Editorial: The challenges of determining natural history

Free access

In the following study, Gross et al. describe the natural history of cavernous malformations (CMs) from their extensive experience with 167 children and 222 CMs.1 Their goal is to describe this lesion’s natural history, and they do this by determining the chance of developing a de novo lesion and the chance of bleeding (or rebleeding). They find that the annual hemorrhage rate for incidental lesions was 0.5% and for hemorrhagic CMs was 11.3% overall (18.2% in the first 3 years). Bleeding tended to occur in the first 3 years, and risk factors for bleeding were hemorrhagic

In the following study, Gross et al. describe the natural history of cavernous malformations (CMs) from their extensive experience with 167 children and 222 CMs.1 Their goal is to describe this lesion’s natural history, and they do this by determining the chance of developing a de novo lesion and the chance of bleeding (or rebleeding). They find that the annual hemorrhage rate for incidental lesions was 0.5% and for hemorrhagic CMs was 11.3% overall (18.2% in the first 3 years). Bleeding tended to occur in the first 3 years, and risk factors for bleeding were hemorrhagic presentation, brainstem location, and associated developmental venous anomaly (DVA). Twenty new CMs developed in 10 patients—a risk of 2.5% per lesion-year or 1.2% per patient-year. This risk was higher in patients with multiple lesions and lowest in patients with single nonfamilial lesions (0.4% per year).

Natural history studies are important: they form the basis for our treatment recommendations. They are also hard to do—especially in surgery. We naturally tend to see the patients who need surgery, and we tend not to see (or follow) the ones who don’t. The authors are to be congratulated for doing this work.

One of the most important components of a natural history study is the establishment of an “inception cohort.”2 This is a group of patients who are captured at a uniform and preferably early stage of their disease. Gross et al. identified patients who were initially observed with lesions that were visible on T2-weighted images and measured at least 4 mm in diameter. The patients’ clinical statuses were variable: most had presented with hemorrhage, but some lesions were found incidentally or the patient presented with seizures. Both solitary lesions and multiple lesions were included in their analysis. What is not clear are the criteria that were used for initial surgical versus nonsurgical treatment. How many patients were treated up front, and how did they differ from those who were selected for observation and followed in this study?

When using natural history data to make decisions and advise patients, we would ideally like data from patients similar to our own. That can be challenging since the large centers with experience (especially in rare conditions) are the ones with the experience and data. In the Gross paper,1 the referral pattern of the group was not specifically described, but it appears that the patients were seen at an academic medical center with a national and international referral practice that likely includes patients seeking second opinions. There is a well-established program in vascular pediatric neurosurgery and an academic track record in CM management. This means that there is a wealth of information to be shared with the pediatric neurosurgical community—but it also means that the patients seen are likely to be different from those seen in many centers. The proportion of patients who presented with symptomatic hemorrhage (62%) seems high, and hemorrhagic presentation is a risk factor for rebleeding. This can result in a higher rebleeding rate than may be encountered in the general population. In addition, it’s possible that CMs associated with DVA would be preferentially referred to a specialty center, raising the proportion of patients with DVA in this group, compared with the general population.

Finally, we should consider the follow-up. The total amount of follow-up expressed in patient-years seems very good (789.5 patient-years), but some patients had very short follow-ups. Patients were included if they had a minimum follow-up of 1 month without treatment. In order to learn about natural history, patients should have adequate follow-up to see the outcome of interest. For CMs, it seems unlikely that a month is sufficient. It would be interesting to know how many patients had a minimum follow-up of 1 year. In addition, we don’t know the proportion of patients who had short follow-ups with the authors and continued their follow-up in the community (with or without hemorrhage).

Overall, this is a very informative study. As surgeons, we are not often in a position to do natural history studies, so this is a welcome addition to the literature. At the same time, we need to keep some of the challenges in mind as we interpret the results and use them in our decision making.

References

  • 1

    Gross BADu ROrbach DBScott RMSmith ER: The natural history of cerebral cavernous malformations in children. J Neurosurg Pediatr epub ahead of printOctober162015.

