Endoscopic Yield in Anemia of Chronic Disease: A Single-Center Retrospective Study of U.S. Veteran Population

Goals: The aim of the study was to evaluate endoscopic yield in patients with Anemia of Chronic Disease (ACD) who had no Gastrointestinal (GI) symptoms. We analyzed the impact of laboratory and clinical variables on endoscopic yield in this cohort.


Introduction
Anemia is broadly defined as abnormally low Hemoglobin (Hb), with an Hb of ≤ 12 mg/dl in women and ≤ 13 mg/dl in men [1][2][3]. The two most common causes of chronic anemia in the world are Iron Deficiency Anemia (IDA) and Anemia of Chronic Disease (ACD) [2][3][4][5]. In some cases, concomitant IDA and ACD may be present and better characterized as a Mixed Anemia (MA).
Categorizing chronic anemia into IDA, ACD and MA can be challenging as the definitions have varied among studies, and precise cutoffs are laboratory dependent [2,6,7]. Many patients may have a combination of IDA and ACD, falling into the category of MA [5,8,9]. A serum ferritin of <30 ng/ml is suggestive of iron deficiency anemia [6]. Conversely, a serum ferritin of >100 ng/ ml makes a diagnosis of iron deficiency unlikely and suggestive of ACD [8,10,11].
Iron deficiency anemia is an established indication for endoscopy to rule out a Gastrointestinal (GI) source of occult blood loss. Endoscopy methods, such as Esophagogastroduodenoscopy (EGD), Men with Hb <13 g/dl and women with Hb <12 g/dl were included in the study and classified into IDA, MA and ACD based on serum ferritin ≤ 30 ng/ml, 30-100 ng/ml, and >100 ng/ml, respectively. To account for scheduling delays, the value for ferritin utilized was the lowest recorded within 3 months preceding the date of endoscopy. Laboratory data for Hb, Fe Sat%, serum ferritin, and Mean Corpuscular Volume (MCV) were within six months prior to endoscopy. Wanting to avoid clinical situations where endoscopy was imperative regardless of the patient's anemia status, our exclusion criteria included: Further patient data collected included were age at time of endoscopy, gender, race, use of antiplatelet and/or anticoagulant agents within 6 months of endoscopy, and use of Non-Steroidal Anti-Inflammatory Drugs (NSAIDs).
Endoscopic data collected were findings reported on esophagogastroduodenoscopy and colonoscopy. Reflecting routine clinical practice, some patients underwent both EGD and colonoscopy and some only one of the procedures depending on clinical assessment by the treating physician, patient's symptoms, and patient's wishes. Findings were considered significant on EGD and colonoscopy if they were thought to be causing occult GI blood loss. These were determined based on known common acceptable causes of anemia. For EGD, significant findings were Arteriovenous Malformation (AVM), erosive esophagitis of any grade, hemorrhagic gastritis, gastric and/or duodenal erosions, Cameron erosions, malignancy, peptic ulcer disease, celiac disease, Gastric Antral Vascular Ectasia (GAVE), portal gastropathy, polyps >2 cm, and submucosal mass. Non-significant findings included hiatal hernia without Cameron erosions, Barrett's esophagus, esophageal varices, incidental duodenal polyps <2 cm, and fundic polyps. Significant findings for colonoscopy were polyp >2 cm, AVM, colitis of any etiology including Inflammatory Bowel Disease (IBD), NSAIDs, microscopic, ischemic, radiation or infectious colitis, ileal ulcers, malignancy, and colorectal ulcers. Non-significant findings included internal or external hemorrhoids, polyps <2 cm, and diverticulosis. Biopsy results from the endoscopic procedures were reviewed to corroborate endoscopic findings and diagnosis when necessary.
The primary outcome was endoscopic yield, which was calculated by dividing the number of patients with one or more significant findings by the total number of patients who underwent EGD or colonoscopy, reported separately. We did not include findings of video capsule endoscopy as less than 4% of our patients underwent VCE for ACD. Additionally, we evaluated variables that may impact endoscopic yield in patients with ACD, i.e., age, race, laboratory parameters, and use of NSAIDs and anti-platelet and anticoagulant agents.
Statistical analysis was conducted using IBM SPSS Statistics 25.0 (IBM, Armonk, NY). Means and standard deviations are reported for continuous variables and counts and percentages for categorical variables. The chi square test was used to compare two variables measured on a categorical scale. Inferences were made at the 0.05 level of significance with no corrections for multiple comparisons.

Results
A total of 1162 patients underwent endoscopy between January 2010 and December 2018 for indication of anemia. Of these, 366 patients were excluded due to inadequate information on iron studies. For the remaining 796 patients, 426 (53.5%) underwent endoscopy for IDA, 164 (14.1%) for MA and 206 (32.4%) for ACD. Of the 206 patients with ACD, 82 met the inclusion criteria. Table 1 showed the baseline characteristics of the cohort. The mean age was 66 ± 9 and 96.3% men. Based on ethnicity, Caucasians accounted for 67.1% and African Americans accounted for 32.9%. Anti-platelet and/or anticoagulant medications were used by 30.5%, and NSAIDs were used by 9.8% of patients. The mean hemoglobin concentration was 9.57 ± 2.2 gm/dl, ferritin was 393.13 ± 508.21 ng/ ml, Fe Sat% was 19.89 ± 11.82%, and MCV was 87.21 ± 7.51 fL.

