Noninvasive biomarkers as surrogate predictors of clinical and endoscopic remission after infliximab induction in patients with refractory ulcerative colitis.

BACKGROUND/AIMS: Treatment of refractory ulcerative colitis (UC) is a clinical challenge, and after biological therapy, monitoring clinical and endoscopic responses is fundamental. We aimed to investigate and compare the predictive power of different noninvasive parameters for clinical remission and mucosal healing after infliximab induction therapy in refractory UC patients. PATIENTS AND METHODS: Serum and fecal biomarkers, including hemoglobin, white blood cells, erythrocyte sedimentation rate, C-reactive protein (CRP), and fecal calprotectin (FC), and colonoscopy were assessed in 44 patients with refractory UC before and after (week 12) infliximab induction. Clinical and endoscopic responses were measured by clinical Mayo score and endoscopic Mayo subscore, respectively.
RESULTS: After infliximab induction, 54.5% and 65.9% had clinical remission and mucosal healing, respectively. Post-induction CRP and FC were significantly lower in clinical responders versus nonresponders (P = 0.01 and 0.001, respectively) and in patients with mucosal healing than without (P < 0.001). Among all the parameters tested, FC had the best predictive value of clinical remission [Area under the curve (AUC = 0.826)] and mucosal healing (AUC = 0.949). Post-induction FC had 87.5% sensitivity and 89% specificity (cut-off <100 μg/g) for predicting clinical remission and 89.7% sensitivity and 93.3% specificity (cut-off <58 μg/g) for predicting mucosal healing.
CONCLUSIONS: Post-infliximab induction FC can be used as a surrogate marker for predicting clinical remission and mucosal healing in refractory UC patients.

INTRODUCTION

Refractory ulcerative colitis (UC) is moderate-to-severe UC that is either refractory to or intolerant of conventional therapy and immunomodulators and its treatment is a clinical challenge that may end at colectomy.[1] Many patients present with inadequate or no response to steroid treatment (steroid-resistant or refractory colitis) or clinical relapse upon withdrawal of steroids (steroid-dependent colitis) or present with active disease or relapse despite thiopurines at an appropriate dose for at least 3 months (immunomodulator-refractory colitis) and their medical treatment may prove ineffective or toxic, leading to surgical intervention.[23]

Early recognition of the severity of colitis, intensive treatment, and monitoring have contributed to improve the outcome.[4] The main endpoint of treatment in UC is the induction and maintenance of disease remission.[5]

In last years, infliximab (IFX), a chimeric monoclonal antibody directed against tumor necrosis factor-alpha (TNFα), has become an appropriate therapeutic approach for refractory UC patients.[6] These biological agents have proved to be effective in inducing and maintaining remission and substantially improving the clinical course.[5] Mucosal healing (MH) is considered an additional highly significant therapeutic target to achieve long-term remission and to change the natural course of UC.[7]

Colonoscopy is the gold standard in the diagnosis of mucosal healing. However, this approach has some limitations as endoscopy is invasive, costly, time-consuming, uncomfortable procedure for patients and carries the risk of perforation particularly in patients with severe disease.[89] To overcome these limitations, the possibility of noninvasive parameters for measuring the response to biological therapy is appealing because it could decrease the necessity of endoscopic evaluation along with reducing healthcare costs.

Numerous laboratory markers for predicting clinical and endoscopic remission, the main goals of therapy with anti-TNF-α agents in inflammatory bowel disease (IBD), have been discussed in previous studies with varying rate of success, i.e., hemoglobin (Hb), erythrocyte sedimentation rate (ESR), C-reactive protein (CRP), and fecal calprotectin (FC).[89] Recent studies explained the role of FC as a reliable surrogate marker to estimate IBD activity and MH.[101112]

A reliable and noninvasive marker for predicting clinical outcome and MH could provide clinicians with crucial information after the introduction of anti-TNFα treatment in these patients.[5] However, previous studies have shown conflicting results for prediction of the outcome of anti-TNF-α treatment in IBD.[131415] The aim of our study was to investigate and compare the predictive power of different noninvasive biomarkers for clinical remission and MH to IFX induction therapy in patients with refractory UC and to identify the best parameter cut-off point to predict responses to treatment.

PATIENTS AND METHODS

Study design

This prospective study was carried out in the Gastroenterology Unit of Farwaniya (FAR) Hospital in Kuwait between January 2016 and December 2016. The study was approved by the ethical committee for medical research in the ministry of health, Kuwait and was conducted in accordance with the previsions of the Declaration of Helsinki. An informed written consent was obtained from each eligible patient who participated in the study.

