Hereditary hemorrhagic telangiectasia (HHT) is a dominant autosomic vascular dysplasia with an estimated prevalence of 1 in 5000.1 Almost 50% of the HHT patients have pulmonary arteriovenous malformations (PAVM) and the development of embolic complications has been widely described.1,2

No data have been published on HHT patients in Argentina. Moreover, it is not clear if poor outcomes with embolic complications are related to PAVM severity or pulmonary symptoms.

This cross-sectional study describes the clinical characteristics of patients with HHT referred to our teaching hospital in Buenos Aires, Argentina. We also evaluated the association between significant PAVM, pulmonary symptoms and embolic complications.

A cross-sectional study was performed on data from the institutional records of HHT in the Hospital Italiano, a tertiary teaching hospital in Buenos Aires, Argentina.3 A total of 108 consecutive patients were evaluated in the HHT unit between 2010 and 2012 and included in the study after informed consent was obtained. Participants had a definitive clinical diagnosis of HHT (defined as three or more Curacao criteria: epistaxis, telangiectasia, visceral arteriovenous malformation [AVM] or family history). The study was approved by the internal review committee of the Hospital Italiano.

Significant PAVM (exposure) was defined as the presence of at least one of the following factors: contrast echocardiography grade 2 or greater,4 bilateral PAVM or feeding artery >3mm, or previous PAVM treatment. The primary composite outcome was defined as: cerebrovascular accident, transient ischemic accident, brain abscess or peripheral embolism.

The baseline variables were compared, depending on the presence or absence of significant PAVM. A t-test was used for unpaired data, a Wilcoxon rank test for continuous variables and the Pearson χ2 for discrete variables. The association between significant PAVM (or pulmonary symptoms) and embolic complications was evaluated using Fisher's exact test. A logistic regression was used to adjust for the effects of potential confounding factors.

Finally, to evaluate the possible bias derived from missing PAVM data, a conservative sensitivity analysis was performed assuming that all patients from the missing data group had significant PAVM, and the change in prevalence of the primary outcome variable in this hypothetical situation was evaluated.

The patients’ baseline characteristics are summarized in Table 1. A total of 35 patients (39.8%) had a significant PAVM. Embolic complications occurred in 17 participants (19.3%); the most common complication was stroke.

The characteristics of patients with or without significant PAVM are shown in Table 2. Patients with PAVM were significantly younger (P=.01), had more pulmonary symptoms (60% versus 11%, P<.001) and more embolic complications (34.3% versus 9.4%, P=.006). This difference remained significant in the sensitivity analysis. Furthermore, patients with pulmonary symptoms had a significantly higher rate of embolic events compared to patients without pulmonary symptoms (91.7% versus 8.3%, P=.003).

In the unadjusted analysis, the odds ratio (OR) for embolic complications in patients with significant PAVM was 6.3 (95% CI 1.8–21.8). In the final model, that included age, sex, age at onset and anemia, the OR rose to 7 (95% CI 1.8–27.3).

Patients with HHT seen in our hospital are relatively young. Onset occurs at an early age and their quality of life is moderately affected. Almost 25% of patients with PAVM have pulmonary symptoms, and up to 20% suffer embolic complications. Compared to previous studies in other geographical areas, the age at onset of patients seen in our hospital is younger and there is a greater prevalence of AVM, but they also have several characteristics in common.

Our study shows that patients with significant PAVM have a higher risk of developing embolic complications than those without significant PAVM. This association is marginally influenced by the covariables of age, sex, anemia and age at onset. While Shovlin et al. suggested that pulmonary symptoms were not related with embolic complications,5 in our study these symptoms were closely associated with such complications. However, the data may be biased, since patients with PAVM and embolic complications may be more likely to describe pulmonary symptoms.

The strengths of the study lie in the fact that it is the first of its kind to address the issue of HHT patients in Argentina, and a close association was detected between significant PAVM, pulmonary symptoms and embolic complications. This can be added to the body of evidence underlining the importance of PAVM in poor clinical outcome. On the other hand, the study also has some limitations. Firstly, the use of electronic clinical records means that variable may be measured and handled differently between the two groups, making them incomparable. Moreover, missing data on exposure may lead to the exclusion of patients that differ in some way from those finally included. Another possible limitation is the fact that patients with grade 2 echocardiography were classified as cases of significant PAVM even though they appear to constitute a heterogeneous group that could include patients with or without significant PAVM. Finally the initial selection of patients may undermine the external validity of our study, since patients referred to our tertiary hospital may be more likely to have a serious disease.

To conclude, significant PAVM and pulmonary symptoms may be associated with poor outcome in terms of embolic complications, a finding that would justify further prospective studies.

FA: Study design, data collection and analysis. Preparation of preliminary manuscript.

BLF: Study design, data collection and analysis. Preparation of preliminary manuscript.

EJW: Study design, data interpretation. Manuscript review.

MMS: Study design, baseline patient evaluation and data interpretation. Manuscript review.

All authors read and approved the final manuscript.

The authors declare no conflict of interests.