- Email: [email protected]
Prevalence of Hyperhomocysteinemia in Adult Gluten-Sensitive Enteropathy at Diagnosis: Role of B12, Folate, and Genetics
CLINICAL GASTROENTEROLOGY AND HEPATOLOGY 2005;3:574 –580
Prevalence of Hyperhomocysteinemia in Adult Gluten-Sensitive Enteropathy at Diagnosis: Role of B12, Folate, and Genetics SIMONE SAIBENI,* ANNA LECCHI,‡ GIANMICHELE MEUCCI,§ MARCO CATTANEO,储 LILIANA TAGLIABUE,‡,¶ EMANUELE RONDONOTTI,* SARA FORMENTI,* ROBERTO DE FRANCHIS,* and MAURIZIO VECCHI* *Gastroenterology and Gastrointestinal Endoscopy Service, ‡A. Bianchi Bonomi Thrombosis and Hemophilia Center, and ¶Fondazione Luigi Villa, IRCCS Maggiore Hospital and University of Milan, Milan, Italy; 储Department of Medicine, Surgery and Dentistry, San Paolo Hospital and University of Milan, Milan, Italy; and §Gastroenterology Division, Valduce Hospital, Como, Italy
Background & Aims: Hyperhomocysteinemia, a risk factor for thrombosis, recurrent miscarriages, and osteoporosis, might derive from acquired folate and vitamin B12 deficiencies and from a C677T mutation in methylenetetrahydrofolate reductase (MTHFR) gene. Undiagnosed gluten-sensitive enteropathy (GSE) is associated with vitamin deficiencies, osteoporosis, and recurrent miscarriages. We evaluated the prevalence and the risk factors for hyperhomocysteinemia in patients with newly diagnosed GSE. Methods: In this prospective study performed in a tertiary care setting, 40 consecutive subjects with newly diagnosed GSE were evaluated for homocysteine, folate, and vitamin B12 levels and for C677T polymorphism. One hundred twenty sex- and agematched healthy control subjects were studied. Nonparametric tests and multiple regression analysis were used to evaluate the risk factors in inducing hyperhomocysteinemia in the GSE population. Results: Hyperhomocysteinemia was more frequent in GSE patients than in control subjects (8/40, 20.0% vs 7/120, 5.8%) (relative risk, 3.4; 95% confidence interval, 1.3– 8.9), as well as folate deficiency (17/40, 42.5% vs 10/120, 8.3%) (relative risk, 5.1; 95% confidence interval, 2.5–10.2). Multiple regression analysis showed that folate and B12 levels were independently and inversely associated with homocysteine levels, whereas homozygosity for the MTHFR thermolabile variant was not. The prevalence of MTHFR variant in GSE population was not different from that reported in racially comparable control groups. Gluten-free diet was able to normalize folate, vitamin B12, and homocysteine levels. Conclusions: Hyperhomocysteinemia is frequent in newly diagnosed GSE. Vitamin deficiencies caused by malabsorption are the most important determinants of this condition. Hyperhomocysteinemia might contribute to the occurrence of common complications of undiagnosed GSE.
he possible role of hyperhomocysteinemia as a risk factor for atherothrombosis was first suggested by McCully,1 who reported autopsy evidence of extensive arterial
T
thrombosis in 2 children with homocystinuria. Subsequent investigations have confirmed this hypothesis, and it is now clear that hyperhomocysteinemia is an independent risk factor for both arterial and venous thrombosis.2,3 More recently, a possible role of hyperhomocysteinemia as an etiologic factor of recurrent miscarriages and congenital heart and neural tube defects has also been reported.4 –9 Hyperhomocysteinemia has also been suggested to play a role in osteoporosis and osteoporotic fractures.10,11 Increased levels of plasma homocysteine are typically caused by genetically determined defects in the enzymes involved in homocysteine metabolism, ie, cystathionine -synthase and methylene-tetrahydrofolate reductase (MTHFR), or by acquired deficiencies in vitamin cofactors, such as folate, vitamin B12, and vitamin B6.2 Homozygous traits for cystathionine -synthase and MTHFR deficiency are very rare and lead to severe hyperhomocysteinemia, recurrent thromboembolism, osteoporosis, and a poor prognosis.12,13 Recently, a frequent thermolabile variant of MTHFR caused by a point mutation (C677T)14 has been suggested to cause an exaggerated hyperhomocysteinemic response, particularly in TT homozygotic patients with concurrent vitamin deficiencies.15 Gluten-sensitive enteropathy (GSE) is characterized by a wide spectrum of clinical presentations.16 At diagnosis, both celiac disease and dermatitis herpetiformis (DH), which are part of this spectrum, might be associated with vitamin deficiencies,17–19 osteoporosis,20 and, in the female population, a history of recurrent miscarriages sometimes attributed to vitamin deficiencies,21–24 even if this association has been debated.25 Abbreviations used in this paper: AGA, anti-gliadin antibodies; CI, confidence interval; DH, dermatitis herpetiformis; EMA, anti-endomysial antibodies; GFD, gluten-free diet; GSE, gluten-sensitive enteropathy; MTHFR, methylene-tetrahydrofolate reductase; RR, relative risk. © 2005 by the American Gastroenterological Association 1542-3565/05/$30.00 PII: 10.1053/S1542-3565(05)00022-4
June 2005
HYPERHOMOCYSTEINEMIA IN GLUTEN-SENSITIVE ENTEROPATHY
575
Table 1. Demographic and Clinical Features of GSE Patients and Healthy Control Subjects
Men/women Age at diagnosis (mean ⫾ SD, y) Duration of symptoms (mean ⫾ SD, mo) Clinical presentationa Diarrhea Abdominal pain Anemia Weight loss Bloating Hypertransaminasemia Osteoporosis Dyspepsia Asymptomatic Aphtous stomatitis Recurrent miscarriages Epilepsia Skin lesions
Celiac disease (n ⫽ 32)
DH (n ⫽ 8)
Control subjects (n ⫽ 120)
7/25 39.7 ⫾ 12.5 44.3 ⫾ 83.5
4/4 37.4 ⫾ 20.7 37.5 ⫾ 27.0
33/87 38.9 ⫾ 10.9 —
13 7 7 6 5 5 4 4 2 1 1 1 —
— — — — — — — — — — — — 8
— — — — — — — — — — — — —
SD, standard deviation. aMore than 1 symptom might be present for each patient.
Several cases of thrombosis occurring in patients with celiac disease before diagnosis have also been reported, most of them suggesting a possible role of hyperhomocysteinemia.26 –35 We thus evaluated the prevalence of hyperhomocysteinemia in GSE patients at the time of diagnosis and investigated the role and the possible relationships of folate and vitamin B12 deficiencies and of the MTHFR thermolabile variant in inducing the risk of hyperhomocysteinemia.
Materials and Methods In this prospective study all adult patients with a new diagnosis of GSE seen at our Gastroenterology and Gastrointestinal Endoscopy Service in the period between January 2002 and December 2003 were consecutively enrolled.
Patients The study involved 40 patients with GSE; of them, 32 had celiac disease, and 8 had DH; as it is commonly observed, patients with celiac disease were more frequently women. At the time of enrollment, no patient was assuming drug treatments or was affected by other diseases capable of raising homocysteine plasma levels. The main demographic and clinical features of GSE patients are reported in Table 1. All GSE patients were enrolled at the time of diagnosis, which was obtained according to standardized serologic and histologic criteria (positivity of anti-endomysial IgA antibodies [EMA] and/or anti-gliadin antibodies [AGA] and histologic duodenal findings of villous atrophy and crypt hyperplasia).36 The degree of duodenal involvement was assessed according to a 4-grade scale (grade 0, normal; grade 1, increased IELs; grade 2, increased IELs and crypt hyperplasia; grade 3, increased IELs,
crypt hyperplasia, and [a] partial, [b] subtotal, [c] total villous atrophy) according to Marsh criteria revised by Rostami et al.37 The diagnosis of DH was made by showing typical IgA granular deposits in the subepidermal papillae by means of immunofluorescence on perilesional skin.38 All GSE patients were on a gluten-containing diet at the time of enrollment. The 8 GSE patients with hyperhomocysteinemia at diagnosis were re-evaluated for vitamin status and plasma homocysteine levels after a mean follow-up of 6.2 months on a strict gluten-free diet (GFD) without any vitamin supplementation.
