SEARCH

SEARCH BY CITATION

Keywords:

  • pediatric;
  • inflammatory bowel disease;
  • thromboembolism

Abstract

  1. Top of page
  2. Abstract
  3. METHODS
  4. RESULTS OF THE SEARCH STRATEGY
  5. INCIDENCE OF TE IN CHILDREN WITH IBD
  6. CHARACTERISTICS OF CHILDREN WITH TE IN IBD
  7. PROTHROMBOTIC RISK FACTORS
  8. CLINICAL CHARACTERISTIC OF TE IN CHILDREN WITH IBD
  9. CEREBRAL TE
  10. RETINIC THROMBOEMBOLISM
  11. ABDOMINAL TE
  12. UPPER AND LOWER LIMBS TE
  13. PULMONARY TE
  14. THERAPY FOR TE
  15. MORTALITY
  16. RISK OF RECURRENCE
  17. DISCUSSION
  18. REFERENCES

Background:

Several studies suggest an increased risk of venous and arterial thromboembolism (TE) in adults with inflammatory bowel disease (IBD) compared to the general population. We performed a systematic review of studies on incidence and characteristic of TE in children with IBD.

Methods:

We searched Medline, LILACS, EMBASE, POPLINE, CINHAL, and reference lists of identified articles, without language restrictions, in August 2010.

Results:

Population studies suggest that there is an increased risk of TE in children with IBD compared to controls. TE occurred in children with IBD in all age ranges, mostly (82.8%) during active disease, and more frequently in children with ulcerative colitis (odds ratio [OR] 3.7, 95% confidence interval [CI] 1.8–7.6). At least one specific risk factor for TE was recognized in 50% of cases; two risk factors were present in 24%. Out of 92 published cases of TE in children with IBD, 54.3% occurred in cerebral site, 26% in the limbs, 13% in the abdominal vessels, and the remaining in the retina and lungs. After a first episode of TE, an early recurrence was observed in 11.4% of children, a late recurrence in 10%. A number of different therapeutic schemes were used. Overall mortality was 5.7% and was mostly associated with cerebral TE.

Conclusions:

Population studies are needed to clarify the risk of TE in children with IBD, the relative weight of other risk factors, the characteristics of the events, and to define guidelines of therapy and prophylaxis. (Inflamm Bowel Dis 2010)

In prospective pediatric registries in North America and Europe, the annual incidence of thromboembolism (TE) was estimated to be 0.7–1.4 per 100,000 in the general pediatric population and 5.3 per 10,000 hospitalizations.1–4 Venous TE is more frequent than arterial TE, accounting for about 75% of total manifestations.1–4 Unlike adults, 90%–95% of children have at least one risk factor for TE.3, 4 Risk factors in children include: prematurity, malignancies, trauma/surgery, cardiac disorders, oral contraceptives, obesity, dehydration, infection, and inflammation.3, 4

The risk of venous and arterial TE in patients with inflammatory bowel diseases (IBD) compared to the general population has been investigated in several large historical cohorts in adults (Table 1). These studies used databases both at the hospital and at population levels; therefore, the absolute incidence of TE varies. Albeit with these limitations, the studies consistently reported a 2–3-fold increase in risk of venous TE in patients with IBD compared to those without IBD.1–6 During a moderate to severe flare this increase in risk was much more prominent (hazard ratio [HR] 8·4; P < 0·0001; 9·0 per 1000 person-years).5 The only study failing to show an increased risk of venous TE in patients with IBD compared to the general population, however, shows that venous TE occurred at a younger age in patients with IBD than in those without IBD6 and this anticipation in age is consistent with other studies.7–9

Table 1. Historic Cohort Studies Reporting on Incidence of Thromboembolism in IBD Patients Compared to Non-IBD Patients
StudyCountryPeriodSource of DataSample SizeIncidence of Events in IBD Population /10,000 Person-yearRelative Risk (95% CI)
IBDNon-IBD
  • CD, Crohn's disease; UC, ulcerative colitis; VT, deep venous thrombosis; PE, pulmonary embolism; IHD, ischemic heart disease; CVD, cerebrovascular disease; ATED, undifferentiated arterial TE; MI, myocardial infarction: MEI, mesenteric ischemia; NS, not significant.

  • a

    Hazard rate;

  • b

    data expressed as number of hospitalizations;

  • c

    odds ratio;

  • d

    incidence rate ratio.

Grainge 5UK1987-2001General Practice Research Database13 75671 672VT 263·4 (2.7-4.3)a
Grip6Sweden1987-1997Hospital database1253387VT 15NS
Nguyen7Canada1998-2004Nationwide Inpatient Sample (NIS) database116842522703VT 139 (CD)b1.4 (1.3-1.6)c (CD)
210 (UC)1.8 (1.7-2.0)c (UC)
Bernstein8Canada1984-1997Manitoba health administrative database41005410050VT 31.4 (CD)4.7 (3.5-6.3)d (CD)
30.0 (UC)2.8 (2.1-3.7)d (UC)
PE 10.3 (CD)2.9 (1.8-4.7)d (CD)
19.8 (UC)3.6 (2.5-5.2)d (UC)
Bernstein9USA1984-2003Manitoba Health administrative databases807280489IHD 211.2 (1.1-1.4)d
CVD 131.16 (0.98-1.36)d
ATED 2.5NS
Ha11USA2001-2006Market-Scan Commercial Claims and Encounters database1748769948VT 362.1 (1.8-2.4)c
PE 142.1 (1.7-2.6)c
MEI 3213.6 (10.7-17.6)c
IHD 122NS
MI 18NS
CVD 36NS

Moreover, mortality from venous TE was higher in IBD patients compared to the general population (odds ratio [OR] 2.50, 95% confidence interval [CI] 1.83–3.43), length of hospitalization was significantly longer, and hospital expenditures were consequently increased.7

