Barnes, Barrett H.; Borowitz, Stephen M.; Saulsbury, Frank T.; Hellems, Martha; Sutphen, James L.
Department of Pediatrics, Divisions of Gastroenterology and Immunology, University of Virginia, Charlottesville, Virginia, U.S.A.
Received May 14, 2003; revised January 13, 2004; accepted January 16, 2004.
Address correspondence and reprint requests to Dr. Stephen M. Borowitz, Division of Pediatric Gastroenterology and Nutrition, University of Virginia, Box 800386 HSC, Charlottesville, VA 22908, U.S.A. (e-mail: Witz@virginia.edu).
Background: Inflammatory bowel disease (IBD) is characterized by periods of relapse and remission. Treatment is aimed at reducing symptoms during relapse and prolonging the duration of remissions. 6-Mercaptopurine (6-MP) and its prodrug azathioprine (AZA) are commonly used to prolong clinical remissions. The erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP) are two widely used laboratory markers of inflammation. The authors observed an unexplained discordance between ESR and CRP in children with asymptomatic IBD who were being treated with AZA or 6-MP.
Objective: To characterize children with IBD in remission treated with 6-MP or AZA who have persistently elevated ESR but normal CRP.
Methods: All patients seen in Pediatric Gastroenterology Clinic between January 1, 1995, and December 31, 2002, with Crohn disease or ulcerative colitis who received AZA or 6-MP continuously for at least 6 months were identified and their medical records reviewed.
Results: One hundred twenty patients met the eligibility criteria. Twelve had an ESR >18 mm/hour on at least three occasions during at least 12 consecutive months with a simultaneous CRP <0.8 mg/dL. Eleven of these 12 had no signs or symptoms of active disease and had Pediatric Crohn Disease Activity Index scores <10 for at least 12 consecutive months while the ESR was elevated. Disease duration was similar in the 11 children with asymptomatic disease and with discordant ESR and CRP and in 108 children with concordant ESR and CRP (69.2 + 22.5 months v 54.3 ± 40.1 months, P = 0.0709). Duration of AZA or 6-MP therapy was greater in the 11 children with asymptomatic disease and discordant ESR and CRP than in those with or without symptoms and with concordant ESR and CRP (58.1 ± 16.4 months v 36.6 ± 24.1 months, P = 0.0043). There were no differences between the groups with respect to diagnosis, location of disease, or age at onset of symptoms. The mean corpuscular volume (MCV) was somewhat larger in the children with discordant ESR and CRP than in the children with concordant ESR and CRP (91.4 ± 6.97 fL v 87.0 ± 7.07 fL, respectively, P = 0.0373); however, in both groups, the MCV was in the normal range. There were no significant differences in hematocrit, white blood cell count, serum albumin, total serum protein, or estimated serum globulin between the groups.
Conclusions: The results suggest that among children treated with AZA or 6-MP, CRP may be a more reliable indirect indicator of inflammation than ESR. This report alerts clinicians that some children taking AZA or 6-MP may have persistent elevation of the ESR with a normal CRP and have no clinical evidence of active disease.
The clinical course of inflammatory bowel disease (IBD) is variable and is characterized by periods of relapse and remission. Treatment of IBD is aimed at reducing symptoms during relapse and prolonging the duration of remission. 6-mercaptopurine (6-MP) and its prodrug azathioprine (AZA) are commonly used in the care of children with IBD. These drugs diminish the need for corticosteroids, prolong the duration of remission, and rarely produce severe side effects (1–3).
The symptoms of IBD are largely attributable to intestinal inflammation. Assessing the level of inflammatory activity based on patient-reported symptoms and observed signs is imperfect. Thus, laboratory indicators of inflammation often are used to aid in the assessment of disease activity. The erythrocyte sedimentation rate (ESR) and the C-reactive protein (CRP) are the most widely used indicators of inflammation (4,5). The ESR is a component the Pediatric Crohn Disease Activity Index (PCDAI), the only formal scoring system for childhood Crohn disease (6). We have observed a discordance between ESR and CRP in a number of children with asymptomatic IBD who were treated with AZA or 6-MP. This report characterizes these children and attempts to identify factors that might lead to an elevated ESR but a normal CRP in children with asymptomatic disease who are presumed to have inactive disease.
All patients seen in the Pediatric Gastroenterology Clinic at the University of Virginia between January 1, 1995, and December 31, 2002, with a principal diagnosis of Crohn disease or ulcerative colitis were identified by a search of our computerized clinical data repository. Patients treated with AZA or 6-MP continuously for at least 6 months were identified and their medical records reviewed. We ascertained the principal diagnosis, location of disease (ileal, colonic, or ileocolonic), age at diagnosis, duration of disease, duration of 6-MP or AZA therapy, and clinical status (e.g., active disease v clinical remission). Laboratory tests recorded included white blood cell count, hematocrit, mean corpuscular volume (MCV), serum albumin, serum total protein, CRP, and ESR. In our laboratory, the normal ESR is <18 mm/hour and the normal CRP is <0.8 mg/dL. The serum globulin level was estimated as the difference between serum total protein and serum albumin. PCDAI scores were calculated as previously described. Scores ≤10 were defined as indicating inactive disease (6).
