- Email: [email protected]
An analysis of the effects of smoking and other cardiovascular risk factors on obliteration rates after arteriovenous malformation radiosurgery
Int. J. Radiation Oncology Biol. Phys., Vol. 51, No. 4, pp. 969 –973, 2001 Copyright © 2001 Elsevier Science Inc. Printed in the USA. All rights reserved 0360-3016/01/$–see front matter
PII S0360-3016(01)01734-5
CLINICAL INVESTIGATION
Brain
AN ANALYSIS OF THE EFFECTS OF SMOKING AND OTHER CARDIOVASCULAR RISK FACTORS ON OBLITERATION RATES AFTER ARTERIOVENOUS MALFORMATION RADIOSURGERY AJAY BHATNAGAR, B.S.,* JOHN C. FLICKINGER, M.D.,*† DOUGLAS KONDZIOLKA, M.D.,*† AJAY NIRANJAN, M.B., B.S.,† AND L. DADE LUNSFORD, M.D.*†‡ Departments of *Radiation Oncology, †Neurological Surgery, and ‡Radiology, Center for Image-Guided Neurosurgery, University of Pittsburgh School of Medicine, Pittsburgh, PA Purpose: To assess the relationships of smoking and other cardiovascular disease risk factors (hypertension, diabetes, hypercholesterolemia, and gender) to rates of radiosurgery-induced obliteration of arteriovenous malformations (AVM). Methods and Materials: We evaluated follow-up imaging and clinical data in 329 AVM patients who received gamma knife radiosurgery at the University of Pittsburgh between 1987 and 1994. There were 113 smokers, 29 hypertensives, 5 diabetics, 4 hypercholesterolemics, 159 male patients, and 170 female patients. All patients had regular clinical or imaging follow-up for a minimum of 3 years after radiosurgery. Results: Multivariate analysis showed that smoking had no effect on AVM obliteration (p > 0.43). Hypertension, diabetes, and hypercholesterolemia had no discernible effect on AVM obliteration in this study (p > 0.78). However, females aged 12– 49 had a statistically significant lower in-field obliteration rate than males (78% vs. 89%, p ⴝ 0.0102). Conclusion: Smoking has no effect on AVM obliteration. Hypertension, diabetes, and hypercholesterolemia had no discernible effect in this study. Further study is needed to establish whether estrogen has a vascular protective effect that could partially limit radiosurgical AVM obliteration, as suggested by this study. © 2001 Elsevier Science Inc. Arteriovenous malformation, Radiosurgery, Cardiovascular risk factors, Smoking.
Stereotactic radiosurgery has been shown to be an effective way to manage patients with cerebral arteriovenous malformations (AVM) (1–14). Depending on the dose and treatment technique, radiosurgery has a complete AVM obliteration rate varying from 60% to 90% (1–14). Our recent study of AVM obliteration by radiosurgery correlated AVM obliteration with a linear-quadratic function of dose (15). The alpha/beta ratio was negative (␣/ ⫽ ⫺49.3 ⫾ 5.3), reflecting a higher in-field obliteration rate with marginal doses of 20 –24 Gy than with 25 Gy (93% vs. 85%, p ⫽ 0.034). Pathologic changes after AVM radiosurgery have been well described (16 –19). The initial effect of high-dose single-fraction radiation on AVMs seems to be through damage to endothelial cells. Subsequently, the intimal layer progressively thickens via proliferation of myofibroblasts that elaborate an extracellular matrix, including type IV collagen. Late changes observed include cellular degeneration, hyaline transformation, fragmentation of the elastic laminae, and mineralization in vessel walls. This his-
topathologic response to high-dose irradiation leads to progressive stenosis, luminal closure, and eventual obliteration of the AVM nidus. Based on this pathophysiologic process of AVM obliteration by radiosurgery, it would seem likely that factors promoting vascular disease (atherosclerosis, stroke) would also help promote AVM obliteration. Therefore, in this study we hypothesized that patients with risk factors for atherosclerosis and stroke would have higher obliteration rates after AVM radiosurgery compared to patients without these risk factors. Smoking, hypertension, diabetes mellitus, hypercholesterolemia, and male gender have been associated with increased risk for myocardial infarction, stroke, and/or peripheral vascular occlusive disease (20 –23). Any or all of these could potentially effect AVM obliteration after radiosurgery. In our prior study, we found that females had a statistically significant lower in-field obliteration rate than men (15). We documented in-field obliteration after radiosurgery in 106/139 or 76% of females compared to 114/125 or 91% of men, (p ⫽ 0.0025). This difference in obliteration by gender suggested that vascular protective effects of estrogen
Reprint requests to: John C. Flickinger, M.D., Radiation Oncology B-300, 200 Lothrop Street, Pittsburgh, PA 15213. Tel: (412) 647-
3600; Fax: (412) 647-6029; E-mail: [email protected] Received Apr 23, 2001. Accepted for publication Jun 11, 2001.
