http://informahealthcare.com/bij ISSN: 0269-9052 (print), 1362-301X (electronic) Brain Inj, 2014; 28(8): 1082–1086 ! 2014 Informa UK Ltd. DOI: 10.3109/02699052.2014.901559
Quantitative assessment of post-operative recurrence of chronic subdural haematoma using mean haematoma density Chai-Ching Lin1, Yu-Mei Lu2,3, Tzu-Hsuan Chen4, Shin-Ping Wang5, Sheng-Huang Hsiao6,7*, & Muh-Shi Lin1,8,9* 1
Department of Biotechnology and Animal Science, National Ilan University, Ilan, Taiwan, 2Department of Health Promotion and Health Education, National Taiwan Normal University, Taipei, Taiwan, 3Department of Public Health, Tzu Chi University, Hualien, Taiwan, 4Department of Physical Medicine and Rehabilitation, Mackay Memorial Hospital, Hsinchu, Taiwan, 5Office of Research and development, National Yang-Ming University, Taipei, Taiwan, 6Department of Neurosurgery, Taipei City Hospital, Renai Branch, Taipei, Taiwan, 7Institute of Traditional Medicine, School of Medicine, National Yang-Ming University, Taipei, Taiwan, 8Department of Surgery, Faculty of Medicine, School of Medicine, National Yang-Ming University, Taipei, Taiwan, and 9Department of Neurosurgery, Taipei City Hospital, Zhong Xiao Branch, Taipei, Taiwan
Abstract
Keywords
Purpose: The purpose of this study was to establish a quantitative method with which to assess the post-operative recurrence of chronic subdural haematoma (CSDH). Methods: CT scans were reviewed from 44 consecutive patients with CSDHs who underwent burr hole drainage between July 2008 and January 2012. The area of the haematoma was quantified according to the mean haematoma density (MHD) using computer-based image analysis of pre-operative brain CT scans. MHD as well as other variables of patients with and without post-operative recurrences was statistically compared. Results: Post-operative recurrence was noted in six of the 44 patients that underwent surgical procedures. Among these variables, high MHD, separated type and bilateral and skull base involvement of CSDHs were shown to be significantly related to post-operative recurrence (p50.05). Controlling for separated type in logistic regression analysis revealed the OR of MHD as statistically significant indicators with a p value of less than 0.05 (OR ¼ 1.243; 95% CI ¼ 1.003–1.54). Conclusion: This study provides statistical proof that MHD is a significant, independent, prognostic factor for the post-operative recurrence of CSDH. As such, consideration of MHD could aid in the prediction of post-operative prognosis of CSDHs.
Computer-based image analyses, chronic subdural haematoma, mean haematoma density, post-operative recurrence
Introduction Since Wepfer’s [1] report of chronic subdural haematoma (CSDH) in 1656, considerable advances have been made in characterizing the pathophysiology, diagnostic methods and management of this condition. CSDH has a high recurrence rate even after surgical treatment, ranging from 3–20% [2–5]. Unfortunately, the reasons for this high rate of recurrence have yet to be fully elucidated. The discovery of a new predictive factor could significantly benefit the diagnosis and treatment of this condition. Although there is little consensus regarding the underlying mechanism of CSDH recurrence [6, 7], a number of studies investigating the pathogenesis of recurrent CSDH have reported risk factors, including advanced age, bilateral CSDH, CSDH with cranial base extension and a tendency for bleeding such as that observed in leukaemia, liver disease and chronic renal failure [8–11].