Study population

Forty-four adult patients were consecutively recruited with an established diagnosis of UC based on clinical, endoscopic, and histological criteria with moderate-to-severe active disease based on Mayo clinical score[16] (≥6 points) and Mayo endoscopic subscore[17] (>2), and who were inadequately or not responding to any conventional therapy.

These patients were eligible to receive IFX induction therapy [(5 mg/Kg) intravenously at weeks 0, 2, and 6]. Patients with contraindications to anti-TNFα therapy, previously exposed to IFX, or any other anti-TNF agent, indeterminate colitis and other causes of colitis such as CMV colitis or patients with hemorrhoids were excluded.

Methods

At study entry (week 0 before IFX induction therapy), a thorough clinical history was obtained. Demographics, duration of the disease since diagnosis, and past and current medications for UC were reviewed. The severity and extent of the colitis at week 0 and at week 12 of IFX therapy were elicited. Patients underwent a complete physical examination. Blood and fecal samples were collected for laboratory investigations including complete blood count, ESR and CRP, and FC. Colonoscopy was performed at the Gastroenterology Unit, FAR hospital, Kuwait, using a CF-FH260AZL/I colonovideoscope (Olympus Inc., Tokyo Japan) by two experienced endoscopists in IBD who were blinded to medications taken by the patients, and the Mayo endoscopic subscore was calculated by consensus agreement of the two endoscopists for each patient. When disagreement between endoscopists occurred, the endoscopic subscore recorded by the senior endoscopist was approved.

Scheduled study visits were carried out and at week 12 after IFX induction therapy, clinical history, physical examination, serum and fecal biomarker assessment, and colonoscopy were reevaluated to measure outcomes (clinical and endoscopic remission).

Disease activity and outcome measures

Clinical disease activity was assessed using Mayo clinical score where patients with a score of ≤2 were considered to be in clinical remission. Endoscopic activity was assessed by Mayo endoscopic subscore, and those with score of ≤1 were defined as having endoscopic remission (MH).

Measurement of fecal calprotectin

Fecal samples obtained prior to colonoscopy (at week 0 and 12 of IFX induction) were stored at −20°C. Fecal calprotectin levels were determined using an ELISA kit for calprotectin (Buhlmann Co., Switzerland) according to the manufacturer's instructions. The upper limit of normal FC in FAR hospital lab is <50 μg/g.

Statistical analysis

All statistical analyses were conducted using SPSS for windows version 16 (SPSS Inc., Chicago, IL. USA). Continuous data were expressed as means ± standard deviation (SD) or median (IQR) and compared using Student's and paired t-tests or Mann–Whitney U and Wilcoxon tests and correlated using Pearson's or Spearman's coefficient for normally or abnormally distributed data, respectively. Categorical variables were expressed as percentage and compared using Chi-square (χ2) test. The area under receiver operating characteristic (AUC) curves were plotted to measure and compare the performance of different parameters in predicting the disease improvement and to select the best cut-off value with the highest accuracy [by calculating the sensitivity, specificity, positive (PPV), and negative predictive values (NPV), positive and negative likelihood ratios (+LR, −RL)]. Multiple regression analysis was used to determine risk factors for clinical and endoscopic remission. Statistical significance was considered when P value was <0.05.

RESULTS

Forty-four UC patients were consecutively and prospectively included in the study. Their mean age was 32.6 ± 1.6 years and 59.1% were men. All patients were previously treated with 5-aminosalicylates (5-ASA) and steroids, and 28 patients (64%) were steroid-dependent and 16 patients (36%) were steroid-resistant. Thiopurines were used in 25 patients (57%) and one patient (who failed to respond to thiopurines) received cyclosporine before IFX and did not respond. Disease extension was left-sided colitis (n = 27, 61.4%) and extensive colitis (n = 17, 38.6%). All patients were treated with IFX. The characteristics of the patients are shown in [Table 1]. Of the 44 patients, 24 (54.5%) had clinical remission, whereas 20 (45.5%) did not. On follow-up colonoscopy, MH was achieved in 29/44 (65.9%) patients [Table 1].