INTRODUCTION
969
970
I. J. Radiation Oncology
● Biology ● Physics
Volume 51, Number 4, 2001
Table 1. Raw data of in-field AVM obliteration results by cardiovascular risk factors Patients
Patent AVM
Obliterated AVM
Total
Nonsmokers Smokers Nonhypertensives Hypertensives Nondiabetics Diabetics Nonhypercholesterolemics Hypercholesterolemic
34 (19%) 17 (15%) 48 (16%) 3 (10%) 51 (16%) 0 (0%) 51 (16%) 0 (0%)
176 (81%) 96 (85%) 246 (84%) 26 (90%) 267 (84%) 5 (100%) 268 (84%) 4 (100%)
210 113 294 29 318 5 319 4
and/or its metabolites might account for the lower obliteration rate in females (24). Progesterone seems to have no protective effect on the vasculature (24). Therefore, we sought to investigate the hormonal status in further detail than just male vs. female, as in the previous study. Because we did not have measurements of the estrogen levels in our patients, we sought to discern the effects of estrogen on AVM obliteration rates through subgroup analysis of females aged 12– 49 (when estrogen levels should be highest) and in postmenopausal women with and without use of estrogen replacement therapy (ERT).
tween the different estrogenic status cohorts (females aged 12– 49, men, females younger than 12 or older than or equal to 50) and outcome (obliteration of either the targeted volume or entire AVM). A subset analysis was performed to identify and model any correlation between estrogen replacement in postmenopausal female patients and outcome, as well as any correlation between oral contraceptives in estrogenic females and outcome. Multivariate logistic regression was performed to identify and model any correlation between the cardiovascular parameters (smoking, hypertension, diabetes mellitus, hypercholesterolemia) and outcome (obliteration of either the targeted volume or entire AVM).
METHODS AND MATERIALS Patients We evaluated follow-up MRI and clinical data in 351 patients who received gamma knife radiosurgery at the University of Pittsburgh between 1987 and 1994. All patients had regular clinical or imaging follow-up for a minimum of 3 years at the time of data analysis. Approval for this retrospective study was granted by the University of Pittsburgh Institutional Review Board. We reviewed each patient’s history and physical examination report to identify documentation of smoking, hypertension, diabetes mellitus, and hypercholesterolemia in the past medical history or social history portion of the history and physical examination report. These data were not available in the records of 22 patients, who were excluded, leaving 329 patients for analysis. In addition, for females the history and physical examination report was reviewed to identify also any history of estrogen replacement or birth control pills as documented in the medications list portion of the report. For each smoker, the quantity of cigarettes smoked, as pack per day and pack-years, was noted. For hypertension, diabetes mellitus, and hypercholesterolemia, the presence of each diagnosis was noted. There were 159 male patients and 170 female patients. There were 136 females aged 12– 49 years. In total, there were 113 patients who were smokers, 29 patients who had hypertension, 5 patients who had diabetes mellitus, and 4 patients who were hypercholesterolemic. Statistical analysis Multivariate logistic regression was performed using SPSS software to identify and model any correlation be-
RESULTS Cross-tabulation Table 1 shows the obliteration results tabulated for all accounted patients, classified by presence or lack of each cardiovascular disease risk factor except gender. We found that the ratio of obliterated to patent AVM is similar for smokers and nonsmokers. These results also show that all the diabetics (n ⫽ 5) and hypercholesterolemia patients (n ⫽ 4) had successful obliterations. Hypertensive patients have an obliteration percentage comparable to that found in nonhypertensive patients. Obliteration results were tabulated for the specific cohorts of patients to further evaluate the estrogenic effects on AVM obliteration; the results are shown in Table 2. We found that the estrogenic females had twice the number of patent AVM compared to men. Men and nonestrogenic females had similar percentages of patent AVM. The table also shows that of the 6 postmenopausal female patients on ERT, 33% have patent AVM, whereas of the postmenopausal patients not on ERT, only 6% have patent AVM. Multivariate analysis of cardiovascular risk factors Table 3 shows the results of multivariate logistic regression analysis of the effects of cardiovascular risk factors on obliteration. Each variable was analyzed independently in separate multivariate models with three other factors: marginal dose, marginal dose squared, and estrogenic-age females (12– 49). These three factors remained significant in each multivariate model: marginal dose (0.0003 ⬍ p ⬍ 0.0167), marginal dose squared (0.0008 ⬍ p ⬍ 0.0189), and estrogenic females (p ⫽ 0.0095 ⬍ p ⬍ 0.0109). There were
Cardiovascular risk factors in AVM obliteration
● A. BHATNAGAR et al.
971
Table 2. Raw data of in-field AVM obliteration results in different estrogen cohorts Patients
Patent AVM
Obliterated AVM
Total
Estrogenic females aged 12–49 Men Nonestrogenic females (females ⬍12, females ⱖ50)* Total Postmenopausal w/ no ERT Postmenopausal w/ ERT
30 (22%) 18 (11%) 4 (12%) 52 (16%) 1 (6%) 2 (33%)
106 (78%) 141 (89%) 30 (88%) 277 (84%) 16 (94%) 4 (66%)
136 159 34 329 17 6
* The subset of postmenopausal patients (females ⱖ50) is further categorized in the last two rows. Abbreviation: ERT ⫽ estrogen replacement therapy.