Control Subjects One hundred twenty age- and sex-matched healthy subjects were also included in the study and served as control subjects. All these subjects were chosen from a larger cohort of healthy individuals (⬃900) with a negative personal and family history of thrombosis, celiac disease, or inflammatory bowel disease who had been previously enrolled as normal control subjects for the study of other hemostatic parameters. All these subjects were matched with GSE patients also for other determinants of homocysteine status such as liver and renal function, diabetes, and coffee consumption.
Blood Collection Blood samples were obtained, on informed consent, from all patients and control subjects. Serum and plasma samples were used for the determination of folate and vitamin B12 levels and of homocysteine levels, respectively. Whole blood was collected in ethylenediaminetetraacetic acid and rapidly frozen at ⫺80°C until the time of DNA extraction, which was obtained by means of a standard salting-out procedure.39
576
SAIBENI ET AL
CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 3, No. 6
Table 2. Histologic Grading and Serologic Status in GSE Patients
Histologic grading 3a 3b 3c Serologic status EMA ⫹, AGA ⫹ EMA ⫹, AGA ⫺ EMA ⫺, AGA ⫹
Celiac disease (n ⫽ 32)
DH (n ⫽ 8)
2 4 26
2 2 4
24 7 1
3 4 1
Inc, San Diego, CA) and then analyzed by means of nonparametric tests. In particular, comparisons between groups were performed by means of Fisher exact test, Mann–Whitney test, Wilcoxon matched pairs test, and Kruskal–Wallis test with Dunn post hoc comparison, as appropriate. Correlations between variables were assessed by means of Spearman nonparametric correlation. Relative risks (RRs) and their 95% confidence intervals (CIs) were also calculated according to standard formulas.42 The association between plasma homocysteine levels and serum vitamin levels, MTHFR allele status, and clinical and demographic features of GSE patients was assessed by means of multiple regression analysis. The significance level was set at P ⬍ .05.
Folate and Vitamin B12 Assay Serum folate and vitamin B12 determinations were performed by means of a commercially available RIA kit (DPC, Los Angeles, CA). Folate and vitamin B12 deficiency were set, according to the manufacturer’s instructions, at ⬍3 ng/mL and ⬍200 pg/mL, respectively.
DNA Analysis DNA analysis for MTHFR gene mutation (C677T) was performed according to a previously described method40; briefly, a fragment of 198 base pairs was amplified by polymerase chain reaction. The fragment was then digested by Hinf I restriction enzyme, and subsequent electrophoresis on ethidium bromide stained 2% agarose gel ⫹ 1% NUSIEVE agarose (Cambrex Corporation, East Rutherford, NJ) was performed.
Plasma Homocysteine Assay The concentration of total homocysteine in plasma was determined by high-performance liquid chromatography and fluorimetric detection.41 Hyperhomocysteinemia was diagnosed when plasma homocysteine levels exceeded the 95th percentile of distribution observed among matched healthy control subjects (13.9 mol/L for women, 18.7 mol/L for men).
Statistical Evaluation For statistical analysis, data were entered into a statistical software package (GraphPad Instat; GraphPad Software
Results All GSE patients showed grade 3 histologic lesions, and all of them had high-titer positive EMA or positivity of both AGA IgA and IgG (Table 2). In patients with GSE, median plasma homocysteine level was significantly higher than in matched control subjects (11.1 mol/L, range 5.3–33.2 mol/L vs 8.9 mol/L, range 4.9 – 42.3 mol/L) (P ⬍ .02). The prevalence of hyperhomocysteinemia, as defined above, was significantly higher in GSE patients (8 of 40, 20.0%) than in matched control subjects (7 of 120, 5.8%) (RR, 3.4; 95% CI, 1.3– 8.9; P ⬍ .02) (Table 3). Median serum folate levels were significantly lower in GSE patients than in healthy control subjects (3.4 ng/ mL, range 0.7–13.5 ng/mL vs 6.0 ng/mL, range 2.0 – 24.0 ng/mL; P ⬍ .0001), whereas median vitamin B12 levels were similar (363.5 pg/mL, range 96.0 – 832.0 pg/mL in GSE patients vs 389.5 pg/mL, range 89.0 – 1096.0 pg/mL in healthy control subjects; P ⫽ .26). Seventeen GSE patients (42.5%) had folate deficiency, and 7 (17.5%) had vitamin B12 deficiency. Folate deficiency was significantly more frequent than in healthy control subjects (10/120, 8.3%) (RR, 5.1; 95% CI, 2.5–
Table 3. Prevalence of Hyperhomocysteinemia, Folate, and Vitamin B12 Deficiency and MTHFR Allele Status in GSE Patients and in Healthy Control Subjects Healthy control subjects (n ⫽ 120)
GSE (n ⫽ 40)
Hyperhomocysteinemia Folate deficiency Vitamin B12 deficiency MTHFR genotype at 677 C/C C/T T/T T allele frequency na ⫽ not assessed.
n
%
n
%
RR (95% CI)
8 17 7
20.0 42.5 17.5
7 10 11
5.8 8.3 9.2
3.4 (1.3–8.9) 5.1 (2.5–10.2) 1.9 (0.8–4.6)
13 21 6
32.5 52.5 15.0 41.2
na na na na
— — — —
June 2005
HYPERHOMOCYSTEINEMIA IN GLUTEN-SENSITIVE ENTEROPATHY
577
Table 4. Homocysteine, Folate, and Vitamin B12 Levels in GSE Patients According to Histologic Grading 3a Folate (ng/mL) Vitamin B12 (pg/mL) Homocysteine (mol/L)
3b
7.2 (6.6–13.5) 401.0 (96.0–832.0) 7.7 (5.3–8.2)
4.6 (1.2–12.5) 468.0 (337.0–580.0) 7.8 (6.8–19.6)
3c 2.1 (0.7–10.4) 337.5 (126.0–708.0) 12.4 (5.0–33.2)
P value ⬍.01a .20 ⬍.02a
NOTE. Results are expressed as median and range. a3a vs 3c.
10.2; P ⬍ .0001), whereas vitamin B12 deficiency was not (11/120, 9.2%) (RR, 1.9; 95% CI, 0.8 – 4.6; P ⫽ .16) (Table 3). Median serum folate levels were significantly lower in the 8 GSE patients with hyperhomocysteinemia than in the 32 patients with normal homocysteine levels (1.8 ng/mL, range 0.7– 4.7 ng/mL vs 3.7 ng/mL, range 0.7– 13.5 ng/mL; P ⬍ .03), whereas the difference did not reach the statistical significance for vitamin B12 levels (265.0 pg/mL, range 158.0 –580.0 vs 386.5 pg/mL, range 96.0 – 832.0 pg/mL; P ⫽ .07). Folate deficiency was significantly more frequent in patients with hyperhomocysteinemia (6 of 8, 75.0%) than in normohomocysteinemic patients (11 of 32, 34.4%) (RR, 2.2; 95% CI, 1.2– 4.1; P ⫽ .05), and vitamin B12 deficiency was more frequent in hyperhomocysteinemic than in normohomocysteinemic GSE patients, although this difference did not reach statistical significance (3/8, 37.5% vs 4/32, 12.5%) (RR, 3.0; 95% CI, 0.8 –10.8; P ⫽ .13). When we stratified GSE patients according to their clinical and demographic features, no correlation was observed between age and homocysteine, folate, and vitamin B12 levels or between duration of symptoms before diagnosis and homocysteine, folate, and vitamin B12 levels (data not shown). In particular, no difference was observed regarding median folate, vitamin B12, and homocysteine levels and prevalence of folate and vitamin B12 deficiencies and hyperhomocysteinemia when patients with celiac disease were compared to patients with DH (data not shown). Also, we did not observe any difference in homocysteine, folate, and vitamin B12 levels when the 32 patients with celiac disease were stratified according to the pres-
ence of a typical (with gastrointestinal symptoms) or atypical (no gastrointestinal symptoms) presentation. However, GSE patients with a more severe duodenal involvement had lower folate and higher homocysteine levels than patients with minor duodenal damage (Table 4). The MTHFR thermolabile variant mutation was observed in 21 patients with heterozygosity (52.5%) and 6 with homozygosity (15.0%), resulting in an allele frequency of 41.2% (Table 3). After stratifying GSE patients according to their allele status, no significant difference was observed as far as folate, vitamin B12, and homocysteine levels were concerned, even if a tendency to have lower folate and higher homocysteine levels was observed in mutated homozygous (T/T) patients (data not shown). Three of the 6 patients homozygous for the MTHFR variant had high homocysteine levels, compared with 5 of the remaining nonhomozygous 34 (50.0% vs 14.7%) (RR, 3.4; 95% CI, 1.1–10.6; P ⫽ .08). Conversely, the prevalence of such homozygosity was 3/8 (37.5%) among hyperhomocysteinemic patients and 3/32 (9.4%) among normohomocysteinemic patients (RR, 4.0; 95% CI, 1.0 – 16.2; P ⫽ .08). The multiple regression analysis showed that folate and B12 levels were independently and inversely associated with homocysteine levels, whereas homozygosity for the MTHFR was very close to statistical significance. On the contrary, gender, age, and duration of symptoms were not correlated with plasma homocysteine levels. Data on the variables that were significant or close to statistical significance are reported in Table 5. During GFD, a significant increase in median folate and vitamin B12 levels and a significant decrease in
Table 5. Association of Homocysteine Levels With Folate and Vitamin B12 Levels and MTHFR Homozygosity for Thermolabile Variant in GSE Patients (Multiple Regression Analysis)
Folate levels MTHFR homozygosity for the thermolabile variant Vitamin B12 levels
Coefficient
95% CI
P value
⫺0.705 ⫹2.568 ⫺0.016
From ⫺1.218 to ⫺0.192 From ⫺0.061 to 5.198 From ⫺0.025 to ⫺0.006
.008 .055 .002
578
SAIBENI ET AL
plasma homocysteine levels were observed (Figure 1). All these patients normalized their vitamin status, and 7 of 8 patients reached normal homocysteine levels. The only patient who did not normalize completely her homocysteine level was a woman homozygous for the MTHFR mutation.
