Fecal Calprotectin Shows Predictive Value for GI-GVHD After Allogeneic Stem Cell Transplant
核心洞察
Fecal calprotectin (搜索) demonstrated meaningful predictive value for gastrointestinal graft-versus-host disease (搜索) between days 14 and 21 post-transplant, with an AUC of 0.77 at day 21 in a prospective study of 165 adult allo-HSCT recipients.
The optimal day 21 threshold of 52.5 µg/g provided 75% sensitivity and 87% specificity, while day 7 measurements showed no discriminative capacity (AUC 0.50).
FC levels correlated significantly with endoscopic severity (r=0.31; p=0.02) and declined after treatment initiation from 120 µg/g to 51.5 µg/g (p=0.04), supporting its role in monitoring therapeutic response.
Fecal calprotectin (搜索) (FC) may serve as a dynamic, non-invasive biomarker for predicting, diagnosing, and monitoring gastrointestinal graft-versus-host disease (搜索) (GI-GVHD) following allogeneic hematopoietic stem cell transplantation (allo-HSCT), according to findings from a prospective observational study conducted at the University Hospital of Salamanca (搜索) and published in Bone Marrow Transplantation.
The study, led by Piñero-Pérez and colleagues, enrolled 165 consecutive adult allo-HSCT recipients between December 2019 and December 2022, representing the largest sample size to date for this clinical question. GI-GVHD developed in 52.7% of patients, with histological confirmation achieved in 90.3% of cases.
Timing Is Critical for Predictive Performance
Serial FC measurements obtained on days 7, 14, and 21 post-transplant revealed that the biomarker's predictive utility is highly dependent on the timing of measurement. On day 7, FC showed no discriminative capacity, yielding an area under the ROC curve (AUC) of 0.50. Performance improved on day 14, with an AUC of 0.69 (p < 0.001; 95% CI: 0.59–0.78). The optimal cut-off on day 14 was 39 µg/g, providing 52% sensitivity and 84% specificity.
The highest discriminative performance was observed on day 21, with an AUC of 0.77 (p < 0.001; 95% CI: 0.65–0.89). The optimal threshold of 52.5 µg/g yielded 75% sensitivity and 87% specificity. FC levels above 50 µg/g on day 14 achieved 90% specificity, albeit with lower sensitivity at 37%.
Patients who developed GI-GVHD had significantly higher median FC levels compared to those without GVHD on day 14 (41.5 µg/g vs. 25 µg/g; p < 0.001) and day 21 (79 µg/g vs. 31 µg/g; p < 0.001), while day 7 values were comparable (11 µg/g vs. 10 µg/g; p = 0.94).
Correlation with Endoscopic Severity and Treatment Response
At the onset of GI-GVHD, median FC levels reached 120 µg/g. Investigators found a moderate and statistically significant correlation between FC concentrations and endoscopic severity, assessed using the Cruz-Correa grading system (Pearson's r = 0.31; p = 0.02). However, FC levels did not correlate significantly with clinical severity as measured by MAGIC criteria (r = 0.15; p = 0.28), nor with histological severity.
Regarding treatment monitoring, FC levels declined significantly seven days after initiation of therapy, dropping from a median of 120 µg/g to 51.5 µg/g (p = 0.04), suggesting utility in tracking therapeutic response.
Differential Diagnosis Challenges
The study highlighted the limited specificity of FC in the post-transplant setting. The highest FC levels were observed in patients with pseudomembranous colitis (mean = 1,317.5 µg/g), and patients with concurrent viral or bacterial infections exhibited substantially higher FC levels than those without infections (1,085.55 vs. 289.6 µg/g). FC levels did not significantly discriminate between patients with isolated GI-GVHD and those with concurrent GI-GVHD plus additional complications.
The authors emphasized that FC should not be used in isolation and that interpretation requires an integrated diagnostic approach incorporating clinical assessment, complementary biomarkers such as REG3α or ST2, microbiological studies, and, when necessary, endoscopic and histopathological confirmation.
Study Limitations
Despite its prospective design and large cohort, the study had several limitations. Subgroup analyses remained underpowered to detect significant differences when stratified by GI-GVHD grade or by the etiology of non-GVHD diarrhea. The presence of concurrent infections or other inflammatory gastrointestinal conditions may have confounded FC levels. Additionally, the single-center design underscores the need for external validation in independent cohorts. The authors noted that because a stable multivariable model was not feasible without substantial risk of overfitting, analyses were intentionally restricted to unadjusted and discriminative approaches, and results should be considered exploratory rather than causal.
The study was approved by the local Ethics Committee at the University Hospital of Salamanca (搜索), and all participants provided written informed consent.