Table 3. Multivariate logistic regression analysis of cardiovascular risk factors in AVM obliteration Variable
p value
Regression coefficient ⫾ SE
Odds ratio (95% CI)
Smoking (yes/no) Smoking (pack per day) Smoking (pack-years) Hypertension Diabetes mellitus Hypercholesterolemia
0.7829 0.4335 0.4475 0.6329 0.7614 0.7801
0.094 ⫾ 0.342 0.214 ⫾ 0.273 0.010 ⫾ 0.014 0.325 ⫾ 0.681 4.931 ⫾ 16.24 5.068 ⫾ 8.156
1.09 (0.562–2.14) 1.24 (0.725–2.12) 1.01 (0.983–1.04) 1.38 (0.364–5.26) 138.5 (2.078E-12–9.23E⫹15) 158.9 (1.813E-05–1.39E⫹09)
no other significant independent variables associated with AVM obliteration in this analysis. Subset analysis Multivariate logistic regression analysis of in-field AVM obliteration was performed, using the subset of females aged 12– 49 (estrogenic females), plus all the males in the study. The patient cohort variable was analyzed, along with marginal dose and marginal dose squared for logistic regression. These two factors were shown to be statistically significant in a prior analysis and remained significant in these multivariate models. The results are shown in Table 4. We found that the females aged 12– 49 had a statistically significant lower obliteration rate (p ⫽ 0.0102) than males. A similar analysis was performed, except the subset was females aged 12– 49 vs. the smaller subset of females younger than 12 and older than or equal to 50. The trend was toward higher obliteration in females aged 12– 49. A subset multivariate logistic regression analysis was performed using only the postmenopausal females as the patient sample and the postmenopausal obliteration rate as the outcome. The variable estrogen replacement was analyzed, along with marginal dose and marginal dose squared for logistic regression. A similar subset analysis was performed for estrogenic females, except the variable used was birth control pills, and the outcome was the estrogenic female
obliteration rate. Table 5 shows these results. Estrogen replacement shows a trend toward statistical significance for a negative effect on AVM obliteration for postmenopausal women, as indicated by the negative value of the regression coefficient. Birth control pills show no statistical significance in AVM obliteration for estrogenic females. DISCUSSION This study found that the cardiovascular risk factors, except for male status, do not have a large effect on the AVM obliteration rate induced by radiosurgery, based on the multivariate logistic regression. The initial hypothesis of the study was that atherosclerosis risk factors would have a beneficial effect on the AVM obliteration rate, based on the pathophysiologic process of AVM obliteration by radiosurgery (16 –19). Specifically, smoking does not have a detectable effect on AVM obliteration, despite the relatively large sample size (n ⫽ 113) in this study. Hypertension also does not seem to have a large effect, but there are some caveats that prevent us from making a solid conclusion, such as with smoking. First of all, the sample size (n ⫽ 29) is relatively small compared to the sample size of smokers. In addition, most patients were being treated for hypertension, which may have also caused hypertension to lose its significance on obliteration.
Table 4. Multivariate analysis of estrogenic effects on AVM obliteration Subset comparison
p value
Regression coefficient ⫾ SE
Odds ratio (95% CI)
Estrogenic females vs. males Estrogenic females vs. Nonestrogenic females
0.0102
⫺0.876 ⫾ 0.341
0.416 (0.213–0.812)
0.1311
⫺1.041 ⫾ 0.690
0.353 (0.091–1.365)
972
I. J. Radiation Oncology
● Biology ● Physics
Volume 51, Number 4, 2001
Table 5. Subset analysis of postmenopausal and estrogenic females Subset
Variable
p value
Regression coefficient ⫾ SE
Odds ratio(95% CI)
Postmenopausal females
Estrogen replacement Birth control pills
0.1482
⫺2.52 ⫾ 1.72
0.080 (0.003–2.342)
Estrogenic females
5.48 ⫾ 18.03
0.7612
In the multivariate logistic regression analysis, diabetes mellitus and hypercholesterolemia did not show statistical significance on AVM obliteration. Although the cross-tabulation showed that all the patients with diabetes mellitus or hypercholesterolemia had successful obliteration, only 5 patients in this study had diabetes documented in their histories, and only 4 patients had hypercholesterolemia. From these current data, it cannot be concluded whether hypercholesterolemia and diabetes have an effect on obliteration, because their lack of statistical significance could possibly be due to the small sample size. Therefore, a larger study with further analysis needs to be performed to see whether these factors actually do have an effect on obliteration. This study found that females aged 12– 49 have a statistically significant lower obliteration rate than males. Using the cohort of females aged 12– 49, we wanted to isolate the patients with the highest levels of estrogen to more clearly identify its effects on AVM obliteration. An important limitation of this study is that we did not have measurements of estrogen levels, but relied instead on subgroup analysis by age as a rough substitute. Because this 12– 49-year-old cohort of females has a lower obliteration rate than other females (although not significantly different in subset analysis), the difference between the obliteration rate in males was even greater, with a lower p value than in the prior comparison of all females vs. males (p ⫽ 0.0102 vs. p ⫽ 0.0180). These findings suggest that estrogen could exert a vascular protective effect that limits AVM obliteration after radiosurgery in some female patients. However, the estro-
239.8 (1.08 E-135.34E⫹17)
genic females showed only a trend toward significance when compared with the “nonestrogenic” females (younger than 12 or older than or equal to 50, who had an obliteration rate similar to males), but this is probably because of the small number of patients in the nonestrogenic cohort. The label “nonestrogenic” was used for brevity, but it is well known that low levels of circulating estrogen are present in premenstrual and postmenopausal females (and in males). Further studies with greater numbers of patients in this cohort are warranted for complete analysis of the estrogenic effects on AVM obliteration. The results from the subset analysis of estrogen replacement therapy indicate a trend toward statistical significance on AVM obliteration, which further corroborates the findings showing the estrogenic effect on obliteration. Only 6 patients in the study had documentation of estrogen replacement therapy. Certainly, further analysis of a larger sample of patients needs to be performed to validate any conclusions about ERT. In summary, smoking has been clearly proven to have no effect on AVM obliteration rate; similarly, in this study hypertension, diabetes mellitus, and hypercholesterolemia had no discernible effect on AVM obliteration. With regard to gender as a risk factor, we found that male patients do have a statistically significant higher AVM obliteration rate compared to females aged 12– 49. As suggested by this study, further study is needed to establish if estrogen has a vascular protective effect that could partially limit radiosurgical AVM obliteration.
