Prophylaxis of Crohn's disease recurrence: A surgeon’s perspective

Abstract Management of inflammatory bowel disease has evolved extensively in the last three decades. We have learnt a lot about the pathophysiology and natural history of the disease. New effective classes of drugs with the associated potential morbidity have been introduced. New surgical techniques have been popularized leading to a better understanding of the optimal timing of surgery. The result is a very complex subspecialty of gastroenterology and colorectal surgery called the "IBDologist." Only if we manage these complex patients in the context of a multi‐disciplinary team will we be able to obtain outstanding outcomes, specifically with high and sustained remission rates for these patients.

Seventy-five to ninety percent of patients with CD will have endoscopic recurrence within 1 year of initial surgery, a majority of which will progress to have recurrence of symptoms clinically and require reoperation. [16][17][18] Of the patients that have undergone a primary excisional surgery, 25% will require a subsequent procedure within 5 years, 35%-60% in 10 years. 12,14,19 Historically, in the mid-20th century, the overall rate of recurrence was 30%-34% with a higher risk associated with involvement of small bowel and early age of onset. 20 Recurrence most often occurs just proximal to or at the anastomotic site and is categorized as subclinical or endoscopic, clinical, and lastly surgical. 17,18,[21][22][23] Endoscopic recurrence is graded by the Rutgeerts score 3,18,23 based on evaluation of the neo-terminal ileal intestinal mucosa after surgical resection for CD. [18][19] Normal ileal mucosa is designated i0; i1 for <5 aphthous ulcers; i2 for >5 aphthous ulcers with presence of normal intervening mucosa; i3 for presence of ulceration with grossly normal intervening mucosa; i4 for severe ulcerations, presence of nodules, cobblestoning, or strictures. 18,23 Endoscopic evidence of subclinical recurrence is typically detected early with the severity indicative of likely progression to clinical recurrence 23 and it is detectable as early as 1 week after surgery. 15,18,23 Clinical recurrence is defined as return of symptoms and is seen in 30% of patients within the first year of initial surgery, 5%-10% requiring reoperation less than a year after primary resection.
Finally, surgical recurrence is the subsequent need for reoperation after initial bowel resection.

| RIS K FAC TOR S FOR RECURREN CE
Several factors have been described in the literature including active smoking, penetrating or perforating behavior, perianal disease, prior intestinal resection, extensive small bowel resection (>50 cm), type of anastomosis, postoperative complications, resection margins, age at diagnosis of disease, gender, location and duration of disease, granuloma, family history, CRP level, myenteric plexitis, NOD2/CARD15 mutation, increased TGF-b, and low II10 mRNA level. 11,12,24 Of these, there are five established predictors of postoperative recurrence: penetrating phenotype, perianal disease, prior intestinal resection, extensive small bowel resection, and active smoking. 11,25

| Smoking
Smoking is the only modifiable proven risk factor associated with more severe complications of penetrating and fibrostenotic CD, increased risk of surgical recurrence. 26 28 showed a significant association between current smokers (CS) and need for first surgery (HR 1.27) when compared to those who had never smoked (NS). Furthermore, Reese et al pooled data from 30 studies from 1966 to 2007 and found CS to have an increased risk of clinical recurrence following initial resection (58.3% vs 39% P < .005) and an increased risk of reoperation within 10 years of the initial resection (55.5% vs 32.1% P < .001) than NS. Recurrence is more likely to be at the prior surgical site in CS than NS. 29 In a cohort of 174 patients, Sutherland et al 30 found 5-and 10-year surgical recurrence rates in smokers to be 36% and 70% respectively when compared to NS (20%, 40%). In a retrospective study of 141 patients with ileocecal disease who underwent ileocolonic resection, NS had a recurrence-free rate of 81% and 64% at 5 and 10 years, respectively, vs 65% and 45% in CS (P = .007). 31 The authors also found the recurrence rate to be higher in heavy smokers, further support the role of active smoking. 31