We identified 120 patients continuously treated with AZA or 6-MP for at least 6 months. Of these, 12 had a CRP <0.8 mg/dL with an ESR >18 mm/hour measured on at least three separate occasions during a period of at least 12 consecutive months. These patients were identified as having discordant ESR and CRP. Eleven of these children had no signs or symptoms of active disease and had a PCDAI score <10 for at least 12 consecutive months while the ESR was elevated. The remaining child had moderate symptoms and PCDAI scores between 20 and 25, despite her normal CRP. We stratified the patients into those without symptoms and with discordant ESR and CRP (11 patients) and those with or without symptoms and with concordant ESR and CRP (108 patients).
Continuous variables were compared using the Wilcoxon rank sum test. Categorical and dichotomous variables were compared using Fisher exact test. Differences were considered statistically different if P < 0.05. The Human Investigation Committee of the University of Virginia approved this study.
Disease duration was similar in the children with discordant or concordant ESR and CRP (69.0 ± 22.5 months v 54.3 ± 40.1 months, respectively, P = 0.07). The duration of AZA or 6-MP therapy was somewhat greater in the children with asymptomatic disease and discordant ESR and CRP than in those with asymptomatic disease and concordant ESR and CRP (58.1 ± 16.5 months v 36.6 ± 24.1 months, respectively; P = 0.004). There were no differences in diagnosis, location of disease, or age at onset of symptoms between the groups. The mean corpuscular volume (MCV) was somewhat larger in the children with discordant ESR and CRP than in the children with concordant ESR and CRP (91.4 ± 7.0 fL v 87.0 ± 7.1 fL, respectively; P = 0.04); however, in both groups, the MCV was in the normal range. There were no significant differences between the mean hematocrit, white cell count, serum albumin, serum total protein, or estimated serum globulin of the two groups (Table 1).
Many attempts have been made to develop indices that reflect IBD activity and predict the occurrence of relapse. Commonly used activity rating systems for adults with IBD are the Crohn Disease Activity Index (CDAI) and the Harvey Bradshaw Index (7,8). The PCDAI was developed for children and adolescents (6). The CDAI and PCDAI are heavily weighted toward subjective criteria, such as pain and well-being; however, some laboratory test results, such as HCT, ESR and serum albumin, are included. None of these indices were specifically designed to predict the likelihood that a patient in remission would experience an exacerbation. Many laboratory tests reflecting systemic consequences of inflammation, including ESR, CRP, platelet count, white blood cell count, and orosomucoid level, have been proposed as predictors of relapse in IBD (9–14). Clinical disease activity indices with or without laboratory tests also have been proposed as a means for identifying patients at risk for impending relapse (9,13). The predictive value of these measures for identifying patients at risk for relapse generally has been disappointing. As a result, most clinicians use a combination of clinical assessment and laboratory values to assess disease activity and potential for relapse.
The ESR is widely used as an indicator of inflammation (4). The other most commonly used laboratory test to indirectly assess the extent of inflammation is CRP (4). CRP levels have been reported to correlate well with clinical and pathologic indices of relapse, remission, and response to therapy (10,11).
The ESR is an indirect reflection of plasma acute-phase protein concentrations, which are influenced by the size, shape, and number of erythrocytes and by other plasma constituents, such as serum immunoglobulins (15,16). The major determinant of ESR is the plasma fibrinogen concentration. As fibrinogen concentration increases in response to acute inflammation, there is greater cohesion of erythrocytes, which causes agglutination, rouleaux formation, and a faster rate of sedimentation (16). ESR is affected by changes in red cell mass, morphology, and aggregability, by changes in plasma proteins other than fibrinogen, by drugs such as salicylates, and by smoking (16,17).
We have identified a group of children with IBD in remission taking AZA or 6-MP with persistently elevated ESR but normal CRP levels. Discordance between ESR and CRP has been observed in from 28% to 55% of adult patients with rheumatoid arthritis (18,19). In most of these cases, the ESR is elevated in the face of a normal CRP (18). In many of these patients the discordance can be partly attributed to elevations in serum immunoglobulin G and A or a low hemoglobin level (18). In none of these reports have authors examined the potential impact of medications on ESR.
A possible explanation for the persistent elevation of ESR in the face of normal CRP seen in this study may be changes in red cell volume. Chronic use of AZA and 6-MP is associated with macrocytosis, presumably secondary to alterations in folate metabolism (20). The ESR may be increased because of the anemia commonly observed in children with IBD (16). We did not find any difference in the mean hematocrit of the group with chronically elevated ESR compared with that of the group with normal ESR. We did note a higher mean MCV in the children with elevated ESR than in those with concordant ESR and CRP. However, this difference was small, and in both groups the MCV was within the normal range. Hyperglobulinemia and hyperfibrinogenemia have been associated with increased ESR (15–17,21). Although we did not routinely measure serum fibrinogen or serum immunoglobulin G in our patients, serum total protein, serum albumin, and estimated serum globulin were not significantly different in the group of children with persistently elevated ESR and the group with normal ESR.