REFERENCES 1. Colombo F, Pozza F, Chierego G, Casentini L, Luca GD, Francescon P. Linear accelerator radiosurgery of cerebral arteriovenous malformations: An update. Neurosurgery 1994; 34:14 –21. 2. Ellis TL, Friedman WA, Bova FJ, Kubilis PS, Buatti JM. Analysis of treatment failure after radiosurgery for arteriovenous malformations. J Neurosurg 1998;89(1):104 –110. 3. Flickinger JC, Pollock BE, Kondziolka D, Lunsford LD. A dose-response analysis of arteriovenous malformation obliteration after radiosurgery. Int J Radiat Oncol Biol Phys 1996; 36:873– 879. 4. Friedman WA, Bova FJ, Mendenhall WM. Linear accelerator radiosurgery for arteriovenous malformations: The relationship of size to outcome. J Neurosurg 1995;82:180 –189. 5. Gobin YP, Laurent A, Merienne L, et al. Treatment of brain arteriovenous malformations by embolization and radiosurgery. J Neurosurg 1996;85(1):19 –28. 6. Karlsson B, Lax I, Soderman M. Can the probability for
7.
8.
9.
10.
11.
obliteration after radiosurgery for arteriovenous malformations be accurately predicted? Int J Radiat Oncol Biol Phys 1999;43(2):313–319. Karlsson B, Lindquist C, Steiner L. Prediction of obliteration after gamma knife surgery for cerebral arteriovenous malformations. Neurosurgery 1997;40(3):425– 430; discussion 430 – 431. Kjellberg R, Hanamura T, Davis K, Lyons S, Butler W, Adams R. Bragg-peak proton-beam therapy for arteriovenous malformations of the brain. N Engl J Med 1983;309:269 –273. Loeffler JS, Alexander E, Siddon R, et al. Stereotactic radiosurgery for intracranial arteriovenous malformations using a standard linear accelerator. Int J Radiat Oncol Biol Phys 1989;17:673– 677. Miyawaki L, Dowd C, Wara W, et al. Five year results of LINAC radiosurgery for arteriovenous malformations: Outcome for large AVMS. Int J Radiat Oncol Biol Phys 1999; 44(5):1089 –1106. Pollock BE, Flickinger JC, Lunsford LD, Maitz A, Kondziolka
Cardiovascular risk factors in AVM obliteration
12. 13. 14.
15.
16. 17.
18.
D. Factors associated with successful arteriovenous malformation radiosurgery. Neurosurgery 1998;42(6):1239 –1247. Schlienger M, Atlan D, Lefkopoulos D, et al. Linac radiosurgery for cerebral arteriovenous malformations: Results in 169 patients. Int J Radiat Oncol Biol Phys 2000;46(5):1135–1142. Steiner L, Lindquist C, Adler JR, Torner JC, Steiner M. Clinical outcome of radiosurgery for cerebral arteriovenous malformations. J Neurosurg 1992;77:1– 8. Touboul E, Al Halabi A, Buffat L, et al. Single-fraction stereotactic radiotherapy: A dose-response analysis of arteriovenous malformation obliteration. Int J Radiat Oncol Biol Phys 1998;41(4):855– 861. Flickinger JC, Maesawa S, Kondziolka D, Lunsford LD. An analysis of the clinical radiobiology of arteriovenous malformation obliteration by radiosurgery (abstract). Int J Radiat Oncol Biol Phys 2000;48,S1:255. Schneider BF, Eberhard DA, Steiner LE. Histopathology of arteriovenous malformations after gamma knife radiosurgery. J Neurosurg 1997;87(3):352–357. Szeifert GT, Kemeny AA, Timperley WR, Forster DM. The potential role of myofibroblasts in the obliteration of arteriovenous malformations after radiosurgery. Neurosurgery 1997; 40(1):61– 65; discussion 65– 66. Chang SD, Shuster DL, Steinberg GK, Levy RP, Frankel K. Stereotactic radiosurgery of arteriovenous malformations:
19.
20.
21.
22.
23.
24.
● A. BHATNAGAR et al.
973
Pathologic changes in resected tissue. Clin Neuropathol 1997; 16(2):111–116. Steinberg GK, Chang SD, Levy RP, et al. Surgical resection of large incompletely treated intracranial arteriovenous malformations following stereotactic radiosurgery. J Neurosurg 1996;84(6):920 –928. Kannel WB. Metabolic risk factors for coronary heart disease in women: Perspective from the Framingham Study. Am Heart J 1987;114:413– 419. Neaton JD, Wentworth D, for the Multiple Risk Factor Intervention Trial Research Group. Serum cholesterol, blood pressure, cigarette smoking, and death from coronary heart disease: Overall findings and differences by age for 316,099 white men. Arch Intern Med 1992;152:56 – 64. Wolf PA, D’Agostino RB, Belanger AJ, Kannel WB. Probability of stroke: A risk profile from the Framingham study. Stroke 1991;22:312–318. Hart C, Davey Smith G, Hole D. Risk factors and 20 year stroke mortality in men and women in the Renfrew/Paisley study in Scotland. Stroke 1999;30:1999 –2007. Writing Group for the PEPI Trial. Effects of estrogen or estrogen/progestin regimens on heart disease risk factors in postmenopausal women: the Postmenopausal Estrogen/Progestin Interventions (PEPI) Trial. JAMA 1995;273:199 –208.
