Repeat Imaging for Intracranial Aneurysms in Patients with Autosomal Dominant Polycystic Kidney Disease with Initially Negative Studies: A Prospective Ten-Year Follow-up

Materials and Methods

A large natural history study of ADPKD subjects has been undertaken since 1985 at the University of Colorado Health Sciences Center (UCHSC). Participating subjects underwent a history and physical examination by a single nurse practitioner or physician in the General Clinical Research Center at the UCHSC. The study was approved by the Colorado Multiple Institutional Review Board. Informed consent was obtained from each subject participating in the study. The presence of vascular events in affected family members was assessed. Definite ruptured ICA required a clinical history of subarachnoid hemorrhage, with documentation of subarachnoid hemorrhage and ruptured ICA by neuroimaging, surgery, or autopsy. Probable ruptured ICA included patients with medical records describing a clinical presentation and course likely to be from a ruptured ICA and not consistent with an alternative diagnosis, such as migraine, trauma, arteriovenous malformation, or ischemic stroke. Patients with possible ruptured ICA met the same criteria, but supporting medical records were not available, and medical information was relayed verbally by the patient or family (18).

The current prospective study was undertaken to evaluate the prognostic significance of a negative ICA screening study in asymptomatic individuals with ADPKD. National Institutes of Health (NIH) funding was provided to have repeat imaging studies performed in the patients with ADPKD who had an initially negative imaging study. The priority was to study patients after 8 yr of follow-up. At the end of the NIH funding period, 76 patients had been restudied a median of 9.7 yr (mean, 9.8 ± 1.4 yr; range, 7 to 14.4 yr) after their initial study, and they constitute the results for this publication. The time of follow-up was defined as the period from the initial imaging study to the time of the repeat imaging study. All imaging studies were performed at the UCHSC and were reviewed by one staff radiologist (DR). There were another 60 patients who had negative initial imaging studies, but funding was not available to repeat their imaging studies.

The methods used in the initial screening were CT angiography in 48 patients, four-vessel cerebral angiography in 23 patients, MRA in 2 patients, and magnetic resonance imaging (MRI) in 3 patients. The methods in the repeat study were MRA in 55 patients, CT angiography in 20 patients, and four-vessel cerebral angiography in only 1 patient. The MRA follow-up studies performed at University Hospital consisted of a three-dimensional time-of-flight MRA with 1-mm axial sections and a T2-weighted three-dimensional fast spin-echo (“black blood”) sequence (19) with 0.8-mm cubic voxels. The CT angiograms consisted of 1-mm axial and coronal sections through the circle of Willis. A new ICA was defined as a previously undetected ICA of more than 2 mm diameter. Possible risk factors for ICA, including a positive family history of ruptured ICA, presence of hypertension, and smoking history, were also assessed.

Values are expressed as the mean ± SD (range). Continuous variables were compared between groups by the rank-sum test. Frequencies were compared between groups by Fisher’s exact test. Values of P < 0.05 are reported. The results were analyzed for all patients and for the patients with and without a family history of ruptured ICA.

Results

The clinical characteristics of the 76 patients are shown in Table 1, including age, gender, family history of ruptured ICA, hypertension, and smoking. The 76 patients were from 63 families. The relationships between family members studied are shown in Table 2. Of the study patients with a relative with a repeat imaging study, the relative was usually a sibling (71%). A family history of ruptured ICA was present in 32 patients (9 possible or probable and 23 definite ICA ruptures). Of these 32 patients, 15 had two or more relatives with a history of ruptured ICA, thus suggesting family clustering. The remaining 17 patients had a single relative with a ruptured ICA. An adequate number of images were not available to assess the prevalence of relatives with unruptured ICA. In Table 3, the characteristics of subjects with and without a family history of ruptured ICA are compared. Although the percentage of subjects with any history of smoking was similar between groups, a greater percentage of subjects without family history of ruptured ICA had quit smoking before the follow-up study.