Arterial TE in patients with IBD is less well documented and the results are not consistent. A systematic review of cardiovascular mortality did not find a significant difference between IBD patients and controls.10 Later, Ha et al11 found that IBD patients had a 20% higher risk for any arterial TE compared with controls (HR = 1.2, P < 0.0001), but when examining individual diagnoses only the risk of acute mesenteric ischemia was significantly increased (HR = 11.2, P < 0.001). On the contrary, Bernstein et al9 found that all IBD patients had an increased risk of ischemic heart disease (incidence rate ratio [IRR] 1.26, 95% CI 1.11–1.44), while cerebrovascular disease was more frequent only in those with Crohn's disease (CD) (IRR 1.32, 95% CI, 1.05–1.66), and undifferentiated arterial TE had a higher risk only in IBD patients aged 0 to 39 years (IRR 19.95, 95% CI 1.81–219.92) and 40 to 59 years (IRR 3.17, 95% CI 1.27–7.91) and in females (IRR 1.96, 95% CI 1.24–3.10).

The objectives of this article are: 1) to review all published data on incidence of TE in IBD patients in pediatric age; 2) to review the characteristics of the events, associated risk factors, and outcomes.

METHODS

  1. Top of page
  2. Abstract
  3. METHODS
  4. RESULTS OF THE SEARCH STRATEGY
  5. INCIDENCE OF TE IN CHILDREN WITH IBD
  6. CHARACTERISTICS OF CHILDREN WITH TE IN IBD
  7. PROTHROMBOTIC RISK FACTORS
  8. CLINICAL CHARACTERISTIC OF TE IN CHILDREN WITH IBD
  9. CEREBRAL TE
  10. RETINIC THROMBOEMBOLISM
  11. ABDOMINAL TE
  12. UPPER AND LOWER LIMBS TE
  13. PULMONARY TE
  14. THERAPY FOR TE
  15. MORTALITY
  16. RISK OF RECURRENCE
  17. DISCUSSION
  18. REFERENCES

Medline, LILACS, EMBASE, POPLINE, and CINHAL were searched in August 2010. The following search strategy was used for Medline: (“Thrombosis”[Mesh] OR Thrombosis OR “thromboemboli*” OR “emboli*” OR “stroke”) AND (“Crohn Disease”[Mesh] OR “Colitis, Ulcerative”[Mesh] OR “Inflammatory Bowel Diseases”[Mesh] OR Crohn OR “ulcerative colitis” OR “inflammatory bowel disease”) AND (“infant”[MeSH Terms] OR “child”[MeSH Terms] OR “adolescent”[MeSH Terms] OR infant OR child OR adolescent OR paediatr*). The same keywords were used for the other databases, without language restrictions. Reference lists of all identified studies were checked. All studies identified by the search strategy were screened by two authors (M.L. and M.B.), and full articles were retrieved for all potentially relevant articles. Authors were contacted for clarification if necessary, and any disagreements were resolved through discussion. Only case reports in children 18 years of age or less were included.

RESULTS OF THE SEARCH STRATEGY

  1. Top of page
  2. Abstract
  3. METHODS
  4. RESULTS OF THE SEARCH STRATEGY
  5. INCIDENCE OF TE IN CHILDREN WITH IBD
  6. CHARACTERISTICS OF CHILDREN WITH TE IN IBD
  7. PROTHROMBOTIC RISK FACTORS
  8. CLINICAL CHARACTERISTIC OF TE IN CHILDREN WITH IBD
  9. CEREBRAL TE
  10. RETINIC THROMBOEMBOLISM
  11. ABDOMINAL TE
  12. UPPER AND LOWER LIMBS TE
  13. PULMONARY TE
  14. THERAPY FOR TE
  15. MORTALITY
  16. RISK OF RECURRENCE
  17. DISCUSSION
  18. REFERENCES

The search strategy retrieved 236 articles. In all 185 articles were excluded as they did not concern TE events in children or were not original reports. In total, two historical cohort studies,5, 7 five retrospective analysis of single hospital databases,12–16 eight case series,17–24 and thirty-six case-reports25–60 were retrieved (Fig. 1). Reports were published between 1971 and 2009. In all cases, data were extracted from full articles, except for three reports published in Finnish, Japanese, and Dutch.58–60

thumbnail image

Figure 1. Flow diagram of study selection.

Download figure to PowerPoint

INCIDENCE OF TE IN CHILDREN WITH IBD

  1. Top of page
  2. Abstract
  3. METHODS
  4. RESULTS OF THE SEARCH STRATEGY
  5. INCIDENCE OF TE IN CHILDREN WITH IBD
  6. CHARACTERISTICS OF CHILDREN WITH TE IN IBD
  7. PROTHROMBOTIC RISK FACTORS
  8. CLINICAL CHARACTERISTIC OF TE IN CHILDREN WITH IBD
  9. CEREBRAL TE
  10. RETINIC THROMBOEMBOLISM
  11. ABDOMINAL TE
  12. UPPER AND LOWER LIMBS TE
  13. PULMONARY TE
  14. THERAPY FOR TE
  15. MORTALITY
  16. RISK OF RECURRENCE
  17. DISCUSSION
  18. REFERENCES