  • 2

    Sackett DLHaynes RBGuyatt GHTugwell PMaking a prognosis. Sackett DLHaynes RBGuyatt GH: Clinical Epidemiology A Basic Science for Clinical Medicine 2BostonLittle, Brown & Co1991. 173185

Response

We thank Dr. Kestle for his thoughtful commentary. As he notes, the fact that we practice at a tertiary care referral center for the resection of CMs probably influenced the high rate of hemorrhagic presentation in our study (62%). This high proportion may have contributed to our ability to detect hemorrhage clustering within 3 years. As a comparison, a study from the University of Michigan acquired a pediatric cohort with CMs via a consecutive review of MRI studies over a 12.5-year period.2 With this approach, the most common presentation modality was incidental (42 [46%] of 92 patients), and the rate of hemorrhagic presentation was 33% (30 of 92 patients). While different rates of presentation may influence the overall annual hemorrhage rate in a study, we attempted to adjust for this by providing bleed rates stratified by presentation and by performing a multivariate analysis of hemorrhage risk factors to control for this variable.

We think it unlikely that a referral pattern specific to our center influenced the proportion of CMs associated with DVAs. In our study, this association was seen in 20% of cases, consistent with findings in adult studies.7 In the pediatric natural history report from the University of Michigan, this association was seen in 35% of cases.2 Interestingly, those authors also found greater hemorrhage rates for CMs associated with DVAs (p = 0.06).

Dr. Kestle appropriately queries our criteria for surgical versus nonsurgical CM treatment. We generally excise symptomatic and/or hemorrhagic CMs with a safe surgical corridor of access.8 Of the 222 CMs evaluated, 99 (45%) were ultimately resected. If our study data were to be prospectively collected with resection as an exclusion criterion, a substantial cohort of patients would be lost, potentially introducing bias. However, as noted in our results, 193 (87%) of 222 CMs had at least 1 month of untreated follow-up and were thus included in our natural history analysis. This high inclusion rate was made possible by the fact that we rarely operate acutely on hemorrhagic vascular malformations. In addition, patients referred to our center for surgery, who had months of untreated follow-up after an initial hemorrhage, were included in our study, allowing us to maximize our total patient-years of follow-up. Addressing Dr. Kestle’s thoughtful question about a longer-term follow-up, we note that 124 (64%) of 193 patients had at least 1 year of untreated follow-up.

While we agree that it would be ideal to compare outcomes between treated and untreated hemorrhagic CMs, prospective studies including observation to evaluate the natural history of symptomatic and/or hemorrhagic CMs in children present ethical concerns, particularly given the known worsened natural history of such lesions as compared with incidental CMs.2,3,4,6,7 The relative safety of CM excision coupled with preexisting literature supporting treatment will probably limit the long-term follow-up of a large cohort of untreated CMs.1,5,8–10 Overcoming the difficulty of the lesion’s relative rarity in a pediatric population may be achieved through the development of multiinstitutional prospective databases; however, these cohorts may largely be composed of incidental and/or exclusively surgically inaccessible lesions buried in eloquent parenchyma. Despite the inherent limitations of retrospective approaches to natural history analysis, we hope our study will help to further our understanding of the natural history of pediatric CM. In conjunction with the efforts of other researchers, these data will add to the physician’s ability to provide better-informed patient counseling with relatively unique data on de novo CM development and radiation-induced lesions and with stratified annual hemorrhage rates.

References

  • 1

    Acciarri NGalassi EGiulioni MPozzati EGrasso VPalandri G: Cavernous malformations of the central nervous system in the pediatric age group. Pediatr Neurosurg 45:811042009

  • 2

    Al-Holou WNO’Lynnger TMPandey ASGemmete JJThompson BGMuraszko KM: Natural history and imaging prevalence of cavernous malformations in children and young adults. J Neurosurg Pediatr 9:1982052012

  • 3

    Al-Shahi Salman RHall JMHorne MAMoultrie FJosephson CBBhattacharya JJ: Untreated clinical course of cerebral cavernous malformations: a prospective, population-based cohort study. Lancet Neurol 11:2172242012

  • 4

    Barker FG IIAmin-Hanjani SButler WELyons SOjemann RGChapman PH: Temporal clustering of hemorrhages from untreated cavernous malformations of the central nervous system. Neurosurgery 49:15242001