We have observed persistent discordance between ESR and CRP only in patients taking AZA or 6-MP. It is conceivable that differences in the metabolism of AZA or 6-MP could affect ESR in certain individuals. We do not routinely measure these metabolites, so it is not possible to test this hypothesis with our current data; however, the lack of any association between serum albumin, transaminases, MCV, white blood cell count, or total neutrophil count and persistently elevated ESR does not favor the theory that accumulation of metabolites is the source of this disparity.
An important goal in the management of IBD is finding a serum marker that will accurately predict relapse. If such a marker could be found, earlier institution of therapy might prevent or attenuate relapse. In adults, serum CRP, orosomucoid, and alpha-1-antitrypsin levels have been shown to increase several months before an acute relapse (9–11). Although many pediatric gastroenterologists use ESR as an indirect measure of intestinal inflammation in IBD, our observations suggest that CRP may be a more reliable indicator of inflammation in patients taking AZA or 6-MP. We do not know the mechanism, but the purpose of this report is to alert clinicians that some children taking AZA or 6-MP may have persistently elevated ESR in the face of a normal CRP and have no clinical evidence of active disease.
1. Markowitz J, Grancher K, Kohn N, et al. A multicenter trial of 6-mercaptopurine and prednisone in children with newly diagnosed Crohn's disease. Gastroenterology 2000;119:895–902.
2. Barabino A, Torrente F, Ventura A, et al. Azathioprine in paediatric inflammatory bowel disease: an Italian multicentre survey. Aliment Pharmacol Ther 2002;16:1125–30.
3. Kirschner BS. Safety of azathioprine and 6-mercaptopurine in pediatric patients with inflammatory bowel disease. Gastroenterology 1998;115:813–21.
4. Hyams JS, Mandel F, Ferry GD, et al. Relationship of common laboratory parameters to the activity of Crohn's disease in children. J Pediatr Gastroenterol Nutr 1992;14:216–22.
5. Jaye DL, Waites KB. Clinical applications of c-reactive protein in pediatrics. Pediatr Infect Dis J 1997;16:735–47.
6. Hyams JS, Ferry BD, Mandel FS, et al. Development and validation of a pediatric Crohn's disease activity index. J Pediatr Gastroenterol Nutr 1991;12:439–47.
7. Best WR, Beckel JM, Singleton JW, et al. Development of a Crohn's disease activity index. National Cooperative Crohn's Disease Study. Gastroenterology 1976;70:439–44.
8. Harvey RF, Bradshaw JM. A simple index of Crohn's disease activity. Lancet 1980;2:514.
9. Wright JP, Yojng GO, Tigler-Wybrandi N. Predictors of acute relapse of Crohn's disease–a laboratory and clinical study. Dig Dis Sci 1987;32:164–70.
10. Andre C, Descos L, Vignal J, et al. C-reactive protein as a predictor of relapse in asymptomatic patients with Crohn's disease. Scott Med J 1983;28:26–9.
11. Boirivant V, Leoni M, Taricotti D, et al. The clinical significance of serum c-reactive protein levels in Crohn's disease. J Clin Gastroenterol 1988;10:401–5.
12. Kjeldsen J, Lauritsen K, De Muckadell OB. Serum concentrations of orosomucoid: improved decision-making for tapering prednisolone therapy in patients with inflammatory bowel disease. Scand J Gastroenterol 1997;32:933–41.
13. Brignola C, Iannone P, Belloli C, et al. Predication of relapse in patients with Crohn's disease in remission: a simplified index using laboratory tests, enhanced by clinical characteristics. Eur J Gastroenterol Hepatol 1994;6:955–61.
14. Brignola C, Campieri M, Bazzocchi G, et al. A laboratory index for predicting relapse in asymptomatic patients with Crohn's disease. Gastroenterol 1986;91:1490–4.
15. Gabay C, Kushner I. Acute-phase proteins and other systemic responses to inflammation. N Engl J Med 1999;340:448–54.
16. Zlonis M. The mystique of the erythrocyte sedimentation rate. Clin Lab Med 1993;13:787–800.
17. Hilliard NJ, Waites KB. C-reactive protein and ESR: what can one test tell you that the other can't?Contemp Pediatr 2002;19:64–74.
18. Wolfe F. Comparative usefulness of c-reactive protein and erythrocyte sedimentation rate in patients with rheumatoid arthritis. J Rheumatol 1997;24:1477–85.
19. Amos RS, Constable TJ, Crockson RA, et al. Rheumatoid arthritis: relation of serum C-reactive protein and erythrocyte sedimentation rates to radiographic changes. BMJ 1977;1:195–7.
20. Bernstein CN, Artinian L, Anton PA, et al. Low-dose 6-mercaptopurine in inflammatory bowel disease is associated with minimal hematologic toxicity. Dig Dis Sci 1994;39:1638–41.
21. Reinhart WH, Nagy C. Albumin affects erythrocyte aggregation and sedimentation. Eur J Clin Invest 1995;25:523–8.
© 2004 Lippincott Williams & Wilkins, Inc.