*These authors contributed equally to this work. Correspondence: Muh-Shi Lin, MD, PhD, and Sheng-Huang Hsiao, MD, PhD, Department of Neurosurgery, Number 87, Tongde Road, Nangang District, Taipei City 115, Taiwan. Tel: +866-2-27861288. Fax: +866-227888492. E-mail: [email protected]
History Received 6 August 2013 Accepted 24 February 2014 Published online 3 April 2014
Various neuroimaging features of CSDH on computed tomography (CT) scans and magnetic resonance imaging are regarded as predictive factors of post-operative recurrence. Cases of high and mixed haematoma densities on CT scans (separated type of CSDH) have demonstrated a higher tendency toward re-bleeding [11–13]. However, these studies provide only qualitative data. No previous quantitative studies have analysed the relevant neuroimaging characteristics or compared the differences between recurrent and non-recurrent CSDHs. This research team previously reported the quantitative image analysis of the brain re-expansion rate in unilateral and bilateral CSDHs [14]. This computerassisted quantitative method has proven practical and simple to use. The current study developed a simple, computer-based method for the analysis of CT images to facilitate the quantification of density characteristics in CSDHs in terms of mean haematoma density (MHD). Then a statistical examination was conducted of its practicability in predicting postoperative recurrence. The results demonstrate the effectiveness of the proposed method of MHD calculation for the prediction of patient outcomes. To the best of the authors’ knowledge, no such method has been previously reported in the literature.
Mean haematoma density for CSDH
DOI: 10.3109/02699052.2014.901559
Materials and methods Patients This study reviewed brain CT scans and the medical reports of 46 consecutive patients with CSDHs who underwent surgery in the institute during the period between July 2008 and January 2012. Patients who had undergone craniotomy or presented with coagulopathy due to liver disease or chronic renal failure were excluded from this study. These patients underwent burr hole craniostomy drainage on one side (unilateral CSDH, n ¼ 34) or both sides (bilateral CSDHs, n ¼ 12). Two of these patients, with homogenous isodense haematoma (two unilateral), were excluded from this study due to difficulties in distinguishing the contours of the haematomas from the brain parenchyma using CT scans. This resulted in a total of 44 patients. This study was approved by the ethics committee of the institute. Follow-up CT scans of every patient were performed 24 hours after surgery to assess the extent of CSDH evacuation and subsequently at monthly intervals (up to 6 months) or until the total disappearance of the subdural collection and clinical recovery to neurologically stable status. In the event of persistent neurological deficit or clinical deterioration, CT scanning was performed earlier. A second operation was indicated for patients who were both radiologically and symptomatically proven to have had a recurrence. Specifically, surgery was recommended when persistent neurological symptoms were present and follow-up CT scans revealed an increase in subdural collection with cerebral compression on the operated side, compared to CT findings obtained 24 hours after surgery. Computer-assisted quantitative analyses of CSDHs The procedures used to trace the margins of the haematoma on brain CT scans are outlined in Figure 1 of a previous report [14]. In brief, the boundary of the haematoma was traced and defined using image analysis software (GE PACS Web System). The density of the traced haematoma was calculated and presented in Hounsfield units (HU) for each axial slice. Assessment of CSDHs in terms of MHDs This study defined MHD as the average density expressed in HU over the entire volume of CSDH. MHD in unilateral CSDH was calculated using the following formula: the mean of Ai, where Ai ¼ HU of the traced haematoma quantified for each axial CT slice (i ¼ serial CT slice number). MHD in bilateral CSDH was calculated using the following formula: the mean of Bi, where Bi ¼ the average HU of the traced haematoma on both sides of each axial CT slice (i ¼ serial CT slice number). Statistical methods Data are presented as the mean for continuous variables and number (percentage) for categorical variables. This study used the chi-square test and Mann-Whitney U-test to assess differences between recurrence and non-recurrence
1083
groups. The relationship between variables and the rate of recurrence was calculated using logistic regression analysis. Variables with an odds ratio (OR) exceeding 1 as well as a 95% confidence interval (CI) other than 1 were considered independent prognostic factors [15]. A p value of 50.05 was considered statistically significant. All analysis was performed using SPSS software (version 15.0; SPSS Inc., Chicago, IL).