PII S0360-3016(01)01734-5
CLINICAL INVESTIGATION
Brain
AN ANALYSIS OF THE EFFECTS OF SMOKING AND OTHER CARDIOVASCULAR RISK FACTORS ON OBLITERATION RATES AFTER ARTERIOVENOUS MALFORMATION RADIOSURGERY AJAY BHATNAGAR, B.S.,* JOHN C. FLICKINGER, M.D.,*† DOUGLAS KONDZIOLKA, M.D.,*† AJAY NIRANJAN, M.B., B.S.,† AND L. DADE LUNSFORD, M.D.*†‡ Departments of *Radiation Oncology, †Neurological Surgery, and ‡Radiology, Center for Image-Guided Neurosurgery, University of Pittsburgh School of Medicine, Pittsburgh, PA Purpose: To assess the relationships of smoking and other cardiovascular disease risk factors (hypertension, diabetes, hypercholesterolemia, and gender) to rates of radiosurgery-induced obliteration of arteriovenous malformations (AVM). Methods and Materials: We evaluated follow-up imaging and clinical data in 329 AVM patients who received gamma knife radiosurgery at the University of Pittsburgh between 1987 and 1994. There were 113 smokers, 29 hypertensives, 5 diabetics, 4 hypercholesterolemics, 159 male patients, and 170 female patients. All patients had regular clinical or imaging follow-up for a minimum of 3 years after radiosurgery. Results: Multivariate analysis showed that smoking had no effect on AVM obliteration (p > 0.43). Hypertension, diabetes, and hypercholesterolemia had no discernible effect on AVM obliteration in this study (p > 0.78). However, females aged 12– 49 had a statistically significant lower in-field obliteration rate than males (78% vs. 89%, p ⴝ 0.0102). Conclusion: Smoking has no effect on AVM obliteration. Hypertension, diabetes, and hypercholesterolemia had no discernible effect in this study. Further study is needed to establish whether estrogen has a vascular protective effect that could partially limit radiosurgical AVM obliteration, as suggested by this study. © 2001 Elsevier Science Inc. Arteriovenous malformation, Radiosurgery, Cardiovascular risk factors, Smoking.
Stereotactic radiosurgery has been shown to be an effective way to manage patients with cerebral arteriovenous malformations (AVM) (1–14). Depending on the dose and treatment technique, radiosurgery has a complete AVM obliteration rate varying from 60% to 90% (1–14). Our recent study of AVM obliteration by radiosurgery correlated AVM obliteration with a linear-quadratic function of dose (15). The alpha/beta ratio was negative (␣/ ⫽ ⫺49.3 ⫾ 5.3), reflecting a higher in-field obliteration rate with marginal doses of 20 –24 Gy than with 25 Gy (93% vs. 85%, p ⫽ 0.034). Pathologic changes after AVM radiosurgery have been well described (16 –19). The initial effect of high-dose single-fraction radiation on AVMs seems to be through damage to endothelial cells. Subsequently, the intimal layer progressively thickens via proliferation of myofibroblasts that elaborate an extracellular matrix, including type IV collagen. Late changes observed include cellular degeneration, hyaline transformation, fragmentation of the elastic laminae, and mineralization in vessel walls. This his-
topathologic response to high-dose irradiation leads to progressive stenosis, luminal closure, and eventual obliteration of the AVM nidus. Based on this pathophysiologic process of AVM obliteration by radiosurgery, it would seem likely that factors promoting vascular disease (atherosclerosis, stroke) would also help promote AVM obliteration. Therefore, in this study we hypothesized that patients with risk factors for atherosclerosis and stroke would have higher obliteration rates after AVM radiosurgery compared to patients without these risk factors. Smoking, hypertension, diabetes mellitus, hypercholesterolemia, and male gender have been associated with increased risk for myocardial infarction, stroke, and/or peripheral vascular occlusive disease (20 –23). Any or all of these could potentially effect AVM obliteration after radiosurgery. In our prior study, we found that females had a statistically significant lower in-field obliteration rate than men (15). We documented in-field obliteration after radiosurgery in 106/139 or 76% of females compared to 114/125 or 91% of men, (p ⫽ 0.0025). This difference in obliteration by gender suggested that vascular protective effects of estrogen
Reprint requests to: John C. Flickinger, M.D., Radiation Oncology B-300, 200 Lothrop Street, Pittsburgh, PA 15213. Tel: (412) 647-
3600; Fax: (412) 647-6029; E-mail: [email protected] Received Apr 23, 2001. Accepted for publication Jun 11, 2001.