Discussion Despite the presence in the literature of isolated case reports of hyperhomocysteinemia occurring in patients with celiac disease, no data were so far available on the true prevalence of this association and on its possible underlying mechanisms. In the present study, which is the first prospectively performed in a consecutive series of subjects with a new diagnosis of GSE, we observed that these subjects have both high median plasma homocysteine levels and an increased prevalence of hyperhomocysteinemia. Indeed, 1 of 5 GSE patients in this series had hyperhomocysteinemia, as defined by plasma homocysteine levels higher than the 95th percentile of values of the control population. Gender, age at diagnosis, duration of symptoms, and type of clinical presentation were not predictive of the presence of hyperhomocysteinemia; however, high levels of homocysteine appear to be strictly related to the degree of vitamin deficiencies, resulting from more severe intestinal damage. The carriage of the MTHFR thermolabile variant gene mutation, leading to reduced activity of this key enzyme in homocysteine metabolism, might also be implicated in some patients. However, because the prevalence of homozygosity for the C677T mutation in our GSE population (15.0%) and the frequency of the mutated allele (41.2%) were not different from previously reported background populations from our group (22.4% and 47.4%, respectively),43 it appears that there is not a relevant role for the genetic variant. Moreover, the ability of GFD to normalize folate, vitamin B12, and homocysteine levels in GSE patients points to the major role played by malabsorptive causes than by the genetic defect. The major role of vitamin deficiencies in inducing hyperhomocysteinemia in GSE patients is not surprising, because folate and vitamin B12 deficiencies are wellknown risk factors for hyperhomocysteinemia2 and are also frequently reported in GSE patients at diagnosis.17–19 The finding of frequent vitamin B12 deficiency (in our series twice that of control subjects) suggests that distal small bowel might also be involved in GSE,19 although a relationship with autoimmune gastritis cannot be excluded in the present series.
CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 3, No. 6
Whatever the mechanism(s) of its induction, the observation that hyperhomocysteinemia is so frequent in GSE appears to provide an important link between this disease and some of its more frequent complications. Indeed, women with a subsequent diagnosis of GSE appear to be characterized by the frequent occurrence of early fetal loss and newborn malformations,24 but recently this association has been debated.25 Although this feature has been correlated to the presence of other factors possibly occurring in GSE and mainly related to malabsorption, increasing evidence supports a major role of hyperhomocysteinemia in these complications of pregnancy in the general population. In fact, recurrent pregnancy loss and congenital heart and neural tube defects
Figure 1. Changes in folate (A), vitamin B12 (B), and homocysteine levels (C) in the 8 hyperhomocysteinemic GSE patients followed up during GFD.
June 2005
are considered a consequence of hyperhomocysteinemia.4 –9 Thus, the present observation might provide an important clue to the understanding of the pathogenetic mechanisms of pregnancy complications in GSE women before diagnosis. The finding that homocysteine levels return to normal in the majority of patients in our series shortly after the beginning of GFD is also in agreement with the observation that the risk for complicated pregnancy reverts to normal in GSE women after diagnosis. High levels of plasma homocysteine have also been associated with osteoporosis,10,11 which is a typical feature of GSE.20 Although several other factors are undoubtedly associated with the development of osteoporosis in GSE (including vitamin D and calcium malabsorption, secondary hyperparathyroidism, inflammatory cytokines),20 hyperhomocysteinemia might represent a further co-factor in a significant proportion of patients. The most widely recognized condition associated with hyperhomocysteinemia is thrombosis. The mechanisms by which hyperhomocysteinemia induces the increased risk for thrombosis are unknown, but direct damage to endothelial cells is suggested, leading to an increased risk at both arterial and venous sites.1–3 The high prevalence of hyperhomocysteinemia observed in our series of patients would support an association of undiagnosed GSE with an increased risk of thrombosis. However, GSE is not usually included in the list of diseases characterized by a pro-thrombotic behavior. This might be due to the usually young age of the patients at the time of diagnosis. In recent years, however, in parallel with the increased recognition of adult GSE cases, there has been a large number of observations reporting thromboses associated with silent or undiagnosed GSE,26 –35 in several of which a role for hyperhomocysteinemia has been advocated. According to these reports and to the present findings, GSE before diagnosis might be regarded as a disease at risk for thrombotic complications. Again, the rapid normalization of homocysteine levels observed in almost all our patients with hyperhomocysteinemia when put on GFD is likely associated with a reduction in their risk for thrombosis once the disease is diagnosed. However, we should also keep in mind that GSE patients adhering to a strict GFD for years might develop vitamin deficiencies, mainly folate deficiency,44 and exhibit high plasma homocysteine levels.45 It is interesting to note that rates of myocardial infarction and stroke do not seem substantially different in adult GSE patients compared to the general population, even if a lower prevalence of hypertension and hypercholesterolemia has been described in GSE.46 Thus, one might think that other risk factors for thrombosis
HYPERHOMOCYSTEINEMIA IN GLUTEN-SENSITIVE ENTEROPATHY
579
should be more prevalent in this clinical setting, and that hyperhomocysteinemia could be one of them. The present findings and the observation that GSE is a frequent disease often characterized by a paucity of typical symptoms, particularly in adulthood,47 also suggest that the serologic screening for GSE should be included in the diagnostic workup of all patients with thrombosis related to hyperhomocysteinemia. In conclusion, the present study shows that GSE patients frequently have hyperhomocysteinemia when on a gluten-containing diet. Hyperhomocysteinemia, which is frequently due to vitamin deficiencies and is rapidly corrected by a GFD, might represent a link between undiagnosed GSE and some of its complications.