| Perforating disease
Perforating phenotype is more aggressive, associated with more complications, and is described as acute/subacute with abscesses, or chronic with fistulae. 31,36,37 Perforating disease is one of the most common indications for surgery and has the highest rate of surgical recurrence. 1,33,38 In 2008, Simillis conducted a meta-analysis of 13 studies comparing surgical recurrence and the need for reoperation between perforating and non-perforating CD patients who have undergone a primary resection. There was a significantly higher rate of surgical recurrence in perforating phenotype (P = .002) with phenotypic concordance between the primary and recurrent presentation and shorter surgical recurrence-free survival (1.7 vs 13 years). 38 A meta-analysis looking at postoperative medical management vs placebo in maintaining clinical remission showed that perforating disease had greater endoscopic recurrence scores in the placebo group (P < .05). 39

| Small bowel disease
Ileocolonic disease is the most common presentation and thus most common site of surgical resection. 21 Bernell et al 5,40 reported increased RR of POR to 1.8 and 1.5 (95% CI: 1.1-2.0) respectively in patients with small bowel or ileocecal disease compared to colonic. Manser et al 41 found ileal disease to be a predictive factor for reoperation (OR 2.42, P = .05). Ileal disease was associated with higher POR in a study by Borley et al 42 , and finally, a Swedish study evaluated the cumulative POR rate for ileal, ileocolonic, and colonic CD at 5, 10, and 15 years and found ileal and ileocolonic disease to have significantly higher POR than colonic disease. 43 Ileocolonic CD confers a shorter interval between onset of symptoms and initial operation associated with higher POR. 44

| Perianal disease
The documented prevalence of perianal/anorectal involvement varies from 10% to 40%. 45,46 Isolated perianal disease is rare and is present in <5% of patients 47 and conversely 15% of patients with ileocolonic disease and over 90% with colonic disease will have concomitant perianal involvement. 46 Presence of perianal disease typically confers a more debilitating course with more frequent and often more severe extraintestinal manifestations and tends to be less responsive to conventional steroid therapy. 24,48,49 Treatment involves both medical and surgical modalities; antibiotic bridge to immunosuppressive therapy and exam under anesthesia with abscess drainage and commonly, seton placement. 50 The success of surgical treatment is dependent upon the presence of active proctitis. In a recent study, patients with rectal disease had a significantly higher rate of proctectomy than patients with rectal sparing (77.6% vs 13.6%, P < .0001). 51

| Duration of disease
Patients are stratified based on the risk of POR in three categories: low-risk patients >10 years of disease, <10 cm segment of stricture, and no history of prior surgery or first surgery; moderate-risk <10 years since diagnosis, >10 cm segment of intestinal stricture or narrowing, non-perforating disease, no history of multiple surgeries; high-risk patients are smokers, have perforating disease and history of multiple surgeries. 52 Similarly, work by Yamamoto 53 supported the theory that a shorter disease course from the time of diagnosis to initial surgery was a strong predictor of POR. However, what constituted a "short" duration of disease has yet to be agreed upon, but is generally thought to be <10 years. Poggioli 54 found that preoperative disease duration of <6 years was associated with higher surgical recurrence rates postoperatively. Another study by Chardavoyne et al 55 found higher recurrence rates between 3 and 10 years of disease duration. In contrast, a pediatric study found that patients who underwent primary resection within 1 year of symptom onset experienced delayed recurrence (30% in 8 years), compared to those who underwent surgery between 1 and 4 years after symptom onset (50% in 4 years) and those who had surgery 4 or more years after symptom onset (50% in 3 years). 56 While it remains unclear whether disease duration plays a role in surgical recurrence, a shorter duration between onset of symptoms and resection indicates a likely more aggressive phenotype. 53

| Age at diagnosis
The association of age at diagnosis and POR has been frequently reported in the literature. The younger a patient the more aggressive and disabling the disease course. 57 Scarpa et al 58 conducted a retrospective review of 120 patients and found younger age at diagnosis was a risk factor for POR. Dombal found the highest recurrence rate in younger children and adolescents, lower in adults, and lowest rates of recurrence in patients >60 years. 19

| Microbiome
The investigations of role of the fecal microbiome in POR is at its infancy. It is believed that the fecal and intestinal microbiota is an antigenic driver of CD recurrence. 59

| Nutritional supplementation with elemental diet
Enteral nutrition with an elemental diet is widely utilized in Japan preoperatively for disease control and maintenance of remission 61,62 and nutritional optimization, even though a recent meta-analysis suggested it to be inferior to corticosteroid treatment. 63 In surgically induced remission enteral nutrition has been shown to prevent POR. 64,65 With a better understanding of the role of the microbiome, it would be interesting to investigate the impact of elemental diet on the gut microbiome to elucidate potential mechanisms of recurrence prevention.