Incidence of TE in children with IBD has been reported in two large cohort studies5, 7 and five retrospective analysis of single hospital databases.12–16 Nguyen and Sam7 in a large historical cohort based on the Nationwide Inpatient Sample (NIS) database reflecting a 20% stratified sample of nonfederal acute-care hospitals in the United States, report 26/79,983 hospitalizations for venous TE in non-IBD children compared to 3/604 in children with CD (RR 20.2, 95% CI 7.0–57.8, P < 0.0001), and 3/671 in children with ulcerative colitis (UC) (RR 13.7, 95% CI 4.1–45.3, P < 0.001) in the age group 0–10 years (Table 2). In the 11–20-year age group, there were 47/24,006 venous TE in non-IBD children compared to 36/3834 in children with CD (RR 4.7, 95% CI 3.0–7.3, P < 0.00001), and 19/7323 in children with UC (RR 1.3, 95% CI 0.7–2.2, P < 0.3). A second cohort study, based on the UK general practitioner database and including 556 IBD and 3506 non-IBD children under 20 years of age, reported only one case of TE in this age group, in a child with UC (unpubl. data provided by the main author).5

Table 2. Age-adjusted Rates of Venous TE Among Children with Crohn's Disease (CD), Ulcerative Colitis (UC), or Non-IBD in the Nationwide Inpatient Sample USA7
  1. Detailed data obtained from main author.

thumbnail image

Five single-hospital retrospective case note reviews estimated that the risk of vascular complication in children with IBD ranged between 1.3% and 3.3%.12–16

CHARACTERISTICS OF CHILDREN WITH TE IN IBD

  1. Top of page
  2. Abstract
  3. METHODS
  4. RESULTS OF THE SEARCH STRATEGY
  5. INCIDENCE OF TE IN CHILDREN WITH IBD
  6. CHARACTERISTICS OF CHILDREN WITH TE IN IBD
  7. PROTHROMBOTIC RISK FACTORS
  8. CLINICAL CHARACTERISTIC OF TE IN CHILDREN WITH IBD
  9. CEREBRAL TE
  10. RETINIC THROMBOEMBOLISM
  11. ABDOMINAL TE
  12. UPPER AND LOWER LIMBS TE
  13. PULMONARY TE
  14. THERAPY FOR TE
  15. MORTALITY
  16. RISK OF RECURRENCE
  17. DISCUSSION
  18. REFERENCES

Eight case series17–24 and 36 single case-reports25–60 were retrieved, accounting for a total of 70 children with TE in IBD. The characteristics of children with thrombosis during IBD are reported in Table 3. UC was significantly more frequently associated with TE events than CD (OR 3.7, 95% CI 1.8–7.6, P < 0.0001). There was no gender association. Thrombosis occurred in children of all ages, from 1 month to 18 years. Children under 12 years accounted for 38% of cases, with 11% of children being under 6 years. Three children were less than 2 years, and one case was reported in a newborn of 4 weeks of age with CD and protein-losing enteropathy without other known risk factors.27

Table 3. Characteristics of the Children with TE in IBD (N = 70)
Type of IBDCD = 22 UC = 44
IC = 3 NS = 1
  1. CD, Crohn's disease; UC, ulcerative colitis; IC, indeterminate colitis;

  2. NR, not reported.

SexF = 36 M = 33 NS = 1
Age-mean12,7 years
Age classes - years12-18: 42 (60%)
6-12: 19 (27%)
0-6: 8 (11%)
NS: 1
Disease activityActive: 58 (82.8%)
Remission: 2 (2.8%)
Recent flare: 4 (5.7%)
NS: 6 (8.5%)
IBD Duration - mean1,75 years
IBD Duration by classes>2 years: 16 (56 %)
≤2 years: 39 (23 %)
NS: 15 (21 %)
Disease site 
 - Small bowel (CD)2
 - Colon (CD)2
 - Ileocolon (CD)9
 - Associated perianal disease (CD)2
 - NS (CD)9
 - Cecum-ascending colon (UC)1
 - Rectum-sigma (UC)2
 - Pancolitis (UC)10
 - NS (UC)31
Therapy at time of thrombosisCDUC
 - Steroid922
 - Azathioprine51
 - Cyclosporine01
 - Infliximab10
 - Enteral nutrition01
 - 5-ASA412
 - Antibiotics23
 - ACTH01
 - None55
 - NR414

The interval between IBD onset and occurrence of TE was on average 1.75 years, but with a wide range (1 month to 8 years). Some cases of TE were presented in the original articles as “first symptom of IBD”, but in most cases the patient description highlighted a story of delayed diagnosis, with long-term typical IBD symptoms such as diarrhea, abdominal pains, weight loss, or vomiting, and therefore they were counted as full-blown active IBD. Overall, cases with a longer history of IBD (more than 2 years) were more frequently associated with TE than cases with shorter IBD duration (OR 5.8, 95% CI 2.5–13.6, P < 0.00001).

TE occurred during clinically active IBD in 58 (82.8%) children. Only two (2.8%) experienced TE during complete clinical remission. Of the four children with TE after a recent IBD flare, two had a concomitant acute infection (influenza, tonsil abscess).

Clinical activity scores and IBD extensions were detailed only in a minority of reports. Although with these limitations, more TE occurred in UC with pancolitis, and in CD presenting both ileal and colonic involvement.

At the time of TE occurrence, 40 (77%) of the 52 children for which this information was available were under immunosuppressive treatment. There was not enough information to establish if the treatment was adequate to the state of the patient. However, 10 children (19.2%) were not given any treatment since a diagnosis of IBD had not been made.