Total Patients=82
Mean Age  Table 4 reported the endoscopic yield for EGD and colonoscopy based on laboratory and clinical variables. Endoscopic yield for EGD was 50% in both Hb ≤ 8 gm/dl and >8 gm/dl (p=1.00) and 40% and 57% for MCV ≤85 and >85, respectively (p=0.16). Ferritin 101-499 ng/ml and ≥ 500 ng/ml had endoscopic yield yields of 49% and 46%, respectively (p=0.80). Endoscopic yield for EGD was 50% in both those on anticoagulants and/or antiplatelet agents within 3 months of endoscopy and those who were not (p=1.00), and endoscopic yield was 75% and 47% for those who were on NSAIDs and those who were not (p=0.26). Thus, none of the clinical or lab variables had an impact on endoscopic yield for EGD.

Discussion and Conclusions
Iron deficiency anemia is a well-established indication for endoscopic investigation. However, there are no guidelines regarding the role of endoscopy in evaluation of anemia of chronic disease. To our knowledge, this study is the first to evaluate yield for EGD and colonoscopy in patients who have ACD without GI symptoms and examine variables that may increase endoscopic yield in these patients. Prior studies reported endoscopic yields ranging from 8% to 41% for EGD and 6% to 37% for colonoscopy [12][13][14][15][16][17]20]. Our study found endoscopic yield to be above this range for EGD (50%) and at the low end of the range for colonoscopy (11%). The variations in defining significant findings for EGD and colonoscopy may explain the discrepancies among our study and others. Another possibility for discrepancies between our study and past studies may due to the patient populations. The VA population is heavily Caucasian males which may limit our studies application to the general public. Based upon laboratory analysis and the demographics of our study population, there was no significant statistical relationship of endoscopic findings and degree of anemia, ferritin level, MCV, and use of antiplatelet and/or anticoagulants or NSAIDs. Of note, the endoscopic yield of colonoscopies were lower compared to prior studies which may be due to the lower than expected number of colonoscopies performed for ACD.
Endoscopic yield for colonoscopy was 6% and 13% for Hb ≤ 8 gm/ dl and >8 gm/dl, respectively (p=0.75), and 15% and 8% for MCV ≤ 85 and >85, respectively (p=0.66). Endoscopic yield for colonoscopy was 12% and 8% for ferritin 101-499 ng/ml and ≥ 500 ng/ml, respectively (p=1.00). Endoscopic yield for colonoscopy was 25% for those using anticoagulants and/or antiplatelet agents and 6% for those not using (p=0.11). Endoscopic yield was 0% and 11% in those who were using NSAIDs and those who were not (p=0.83). Consequently, as with EGD, none of the clinical or laboratory variables had an impact on endoscopic yield for colonoscopy.
Our study has several limitations. First, the study was conducted at a single medical facility.
Our largely homogeneous sample was predominantly older white male veterans.
Consequently, generalizability to other populations (e.g., younger ages and other races, females, and broader community settings) should be done with caution. Second, in retrospective chart reviews, data omissions are more likely and the accuracy of data collection may be more challenging. For example, approximately one-third of patients diagnosed with anemia did not have the serum ferritin lab necessary to determine if a patient had anemia of chronic disease. Third, because our facility followed stringent criteria to perform endoscopy on patients with anemia, our sample size was restricted. Further studies will need to be conducted to assess if underlying comorbidities such as age, obesity, tobacco use, and cause of anemia of chronic disease have any statistical significance in regards to endoscopic yield. Of note, there is limited data on endoscopic yield of VCE for ACD. Consideration should be taken to evaluate if there is a role of VCE for adequate evaluate of ACD.
Our study suggests that EGD is beneficial for evaluation of all patients with ACD without GI symptoms, regardless of underlying    lab or clinical findings. The current standard of practice recommends complete endoscopic evaluation for patients with iron deficiency anemia only. However, there are no such recommendations for formal endoscopic investigation for patients that fall into the anemia of chronic disease. Given the endoscopic yield for EGDs in our study group that fit the criteria of anemia of chronic disease, there should be consideration if these patients would benefit from at least an EGD to assess the anemia. Often the endoscopic findings can be treated either endoscopically or which medication which may improve a patient's anemia. On the other, due to limited yield of colonoscopy in patients with ACD, colonoscopy should be performed on a caseby-case basis as the number of colonoscopies was low. Further largescale studies would be needed to investigate if patients with anemia of chronic disease could benefit from full endoscopic work up similar to patients with iron deficiency anemia. Large-scale prospective studies are needed to draw more definitive conclusions about the use of endoscopy in patients with anemia of chronic disease in the absence of GI symptoms and classic iron deficiency.