The 60 patients who had not been restudied were contacted, except 3 who had died. None of the 60 patients had suffered a ruptured ICA after a median of 8.5 yr. The three deaths were due to suicide, cancer, and cardiac failure. Two subjects reported strokes that did not meet the definition of a ruptured ICA. There were no significant differences between the 76 patients who were restudied and the remaining 60 patients with respect to age at initial study, gender, family history of ruptured ICA, hypertension, and smoking.

On restudy, of the 76 subjects who had negative initial imaging studies, only 2 demonstrated ICA. The initial imaging study of one of these subjects was reviewed and was retrospectively thought to be positive. This subject, age 36 at initial study, had no family history of ICA, and the 2-mm ICA, which was retrospectively seen on his initial CT angiography study, had not changed on follow-up by MRA 11.3 yr later.

The second subject with a negative initial imaging study who was positive on follow-up had a positive family history of ruptured ICA in her uncle (age 44), and her mother died suddenly of unknown cause at the age of 37 yr. The uncle and mother both were diagnosed with ADPKD. This subject had a four-vessel cerebral angiography initially at age 29 and 9.6 yr later underwent CT angiography screening, followed by four-vessel cerebral angiography. The follow-up studies demonstrated a 10-mm left vertebral artery ICA, a 10-mm extracranial left internal carotid artery aneurysm, and a 4-mm right middle cerebral artery ICA (Figure 1). The 4-mm aneurysm was retrospectively seen on the initial images; it may have enlarged during follow-up (Figure 1). In addition to the family history, the patient had a 12.5 pack-year history of smoking at the time of the repeat imaging study, and she was diagnosed with hypertension at the time of the repeat study. She was normotensive, with a 6 pack-year history of smoking at the initial study. The subject underwent surgical clipping of the right middle cerebral artery aneurysm and coiling of the left vertebral artery aneurysm without complications.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1. Initial and follow-up four-vessel cerebral angiographic images on a subject with autosomal dominant polycystic kidney disease (ADPKD). The follow-up images were obtained 9.6 yr after the initial study. The follow-up images reveal a new 10-mm left vertebral artery aneurysm (A), a new 10-mm extracranial left internal carotid artery aneurysm (B), and a 4-mm right middle cerebral artery aneurysm that may have enlarged and was not detected on the initial study because of lower spatial resolution of early digital subtraction angiography (C).

Discussion

A ruptured ICA is a devastating event, with significant mortality and neurologic impairment in survivors. The prevalence of ICA is significantly increased in subjects with ADPKD (10–13⇓⇓⇓), and ICA rupture occurs at a younger age in patients with ADPKD than in the general U.S. population (20). In addition, the morbidity and mortality due to ruptured ICA are greater in subjects with ADPKD (20). Thus, the years of productive and healthy life lost with a ruptured ICA are substantially greater in ADPKD subjects than in the general population.

Therefore, the clinical decision regarding screening patients with ADPKD for the presence of ICA is an important issue. The relevance of the screening is demonstrated by a recent prospective study that found new aneurysms in 5 of 20 patients with ADPKD with a previously documented ICA (11 of which had ruptured) during a mean follow-up period of 15.2 yr. Although there was no mortality in these patients during that period (16), 50% of patients required surgical clipping of the ICA. These results suggest that follow-up imaging studies in patients with known ICA is worthwhile.

The question asked in the study presented here is whether ADPKD subjects who have had a negative imaging study in search of ICA should have a repeat imaging study within the next 10 yr. In the present prospective study, 76 patients with ADPKD who had a negative imaging study consented to have a repeat imaging study a mean of 9.8 yr later. The results indicated that two subjects had a positive repeat imaging study (2.6%), although only one subject required surgical intervention. It is likely that lifelong screening—for example, every 10 yr—would continue to yield a small percentage of new findings, including findings that require intervention. It remains uncertain whether such a screening could alter the lifelong risk of death and disability in patients with ADPKD. The one patient identified in this rescreening who required surgical intervention had a positive family history for a ruptured ICA and underwent successful clipping of two ICA; the 10-mm extracranial internal carotid aneurysm was not clipped. Of the other patients, the negative repeat study occurred despite the fact that 41% had a family history of ruptured ICA, 79% had hypertension, and 49% had a history of smoking. In addition, none of the 60 patients who were not rescreened reported an ICA event or died of a known ICA. However, we cannot exclude the presence of an asymptomatic ICA in these patients.