PROTHROMBOTIC RISK FACTORS

  1. Top of page
  2. Abstract
  3. METHODS
  4. RESULTS OF THE SEARCH STRATEGY
  5. INCIDENCE OF TE IN CHILDREN WITH IBD
  6. CHARACTERISTICS OF CHILDREN WITH TE IN IBD
  7. PROTHROMBOTIC RISK FACTORS
  8. CLINICAL CHARACTERISTIC OF TE IN CHILDREN WITH IBD
  9. CEREBRAL TE
  10. RETINIC THROMBOEMBOLISM
  11. ABDOMINAL TE
  12. UPPER AND LOWER LIMBS TE
  13. PULMONARY TE
  14. THERAPY FOR TE
  15. MORTALITY
  16. RISK OF RECURRENCE
  17. DISCUSSION
  18. REFERENCES

Risk factors for TE in children with IBD are shown in Table 4. One or more nonspecific alterations of the coagulative state were present in 51/70 (72.8%) children: 26 had thrombocytosis; five had AT III deficit; two hyperfibrinogenemia; one hyperhomocysteinemia; 16 had either PT, PTT, or INR alterations; three had low levels of antithrombin; two had activated VIII factor; one had hyperfibrinogemia; one had increased antiplasmine; one had hyperomocystinemia. More than one nonspecific alteration of the coagulative state was present in about one-third of the children.

Table 4. Procoagulative Risk Factors in Children with TE and IBD (N = 70)
 Risk Factors
  1. MTHFR, methylenetetrahydrofolate reductase.

Coagulation disorders 
 - Specific disorders18
  - MTHFR mutation5
  - Prothrombin gene mutation3
  - Protein C deficit3
  - Factor V Leiden heterozygote mutation3
  - Protein S deficit2
  - PAI mutation1
  - Anti-phospholipids antibodies2
 - Other unspecific disorders51
General risk factors28
 - Surgery8
 - Immobilisation5
 - Infection5
 - Endocarditis or myocarditis2
 - Hypertension2
 - Central venous catheter2
 - Dehydration1
 - Oral contraceptive1
 - Obesity1
 - Tagayasu vasculitis1

Inherited thrombophilia was properly investigated only in a minority of cases. With this limitation, 13/70 (18.6%) children were recognized as having one or more specific coagulation disorders.

General risk factors for TE (one or more) were reported in 28/70 (40%) children. The main general risk factors were either recent surgery or immobilization, reported in 13/70 (18.5%) children. Overall, 7/70 (10%) children had an acute infection at the time of TE occurrence. A central venous catheter was in place in only two cases (a third case had a central catheter at the time of TE recurrence). Smoking was not reported in any case.

Overall, a known risk factor for TE, such as an inherited coagulative disorder, or a generally recognized risk factor were observed in 35/70 (50%) children. Two concomitant specific risk factors were reported in 17/70 (24.2%) children.

CLINICAL CHARACTERISTIC OF TE IN CHILDREN WITH IBD

  1. Top of page
  2. Abstract
  3. METHODS
  4. RESULTS OF THE SEARCH STRATEGY
  5. INCIDENCE OF TE IN CHILDREN WITH IBD
  6. CHARACTERISTICS OF CHILDREN WITH TE IN IBD
  7. PROTHROMBOTIC RISK FACTORS
  8. CLINICAL CHARACTERISTIC OF TE IN CHILDREN WITH IBD
  9. CEREBRAL TE
  10. RETINIC THROMBOEMBOLISM
  11. ABDOMINAL TE
  12. UPPER AND LOWER LIMBS TE
  13. PULMONARY TE
  14. THERAPY FOR TE
  15. MORTALITY
  16. RISK OF RECURRENCE
  17. DISCUSSION
  18. REFERENCES

This review identified 70 reported children with TE in IBD. In addition, 8/70 (11.4%) children experienced a second TE event during the same hospitalization, while 7/70 (10%) experienced one or more separate TE recurrences. Therefore, a total of 92 TE events are described in this review. The localization of TE events was as follows: 50/92 (54.3%) cerebral, 24/92 (26%) limbs, 12/94 (13%) abdominal, 3/92 (3.3%) retinal, 3/92 (3.3%) pulmonary. Venous TE accounted for 67/92 cases (72.8%), arterial TE for 10/92 (9.8%), while two cases (2.3%) were referred as mixed arterovenous, and 13 were not specified. Arterial TE occurred in cerebral arteries (six cases), retinal arteries (three cases), and colonic vessels (one case).

CEREBRAL TE

  1. Top of page
  2. Abstract
  3. METHODS
  4. RESULTS OF THE SEARCH STRATEGY
  5. INCIDENCE OF TE IN CHILDREN WITH IBD
  6. CHARACTERISTICS OF CHILDREN WITH TE IN IBD
  7. PROTHROMBOTIC RISK FACTORS
  8. CLINICAL CHARACTERISTIC OF TE IN CHILDREN WITH IBD
  9. CEREBRAL TE
  10. RETINIC THROMBOEMBOLISM
  11. ABDOMINAL TE
  12. UPPER AND LOWER LIMBS TE
  13. PULMONARY TE
  14. THERAPY FOR TE
  15. MORTALITY
  16. RISK OF RECURRENCE
  17. DISCUSSION
  18. REFERENCES

Cerebral TEs were the most frequently reported TE in children with IBD. For the diagnosis, computed tomography was used in 26% of cases, magnetic resonance in 60%, arteriography in 15%. Out of the 50 reported cerebral TE, 31 (62%) were venous TE, six (12%) were arterial TE, one was mixed, and the remaining 12 were unspecified. The lateral sinus and the superior sagittal sinus were the most frequent venous sites (48% and 35.4%, respectively). Arterial cerebral TE occurred in the middle cerebral artery (three), basilar artery (two), and posterior cerebral artery (one).

Clinical manifestation usually included a mix of signs and symptoms. Notably, seizure occurred as an isolated sign in 6/50 (12%) of children. Although half of the children had a headache at the time of TE, this was an isolated symptom only in one case. Hemiplegia or hemiparesis were the only presenting signs of TE in two children, while they were associated with other manifestations in 49% of cases. Sensitive signs (hemiparesthesias, paresthesias, propioceptive defects) occurred in 12% of cases, but were never isolated. Other signs included speech disorders and nistagmus. Impaired consciousness varied from obtundation to rapidly progressing coma, and was mostly associated with other signs (26%).