References 1. McCully KS. Vascular pathology of homocysteinemia: implications for the pathogenesis of arteriosclerosis. Am J Pathol 1969; 56:111–128. 2. Welch GN, Loscalzo J. Homocysteine and atherothrombosis. N Engl J Med 1998;338:1042–1050. 3. Cattaneo M, Martinelli I, Mannucci PM. Hyperhomocysteinemia as a risk factor for deep-vein thrombosis. N Engl J Med 1996; 335:974 –975. 4. Kumar KS, Govindaiah V, Naushad SE, et al. Plasma homocysteine levels correlated to interactions between folate status and methylene tetrahydrofolate reductase gene mutation in women with unexplained recurrent pregnancy loss. J Obstet Gynaecol 2003;23:55–58. 5. Nelen WL, Blom HJ, Steegers EA, et al. Hyperhomocysteinemia and recurrent early pregnancy loss: a meta-analysis. Fertil Steril 2000;74:1196 –1199. 6. Quere I, Bellet H, Hoffet M, et al. A woman with five consecutive fetal deaths: case report and retrospective analysis of hyperhomocysteinemia prevalence in 100 consecutive women with recurrent miscarriages. Fertil Steril 1998;69:152–154. 7. Ray JG, Laskin CA. Folic acid and homocysteine metabolic defects and the risk of placental abruption, pre-eclampia and spontaneous pregnancy loss: a systematic review. Placenta 1999; 20:519 –529. 8. Goddijn-Wessel TA, Wouters MG, van de Molen EF, et al. Hyperhomocysteinemia: a risk factor for placental abruption or infarction. Eur J Obstet Gynecol Reprod Biol 1996;66:23–29. 9. Rosenquist TH, Ratashak SA, Selhub J. Homocysteine induces congenital defects of the heart and neural tube: effect of folate. Proc Natl Acad Sci U S A 1996;93:15227–15232. 10. van Meurs JBJ, Dhonukshe-Rutten RAM, Plujim SMF, et al. Homocysteine levels and the risk of osteoporotic fracture. N Engl J Med 2004;350:2033–2041. 11. McLean RR, Jacques PF, Selhub J, et al. Homocysteine as a predictive factor for hip fracture in older persons. N Engl J Med 2004;350:2042–2049. 12. Mudd SH, Skovby F, Levy HL, et al. The natural history of homocystinuria due to cystathionine -synthase deficiency. Am J Hum Genet 1985;37:1–31. 13. D’Angelo A, Selhub J. Homocysteine and thrombotic disease. Blood 1997;90:1–11. 14. Frosst P, Blom HJ, Milos R, et al. A candidate genetic risk factor for vascular disease: a common mutation in methylene-tetrahydrofolate reductase. Nat Genet 1995;10:111–113. 15. Deloughery TG, Evans A, Sadeghi A, et al. Common mutation in methylenetetrahydrofolate reductase: correlation with homocys-
580
16.
17. 18. 19.
20. 21.
22. 23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
SAIBENI ET AL
teine metabolism and late-onset vascular disease. Circulation 1996;94:3074 –3078. Fasano A, Catassi C. Current approaches to diagnosis and treatment of celiac disease: an evolving spectrum. Gastroenterology 2001;120:636 – 651. Kennedy NP, Feighery C. Clinical features of celiac disease today. Biomed Pharmacother 2000;54:373–380. Parnell ND, Ciclitira PJ. Review article: coeliac disease and its management. Aliment Pharmacol Ther 1999;13:1–13. Dickey W. Low serum vitamin B12 is common in coeliac disease and is not due to autoimmune gastritis. Eur J Gastroenterol Hepatol 2002;14:425– 427. Kemppainen T, Kroger H, Janatuinen E, et al. Osteoporosis in adult patients with celiac disease. Bone 1999;24:249 –255. Gasbarrini A, Torre ES, Trivellini C, et al. Recurrent spontaneous abortion and intrauterine fetal growth retardation as symptoms of coeliac disease. Lancet 2000;356:399 – 400. Martinelli P, Troncone R, Paparo F, et al. Coeliac disease and unfavourable outcome of pregnancy. Gut 2000;46:332–335. Sher KS, Mayberry JF. Female fertility, obstetrics and gynaecological history in celiac disease: a case-control study. Digestion 1994;55:243–246. Ciacci C, Cirillo M, Auriemma G, et al. Celiac disease and pregnancy outcome: celiac disease and pregnancy outcome. Am J Gastroenterol 1996;91:718 –722. Greco L, Veneziano A, Di Donato L, et al. Undiagnosed coeliac disease does not appear to be associated with unfavourable outcome of pregnancy. Gut 2004;53:149 –151. Gabrielli M, Santoliquido A, Gasbarrini G, et al. Latent coeliac disease, hyperhomocysteinemia and pulmonary thromboembolism: a close link. Thromb Haemost 2003;89:203–204. Marteau P, Cadranel JF, Messing B, et al. Association of hepatic vein obstruction and coeliac disease in North African subjects. J Hepatol 1994;20:650 – 653. Zenjari T, Boruchowicz A, Desreumaux P, et al. Association of coeliac disease and portal venous thrombosis. Gastroenterol Clin Biol 1995;19:953–954. Grigg AP. Deep venous thrombosis as the presenting feature in a patient with coeliac disease and homocysteinemia. Aust N Z J Med 1999;29:566 –567. Hozyasz K, Milanovski A. Hyperhomocysteinemia in a coeliac disease heterozygote for the two common mutations (677C-⬎T and 1298A-⬎C) of the methylenetetrahydrofolate reductase gene: case report. Med Wieku Rozwoj 2002;6:57– 61. Manzano ML, Garfia C, Manzanares J, et al. Celiac disease and Budd-Chiari syndrome: an uncommon association. Gastroenterol Hepatol 2002;25:159 –161. Gefel D, Doncheva M, Ben-Valid E, et al. Recurrent stroke in a young patient with celiac disease and hyperhomocysteinemia. Isr Med Assoc J 2002;4:222–223.
CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 3, No. 6
33. Andres E, Pflumio F, Knab MC, et al. Splenic thrombosis and celiac disease: a fortuitous association? Presse Med 2000; 29:1933–1934. 34. Kremer Hovinga JA, Baerlocher G, Wuillemin WA, et al. Deep venous thrombosis of the leg in acquired thrombophilia: hyperhomocysteinemia as a sequela of undetected celiac disease. Ther Umsch 1999;56:519 –522. 35. Hida M, Erreimi N, Ettari S, et al. Associated celiac disease and venous thrombosis. Arch Pediatr 2000;7:215–216. 36. Working Group of the United European Gastroenterology Week. When is a coeliac a coeliac? Eur J Gastroenterol Hepatol 2001; 13:1123–1128. 37. Rostami K, Kerckhaert J, von Blomberg BM, et al. SAT and serology in adult coeliacs: seronegative coeliac disease seems a reality. Neth J Med 1998;53:15–19. 38. de Franchis R, Primignani M, Monti M, et al. Small bowel involvement in dermatitis herpetiformis and in linear IgA bullous dermatitis. J Clin Gastronterol 1983;5:429 – 436. 39. Miller SA, Dykes DD, Polesky HF. A simple salting-out procedure for extracting DNA from human mutated cells. Nucleic Acids Res 1988;16:1215–1218. 40. Cattaneo M, Tsai MY, Bucciarelli P, et al. A common mutation in the methylenetetrahydrofolate reductase gene (C677T) increases the risk for deep-vein thrombosis in patients with mutant factor V (factor V:Q506). Arterioscler Thromb Vasc Biol 1997;17: 1662–1666. 41. Zighetti ML, Cattaneo M, Falcon CR, et al. Absence of hyperhomocysteinemia in ten patients with primary pulmonary hypertension. Thromb Res 1997;85:279 –282. 42. Armitage P, Berry G. Statistical methods in medical research. Oxford, UK: Blackwell Science Ltd, 1994. 43. Vecchi M, Sacchi E, Saibeni S, et al. Inflammatory bowel diseases are not associated with major hereditary conditions predisposing to thrombosis. Dig Dis Sci 2000;45:1465–1469. 44. Thompson T. Folate, iron, and dietary fiber contents of the glutenfree diet. J Am Diet Assoc 2000;100:1389 –1396. 45. Hallert C, Grant C, Grehn S, et al. Evidence of poor vitamin status in celiac patients on a gluten-free diet for 10 years. Aliment Pharmacol Ther 2002;16:1333–1339. 46. West J, Logan RFA, Card TR, et al. Risk of vascular disease in adults with diagnosed coeliac disease: a population-based study. Aliment Pharmacol Ther 2004;20:73–79. 47. Shamir R. Advances in celiac disease. Gastroenterol Clin North Am 2003;32:931–947.
Address requests for reprints to: Maurizio Vecchi, M.D., Gastroenterology and Gastrointestinal Endoscopy Service, Department of Internal Medicine, IRCCS Maggiore Hospital, Via Pace 9, 20122 Milan, Italy. e-mail: [email protected]; fax: ⴙ39-0250320747.