Results Patient characteristics Between July 2008 and January 2012, 44 consecutive patients presenting with CSDHs underwent burr hole drainage. This included 30 males and 14 females with a mean age of 45.14 ± 5.90 years (range ¼ 34–56 years). Among these, 32 patients required unilateral burr hole drainage (72.7%) and 12 required bilateral operations (27.3%). Among the 44 patients, six patients (13.6%) presented post-operative recurrence in the presence of neurological deficit and CT scan findings [12, 16]. The average time to recurrence after burr hole drainage was 36.8 days. Patient-related data are presented in Table I. Univariate analysis for post-operative recurrence Univariate analysis was performed to examine factors associated with the post-operative recurrence of CSDH. Several factors were shown to be significantly related to the risk of recurrence. These included separated type, bilateral, skull base involvement of CSDH and MHD (Table II). The impacts of bilateral CSDHs, separated type CSDH and CSDH with skull base involvement on recurrence have previously been reported as important prognostic factors for CSDHs [9, 12, 13, 17, 18], thereby corroborating previous findings [14]. The current study, however, is unique in its identification of MHD as a significant factor closely associated with the recurrence of CSDH.
Table I. Patient demographic and clinical data (n ¼ 44). Age, years Gender, n (%) Male Female Burr hole drainage of CSDH, n (%) Unilateral burr hole drainage Bilateral burr hole drainage Pre-operative characteristics of CSDH Laterality, n (%) Unilateral Bilateral Definitive CT number (HU) Skull base involvement, n (%) Present Absent Separate type,* n (%) Present Absent Post-operative recurrence, n (%) Present Absent Duration of follow-up (months) *Nakaguchi’s classification of CSDH.
45.14 ± 5.901 (34–56) 30 (68.2%) 14 (31.8%) 32 (72.7%) 12 (27.3%) 32 (72.7%) 12 (27.3%) 21.23 ± 7.917 (9.5–38.82) 11 (25%) 33 (75%) 5 (11.4%) 39 (88.6%) 6 (13.6%) 38 (86.4%) 25.73 ± 7.934 (10–39)
1084
C.-C. Lin et al.
Brain Inj, 2014; 28(8): 1082–1086
Table II. Variable factors and clinical outcomes.
Variable Age, years Gender Female Male Characteristics of CSDH Mean haematoma density (HU) Skull base involvement Present Absent Bilateral Present Absent Separate type Present Absent
Recurrence (%) (n ¼ 6)
No recurrence (%) (n ¼ 38)
48.17 ± 6.676
44.66 ± 5.720
2 (33.3) 4 (66.7)
12 (31.6) 26 (68.4)
33 ± 5.98
19.37 ± 6.48
6 (100) 0 (0)
5 (13.2) 33 (86.8)
5 (83.3) 1 (16.7)
7 (18.4) 31 (81.6)
4 (66.7) 2 (33.3)
1 (2.6) 37 (97.4)
p Value 0.17 0.93 ( 2 ¼ 0.007)
0.0037* 0.000* ( 2 ¼ 20.84) 0.001* ( 2 ¼ 11.01) 0.000* ( 2 ¼ 21.10)
*Significant at p50.05.
Table III. Logistic analyses of factors with impact on recurrence. Variable Mean haematoma density (HU) Separate type CSDH
Odds ratio
95% CI
p Value
1.243 14.966
1.00–1.54 0.61–367.48
0.047* 0.098
*Significant at p50.05. Reference group is non-recurrence patients.
Logistic analysis of risk of post-operative recurrence in CSDH Following univariate analysis, a forward stepwise logistic model was assembled to identify independent prognostic factors. Separated type, bilateral, skull base involvement of CSDH and high MHD were closely related to recurrence. These factors were further evaluated using logistic regression analysis. Following analysis of the two variants (separated type and MHD) in a logistic regression model, it was found that, after controlling for separated type, the OR of MHD was statistically significant with a p value of less than 0.05 (OR ¼ 1.243; 95% CI ¼ 1.003–1.54). It is postulated that a one-unit increase in the MHD of CSDH could be expected to increase the odds of recurrence by a factor of 1.243 (Table III). These findings provide statistical evidence to support the idea that MHD generated using the proposed computer-assisted quantitative method can be used as a significant independent prognostic factor for the assessment of post-operative recurrence of CSDH.