Complete recovery was observed in 17/50 (34%), partial recovery in 16/50 (32%), and death in 3/50 (6%). About two-thirds of residual deficit involved motricity, with seizures, speech disorders, learning difficulties, pseudotumor, and emotional disorders being the other deficits.

RETINIC THROMBOEMBOLISM

  1. Top of page
  2. Abstract
  3. METHODS
  4. RESULTS OF THE SEARCH STRATEGY
  5. INCIDENCE OF TE IN CHILDREN WITH IBD
  6. CHARACTERISTICS OF CHILDREN WITH TE IN IBD
  7. PROTHROMBOTIC RISK FACTORS
  8. CLINICAL CHARACTERISTIC OF TE IN CHILDREN WITH IBD
  9. CEREBRAL TE
  10. RETINIC THROMBOEMBOLISM
  11. ABDOMINAL TE
  12. UPPER AND LOWER LIMBS TE
  13. PULMONARY TE
  14. THERAPY FOR TE
  15. MORTALITY
  16. RISK OF RECURRENCE
  17. DISCUSSION
  18. REFERENCES

The three cases of retinal TE were arterial cases. Symptoms included acute loss of vision, hemianopsia, proptosis, and headache. Two children experienced a cerebral infarction either before or after the retinal TE.12, 28 Outcome was reported for two children: one progressed to complete blindness, the second improved after acute impairment.

ABDOMINAL TE

  1. Top of page
  2. Abstract
  3. METHODS
  4. RESULTS OF THE SEARCH STRATEGY
  5. INCIDENCE OF TE IN CHILDREN WITH IBD
  6. CHARACTERISTICS OF CHILDREN WITH TE IN IBD
  7. PROTHROMBOTIC RISK FACTORS
  8. CLINICAL CHARACTERISTIC OF TE IN CHILDREN WITH IBD
  9. CEREBRAL TE
  10. RETINIC THROMBOEMBOLISM
  11. ABDOMINAL TE
  12. UPPER AND LOWER LIMBS TE
  13. PULMONARY TE
  14. THERAPY FOR TE
  15. MORTALITY
  16. RISK OF RECURRENCE
  17. DISCUSSION
  18. REFERENCES

Of the 12 cases of abdominal TE, 11 were venous (five Budd–Chiari syndromes,29–32, 58 three portal veins,22, 29, 33 two superior mesenteric vein,33, 34 one splenic vein29) and one arterial (multiple colon vessels).31 Abdominal thrombosis in children with IBD occurred significantly later in life compared to cerebral thrombosis (mean age 15.1 ± 2.2 years for abdominal TE versus 11.7 ± 4.8 years for cerebral TE, P = 0.01). Reported symptoms included: pain, ascitis, hepatomegaly, abdominal distension, bloody stools, vomiting, fever, and elevation of transaminases. Four cases were treated surgically. Overall, six children (50%) fully recovered, one child with concomitant sepsis died.

UPPER AND LOWER LIMBS TE

  1. Top of page
  2. Abstract
  3. METHODS
  4. RESULTS OF THE SEARCH STRATEGY
  5. INCIDENCE OF TE IN CHILDREN WITH IBD
  6. CHARACTERISTICS OF CHILDREN WITH TE IN IBD
  7. PROTHROMBOTIC RISK FACTORS
  8. CLINICAL CHARACTERISTIC OF TE IN CHILDREN WITH IBD
  9. CEREBRAL TE
  10. RETINIC THROMBOEMBOLISM
  11. ABDOMINAL TE
  12. UPPER AND LOWER LIMBS TE
  13. PULMONARY TE
  14. THERAPY FOR TE
  15. MORTALITY
  16. RISK OF RECURRENCE
  17. DISCUSSION
  18. REFERENCES

Of the 24 reported TE in limbs, all were venous, except for one child with an arterial TE of a finger.16 Nineteen occurred in lower limbs, five in upper limbs. Six cases occurred as complications of a recent TE in a different site. All children fully recovered, except one, for which the follow-up was only 3 weeks.35

PULMONARY TE

  1. Top of page
  2. Abstract
  3. METHODS
  4. RESULTS OF THE SEARCH STRATEGY
  5. INCIDENCE OF TE IN CHILDREN WITH IBD
  6. CHARACTERISTICS OF CHILDREN WITH TE IN IBD
  7. PROTHROMBOTIC RISK FACTORS
  8. CLINICAL CHARACTERISTIC OF TE IN CHILDREN WITH IBD
  9. CEREBRAL TE
  10. RETINIC THROMBOEMBOLISM
  11. ABDOMINAL TE
  12. UPPER AND LOWER LIMBS TE
  13. PULMONARY TE
  14. THERAPY FOR TE
  15. MORTALITY
  16. RISK OF RECURRENCE
  17. DISCUSSION
  18. REFERENCES

Only three cases of pulmonary TE in children with IBD were retrieved.14, 36, 37 All of them also had cerebral TE. The prognosis of two of the cases is not reported, while one child with cerebral TE and pulmonary embolism died from concomitant probable meningitis.

THERAPY FOR TE

  1. Top of page
  2. Abstract
  3. METHODS
  4. RESULTS OF THE SEARCH STRATEGY
  5. INCIDENCE OF TE IN CHILDREN WITH IBD
  6. CHARACTERISTICS OF CHILDREN WITH TE IN IBD
  7. PROTHROMBOTIC RISK FACTORS
  8. CLINICAL CHARACTERISTIC OF TE IN CHILDREN WITH IBD
  9. CEREBRAL TE
  10. RETINIC THROMBOEMBOLISM
  11. ABDOMINAL TE
  12. UPPER AND LOWER LIMBS TE
  13. PULMONARY TE
  14. THERAPY FOR TE
  15. MORTALITY
  16. RISK OF RECURRENCE
  17. DISCUSSION
  18. REFERENCES

Several different therapeutic schemes were used (Table 5). In particular, 21% of children did not receive any specific treatment and 7% received only symptomatic treatment. More than one specific drug was given to 25% of children, while 7% underwent surgery.