INTRODUCTION
969
970
I. J. Radiation Oncology
● Biology ● Physics
Volume 51, Number 4, 2001
Table 1. Raw data of in-field AVM obliteration results by cardiovascular risk factors Patients
Patent AVM
Obliterated AVM
Total
Nonsmokers Smokers Nonhypertensives Hypertensives Nondiabetics Diabetics Nonhypercholesterolemics Hypercholesterolemic
34 (19%) 17 (15%) 48 (16%) 3 (10%) 51 (16%) 0 (0%) 51 (16%) 0 (0%)
176 (81%) 96 (85%) 246 (84%) 26 (90%) 267 (84%) 5 (100%) 268 (84%) 4 (100%)
210 113 294 29 318 5 319 4
and/or its metabolites might account for the lower obliteration rate in females (24). Progesterone seems to have no protective effect on the vasculature (24). Therefore, we sought to investigate the hormonal status in further detail than just male vs. female, as in the previous study. Because we did not have measurements of the estrogen levels in our patients, we sought to discern the effects of estrogen on AVM obliteration rates through subgroup analysis of females aged 12– 49 (when estrogen levels should be highest) and in postmenopausal women with and without use of estrogen replacement therapy (ERT).
tween the different estrogenic status cohorts (females aged 12– 49, men, females younger than 12 or older than or equal to 50) and outcome (obliteration of either the targeted volume or entire AVM). A subset analysis was performed to identify and model any correlation between estrogen replacement in postmenopausal female patients and outcome, as well as any correlation between oral contraceptives in estrogenic females and outcome. Multivariate logistic regression was performed to identify and model any correlation between the cardiovascular parameters (smoking, hypertension, diabetes mellitus, hypercholesterolemia) and outcome (obliteration of either the targeted volume or entire AVM).
METHODS AND MATERIALS Patients We evaluated follow-up MRI and clinical data in 351 patients who received gamma knife radiosurgery at the University of Pittsburgh between 1987 and 1994. All patients had regular clinical or imaging follow-up for a minimum of 3 years at the time of data analysis. Approval for this retrospective study was granted by the University of Pittsburgh Institutional Review Board. We reviewed each patient’s history and physical examination report to identify documentation of smoking, hypertension, diabetes mellitus, and hypercholesterolemia in the past medical history or social history portion of the history and physical examination report. These data were not available in the records of 22 patients, who were excluded, leaving 329 patients for analysis. In addition, for females the history and physical examination report was reviewed to identify also any history of estrogen replacement or birth control pills as documented in the medications list portion of the report. For each smoker, the quantity of cigarettes smoked, as pack per day and pack-years, was noted. For hypertension, diabetes mellitus, and hypercholesterolemia, the presence of each diagnosis was noted. There were 159 male patients and 170 female patients. There were 136 females aged 12– 49 years. In total, there were 113 patients who were smokers, 29 patients who had hypertension, 5 patients who had diabetes mellitus, and 4 patients who were hypercholesterolemic. Statistical analysis Multivariate logistic regression was performed using SPSS software to identify and model any correlation be-
RESULTS Cross-tabulation Table 1 shows the obliteration results tabulated for all accounted patients, classified by presence or lack of each cardiovascular disease risk factor except gender. We found that the ratio of obliterated to patent AVM is similar for smokers and nonsmokers. These results also show that all the diabetics (n ⫽ 5) and hypercholesterolemia patients (n ⫽ 4) had successful obliterations. Hypertensive patients have an obliteration percentage comparable to that found in nonhypertensive patients. Obliteration results were tabulated for the specific cohorts of patients to further evaluate the estrogenic effects on AVM obliteration; the results are shown in Table 2. We found that the estrogenic females had twice the number of patent AVM compared to men. Men and nonestrogenic females had similar percentages of patent AVM. The table also shows that of the 6 postmenopausal female patients on ERT, 33% have patent AVM, whereas of the postmenopausal patients not on ERT, only 6% have patent AVM. Multivariate analysis of cardiovascular risk factors Table 3 shows the results of multivariate logistic regression analysis of the effects of cardiovascular risk factors on obliteration. Each variable was analyzed independently in separate multivariate models with three other factors: marginal dose, marginal dose squared, and estrogenic-age females (12– 49). These three factors remained significant in each multivariate model: marginal dose (0.0003 ⬍ p ⬍ 0.0167), marginal dose squared (0.0008 ⬍ p ⬍ 0.0189), and estrogenic females (p ⫽ 0.0095 ⬍ p ⬍ 0.0109). There were
Cardiovascular risk factors in AVM obliteration
● A. BHATNAGAR et al.
971
Table 2. Raw data of in-field AVM obliteration results in different estrogen cohorts Patients
Patent AVM
Obliterated AVM
Total
Estrogenic females aged 12–49 Men Nonestrogenic females (females ⬍12, females ⱖ50)* Total Postmenopausal w/ no ERT Postmenopausal w/ ERT
30 (22%) 18 (11%) 4 (12%) 52 (16%) 1 (6%) 2 (33%)
106 (78%) 141 (89%) 30 (88%) 277 (84%) 16 (94%) 4 (66%)
136 159 34 329 17 6
* The subset of postmenopausal patients (females ⱖ50) is further categorized in the last two rows. Abbreviation: ERT ⫽ estrogen replacement therapy.