Table 1

Characteristics of the study patients at baseline (week 0) and after infliximab induction (follow up at week 12)

Assessment of laboratory parameters and their relation to clinical and endoscopic remission

Figure 1 shows that, in the cohort of patients with clinical remission at week 12 (follow-up), the median FC value reduced from 507 μg/g to 23 μg/g after IFX induction therapy (P < 0.001). In contrast, in patients who did not achieve clinical remission, the difference in FC values that reduced from 312 μg/g to 204 μg/g was not significant. At baseline, the FC values did not differ between patients who achieved clinical remission and those who did not (507 μg/g vs. 312 μg/g, P = 0.08), whereas after IFX induction, the latter group had significantly higher median FC values compared to the former group (204 μg/g vs. 23 μg/g, P = 0.001). In addition, the median CRP levels significantly decreased after anti-TNF α induction in the group of patients with clinical remission at week 12 (37 mg/L vs. 16 mg/L, P = 0.04) and when compared with patients without clinical remission (42 mg/L vs. 16 mg/L, P = 0.006). No significant differences between patients regarding other parameters [Figure 1].

Figure 1

Laboratory parameters at baseline (Week 0) and after infliximab induction (follow-up at week 12). Variable at baseline (Week 0); Variable at follow-up (Week 12). Hb: Hemoglobin; WBC: White blood cells; ESR: Erythrocyte sedimentation rate; CRP: C-reactive protein; FC: Fecal calprotectin

We found that CRP and FC values at follow-up colonoscopy (week 12 after IFX induction) were significantly lower in patients with MH compared to patients without MH (18 mg/L vs. 46 mg/L, P = 0.02, and 23 μg/g vs. 386 μg/g, P < 0.001, respectively), whereas CRP and FC values at baseline colonoscopy were not significantly different (40 mg/L vs. 38 mg/L, P = 0.347, 358 μg/g vs. 347 μg/g, P = 0.197, respectively). In patients with MH, CRP, and FC values at follow-up colonoscopy were significantly lower compared to CRP and FC values at baseline colonoscopy (38 mg/L vs. 18 mg/L, P = 0.023, and 347 μg/g vs. 23 μg/g, P < 0.001, respectively), whereas there was no significant difference of CRP and FC values from baseline to follow-up colonoscopy in patients without MH (40 mg/L vs. 46 mg/L, P = 0.51 and 358 μg/g vs. 386 μg/g, P = 0.487). There were no significant differences between patients with and without MH regarding other parameters [Figure 1].

Comparison of predictive accuracy and determination of the best cut-off value of laboratory parameters for clinical and endoscopic remission

Based on the ROC curves, Hb, WBCs, ESR, CRP, and FC had good prognostic accuracy for the prediction of clinical remission after IFX induction [Figure 2a], where FC yielded the highest AUC (0.826) and 95% confidence interval (CI) (0.682–0.923, P < 0.001), with 87.5% sensitivity, 89% specificity, 86.9% PPV, 89.5% NPV, and 8 +LR at cut-off of <100 μg/g [Table 2].

Figure 2

Area under the receiver operating characteristic curve (AUC) of laboratory parameters to predict clinical (a) and endoscopic (b) remission. FC had the highest AUC in predicting clinical remission (AUC = 0.826) and mucosal healing (AUC = 0.949). Hb: Hemoglobin; WBC: White blood cells; ESR: Erythrocyte sedimentation rate; CRP: C-reactive protein; FC: Fecal calprotectin

Table 2

Diagnostic accuracy of laboratory parameters to predict clinical and endoscopic remission with the best predictive cut-offs

In addition, by using the ROC curve, the ability of the post-induction FC values to predict mucosal healing at follow up colonoscopy (week 12 of IFX induction therapy) [Figure 2b] revealed the highest AUC (0.949) and 95%CI (0.838–0.992) with 89.7% sensitivity, 93.3% specificity, 96.3% PPV, 82.4% NPV and 13.4 + LR with a cut-off of <58 μg/g (P < 0.001) [Table 2].

Risk factor analysis for clinical and endoscopic remission

Multiple regression analysis was performed for patients with clinical remission and MH, and the variables included were age, gender, disease duration and extension, patients' medication, and laboratory data after IFX induction therapy. The analysis showed that the only variable retained in the models was the FC level, and it was the only variable that may predict clinical remission and mucosal healing (P = 0.041 and 0.02, respectively) [Table 3].

Table 3

Multiple regression analysis of risk factors affected clinical and endoscopic remission in the studied sample

DISCUSSION

UC is a lifelong disease with generally unpredictable clinical course that requires colonoscopy for diagnosis and disease monitoring. Identifying markers that allow noninvasive diagnosis and disease monitoring would be of significant advantage, especially in refractory UC. Several studies have investigated the use of noninvasive markers to predict or monitor the response to medical treatment.[89101112]

This study evaluated only short-term (90 days) response to IFX in patients with refractory UC because almost always long-term remission is established in this period of time, and when a patient does not respond after three drug infusions, this predicts that he will not respond to further administration of the drug.[1819] In the present study, clinical remission and MH were achieved in 54.5% and 65.9% after IFX induction course, respectively.