Discussion The high rate of post-operative recurrence in CSDHs is a critical issue. To overcome this clinical problem, a simple model was established for image processing to facilitate the quantification of density characteristics in CSDHs using MHD. This study retrospectively reviewed the results of 44 CSDH patients who had undergone burr hole drainage. The results show that MHD confers significant benefits in predicting post-operative recurrence. The findings indicate the following: (1) MHD is a significant factor associated with the recurrence of CSDH as well as separated type, bilateral
and skull base involvement in CSDH; and (2) when controlling for separated type CSDH, MHD is independently correlated with recurrence. It is postulated that a one-unit increase in the MHD of CSDH would increase the odds of recurrence by a factor of 1.243. These findings corroborate the findings obtained in a previous study [14] and provide statistical proof that the calculated MHD is independently associated with recurrence. The data demonstrate that this quantitative technique is suitable for the prediction of outcome in patients with CSDHs. Nakaguchi et al. [12] described the respective stages in the pathogenesis of CSDHs. During the initial formation (homogenous type), the tendency toward re-bleeding is low, due to the balance between fibrinolysis and coagulation. The following stage showed greater vascularity than the homogenous stage and brain CT scans revealed a high density layer along the inner membrane (laminar type). Following the maturation of CSDHs, the tendency for re-bleeding from the neomembrane of the outer membrane was high and the blood components were separated by gravity (separated type). As the outer membrane underwent fibrosis, CSDHs resolved in the trabecular stage and the tendency for bleeding was relatively low [11, 12]. The tendency toward re-bleeding depends on the balance between coagulative and fibrinolytic activities. Furthermore, according to Kurokawa et al. [17], mixed high and low intensity signals in T2WI or low intensity signals in T1WI are the most reliable predictive factors in revealing the rapid aggravation of bilateral CSDHs. These clinical findings and qualitative analysis prompt analytical quantification. Despite the high rate of post-operative CSDH recurrence, the underlying reasons have yet to be elucidated. Many researchers have investigated the risk factors related to postoperative CSDH recurrence. Stanisic et al. [19] reported a number of risk factors including separated type CSDH, midline displacement of more than 5 mm and the appearance of acute subdural clots in CT scans. The current study demonstrates that patients with bilateral, separated type or CSDH with skull base involvement are more prone to recurrence. These findings are in agreement with those of a previous report [14] as well as other clinical reports
Mean haematoma density for CSDH
DOI: 10.3109/02699052.2014.901559
[12, 13, 17, 20]. This study moved beyond qualitative analysis by quantifying the area of pre-operative haematomas using computer-based image analysis of CT scans in terms of MHD, which was shown to be a significant independent prognostic factor associated with recurrence. Patients with CSDHs may differ with regard to the volumes of the hyperdense and hypodense components of haematoma observed in CT imaging. In addition, these differences in the density of haematomas may present different clinical manifestations. CSDHs of higher density were not fully matured where repeated micro-haemorrhages from the neocapillary network in the outer membrane took place. The active growth of blood vessels into the membrane of CSDH was shown to be proportional to the amount of re-bleeding in the haematoma cavity, resulting in CSDHs of higher density. In stages associated with a higher rate of recurrence (i.e. separated type or high density homogenous type) [11–13], the ratio of hyperdense to hypodense components within a haematoma tends to be higher. The proposed quantitative method determines the MHD of an entire subdural haematoma, providing an overall average value based on HU as a standard. Thus, patients with a majority of hyperdense components within a haematoma (as determined via quantitative