In recent years, noninvasive central nervous system imaging techniques have emerged. For example, MRA was used in the initial study for only 2 patients, whereas 55 of the 76 patients had an MRA as the technique for the repeat study. Unlike CT or four-vessel cerebral angiography, MRA does not require contrast media, making it a more desirable modality in patients with impaired renal function. However, MRA cannot be used in patients with certain types of surgical clips on previous aneurysms. The use of CT angiography decreased from 48 patients in the initial screening to 20 patients in the repeat screening. The sensitivity of CT angiography and MRA ranges from 60% to 95% in the literature. It is clear that the sensitivity is less for smaller aneurysms (21–25⇓⇓⇓⇓). It is also clear that these noninvasive studies can demonstrate aneurysms not seen on catheter digital subtraction angiograms (23,26⇓). The gold standard, however, is still considered four-vessel cerebral angiography for ICA screening. In an earlier study, 25% of patients with ADPKD compared with 10% of patients without ADPKD had side effects with the use of four-vessel cerebral angiography (10). There was no explanation for this difference in side effects, but increased vascular sensitivity in patients with ADPKD was suggested. The relative costs of MRA, CT angiography, and four-vessel cerebral angiography at our institution are $2650, $1800, and $10,000, respectively. Therefore, there are both medical and economic implications for these ICA screening studies.

Because of the high cost of repeating imaging studies, a tentative conclusion could be that repeat imaging studies for an ICA are not necessary in an asymptomatic patient with ADPKD with an initially negative study. However, such a recommendation would have missed the 1 out of 76 subjects in this study who developed a clinically significant new aneurysm. This subject had a family history of ruptured ICA, was hypertensive at the time of rescreening, and was a smoker. However, these factors were also present in many of the other subjects as well. Ruptured ICA have been shown to cluster in certain ADPKD families (18). Therefore, perhaps a recommendation of rescreening only those with a family history of ruptured ICA would be appropriate—in which case, 44 of 76 patients with ADPKD with a negative initial imaging study would have been able to forego rescreening. This recommendation is consonant with a proposal in a recent review about the management of cerebral aneurysms in patients with ADPKD (5).

A novel PKD mutation has been identified in three families with ICA and/or very early onset disease (27). In addition, an association has recently been found between ADPKD families with vascular complications and the position of the PKD1 mutation, with such families being more likely to have a mutation toward the 5′ end of the gene (28). With future progress in mutation screening, it may be possible to identify families at increased risk for ICA, thus facilitating the decision for initial as well as repeat ICA screening.

In summary, the present prospective study reports that over a mean of 9.8 yr of follow-up of patients with ADPKD with an initially negative imaging study, only 2 of 76 patients had a positive follow-up study, with only 1 of 76 demonstrating new, clinically significant ICA. This single occurrence, however, was clinically relevant and led to successful aneurysmal clipping with an uneventful follow-up. The potential risk factors of family history of ruptured ICA, hypertension, and smoking all were present in the subject developing an aneurysm, but were also present in many of the subjects with negative repeat imaging studies. Because ruptured ICA are known to cluster within ADPKD families, and the only subject who developed a new ICA had a family history of ruptured ICA, one approach may be to rescreen only those subjects with a family history of ruptured ICA after an interval of 10 yr. However, it may be equally appropriate to not rescreen any asymptomatic patients with a previous negative imaging study for ICA. Advances in mutation screening may lead to a greater ability to assign probabilities of ICA development and rupture dependent on the particular mutation present in the individual with ADPKD.