Table 5. Therapy in Children with TE in IBD (N=70)
 Children (% of Total)
  1. LMWH, low molecular weight heparin; UFH, unfractionated heparin;

  2. NR, not reported.

Only ASA6 (8.5)
Only LMWH/UHF6 (8.5)
LMWH/UHF + other18 (25)
 + warfarin14
 + warfarin+ thrombolytic2
 + thrombolytic1
 + Ticlopidine+ tetracosactide1
Thrombolytic + other1 (1.4)
Surgery5 (7)
Only mannitol or other diuretics2 (2.8)
Only anticonvulsants2 (2.8)
Only antibiotics1 (1.4)
None15 (21)
NR14 (20)

When only cases published after 2000 were considered, no significant changes were observed in the type of therapy administered to children with TE, except for the use of thrombolytic agents.

MORTALITY

  1. Top of page
  2. Abstract
  3. METHODS
  4. RESULTS OF THE SEARCH STRATEGY
  5. INCIDENCE OF TE IN CHILDREN WITH IBD
  6. CHARACTERISTICS OF CHILDREN WITH TE IN IBD
  7. PROTHROMBOTIC RISK FACTORS
  8. CLINICAL CHARACTERISTIC OF TE IN CHILDREN WITH IBD
  9. CEREBRAL TE
  10. RETINIC THROMBOEMBOLISM
  11. ABDOMINAL TE
  12. UPPER AND LOWER LIMBS TE
  13. PULMONARY TE
  14. THERAPY FOR TE
  15. MORTALITY
  16. RISK OF RECURRENCE
  17. DISCUSSION
  18. REFERENCES

Outcomes were specified in only 41/70 (59%) children. With this limitation, mortality was reported in 4/70 (5.7%) children with TE. Three of these children had cerebral TE (one with pulmonary embolism). The fourth child had an abdominal TE with concomitant sepsis.

RISK OF RECURRENCE

  1. Top of page
  2. Abstract
  3. METHODS
  4. RESULTS OF THE SEARCH STRATEGY
  5. INCIDENCE OF TE IN CHILDREN WITH IBD
  6. CHARACTERISTICS OF CHILDREN WITH TE IN IBD
  7. PROTHROMBOTIC RISK FACTORS
  8. CLINICAL CHARACTERISTIC OF TE IN CHILDREN WITH IBD
  9. CEREBRAL TE
  10. RETINIC THROMBOEMBOLISM
  11. ABDOMINAL TE
  12. UPPER AND LOWER LIMBS TE
  13. PULMONARY TE
  14. THERAPY FOR TE
  15. MORTALITY
  16. RISK OF RECURRENCE
  17. DISCUSSION
  18. REFERENCES

Apart from the 8/70 (11.4%) children with early TE recurrence within the same hospitalization,12, 15, 18, 30, 36–38 7/70 (10%) experienced a separate TE recurrence,17, 18, 20, 28, 29, 40 with two presenting more than one recurrence.29, 40 The interval between the first and the second episode of TE ranged from 1 month to some years. Sites of recurrence varied. All but one of the children with recurrence had risk factors for TE.

DISCUSSION

  1. Top of page
  2. Abstract
  3. METHODS
  4. RESULTS OF THE SEARCH STRATEGY
  5. INCIDENCE OF TE IN CHILDREN WITH IBD
  6. CHARACTERISTICS OF CHILDREN WITH TE IN IBD
  7. PROTHROMBOTIC RISK FACTORS
  8. CLINICAL CHARACTERISTIC OF TE IN CHILDREN WITH IBD
  9. CEREBRAL TE
  10. RETINIC THROMBOEMBOLISM
  11. ABDOMINAL TE
  12. UPPER AND LOWER LIMBS TE
  13. PULMONARY TE
  14. THERAPY FOR TE
  15. MORTALITY
  16. RISK OF RECURRENCE
  17. DISCUSSION
  18. REFERENCES

This review is an attempt to better describe the phenomenon of TE in children with IBD. We used a wide search strategy in several databases without language restrictions. Nevertheless, only two cohort studies, five single-hospital database reviews, eight case series, and 36 case reports were identified. Therefore, the first limitation of this review is the scantiness of published cases. Underreporting and selective reporting of severe cases (cerebral TE) are the probable biases of this review. Nevertheless, this series of 92 cases of TE in children with IBD provides an interesting basis for new hypotheses of research. The second limitation of this review is that diagnostic accuracy, treatment strategies, and investigations for thrombophilic disorders have improved over time. All these factors cause heterogeneity in the reports.

Single-center retrospective hospital database analysis estimated a risk of vascular complication in children with IBD ranging from 1.4%–3.3%.14 Retrospective hospital database analysis, however, represent a cumulative incidence over a long observation period and may therefore be biased by patient selection in specialized IBD centers, refer to relatively small sample size, and lack a control group. Thus, single-center series may falsely elevate the prevalence of TE. Based on the larger cohort available, an incidence of venous TE of ≈0.5% of hospitalized children with IBD is estimated, with a 4–20-fold increased risk compared to children without IBD.7 However, more large-population studies are needed to confirm these data.