Prevalence of Hyperhomocysteinemia in Adult Gluten-Sensitive Enteropathy at Diagnosis: Role of B12, Folate, and Genetics SIMONE SAIBENI,* ANNA LECCHI,‡ GIANMICHELE MEUCCI,§ MARCO CATTANEO,储 LILIANA TAGLIABUE,‡,¶ EMANUELE RONDONOTTI,* SARA FORMENTI,* ROBERTO DE FRANCHIS,* and MAURIZIO VECCHI* *Gastroenterology and Gastrointestinal Endoscopy Service, ‡A. Bianchi Bonomi Thrombosis and Hemophilia Center, and ¶Fondazione Luigi Villa, IRCCS Maggiore Hospital and University of Milan, Milan, Italy; 储Department of Medicine, Surgery and Dentistry, San Paolo Hospital and University of Milan, Milan, Italy; and §Gastroenterology Division, Valduce Hospital, Como, Italy
Background & Aims: Hyperhomocysteinemia, a risk factor for thrombosis, recurrent miscarriages, and osteoporosis, might derive from acquired folate and vitamin B12 deficiencies and from a C677T mutation in methylenetetrahydrofolate reductase (MTHFR) gene. Undiagnosed gluten-sensitive enteropathy (GSE) is associated with vitamin deficiencies, osteoporosis, and recurrent miscarriages. We evaluated the prevalence and the risk factors for hyperhomocysteinemia in patients with newly diagnosed GSE. Methods: In this prospective study performed in a tertiary care setting, 40 consecutive subjects with newly diagnosed GSE were evaluated for homocysteine, folate, and vitamin B12 levels and for C677T polymorphism. One hundred twenty sex- and agematched healthy control subjects were studied. Nonparametric tests and multiple regression analysis were used to evaluate the risk factors in inducing hyperhomocysteinemia in the GSE population. Results: Hyperhomocysteinemia was more frequent in GSE patients than in control subjects (8/40, 20.0% vs 7/120, 5.8%) (relative risk, 3.4; 95% confidence interval, 1.3– 8.9), as well as folate deficiency (17/40, 42.5% vs 10/120, 8.3%) (relative risk, 5.1; 95% confidence interval, 2.5–10.2). Multiple regression analysis showed that folate and B12 levels were independently and inversely associated with homocysteine levels, whereas homozygosity for the MTHFR thermolabile variant was not. The prevalence of MTHFR variant in GSE population was not different from that reported in racially comparable control groups. Gluten-free diet was able to normalize folate, vitamin B12, and homocysteine levels. Conclusions: Hyperhomocysteinemia is frequent in newly diagnosed GSE. Vitamin deficiencies caused by malabsorption are the most important determinants of this condition. Hyperhomocysteinemia might contribute to the occurrence of common complications of undiagnosed GSE.
he possible role of hyperhomocysteinemia as a risk factor for atherothrombosis was first suggested by McCully,1 who reported autopsy evidence of extensive arterial
T
thrombosis in 2 children with homocystinuria. Subsequent investigations have confirmed this hypothesis, and it is now clear that hyperhomocysteinemia is an independent risk factor for both arterial and venous thrombosis.2,3 More recently, a possible role of hyperhomocysteinemia as an etiologic factor of recurrent miscarriages and congenital heart and neural tube defects has also been reported.4 –9 Hyperhomocysteinemia has also been suggested to play a role in osteoporosis and osteoporotic fractures.10,11 Increased levels of plasma homocysteine are typically caused by genetically determined defects in the enzymes involved in homocysteine metabolism, ie, cystathionine -synthase and methylene-tetrahydrofolate reductase (MTHFR), or by acquired deficiencies in vitamin cofactors, such as folate, vitamin B12, and vitamin B6.2 Homozygous traits for cystathionine -synthase and MTHFR deficiency are very rare and lead to severe hyperhomocysteinemia, recurrent thromboembolism, osteoporosis, and a poor prognosis.12,13 Recently, a frequent thermolabile variant of MTHFR caused by a point mutation (C677T)14 has been suggested to cause an exaggerated hyperhomocysteinemic response, particularly in TT homozygotic patients with concurrent vitamin deficiencies.15 Gluten-sensitive enteropathy (GSE) is characterized by a wide spectrum of clinical presentations.16 At diagnosis, both celiac disease and dermatitis herpetiformis (DH), which are part of this spectrum, might be associated with vitamin deficiencies,17–19 osteoporosis,20 and, in the female population, a history of recurrent miscarriages sometimes attributed to vitamin deficiencies,21–24 even if this association has been debated.25 Abbreviations used in this paper: AGA, anti-gliadin antibodies; CI, confidence interval; DH, dermatitis herpetiformis; EMA, anti-endomysial antibodies; GFD, gluten-free diet; GSE, gluten-sensitive enteropathy; MTHFR, methylene-tetrahydrofolate reductase; RR, relative risk. © 2005 by the American Gastroenterological Association 1542-3565/05/$30.00 PII: 10.1053/S1542-3565(05)00022-4
June 2005
HYPERHOMOCYSTEINEMIA IN GLUTEN-SENSITIVE ENTEROPATHY
575
Table 1. Demographic and Clinical Features of GSE Patients and Healthy Control Subjects
Men/women Age at diagnosis (mean ⫾ SD, y) Duration of symptoms (mean ⫾ SD, mo) Clinical presentationa Diarrhea Abdominal pain Anemia Weight loss Bloating Hypertransaminasemia Osteoporosis Dyspepsia Asymptomatic Aphtous stomatitis Recurrent miscarriages Epilepsia Skin lesions
Celiac disease (n ⫽ 32)
DH (n ⫽ 8)
Control subjects (n ⫽ 120)
7/25 39.7 ⫾ 12.5 44.3 ⫾ 83.5
4/4 37.4 ⫾ 20.7 37.5 ⫾ 27.0
33/87 38.9 ⫾ 10.9 —
13 7 7 6 5 5 4 4 2 1 1 1 —
— — — — — — — — — — — — 8
— — — — — — — — — — — — —
SD, standard deviation. aMore than 1 symptom might be present for each patient.
Several cases of thrombosis occurring in patients with celiac disease before diagnosis have also been reported, most of them suggesting a possible role of hyperhomocysteinemia.26 –35 We thus evaluated the prevalence of hyperhomocysteinemia in GSE patients at the time of diagnosis and investigated the role and the possible relationships of folate and vitamin B12 deficiencies and of the MTHFR thermolabile variant in inducing the risk of hyperhomocysteinemia.
Materials and Methods In this prospective study all adult patients with a new diagnosis of GSE seen at our Gastroenterology and Gastrointestinal Endoscopy Service in the period between January 2002 and December 2003 were consecutively enrolled.
Patients The study involved 40 patients with GSE; of them, 32 had celiac disease, and 8 had DH; as it is commonly observed, patients with celiac disease were more frequently women. At the time of enrollment, no patient was assuming drug treatments or was affected by other diseases capable of raising homocysteine plasma levels. The main demographic and clinical features of GSE patients are reported in Table 1. All GSE patients were enrolled at the time of diagnosis, which was obtained according to standardized serologic and histologic criteria (positivity of anti-endomysial IgA antibodies [EMA] and/or anti-gliadin antibodies [AGA] and histologic duodenal findings of villous atrophy and crypt hyperplasia).36 The degree of duodenal involvement was assessed according to a 4-grade scale (grade 0, normal; grade 1, increased IELs; grade 2, increased IELs and crypt hyperplasia; grade 3, increased IELs,
crypt hyperplasia, and [a] partial, [b] subtotal, [c] total villous atrophy) according to Marsh criteria revised by Rostami et al.37 The diagnosis of DH was made by showing typical IgA granular deposits in the subepidermal papillae by means of immunofluorescence on perilesional skin.38 All GSE patients were on a gluten-containing diet at the time of enrollment. The 8 GSE patients with hyperhomocysteinemia at diagnosis were re-evaluated for vitamin status and plasma homocysteine levels after a mean follow-up of 6.2 months on a strict gluten-free diet (GFD) without any vitamin supplementation.
