In the last scientific update (June 2001 Issue), screening for pulmonary arteriovenous malformations (PAVMs) was updated. As discussed in the June Issue, the two articles from the Danish and Canadian Centers validated the echobubble technique as a more sensitive test (”almost” 100 % of HHT patients with PAVMs detected), compared with other screening methods used prior to 1999.

In this Issue we will review a more controversial article and subject, i.e. screening for cerebral arteriovenous malformations (CAVM). For purposes of this discussion, the CAVM is limited to the brain and less frequently the spinal cord. Our Scientific and Medical Advisory Board is still divided about whether to screen all patients with HHT for CAVM. The five North American Centers are screening for CAVM at least once in adulthood and earlier in children if there is a family history of CAVM or brain hemorrhage. These recommendations differ from some of the Centers in Europe so there is not a complete consensus as yet. Much of the discussion, pro and con about the value of brain screening, indications for treatment, and outcomes are discussed in the Scientific Update (September 2000 Issue).

What no one will contest is that brain screening with modern Magnetic Resonance Imaging (MRI) is quite precise! What is not clear as yet is what size CAVM poses a risk for hemorrhage? What size CAVM needs treatment? What type of treatment is best for the CAVM associated with HHT? Finally, what is the outcome of treatment or lack of treatment?

Title:Cerebrovascular Manifestations in 321 Cases of Hereditary Hemorrhagic Telangiectasia.
Authors: Maher, CO, Piepgras, DG, Brown, RD, Friedman, JA, Pollock, BE.
Journal: Stroke 2001; 32:877-882.

The authors reviewed retrospectively, 321 consecutive patients with HHT evaluated at the Mayo Clinic over the past 20 years. Twelve (12) patients, representing 3.7% of the 321 patients evaluated, had symptomatic CAVM (hemorrhage or seizure). Unfortunately, we are not provided with how many of the 321 patients had family members who were not seen or, who died or were disabled by brain hemorrhage. Only the patients who were seen in person in their clinic were examined and brain imaging was performed in only 46 of the 321 patients. This paper led us to submit a letter to the Editor of the Journal, Stroke. This letter is printed in prepublication form below:

COMPELLING REASONS TO SCREEN BRAIN IN HHT - SUBMITTED AS A LETTER TO STROKE 2001 BY:

J. Mandzia, Ph.D. Candidate, K. Henderson, M.S., M. Faughnan, M.D., R.I. White, Jr., M.D.

Compelling Reasons to Screen Brain in HHT
To the Editor:

We read with great interest the article by Maher et al. (2001): Cerebrovascular Manifestations in 321 Cases of Hereditary Hemorrhagic Telangiectasia (HHT) (1). The authors reviewed the medical records from HHT patients seen at their institution over 20 years. However, we disagree with the authors’ final conclusion that HHT patients should not undergo routine screening for cerebrovascular malformations (CVMs) if asymptomatic. We advocate a single baseline magnetic resonance imaging (MRI) screen in all patients with HHT, as CVMs can cause devastating neurological complications.
Maher et al.’s finding of a prevalence rate of CVMs of 12/321 (3.7%) in their patient group is rather low in comparison to other studies (2,3). The authors acknowledge this and attribute it to screening only symptomatic patients. We believe that neurological symptoms are not a good indicator of who should receive MRI screening and based on our experience we have found that neurological symptoms are not necessarily present in individuals with CVMs, PRIOR TO HEMORRHAGE (2-4). The consequences of a CVM related hemorrhage could be devastating.
Other studies have described CVMs in HHT and have reported micro arteriovenous malformations (AVMs), small AVMs and macro AVMs, as well as cerebral telangiectases and cavernomas (2-4). Fulbright et al. and Willemse et al. who used non-contrast and gadolinium enhanced MRI and digital subtraction angiography (DSA) respectively found a prevalence of CVMs of 22 % and 12.2 %. The difference in prevalence rates between these two studies may be due to increased detection of other CVMs such as micro AVMs and cerebral telangiectases found on contrast enhanced MRI in the Fulbright et al. study. However, in terms of AVMs (� 10 mm in diameter) which are predisposed to symptoms, both studies respectively found a prevalence rate of 11-12 % in consecutive patients with HHT imaged by these techniques. In Maher et al’s retrospective analysis only 46 of 322 patients had MRI imaging. If Maher et al. had performed MR imaging of all their patients, we would have predicted that at least 32/321 (10%) of their patients rather than the 12 that were symptomatic, would have had a CVM, which could potentially cause symptoms (2-4).
The Maher et al. study did not include the clinical characteristics of their patient population or a detailed pedigree analysis of the family members. However, patients did fill out a questionnaire, which included family history questions, but the findings were not reported in the paper. The authors only included living patients therefore, likely failed to report deaths due to CVMs in family members reducing the actual prevalence of CVMS in their group (survivor bias). A subset of HHT families may be more prone to harbouring CVMs and there is some evidence that suggests that families who have a high prevalence of PAVMs may have an increased likelihood of also having a CVM (5). The increased prevalence rate of CVMs in the other studies previously discussed (2, 3) may be related to this finding, as both centres have a strong referral bias for PAVMs. In the future, genotype-phenotype analyses will advance our understanding of the relationship of PAVMs and CAVMs.
There are still many unanswered questions with regards to CVMs and screening in HHT. The current gold standard for CVM screening is four-vessel angiogram, but it remains too costly and invasive from being a standard screening technique. To better understand the relationship between CVMs and HHT, we need to combine data from multiple centres using our current available screening techniques such as non-contrast and contrast enhanced MRI, to determine the actual prevalence of CVMs in HHT and to characterize the morphology of HHT related CVMs. Studies such as Matsubara et al. have characterized CVMs in HHT in a small group of patients, but pooled centre data will further expand this knowledge and also permit to characterize prospectively the risk of micro, small and macro AVMs in a large, unselected group of HHT patients (4).
Maher et al.’s findings support those of Willemse et al. who found that HHT CVMs have a lower risk of hemorrhage than sporadic CVMs, but these findings alone are not persuasive enough to suggest that routine screening in HHT patients should not be practiced. In light of this, we feel strongly that all HHT patients should undergo baseline MRI screening, as neurological complications can be devastating. The risk of neurological events in HHT patients outweighs the costs generated by prospective screening. Family members of families who have had a serious neurological event should be given priority for screening. Pooled centre data will also permit development of guidelines for treatment of specific aspects of HHT related CVMs, such as their multiplicity and size. The goal of MRI screening in HHT is to prevent life threatening complications from CVMs.
References

1. Maher C.O., Piepgras D.G., Brown R.D., Friedman J.A., Pollock B.E. Cerebrovascular manifestations in 321 cases of Hereditary Hemorrhagic Telangiectasia. Stroke 2001; 32: 877-882
2. Willemse R.B., Mager J.J., Westerman C.J.J., Overtoom T.T.C., Mauser H., Wolberg J.G. Bleeding risk of cerebrovascular malformations in hereditary hemorrhagic telangiectasia. J Neurosurg 2000; 92: 779-784
3. Fullbright R.K., Chaloupka J.C., Putman C.M., Sze G.K., Merriam M.M, Lee G.K., Fayad P.B., Awad I.A., White R.I. MR of Hereditary Hemorrhagic Telangiectasia: Prevalence and Spectrum of Cerebrovascular Malformations. AJNR Am J Neuroradiol 1998; 19:477-484
4. Matsubara S., Mandzia J.L, ter Brugge K, Willinsky R.A., Montanera W, Faughnan M.E. Angiographic and clinical characteristics of patients with cerebral arteriovenous malformations associated with hereditary hemorrhagic telangiectasia. AJNR Am J Neuroradiol 2000; 21:1016-1020
5. Cymerman U, Vera S, Pece-Barbara N, Bourdeau A., White R.I., Dunn J., Letarte M. Identification of hereditary hemorrhagic telangiectasia type 1 in newborns by protein expression and mutation analysis of endoglin. Pediatr Res 2000; 47:24-35
Comments:

As you have gathered by now it is not uncommon for physicians and scientists to disagree about issues, but with time, a consensus is reached. As Dr. Letarte and her colleagues documented in an article we co-authored in Pediatric Research, January 2000, families with pulmonary arteriovenous malformations often harbor CAVMs as well. Since the incidence of large CAVMs (10-mm nidus or fistula) only number 10-12%, they are not as frequent as PAVMs. Still, with modern treatment techniques, they are very approachable and we believe that with pooled data from many of the Centers, we will be able to sort out which ones are at greatest risk and which is the best modality for treatment.

Robert I. White, Jr., M.D., Chair
Scientific Medical Advisory Board