Table 3. Multivariate logistic regression analysis of cardiovascular risk factors in AVM obliteration Variable
p value
Regression coefficient ⫾ SE
Odds ratio (95% CI)
Smoking (yes/no) Smoking (pack per day) Smoking (pack-years) Hypertension Diabetes mellitus Hypercholesterolemia
0.7829 0.4335 0.4475 0.6329 0.7614 0.7801
0.094 ⫾ 0.342 0.214 ⫾ 0.273 0.010 ⫾ 0.014 0.325 ⫾ 0.681 4.931 ⫾ 16.24 5.068 ⫾ 8.156
1.09 (0.562–2.14) 1.24 (0.725–2.12) 1.01 (0.983–1.04) 1.38 (0.364–5.26) 138.5 (2.078E-12–9.23E⫹15) 158.9 (1.813E-05–1.39E⫹09)
no other significant independent variables associated with AVM obliteration in this analysis. Subset analysis Multivariate logistic regression analysis of in-field AVM obliteration was performed, using the subset of females aged 12– 49 (estrogenic females), plus all the males in the study. The patient cohort variable was analyzed, along with marginal dose and marginal dose squared for logistic regression. These two factors were shown to be statistically significant in a prior analysis and remained significant in these multivariate models. The results are shown in Table 4. We found that the females aged 12– 49 had a statistically significant lower obliteration rate (p ⫽ 0.0102) than males. A similar analysis was performed, except the subset was females aged 12– 49 vs. the smaller subset of females younger than 12 and older than or equal to 50. The trend was toward higher obliteration in females aged 12– 49. A subset multivariate logistic regression analysis was performed using only the postmenopausal females as the patient sample and the postmenopausal obliteration rate as the outcome. The variable estrogen replacement was analyzed, along with marginal dose and marginal dose squared for logistic regression. A similar subset analysis was performed for estrogenic females, except the variable used was birth control pills, and the outcome was the estrogenic female
obliteration rate. Table 5 shows these results. Estrogen replacement shows a trend toward statistical significance for a negative effect on AVM obliteration for postmenopausal women, as indicated by the negative value of the regression coefficient. Birth control pills show no statistical significance in AVM obliteration for estrogenic females. DISCUSSION This study found that the cardiovascular risk factors, except for male status, do not have a large effect on the AVM obliteration rate induced by radiosurgery, based on the multivariate logistic regression. The initial hypothesis of the study was that atherosclerosis risk factors would have a beneficial effect on the AVM obliteration rate, based on the pathophysiologic process of AVM obliteration by radiosurgery (16 –19). Specifically, smoking does not have a detectable effect on AVM obliteration, despite the relatively large sample size (n ⫽ 113) in this study. Hypertension also does not seem to have a large effect, but there are some caveats that prevent us from making a solid conclusion, such as with smoking. First of all, the sample size (n ⫽ 29) is relatively small compared to the sample size of smokers. In addition, most patients were being treated for hypertension, which may have also caused hypertension to lose its significance on obliteration.
Table 4. Multivariate analysis of estrogenic effects on AVM obliteration Subset comparison
p value
Regression coefficient ⫾ SE
Odds ratio (95% CI)
Estrogenic females vs. males Estrogenic females vs. Nonestrogenic females
0.0102
⫺0.876 ⫾ 0.341
0.416 (0.213–0.812)
0.1311
⫺1.041 ⫾ 0.690
0.353 (0.091–1.365)
972
I. J. Radiation Oncology
● Biology ● Physics
Volume 51, Number 4, 2001
Table 5. Subset analysis of postmenopausal and estrogenic females Subset
Variable
p value
Regression coefficient ⫾ SE
Odds ratio(95% CI)
Postmenopausal females
Estrogen replacement Birth control pills
0.1482
⫺2.52 ⫾ 1.72
0.080 (0.003–2.342)
Estrogenic females
5.48 ⫾ 18.03
0.7612
In the multivariate logistic regression analysis, diabetes mellitus and hypercholesterolemia did not show statistical significance on AVM obliteration. Although the cross-tabulation showed that all the patients with diabetes mellitus or hypercholesterolemia had successful obliteration, only 5 patients in this study had diabetes documented in their histories, and only 4 patients had hypercholesterolemia. From these current data, it cannot be concluded whether hypercholesterolemia and diabetes have an effect on obliteration, because their lack of statistical significance could possibly be due to the small sample size. Therefore, a larger study with further analysis needs to be performed to see whether these factors actually do have an effect on obliteration. This study found that females aged 12– 49 have a statistically significant lower obliteration rate than males. Using the cohort of females aged 12– 49, we wanted to isolate the patients with the highest levels of estrogen to more clearly identify its effects on AVM obliteration. An important limitation of this study is that we did not have measurements of estrogen levels, but relied instead on subgroup analysis by age as a rough substitute. Because this 12– 49-year-old cohort of females has a lower obliteration rate than other females (although not significantly different in subset analysis), the difference between the obliteration rate in males was even greater, with a lower p value than in the prior comparison of all females vs. males (p ⫽ 0.0102 vs. p ⫽ 0.0180). These findings suggest that estrogen could exert a vascular protective effect that limits AVM obliteration after radiosurgery in some female patients. However, the estro-
239.8 (1.08 E-135.34E⫹17)
genic females showed only a trend toward significance when compared with the “nonestrogenic” females (younger than 12 or older than or equal to 50, who had an obliteration rate similar to males), but this is probably because of the small number of patients in the nonestrogenic cohort. The label “nonestrogenic” was used for brevity, but it is well known that low levels of circulating estrogen are present in premenstrual and postmenopausal females (and in males). Further studies with greater numbers of patients in this cohort are warranted for complete analysis of the estrogenic effects on AVM obliteration. The results from the subset analysis of estrogen replacement therapy indicate a trend toward statistical significance on AVM obliteration, which further corroborates the findings showing the estrogenic effect on obliteration. Only 6 patients in the study had documentation of estrogen replacement therapy. Certainly, further analysis of a larger sample of patients needs to be performed to validate any conclusions about ERT. In summary, smoking has been clearly proven to have no effect on AVM obliteration rate; similarly, in this study hypertension, diabetes mellitus, and hypercholesterolemia had no discernible effect on AVM obliteration. With regard to gender as a risk factor, we found that male patients do have a statistically significant higher AVM obliteration rate compared to females aged 12– 49. As suggested by this study, further study is needed to establish if estrogen has a vascular protective effect that could partially limit radiosurgical AVM obliteration.