Control Subjects One hundred twenty age- and sex-matched healthy subjects were also included in the study and served as control subjects. All these subjects were chosen from a larger cohort of healthy individuals (⬃900) with a negative personal and family history of thrombosis, celiac disease, or inflammatory bowel disease who had been previously enrolled as normal control subjects for the study of other hemostatic parameters. All these subjects were matched with GSE patients also for other determinants of homocysteine status such as liver and renal function, diabetes, and coffee consumption.
Blood Collection Blood samples were obtained, on informed consent, from all patients and control subjects. Serum and plasma samples were used for the determination of folate and vitamin B12 levels and of homocysteine levels, respectively. Whole blood was collected in ethylenediaminetetraacetic acid and rapidly frozen at ⫺80°C until the time of DNA extraction, which was obtained by means of a standard salting-out procedure.39
576
SAIBENI ET AL
CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 3, No. 6
Table 2. Histologic Grading and Serologic Status in GSE Patients
Histologic grading 3a 3b 3c Serologic status EMA ⫹, AGA ⫹ EMA ⫹, AGA ⫺ EMA ⫺, AGA ⫹
Celiac disease (n ⫽ 32)
DH (n ⫽ 8)
2 4 26
2 2 4
24 7 1
3 4 1
Inc, San Diego, CA) and then analyzed by means of nonparametric tests. In particular, comparisons between groups were performed by means of Fisher exact test, Mann–Whitney test, Wilcoxon matched pairs test, and Kruskal–Wallis test with Dunn post hoc comparison, as appropriate. Correlations between variables were assessed by means of Spearman nonparametric correlation. Relative risks (RRs) and their 95% confidence intervals (CIs) were also calculated according to standard formulas.42 The association between plasma homocysteine levels and serum vitamin levels, MTHFR allele status, and clinical and demographic features of GSE patients was assessed by means of multiple regression analysis. The significance level was set at P ⬍ .05.
Folate and Vitamin B12 Assay Serum folate and vitamin B12 determinations were performed by means of a commercially available RIA kit (DPC, Los Angeles, CA). Folate and vitamin B12 deficiency were set, according to the manufacturer’s instructions, at ⬍3 ng/mL and ⬍200 pg/mL, respectively.
DNA Analysis DNA analysis for MTHFR gene mutation (C677T) was performed according to a previously described method40; briefly, a fragment of 198 base pairs was amplified by polymerase chain reaction. The fragment was then digested by Hinf I restriction enzyme, and subsequent electrophoresis on ethidium bromide stained 2% agarose gel ⫹ 1% NUSIEVE agarose (Cambrex Corporation, East Rutherford, NJ) was performed.
Plasma Homocysteine Assay The concentration of total homocysteine in plasma was determined by high-performance liquid chromatography and fluorimetric detection.41 Hyperhomocysteinemia was diagnosed when plasma homocysteine levels exceeded the 95th percentile of distribution observed among matched healthy control subjects (13.9 mol/L for women, 18.7 mol/L for men).
Statistical Evaluation For statistical analysis, data were entered into a statistical software package (GraphPad Instat; GraphPad Software
Results All GSE patients showed grade 3 histologic lesions, and all of them had high-titer positive EMA or positivity of both AGA IgA and IgG (Table 2). In patients with GSE, median plasma homocysteine level was significantly higher than in matched control subjects (11.1 mol/L, range 5.3–33.2 mol/L vs 8.9 mol/L, range 4.9 – 42.3 mol/L) (P ⬍ .02). The prevalence of hyperhomocysteinemia, as defined above, was significantly higher in GSE patients (8 of 40, 20.0%) than in matched control subjects (7 of 120, 5.8%) (RR, 3.4; 95% CI, 1.3– 8.9; P ⬍ .02) (Table 3). Median serum folate levels were significantly lower in GSE patients than in healthy control subjects (3.4 ng/ mL, range 0.7–13.5 ng/mL vs 6.0 ng/mL, range 2.0 – 24.0 ng/mL; P ⬍ .0001), whereas median vitamin B12 levels were similar (363.5 pg/mL, range 96.0 – 832.0 pg/mL in GSE patients vs 389.5 pg/mL, range 89.0 – 1096.0 pg/mL in healthy control subjects; P ⫽ .26). Seventeen GSE patients (42.5%) had folate deficiency, and 7 (17.5%) had vitamin B12 deficiency. Folate deficiency was significantly more frequent than in healthy control subjects (10/120, 8.3%) (RR, 5.1; 95% CI, 2.5–
Table 3. Prevalence of Hyperhomocysteinemia, Folate, and Vitamin B12 Deficiency and MTHFR Allele Status in GSE Patients and in Healthy Control Subjects Healthy control subjects (n ⫽ 120)
GSE (n ⫽ 40)
Hyperhomocysteinemia Folate deficiency Vitamin B12 deficiency MTHFR genotype at 677 C/C C/T T/T T allele frequency na ⫽ not assessed.
n
%
n
%
RR (95% CI)
8 17 7
20.0 42.5 17.5
7 10 11
5.8 8.3 9.2
3.4 (1.3–8.9) 5.1 (2.5–10.2) 1.9 (0.8–4.6)
13 21 6
32.5 52.5 15.0 41.2
na na na na
— — — —
June 2005
HYPERHOMOCYSTEINEMIA IN GLUTEN-SENSITIVE ENTEROPATHY
577
Table 4. Homocysteine, Folate, and Vitamin B12 Levels in GSE Patients According to Histologic Grading 3a Folate (ng/mL) Vitamin B12 (pg/mL) Homocysteine (mol/L)
3b
7.2 (6.6–13.5) 401.0 (96.0–832.0) 7.7 (5.3–8.2)
4.6 (1.2–12.5) 468.0 (337.0–580.0) 7.8 (6.8–19.6)
3c 2.1 (0.7–10.4) 337.5 (126.0–708.0) 12.4 (5.0–33.2)
P value ⬍.01a .20 ⬍.02a
NOTE. Results are expressed as median and range. a3a vs 3c.
10.2; P ⬍ .0001), whereas vitamin B12 deficiency was not (11/120, 9.2%) (RR, 1.9; 95% CI, 0.8 – 4.6; P ⫽ .16) (Table 3). Median serum folate levels were significantly lower in the 8 GSE patients with hyperhomocysteinemia than in the 32 patients with normal homocysteine levels (1.8 ng/mL, range 0.7– 4.7 ng/mL vs 3.7 ng/mL, range 0.7– 13.5 ng/mL; P ⬍ .03), whereas the difference did not reach the statistical significance for vitamin B12 levels (265.0 pg/mL, range 158.0 –580.0 vs 386.5 pg/mL, range 96.0 – 832.0 pg/mL; P ⫽ .07). Folate deficiency was significantly more frequent in patients with hyperhomocysteinemia (6 of 8, 75.0%) than in normohomocysteinemic patients (11 of 32, 34.4%) (RR, 2.2; 95% CI, 1.2– 4.1; P ⫽ .05), and vitamin B12 deficiency was more frequent in hyperhomocysteinemic than in normohomocysteinemic GSE patients, although this difference did not reach statistical significance (3/8, 37.5% vs 4/32, 12.5%) (RR, 3.0; 95% CI, 0.8 –10.8; P ⫽ .13). When we stratified GSE patients according to their clinical and demographic features, no correlation was observed between age and homocysteine, folate, and vitamin B12 levels or between duration of symptoms before diagnosis and homocysteine, folate, and vitamin B12 levels (data not shown). In particular, no difference was observed regarding median folate, vitamin B12, and homocysteine levels and prevalence of folate and vitamin B12 deficiencies and hyperhomocysteinemia when patients with celiac disease were compared to patients with DH (data not shown). Also, we did not observe any difference in homocysteine, folate, and vitamin B12 levels when the 32 patients with celiac disease were stratified according to the pres-
ence of a typical (with gastrointestinal symptoms) or atypical (no gastrointestinal symptoms) presentation. However, GSE patients with a more severe duodenal involvement had lower folate and higher homocysteine levels than patients with minor duodenal damage (Table 4). The MTHFR thermolabile variant mutation was observed in 21 patients with heterozygosity (52.5%) and 6 with homozygosity (15.0%), resulting in an allele frequency of 41.2% (Table 3). After stratifying GSE patients according to their allele status, no significant difference was observed as far as folate, vitamin B12, and homocysteine levels were concerned, even if a tendency to have lower folate and higher homocysteine levels was observed in mutated homozygous (T/T) patients (data not shown). Three of the 6 patients homozygous for the MTHFR variant had high homocysteine levels, compared with 5 of the remaining nonhomozygous 34 (50.0% vs 14.7%) (RR, 3.4; 95% CI, 1.1–10.6; P ⫽ .08). Conversely, the prevalence of such homozygosity was 3/8 (37.5%) among hyperhomocysteinemic patients and 3/32 (9.4%) among normohomocysteinemic patients (RR, 4.0; 95% CI, 1.0 – 16.2; P ⫽ .08). The multiple regression analysis showed that folate and B12 levels were independently and inversely associated with homocysteine levels, whereas homozygosity for the MTHFR was very close to statistical significance. On the contrary, gender, age, and duration of symptoms were not correlated with plasma homocysteine levels. Data on the variables that were significant or close to statistical significance are reported in Table 5. During GFD, a significant increase in median folate and vitamin B12 levels and a significant decrease in
Table 5. Association of Homocysteine Levels With Folate and Vitamin B12 Levels and MTHFR Homozygosity for Thermolabile Variant in GSE Patients (Multiple Regression Analysis)
Folate levels MTHFR homozygosity for the thermolabile variant Vitamin B12 levels
Coefficient
95% CI
P value
⫺0.705 ⫹2.568 ⫺0.016
From ⫺1.218 to ⫺0.192 From ⫺0.061 to 5.198 From ⫺0.025 to ⫺0.006
.008 .055 .002
578
SAIBENI ET AL
plasma homocysteine levels were observed (Figure 1). All these patients normalized their vitamin status, and 7 of 8 patients reached normal homocysteine levels. The only patient who did not normalize completely her homocysteine level was a woman homozygous for the MTHFR mutation.