Overall, several of the findings of this review in children are consistent with what has been previously observed in adults.61 This review highlighted a higher incidence of TE in patients with UC compared to those with CD (OR 3.7, P < 0.0001), which is consistent with what is reported in adults,7, 8, 11 although with some exceptions.6 Both in children and adults there is no gender association.7–8

TE during IBD occurred in children with IBD in all age ranges, while in the general population the most affected are newborns and teenagers.4 Although children with a longer history of IBD had a greater risk (OR 5.8, P < 0.00001) TE was observed also soon after IBD onset, and the history of delay in diagnosis suggested in some reports highlights the need for early recognition of the signs of IBD in children.

Only two cases of TE in children occurred during nonactive IBD. Recently, a large cohort study based on the UK general practitioner database found that adults with IBD are at increased risk of venous TE compared to the general population even during the chronic phases (HR 6.5, 95% CI 4.6–9.2, P < 0.0001).5 Protocols of IBD therapy differ between adults and children, as in the latter the chronic phases are less tolerated. Also, the prevalence of IBD-associated risk factors such as surgery and malignancies are lower in children compared to adults. Even if the evidence is insufficient to show conclusively that children without active IBD do not have an increased risk of TE compared to the general population, the cases reported so far support this hypothesis.

Two children experienced thrombosis soon after a recent flare of IBD, during an acute infection. Other studies showed that acute infections, even after adjustment for immobilization, are associated with a transient increased risk of venous TE.62 In a review of over 200 children with ischemic stroke, about a third presented coryza, cough, or flu-like symptoms.63 Infection may augment the risk of TE by directly activating the inflammation cascade, and by inducing a reactivation of IBD, although this last hypothesis is controversial.64

Colonic involvement has been previously considered as a predisposing factor for venous TE in adults with UC65 and CD,7 even though pancolitis was not associated with a statistically significant higher rate of venous TE.7 Unfortunately, data on the endoscopic severity are lacking and it is not fully clear if disease localization and extension are independent risk factors for TE. Similarly, the literature does not clarify whether the presence of abscesses and stenoses are independent risk factors with an additive effect over disease activity, while in a large population study active abdominal fistulas were independently associated with greater likelihood of venous TE (OR 1.39, 95% CI 1.13–1.70).7

Steroids are generally considered as prothrombotic agents. More recently tumor necrosis factor alpha (TNF-α) blockers have been associated with cases of venous TE without other risk factors.61, 66 However, steroids and TNF-α blockers can also play a major role in controlling IBD activity, which is considered the major determinant in the genesis of TE. So far there are insufficient data to further assess the role of these drugs in the genesis of TE in IBD patients. Given their possible effectiveness in controlling inflammation, it is probable that their relative role in the genesis of TE depends on the balance of benefits and risks, and varies individually among IBD patients.

In this review, 18% of children with TE during IBD presented a congenital procoagulative disorder. A recent meta-analysis found a statistically significant association with venous TE for each inherited trait evaluated.67 A second meta-analysis on arterial ischemic stroke showed that all inherited coagulation disorders examined were more common in children with TE than in controls.68 Overall, an inherited thrombophilic disorder was detected in 36% of all children with venous TE,69 in 44% of children with ischemic stroke,68 and in 33% of patients with IBD and venous TE.61 This review found an inherited coagulation disorders in 18% of children with IBD, but this can easily represent an underestimation of the real prevalence, as children where not systematically investigated for thrombophilia. In the published literature, although with some discordances, the majority of the studies observed that the prevalence of inherited thrombophilia neither differs in IBD compared to healthy controls nor in patients with TE with or without IBD.69

In this review, 50% of children had no other risk factor for TE than IBD. This differs from what was observed in prospective TE registers in the general population of children, where at least one risk factor was present in 90% of cases3, 4 and reinforces the hypothesis that children with IBD have per se a risk of TE, although in many case they also present additional risk factors.61 The mechanisms of the thrombotic tendency in IBD relate to systemic inflammation and include: 1) upregulation of procoagulants such as factors V, VII, VIII, prothrombin, fibrinogen, thromboplastin, and inhibition of fibrinolysis61; 2) increased tissue factor induction11, 61, 70; 3) alterations of the intestinal vasculature such as microvascular stenosis, diminished intestinal blood flow, increased leukocyte and platelet aggregation leading to thrombosis, ischemia, and infarction11, 68, 70; 4) association with other autoimmunities such as anticardiolipin (aCL) antibodies.61 A fifth mechanism might involve the metabolic changes associated with IBD: deficiencies of vitamin B12, B6, and folic acid; hyperhomocysteinemia (either genetically caused, or due to nutritional deficiencies, or to the use of sulfasalazine and methotrexate); anemia (risk factor for cerebral ischemia), changes in the lipid spectrum.61, 69, 70 Finally, there is an association between IBD and general risk factors for TE such as smoking and the use of oral contraceptives. Surgical treatment, prolonged immobilization, fluid depletion, and central venous catheters are also associated with IBD as a consequence of the disease.61 Having undergone an IBD-related bowel surgery has been shown to augment the risk of venous TE in adults (the rate of venous TE increased from 19.5–42.3 per 1000 after colectomy in UC).7 In this review general risk factors were reported in the 25.7% of cases of TE and, among these, surgery, immobilization, and infections were the most frequent.

What is still not clear in the literature is the relative weight of the different risk factors in the genesis of TE in patients with IBD, how patients should be scored for risk, and if pharmacological prophylaxis should be considered for children.

According to published cases, it appears that cerebral sites are the most common sites of TE in children with IBD. This differs from what has been reported in adults, where the most common TE manifestations in IBD were venous TE of the legs and pulmonary embolism,5–7, 61 although unusual sites, including intracardiac, cerebral, retinal, hepatic, mesenteric, and unidentified veins were also affected.7, 61 Large epidemiological studies are needed before confirming that cerebral TE are more frequent in children with IBD compared to the general population of children, and to study the determinants of these events.