Our study assessed different noninvasive markers (Hb, WBCs, ESR, CRP, and FC) to elaborate reliable methods to predict outcome, and we demonstrated that among these markers, FC levels measured after IFX induction course can be used as a surrogate marker to estimate the efficacy of IFX as well as to predict the clinical response and MH. Our findings were compatible with that of D'Haens et al.[12] who found a significant correlation between FC values and endoscopic disease scores (Mayo subscores) for predicting endoscopic remission. Zittan et al.[20] also demonstrated that the FC test highly correlated with clinical symptoms in UC but less so in colonic CD. However, Tursi et al.[21] observed that FC correlated with clinical and endoscopic severity of IBD.

The present work showed that CRP and FC levels were significantly lower in patients with clinical remission than those without; however, only FC was significantly lower in patients with endoscopic remission than those without. This finding can be explained by the fact that FC appears to be a better indicator of mucosal inflammation than serum biomarkers, including CRP.[101122]

On monitoring the response to IFX, FC had the highest AUC for prediction of clinical and endoscopic remission compared to serum markers. Regarding clinical remission, FC had 87.5% sensitivity, 89% specificity, and 86.9% PPV at cut-off <100 μg/g. Our results are consistent with the previously published data in this field. Molander et al.[14] reported that FC level is a useful marker for predicting clinical outcome after anti-TNF-α treatment. On the other hand, Tursiet al.[23] showed that FC test had lower diagnostic accuracy (66.7% sensitivity, 56.1% specificity, and 18.2% PPV) in predicting clinical remission in UC patients under treatment with anti-TNFα; moreover, they concluded that rapid FC seemed better in predicting persistence of endoscopic lesions than clinical remission.

In addition, many authors have observed that patients in clinical remission exhibited low FC levels, whereas patients with high levels of this biomarker were at an increased risk of relapsing, suggesting that the FC level might be used not only as a marker of clinical remission but also to predict relapse.[2425]

Most physicians still use a clinical activity score in their everyday practice for patient monitoring, favoring it over endoscopy and biomarkers in the management of their patients.[26] However, the concept of attaining MH as the primary endpoint of treatment in IBD has been established that may be associated with lower relapse rates, hospitalization rates, and reduced need for surgery.[27]

For prediction of MH in our study, FC had 89.7% sensitivity, 93.3% specificity, and 96.3% PPV with cut-off <58 μg/g. In addition, about two-thirds of UC patients showed normalization of FC levels at week 12 of follow-up, and all of them achieved MH according to endoscopic Mayo subscore; hence, no discrepancies between MH and FC level normalization were observed in our study. Our results are consistent with De Vos et al.[13] who showed that FC value <50 mg/kg was a very good predictor of mucosal healing at week 10 of IFX induction. Szczepański et al.[28] stated that FC is a good biomarker of MH with values below 54 μg/g enabling to select 77% patients with full MH.

Unlike our results, several studies experienced different FC cut-off values for predicting MH in UC; Lin et al.[29] and Theede et al.[30] demonstrated that a calprotectin level below 190 μg/g was associated with mucosal healing in UC (UCEIS and Mayo Endoscopic Score). These findings confirmed that FC can reflect endoscopic disease activity with different endoscopic scores.

Calprotectin is a marker of neutrophil migration to the intestinal lumen,[31] and its high level is associated with increased inflammation in the intestine.[3233] Further, FC is a better predictor of mucosal inflammation than other acute phase reactants such as ESR and CRP, which is consistent with data from other studies.[3134] Therefore, monitoring FC may decrease the need for invasive procedures to distinguish patients with quiescent disease who develop functional symptoms from those with active inflammation and who may benefit from escalation in therapy.[20] FC is a reliable, simple, inexpensive, and safe test that correlates closely with endoscopic findings in UC.[3536] It provides more objective results than fecal microscopy and bypasses the difficulty of identifying stool leukocytes.[37]

Our limitations were being a single-center and small sample sized study, as well as lack of histological assessment, and hence, lack of estimation of biomarker cut-off values for prediction of histological remission.

To our knowledge, it is one of the first studies in Gulf countries to address IFX therapy in refractory colitis. Therefore, these findings need to be confirmed by more prospective studies with larger populations and to determine whether these biomarkers will be of benefit for the more severely affected patients who are unable to undergo endoscopic screening. In addition, further studies are needed to assess long-term outcome after the induction therapy of IFX.

CONCLUSION

Post-IFX induction FC can be used as a surrogate marker for predicting clinical remission and MH in patients with refractory UC.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.