Discussion Despite the presence in the literature of isolated case reports of hyperhomocysteinemia occurring in patients with celiac disease, no data were so far available on the true prevalence of this association and on its possible underlying mechanisms. In the present study, which is the first prospectively performed in a consecutive series of subjects with a new diagnosis of GSE, we observed that these subjects have both high median plasma homocysteine levels and an increased prevalence of hyperhomocysteinemia. Indeed, 1 of 5 GSE patients in this series had hyperhomocysteinemia, as defined by plasma homocysteine levels higher than the 95th percentile of values of the control population. Gender, age at diagnosis, duration of symptoms, and type of clinical presentation were not predictive of the presence of hyperhomocysteinemia; however, high levels of homocysteine appear to be strictly related to the degree of vitamin deficiencies, resulting from more severe intestinal damage. The carriage of the MTHFR thermolabile variant gene mutation, leading to reduced activity of this key enzyme in homocysteine metabolism, might also be implicated in some patients. However, because the prevalence of homozygosity for the C677T mutation in our GSE population (15.0%) and the frequency of the mutated allele (41.2%) were not different from previously reported background populations from our group (22.4% and 47.4%, respectively),43 it appears that there is not a relevant role for the genetic variant. Moreover, the ability of GFD to normalize folate, vitamin B12, and homocysteine levels in GSE patients points to the major role played by malabsorptive causes than by the genetic defect. The major role of vitamin deficiencies in inducing hyperhomocysteinemia in GSE patients is not surprising, because folate and vitamin B12 deficiencies are wellknown risk factors for hyperhomocysteinemia2 and are also frequently reported in GSE patients at diagnosis.17–19 The finding of frequent vitamin B12 deficiency (in our series twice that of control subjects) suggests that distal small bowel might also be involved in GSE,19 although a relationship with autoimmune gastritis cannot be excluded in the present series.
CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 3, No. 6
Whatever the mechanism(s) of its induction, the observation that hyperhomocysteinemia is so frequent in GSE appears to provide an important link between this disease and some of its more frequent complications. Indeed, women with a subsequent diagnosis of GSE appear to be characterized by the frequent occurrence of early fetal loss and newborn malformations,24 but recently this association has been debated.25 Although this feature has been correlated to the presence of other factors possibly occurring in GSE and mainly related to malabsorption, increasing evidence supports a major role of hyperhomocysteinemia in these complications of pregnancy in the general population. In fact, recurrent pregnancy loss and congenital heart and neural tube defects
Figure 1. Changes in folate (A), vitamin B12 (B), and homocysteine levels (C) in the 8 hyperhomocysteinemic GSE patients followed up during GFD.
June 2005
are considered a consequence of hyperhomocysteinemia.4 –9 Thus, the present observation might provide an important clue to the understanding of the pathogenetic mechanisms of pregnancy complications in GSE women before diagnosis. The finding that homocysteine levels return to normal in the majority of patients in our series shortly after the beginning of GFD is also in agreement with the observation that the risk for complicated pregnancy reverts to normal in GSE women after diagnosis. High levels of plasma homocysteine have also been associated with osteoporosis,10,11 which is a typical feature of GSE.20 Although several other factors are undoubtedly associated with the development of osteoporosis in GSE (including vitamin D and calcium malabsorption, secondary hyperparathyroidism, inflammatory cytokines),20 hyperhomocysteinemia might represent a further co-factor in a significant proportion of patients. The most widely recognized condition associated with hyperhomocysteinemia is thrombosis. The mechanisms by which hyperhomocysteinemia induces the increased risk for thrombosis are unknown, but direct damage to endothelial cells is suggested, leading to an increased risk at both arterial and venous sites.1–3 The high prevalence of hyperhomocysteinemia observed in our series of patients would support an association of undiagnosed GSE with an increased risk of thrombosis. However, GSE is not usually included in the list of diseases characterized by a pro-thrombotic behavior. This might be due to the usually young age of the patients at the time of diagnosis. In recent years, however, in parallel with the increased recognition of adult GSE cases, there has been a large number of observations reporting thromboses associated with silent or undiagnosed GSE,26 –35 in several of which a role for hyperhomocysteinemia has been advocated. According to these reports and to the present findings, GSE before diagnosis might be regarded as a disease at risk for thrombotic complications. Again, the rapid normalization of homocysteine levels observed in almost all our patients with hyperhomocysteinemia when put on GFD is likely associated with a reduction in their risk for thrombosis once the disease is diagnosed. However, we should also keep in mind that GSE patients adhering to a strict GFD for years might develop vitamin deficiencies, mainly folate deficiency,44 and exhibit high plasma homocysteine levels.45 It is interesting to note that rates of myocardial infarction and stroke do not seem substantially different in adult GSE patients compared to the general population, even if a lower prevalence of hypertension and hypercholesterolemia has been described in GSE.46 Thus, one might think that other risk factors for thrombosis
HYPERHOMOCYSTEINEMIA IN GLUTEN-SENSITIVE ENTEROPATHY
579
should be more prevalent in this clinical setting, and that hyperhomocysteinemia could be one of them. The present findings and the observation that GSE is a frequent disease often characterized by a paucity of typical symptoms, particularly in adulthood,47 also suggest that the serologic screening for GSE should be included in the diagnostic workup of all patients with thrombosis related to hyperhomocysteinemia. In conclusion, the present study shows that GSE patients frequently have hyperhomocysteinemia when on a gluten-containing diet. Hyperhomocysteinemia, which is frequently due to vitamin deficiencies and is rapidly corrected by a GFD, might represent a link between undiagnosed GSE and some of its complications.