This review reports a 5.7% mortality rate overall in children with TE in IBD, incomplete recovery in 34% of children with cerebral TE, early recurrence in 11%, and late recurrence in 10%. Treatment was extremely heterogeneous, as observed by others both in adults with IBD and in the general population of children.3, 61, 71 These data underline the importance of guidelines for diagnosis, treatment, and prevention of TE in children with IBD.

Pharmacological prophylaxis is recommended in some guidelines on IBD management in adults. The British Society of Gastroenterology recommends subcutaneous heparin for all severe UC,72 whereas the American College of Chest Physicians suggests heparin for all nonambulating hospitalized IBD patients, or graduated compression stockings or intermittent pneumatic compression in the case of contraindications to anticoagulants.4 Pharmacological prophylaxis for all nonhospitalized adults is questioned, as more than 300 patients would need to be treated to prevent one venous TE event.73

At the moment, there are no specific guidelines for the management of TE in children with IBD. Guidelines on the management of TE and stroke in the general pediatric population have been published recently by the United Kingdom Royal College of Physicians (UK-RCP),74 the American College of Chest Physicians (ACCP),75, 76 and the American Heart Association (AHA).77 The pediatric literature on TE lacks systematic randomized trials to establish an evidence-based management of TE, and the guidelines diverge in some of their recommendations.77 Regarding acute treatment, in cerebral sinovenous thrombosis there is agreement among the guidelines to treat non-neonates with anticoagulants.75–77 In the case of acute ischemic stroke, AHA and ACCP76, 77 recommend initial treatment with unfractionated heparin (UFH) or low molecular weight heparin (LMWH) until cardioembolism and craniocervical dissection are ruled out. This differs from the UK-RCP guidelines, which advise starting with aspirin 5 mg/kg/die.74 All guidelines agree on not recommending fibrinolytic therapy with tPA or alteplase for childhood arterial ischemic stroke, except for cases occurring within 3 hours after stroke.79

As prevention of stroke, AHA77 and ACCP76 recommend anticoagulation therapy only for selected prothrombotic states, although IBD is not specifically cited; UK-RCP74 advises referral to a hematologist. Thrombophilic screening and alternatives to oral contraception are recommended only after a first TE event,77 and they are not specifically discussed for IBD patients. In children with central venous lines, the CHEST guideline advises against the use of routine systemic thromboprophylaxis, while for children receiving long-term home total parenteral nutrition thromboprophylaxis with vitamin K antagonist is advised.75

Recently, it has been suggested starting LMWH in children with IBD “and additional risk,” as inpatient/immobilized, thrombophilia, thrombocytosis >750/mm3, or severe disease (Pediatric Crohn's Disease Activity Index >30, or Pediatric Ulcerative Colitis Disease Activity Index >45). Outpatient pharmacological prophylaxis (e.g., dypiridamol) was suggested for selected “additional risk” patients.15

Some general recommendations on the management of TE in children with IBD are given in Table 6. Based on the existing body of research it is difficult to provide more specific advice for pharmacoprophylaxis in children. Although some of the above-reported advice may be valid, there are many questions still open to discussion. One major issue regards the safety of heparin in severe UC, which still needs to be adequately assessed in patients at higher risk for TE, since most of the trials on heparins in IBD excluded severe cases of UC.80

Table 6. Key Points for the Management of TE in Children with IBD
1) Diagnose IBD in children promptly.
2) Control IBD activity with an adequate therapy.
3) Perform a risk assessment for TE in all children with IBD:
 • Family history: previous TE, recurrent miscarriage after the first semesters, heart attack or stroke before the age of 50 years, autoimmunities.
 • Individual risk factors: central catheter, smoke, oral contraceptive, use of prothrombotic drugs, surgery/trauma, malnutrition/malabsorption, cardiac diseases, nephrotic syndrome, malignancies, obesity, sickle cell disease, and other intercurrent risk factors as immobilization, infection, dehydration (see point 5).
4) Consider screening for inherited thrombophilia if there is a family history of TE, or other risk factors.
5) Educate patients and prevent risk factors for TE:
 • Stop smoking.
 • Consider alternative contraceptive methods (non-ormonal or progestin-only).
 • Reduce immobilization, use support stockings.
 • Treat dehydration.
 • Treat nutritional deficiencies.
 • Treat infections.
 • Treat associated conditions.
6) Considered pharmacoprophylaxis when one or more risk factors are present.
7) Consider mechanical prophylaxis when there are contraindications to anticoagulants.
8) Keep in mind a possible diagnosis of TE in children with IBD.

More epidemiological studies are needed to understand the real incidence of TE events in children with IBD, to define the relative weight of the different risk factors, to define the characteristics of the events, and to evaluate the safety profile of anticoagulants, including new agents, in severe IBD cases. Pediatric registers of IBD cases could have a role in prospectively collecting this type of information.

REFERENCES

  1. Top of page
  2. Abstract
  3. METHODS
  4. RESULTS OF THE SEARCH STRATEGY
  5. INCIDENCE OF TE IN CHILDREN WITH IBD
  6. CHARACTERISTICS OF CHILDREN WITH TE IN IBD
  7. PROTHROMBOTIC RISK FACTORS
  8. CLINICAL CHARACTERISTIC OF TE IN CHILDREN WITH IBD
  9. CEREBRAL TE
  10. RETINIC THROMBOEMBOLISM
  11. ABDOMINAL TE
  12. UPPER AND LOWER LIMBS TE
  13. PULMONARY TE
  14. THERAPY FOR TE
  15. MORTALITY
  16. RISK OF RECURRENCE
  17. DISCUSSION
  18. REFERENCES