  • 5

    Di Rocco CIannelli ATamburrini G: Cavernomas of the central nervous system in children. A report of 22 cases. Acta Neurochir (Wien) 138:126712741996

  • 6

    Flemming KDLink MJChristianson TJBrown RD: Prospective hemorrhage risk of intracerebral cavernous malformations. Neurology 78:6326362012

  • 7

    Gross BALin NDu RDay AL: The natural history of intracranial cavernous malformations. Neurosurg Focus 30:6E242011

  • 8

    Gross BASmith ERGoumnerova LProctor MRMadsen JRScott RM: Resection of supratentorial lobar cavernous malformations in children. J Neurosurg Pediatr 12:3673732013

  • 9

    Hugelshofer MAcciarri NSure UGeorgiadis DBaumgartner RWBertalanffy H: Effective surgical treatment of cerebral cavernous malformations: a multicenter study of 79 pediatric patients. J Neurosurg Pediatr 8:5225252011

  • 10

    Mottolese CHermier MStan HJouvet ASaint-Pierre GFroment JC: Central nervous system cavernomas in the pediatric age group. Neurosurg Rev 24:55732001

If the inline PDF is not rendering correctly, you can download the PDF file here.

Article Information

ACCOMPANYING ARTICLE DOI: 10.3171/2015.2.PEDS14541.

INCLUDE WHEN CITING Published online October 16, 2015; DOI: 10.3171/2015.4.PEDS15115.

Disclosure The author reports no conflict of interest.

© AANS, except where prohibited by US copyright law.

Headings

References

1

Gross BADu ROrbach DBScott RMSmith ER: The natural history of cerebral cavernous malformations in children. J Neurosurg Pediatr epub ahead of printOctober162015.

2

Sackett DLHaynes RBGuyatt GHTugwell PMaking a prognosis. Sackett DLHaynes RBGuyatt GH: Clinical Epidemiology A Basic Science for Clinical Medicine 2BostonLittle, Brown & Co1991. 173185

1

Acciarri NGalassi EGiulioni MPozzati EGrasso VPalandri G: Cavernous malformations of the central nervous system in the pediatric age group. Pediatr Neurosurg 45:811042009

2

Al-Holou WNO’Lynnger TMPandey ASGemmete JJThompson BGMuraszko KM: Natural history and imaging prevalence of cavernous malformations in children and young adults. J Neurosurg Pediatr 9:1982052012

3

Al-Shahi Salman RHall JMHorne MAMoultrie FJosephson CBBhattacharya JJ: Untreated clinical course of cerebral cavernous malformations: a prospective, population-based cohort study. Lancet Neurol 11:2172242012

4

Barker FG IIAmin-Hanjani SButler WELyons SOjemann RGChapman PH: Temporal clustering of hemorrhages from untreated cavernous malformations of the central nervous system. Neurosurgery 49:15242001

5

Di Rocco CIannelli ATamburrini G: Cavernomas of the central nervous system in children. A report of 22 cases. Acta Neurochir (Wien) 138:126712741996

6

Flemming KDLink MJChristianson TJBrown RD: Prospective hemorrhage risk of intracerebral cavernous malformations. Neurology 78:6326362012

7

Gross BALin NDu RDay AL: The natural history of intracranial cavernous malformations. Neurosurg Focus 30:6E242011

8

Gross BASmith ERGoumnerova LProctor MRMadsen JRScott RM: Resection of supratentorial lobar cavernous malformations in children. J Neurosurg Pediatr 12:3673732013

9

Hugelshofer MAcciarri NSure UGeorgiadis DBaumgartner RWBertalanffy H: Effective surgical treatment of cerebral cavernous malformations: a multicenter study of 79 pediatric patients. J Neurosurg Pediatr 8:5225252011

10

Mottolese CHermier MStan HJouvet ASaint-Pierre GFroment JC: Central nervous system cavernomas in the pediatric age group. Neurosurg Rev 24:55732001

TrendMD

Metrics

Metrics

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 78 78 14
PDF Downloads 86 86 10
EPUB Downloads 0 0 0

PubMed

Google Scholar