References 1. McCully KS. Vascular pathology of homocysteinemia: implications for the pathogenesis of arteriosclerosis. Am J Pathol 1969; 56:111–128. 2. Welch GN, Loscalzo J. Homocysteine and atherothrombosis. N Engl J Med 1998;338:1042–1050. 3. Cattaneo M, Martinelli I, Mannucci PM. Hyperhomocysteinemia as a risk factor for deep-vein thrombosis. N Engl J Med 1996; 335:974 –975. 4. Kumar KS, Govindaiah V, Naushad SE, et al. Plasma homocysteine levels correlated to interactions between folate status and methylene tetrahydrofolate reductase gene mutation in women with unexplained recurrent pregnancy loss. J Obstet Gynaecol 2003;23:55–58. 5. Nelen WL, Blom HJ, Steegers EA, et al. Hyperhomocysteinemia and recurrent early pregnancy loss: a meta-analysis. Fertil Steril 2000;74:1196 –1199. 6. Quere I, Bellet H, Hoffet M, et al. A woman with five consecutive fetal deaths: case report and retrospective analysis of hyperhomocysteinemia prevalence in 100 consecutive women with recurrent miscarriages. Fertil Steril 1998;69:152–154. 7. Ray JG, Laskin CA. Folic acid and homocysteine metabolic defects and the risk of placental abruption, pre-eclampia and spontaneous pregnancy loss: a systematic review. Placenta 1999; 20:519 –529. 8. Goddijn-Wessel TA, Wouters MG, van de Molen EF, et al. Hyperhomocysteinemia: a risk factor for placental abruption or infarction. Eur J Obstet Gynecol Reprod Biol 1996;66:23–29. 9. Rosenquist TH, Ratashak SA, Selhub J. Homocysteine induces congenital defects of the heart and neural tube: effect of folate. Proc Natl Acad Sci U S A 1996;93:15227–15232. 10. van Meurs JBJ, Dhonukshe-Rutten RAM, Plujim SMF, et al. Homocysteine levels and the risk of osteoporotic fracture. N Engl J Med 2004;350:2033–2041. 11. McLean RR, Jacques PF, Selhub J, et al. Homocysteine as a predictive factor for hip fracture in older persons. N Engl J Med 2004;350:2042–2049. 12. Mudd SH, Skovby F, Levy HL, et al. The natural history of homocystinuria due to cystathionine -synthase deficiency. Am J Hum Genet 1985;37:1–31. 13. D’Angelo A, Selhub J. Homocysteine and thrombotic disease. Blood 1997;90:1–11. 14. Frosst P, Blom HJ, Milos R, et al. A candidate genetic risk factor for vascular disease: a common mutation in methylene-tetrahydrofolate reductase. Nat Genet 1995;10:111–113. 15. Deloughery TG, Evans A, Sadeghi A, et al. Common mutation in methylenetetrahydrofolate reductase: correlation with homocys-
580
16.
17. 18. 19.
20. 21.
22. 23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
SAIBENI ET AL
teine metabolism and late-onset vascular disease. Circulation 1996;94:3074 –3078. Fasano A, Catassi C. Current approaches to diagnosis and treatment of celiac disease: an evolving spectrum. Gastroenterology 2001;120:636 – 651. Kennedy NP, Feighery C. Clinical features of celiac disease today. Biomed Pharmacother 2000;54:373–380. Parnell ND, Ciclitira PJ. Review article: coeliac disease and its management. Aliment Pharmacol Ther 1999;13:1–13. Dickey W. Low serum vitamin B12 is common in coeliac disease and is not due to autoimmune gastritis. Eur J Gastroenterol Hepatol 2002;14:425– 427. Kemppainen T, Kroger H, Janatuinen E, et al. Osteoporosis in adult patients with celiac disease. Bone 1999;24:249 –255. Gasbarrini A, Torre ES, Trivellini C, et al. Recurrent spontaneous abortion and intrauterine fetal growth retardation as symptoms of coeliac disease. Lancet 2000;356:399 – 400. Martinelli P, Troncone R, Paparo F, et al. Coeliac disease and unfavourable outcome of pregnancy. Gut 2000;46:332–335. Sher KS, Mayberry JF. Female fertility, obstetrics and gynaecological history in celiac disease: a case-control study. Digestion 1994;55:243–246. Ciacci C, Cirillo M, Auriemma G, et al. Celiac disease and pregnancy outcome: celiac disease and pregnancy outcome. Am J Gastroenterol 1996;91:718 –722. Greco L, Veneziano A, Di Donato L, et al. Undiagnosed coeliac disease does not appear to be associated with unfavourable outcome of pregnancy. Gut 2004;53:149 –151. Gabrielli M, Santoliquido A, Gasbarrini G, et al. Latent coeliac disease, hyperhomocysteinemia and pulmonary thromboembolism: a close link. Thromb Haemost 2003;89:203–204. Marteau P, Cadranel JF, Messing B, et al. Association of hepatic vein obstruction and coeliac disease in North African subjects. J Hepatol 1994;20:650 – 653. Zenjari T, Boruchowicz A, Desreumaux P, et al. Association of coeliac disease and portal venous thrombosis. Gastroenterol Clin Biol 1995;19:953–954. Grigg AP. Deep venous thrombosis as the presenting feature in a patient with coeliac disease and homocysteinemia. Aust N Z J Med 1999;29:566 –567. Hozyasz K, Milanovski A. Hyperhomocysteinemia in a coeliac disease heterozygote for the two common mutations (677C-⬎T and 1298A-⬎C) of the methylenetetrahydrofolate reductase gene: case report. Med Wieku Rozwoj 2002;6:57– 61. Manzano ML, Garfia C, Manzanares J, et al. Celiac disease and Budd-Chiari syndrome: an uncommon association. Gastroenterol Hepatol 2002;25:159 –161. Gefel D, Doncheva M, Ben-Valid E, et al. Recurrent stroke in a young patient with celiac disease and hyperhomocysteinemia. Isr Med Assoc J 2002;4:222–223.
CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 3, No. 6
33. Andres E, Pflumio F, Knab MC, et al. Splenic thrombosis and celiac disease: a fortuitous association? Presse Med 2000; 29:1933–1934. 34. Kremer Hovinga JA, Baerlocher G, Wuillemin WA, et al. Deep venous thrombosis of the leg in acquired thrombophilia: hyperhomocysteinemia as a sequela of undetected celiac disease. Ther Umsch 1999;56:519 –522. 35. Hida M, Erreimi N, Ettari S, et al. Associated celiac disease and venous thrombosis. Arch Pediatr 2000;7:215–216. 36. Working Group of the United European Gastroenterology Week. When is a coeliac a coeliac? Eur J Gastroenterol Hepatol 2001; 13:1123–1128. 37. Rostami K, Kerckhaert J, von Blomberg BM, et al. SAT and serology in adult coeliacs: seronegative coeliac disease seems a reality. Neth J Med 1998;53:15–19. 38. de Franchis R, Primignani M, Monti M, et al. Small bowel involvement in dermatitis herpetiformis and in linear IgA bullous dermatitis. J Clin Gastronterol 1983;5:429 – 436. 39. Miller SA, Dykes DD, Polesky HF. A simple salting-out procedure for extracting DNA from human mutated cells. Nucleic Acids Res 1988;16:1215–1218. 40. Cattaneo M, Tsai MY, Bucciarelli P, et al. A common mutation in the methylenetetrahydrofolate reductase gene (C677T) increases the risk for deep-vein thrombosis in patients with mutant factor V (factor V:Q506). Arterioscler Thromb Vasc Biol 1997;17: 1662–1666. 41. Zighetti ML, Cattaneo M, Falcon CR, et al. Absence of hyperhomocysteinemia in ten patients with primary pulmonary hypertension. Thromb Res 1997;85:279 –282. 42. Armitage P, Berry G. Statistical methods in medical research. Oxford, UK: Blackwell Science Ltd, 1994. 43. Vecchi M, Sacchi E, Saibeni S, et al. Inflammatory bowel diseases are not associated with major hereditary conditions predisposing to thrombosis. Dig Dis Sci 2000;45:1465–1469. 44. Thompson T. Folate, iron, and dietary fiber contents of the glutenfree diet. J Am Diet Assoc 2000;100:1389 –1396. 45. Hallert C, Grant C, Grehn S, et al. Evidence of poor vitamin status in celiac patients on a gluten-free diet for 10 years. Aliment Pharmacol Ther 2002;16:1333–1339. 46. West J, Logan RFA, Card TR, et al. Risk of vascular disease in adults with diagnosed coeliac disease: a population-based study. Aliment Pharmacol Ther 2004;20:73–79. 47. Shamir R. Advances in celiac disease. Gastroenterol Clin North Am 2003;32:931–947.
Address requests for reprints to: Maurizio Vecchi, M.D., Gastroenterology and Gastrointestinal Endoscopy Service, Department of Internal Medicine, IRCCS Maggiore Hospital, Via Pace 9, 20122 Milan, Italy. e-mail: [email protected]; fax: ⴙ39-0250320747.