| SURG IC AL PROPHYL A XIS
In a case-controlled comparative analysis, 138 patients were divided between wide-lumen stapled side-to-side and hand-sewn end-to-end anastomoses. 66 Clinical recurrence was noted in 24% of side-to-side anastomoses and 57% of end-to-end anastomoses with a cumulative surgical recurrence rate of 11% and 20% at 5 years (P = .017). The authors concluded that the side-to-side anastomosis created a wider diameter lumen that reduced clinical recurrence by limiting fecal stasis and reducing recurrence secondary to ischemia. A subsequent meta-analysis of eight studies, with 661 patients, and 712 anastomosis compared multiple types of anastomoses. Side-to-side anastomosis had fewer leaks and reduced overall postoperative complications and length of hospital stay. The overall and surgical recurrence rates of these groups, however, were similar. 67 Several retrospective studies were subsequently performed and confirmed no difference in anastomotic leak rates or surgical recurrence between the two groups 68,69  in the conventional group (P < .0013). 72 Two large multicenter studies reported only two surgical recurrences at 65 months and a 98.6% recurrence-free survival for 5 and 10 years in the Kono-S group. 73,74 Recently, the first randomized controlled trial to compare Kono-S and side-to-side anastomoses confirmed a reduction in the endoscopic and clinical recurrence in favor of the Kono-S. 75 Additionally, the Kono-S patients who developed endoscopic recurrence had significantly lower Rutgeerts scores (P = .03). 75 The opposite approach of a wide mesenteric resection was recently published in a small cohort study of 64 patients. 76 The difference in POR was 2.9% vs 40% (P = .003) in favor of the wide mesenteric excision. While details of the type of anastomoses utilized are not available in the study, wide excision should be compared in a prospective and randomized fashion with the Kono-S anastomosis.

| MEDIC AL PROPHYL A XIS
Medical therapy-immunosuppressive modulators and biologic agents-have been associated with decreased need for operation in the short-term, without definitive long-term evidence. 77 In addition to antibiotics and immunomodulators, several studies have demonstrated anti-tumor necrosis factor to be efficacious in reducing POR in CD, 52,78-85 with a more pronounced effect noted in anti-TNF naïve patients. 86  trial to compare early endoscopic surveillance in patients and escalation of medical therapy based on preoperative risk stratification with standard conservative management. 92 Patients were categorized as high-or low-risk then randomized for 6-vs 18-month initial colonoscopy following surgery. High-risk patients were treated with either thiopurine or and all patients were treated with 3 months of metronidazole. If endoscopic recurrence was observed in any of the patients in the 6-month group, they received additional treatment. At 18 months, the patients who underwent initial ileocolonoscopy at 6 months and additional treatment if indicated had an endoscopic recurrence rate of 49% vs 67%, an 18% reduction. Endoscopic recurrence in the high-risk group at 6 months was 45% for those receiving thiopurine and 21% with adalimumab. Two conclusions can be drawn from these findings: patients considered to be high-risk benefit from shorter intervals of postoperative surveillance; if postoperative medical prophylaxis is indicated, anti-TNF therapy should be considered in high-risk patients. 87,[92][93][94]

| D ISCUSS I ON
Surgical resection in CD is not curative and recurrence is common; thus, maximizing preventative efforts to reduce POR is essential.
Recurrence occurs endoscopically first, followed by return of symptoms resulting in a second operation. Identifying high-risk patients is critical to initiate or resume medical therapy in a timely manner. It is important to look at the disease characteristics that are proven risk factors for POR and intervene on the modifiable factors like smoking. Surgery has to be performed following the principles of wide lumen anastomosis and avoidance of postoperative complications.
The initial step in postoperative CD management is risk stratification. Early postoperative endoscopy for high-risk patients with intensification of medical therapy is associated with 18% lower rate of endoscopic recurrence. By targeting high-risk patients with early colonoscopy, unnecessary and costly treatments can be avoided in low-risk patients. While there is no concrete evidence that smoking decreased therapeutic response to biologic therapy, there is a two times increased risk of POR recurrence in active smokers either incapable or unwilling to stop. 95

D I SCLOS U R E
Conflict of Interest: The authors declare no conflict of interests related to this article.