REFERENCES 1. Colombo F, Pozza F, Chierego G, Casentini L, Luca GD, Francescon P. Linear accelerator radiosurgery of cerebral arteriovenous malformations: An update. Neurosurgery 1994; 34:14 –21. 2. Ellis TL, Friedman WA, Bova FJ, Kubilis PS, Buatti JM. Analysis of treatment failure after radiosurgery for arteriovenous malformations. J Neurosurg 1998;89(1):104 –110. 3. Flickinger JC, Pollock BE, Kondziolka D, Lunsford LD. A dose-response analysis of arteriovenous malformation obliteration after radiosurgery. Int J Radiat Oncol Biol Phys 1996; 36:873– 879. 4. Friedman WA, Bova FJ, Mendenhall WM. Linear accelerator radiosurgery for arteriovenous malformations: The relationship of size to outcome. J Neurosurg 1995;82:180 –189. 5. Gobin YP, Laurent A, Merienne L, et al. Treatment of brain arteriovenous malformations by embolization and radiosurgery. J Neurosurg 1996;85(1):19 –28. 6. Karlsson B, Lax I, Soderman M. Can the probability for
7.
8.
9.
10.
11.
obliteration after radiosurgery for arteriovenous malformations be accurately predicted? Int J Radiat Oncol Biol Phys 1999;43(2):313–319. Karlsson B, Lindquist C, Steiner L. Prediction of obliteration after gamma knife surgery for cerebral arteriovenous malformations. Neurosurgery 1997;40(3):425– 430; discussion 430 – 431. Kjellberg R, Hanamura T, Davis K, Lyons S, Butler W, Adams R. Bragg-peak proton-beam therapy for arteriovenous malformations of the brain. N Engl J Med 1983;309:269 –273. Loeffler JS, Alexander E, Siddon R, et al. Stereotactic radiosurgery for intracranial arteriovenous malformations using a standard linear accelerator. Int J Radiat Oncol Biol Phys 1989;17:673– 677. Miyawaki L, Dowd C, Wara W, et al. Five year results of LINAC radiosurgery for arteriovenous malformations: Outcome for large AVMS. Int J Radiat Oncol Biol Phys 1999; 44(5):1089 –1106. Pollock BE, Flickinger JC, Lunsford LD, Maitz A, Kondziolka
Cardiovascular risk factors in AVM obliteration
12. 13. 14.
15.
16. 17.
18.
D. Factors associated with successful arteriovenous malformation radiosurgery. Neurosurgery 1998;42(6):1239 –1247. Schlienger M, Atlan D, Lefkopoulos D, et al. Linac radiosurgery for cerebral arteriovenous malformations: Results in 169 patients. Int J Radiat Oncol Biol Phys 2000;46(5):1135–1142. Steiner L, Lindquist C, Adler JR, Torner JC, Steiner M. Clinical outcome of radiosurgery for cerebral arteriovenous malformations. J Neurosurg 1992;77:1– 8. Touboul E, Al Halabi A, Buffat L, et al. Single-fraction stereotactic radiotherapy: A dose-response analysis of arteriovenous malformation obliteration. Int J Radiat Oncol Biol Phys 1998;41(4):855– 861. Flickinger JC, Maesawa S, Kondziolka D, Lunsford LD. An analysis of the clinical radiobiology of arteriovenous malformation obliteration by radiosurgery (abstract). Int J Radiat Oncol Biol Phys 2000;48,S1:255. Schneider BF, Eberhard DA, Steiner LE. Histopathology of arteriovenous malformations after gamma knife radiosurgery. J Neurosurg 1997;87(3):352–357. Szeifert GT, Kemeny AA, Timperley WR, Forster DM. The potential role of myofibroblasts in the obliteration of arteriovenous malformations after radiosurgery. Neurosurgery 1997; 40(1):61– 65; discussion 65– 66. Chang SD, Shuster DL, Steinberg GK, Levy RP, Frankel K. Stereotactic radiosurgery of arteriovenous malformations:
19.
20.
21.
22.
23.
24.
● A. BHATNAGAR et al.
973
Pathologic changes in resected tissue. Clin Neuropathol 1997; 16(2):111–116. Steinberg GK, Chang SD, Levy RP, et al. Surgical resection of large incompletely treated intracranial arteriovenous malformations following stereotactic radiosurgery. J Neurosurg 1996;84(6):920 –928. Kannel WB. Metabolic risk factors for coronary heart disease in women: Perspective from the Framingham Study. Am Heart J 1987;114:413– 419. Neaton JD, Wentworth D, for the Multiple Risk Factor Intervention Trial Research Group. Serum cholesterol, blood pressure, cigarette smoking, and death from coronary heart disease: Overall findings and differences by age for 316,099 white men. Arch Intern Med 1992;152:56 – 64. Wolf PA, D’Agostino RB, Belanger AJ, Kannel WB. Probability of stroke: A risk profile from the Framingham study. Stroke 1991;22:312–318. Hart C, Davey Smith G, Hole D. Risk factors and 20 year stroke mortality in men and women in the Renfrew/Paisley study in Scotland. Stroke 1999;30:1999 –2007. Writing Group for the PEPI Trial. Effects of estrogen or estrogen/progestin regimens on heart disease risk factors in postmenopausal women: the Postmenopausal Estrogen/Progestin Interventions (PEPI) Trial. JAMA 1995;273:199 –208.