of whether patients were anti-TNF-naïve or had previously experienced anti-TNF failure (see Fig. 2)28.

Figure 2. Patient outcomes after 5-year vedolizumab exposure

The effects of vedolizumab in treating UC were demonstrated in GEMINI I. The outcomes of the vedolizumab group were significantly better than placebo regardless of whether patients had received prior anti-TNF therapy. However, anti-TNF-naïve patients showed better response compared to patients who had previously been exposed to anti-TNF therapy (clinical response rates of 53.1% vs. 39.0% and clinical remission rates of 23.1% vs. 9.8%)29. Meanwhile, the effects of vedolizumab on clinical response, mucosal healing, and clinical remission were all sustained throughout the 52 weeks of the study 30. No serious adverse events or infections occurred when treating IBD with vedolizumab11.


Considerable variability exists in the success rates of different biologic therapies used for treating IBD. However, we are still on the way of finding predictors for personalized medicine With the reduced immunosupressive side effects and improved long-term efficacy, new generation biologics, such as vedolizumab, may change the need for predictors. Future benchmark trials comparing different biologics are needed to establish better algorithms for treating IBD.


  1. Siegel CA, et al. Inflamm Bowel Dis. 2011 Jan; 17 (1): 30-8.
  2. Schreiber S, et al. Z Gastroenterol. 2015 Jun; 53 (6): 591-602.
  3. Feagan BG, et al. United European Gastroenterol J. 2015; 3 (5 suppl): Presentation LB4.
  4. Hanauer SB, et al. Lancet. 2002 May 4;359(9317):1541-9.
  5. Colombel JF, et al. Gastroenterology. 2007 Jan;132(1):52-65.
  6. Schreiber S, et al. Gut 2006; 55 (Suppl V): A131.
  7. Colombel JF, et al. Gastroenterology. 2011 Oct;141(4):1194-201.
  8. Frøslie KF, et al. Gastroenterology. 2007 Aug;133(2):412-22.
  9. Dulai PS, et al. Am J Gastroenterol. 2015; 110: S809. Abstract 1904.
  10. Jameson JL, Longo DL. N Engl J Med. 2015 Jun 4;372(23):2229-34.
  11. Ansari A, et al. Aliment Pharmacol Ther. 2008 Oct 15;28(8):973-83.
  12. Bek S, et al. Aliment Pharmacol Ther. 2016 Sep;44(6):554-67.
  13. Billiet T, et al. J Crohns Colitis. 2015 Dec;9(12):1120-6.
  14. Peck BC, et al. Inflamm Bowel Dis. 2015 Sep;21(9):2178-87.
  15. Schreiber S, et al. Am J Gastroenterol. 2010 Jul;105(7):1574-82.
  16. Colombel JF, et al. Inflamm Bowel Dis. 2014;20:S42.
  17. Colombel JF, et al. 2014 Meeting of the CCFA. December 4-6, 2014; Orlando, FL; Poster P-039.
  18. Irwin JR, et al. Gastroenterology. 2015;148(4 suppl 1):S-459.
  19. Irwin JR, et al. 2015 Digestive Disease Week Meeting; May 16-19, 2015; Washington, DC; Poster Su1277.
  20. EMA Dosier, Figure 14, page 111 CSR T18.
  21. Feagan BG, et al. N Engl J Med. 2013 Aug 22;369(8):699-710.
  22. Colombel, JF, et al. Gut. 2016 Feb 18. pii: gutjnl-2015-311079. doi: 10.1136/gutjnl-2015-311079.
  23. de Silva PS, et al. Aliment Pharmacol Ther. 2012 Sep;36(5):459-66.
  24. Melmed G, et al. 2015 Digestive Disease Week Meeting; May 16-19, 2015; Washington, DC; Poster Sa1190.
  25. Iwasa R, et al. BMC Gastroenterol. 2015 Aug 14;15:103.
  26. Sandborn WJ, et al. N Engl J Med. 2013 Aug 22; 369 (8): 711-21.
  27. Sands BE, et al. Gastroenterology. 2014 Sep;147(3):618-627.e3.
  28. Vermeire S, et al. J Crohns Colitis. 2016 Sep 28. pii: jjw176.
  29. Feagan BG, et al. Clin Gastroenterol Hepatol. 2017 Feb; 15 (2): 229-239.e5.

Feagan BG et al. N Engl J Med. 2013 Aug 22; 369 (8): 699-710.