methods) tend to have higher MHD. When the two variants (separated type and MHD) were analysed in a logistic regression model, it was found that, after controlling for separated type, the OR of MHD was statistically significant with a p value of less than 0.05 (OR ¼ 1.243; 95% CI ¼ 1.003–1.54). It is postulated that a one-unit increase in the MHD of CSDH would increase the odds of recurrence by a factor of 1.243. The results indicate that this simple, quantitative method could be effectively implemented in daily clinical practice. Furthermore, histogram analysis using HU is currently used to predict lymph node metastasis and tumour invasiveness in lung adenocarcinoma [21]. This study was subject to a number of limitations. First, this method is not applicable to patients with homogeneous isodense CSDHs, because the contours of the haematoma cannot be clearly distinguished from brain parenchyma in CT scans, which can affect the accuracy of MHD calculation [14]. This study excluded patients with homogeneous isodense CSDHs (the most common). One previous study reported that the incidence of CSDH recurrence in high and mixed density groups was significantly higher than that observed in low and isodensity groups (p50.001) [11]. A key objective in this study was the development of a simple, practical method with which to quantify the density of CSDHs in terms of MHD. Specifically, the level of MHD tends to be lower in low/ isodensity CSDHs and higher in high/mixed density CSDHs. Thus, patients with high and mixed density CSDHs should be aware of the high incidence of CSDH recurrence and be cautious with regard to high MHD levels. Although the data obtained is statistically significant, that study included a relatively small number of patients. Further quantitative studies of patients with CSDHs should be undertaken. The use of the proposed model would enable clinicians to alert physicians to the possibility of CSDHs prior to surgery. For patients without neurological deficits but high MHD, the operation could be delayed until MHD levels have subsided. However, in these situations, neurological observation should
1085
be strictly followed because high MHD is significantly related to the likelihood of post-operative recurrence. Moreover, patients with high MHD who undergo surgical treatments should be monitored particularly closely during peri-operative haemostasis, post-operative critical care and long term followup. It is believed that this pilot study deserves application and assessment in clinical practice. Further longitudinal studies are warranted to examine its practicability in the prediction of post-operative recurrence.
Conclusion This study provided quantitative evidence related to the efficacy of using MHD for the evaluation of CSDH. Patients with higher MHD are more likely to undergo post-operative recurrence; therefore, these patients deserve greater clinical attention. To that end, the proposed image processing method enables the rapid screening of patients with CSDHs.
Declaration of interest The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.
References 1. Sambasivan M. An overview of chronic subdural hematoma: Experience with 2300 cases. Surgical Neurology 1997;47:418–422. 2. Laumer R. Implantation of a reservoir for refractory chronic subdural hematoma. Neurosurgery 2002;50:672. 3. Arbit E, Patterson Jr RH, Fraser RA. An implantable subdural drain for treatment of chronic subdural hematoma. Surgical Neurology 1981;15:175–177. 4. Ernestus RI, Beldzinski P, Lanfermann H, Klug N. Chronic subdural hematoma: Surgical treatment and outcome in 104 patients. Surgical Neurology 1997;48:220–225. 5. Mellergard P, Wisten O. Operations and re-operations for chronic subdural haematomas during a 25-year period in a well defined population. Acta Neurochirurgica (Wien) 1996;138: 708–713. 6. Lee KS, Bae WK, Doh JW, Bae HG, Yun IG. Origin of chronic subdural haematoma and relation to traumatic subdural lesions. Brain Injury 1998;12:901–910. 7. Markwalder TM, Seiler RW. Chronic subdural hematomas: To drain or not to drain? Neurosurgery 1985;16:185–188. 8. Frati A, Salvati M, Mainiero F, Ippoliti F, Rocchi G, Raco A, Caroli E, Cantore G, Delfini R. Inflammation markers and risk factors for recurrence in 35 patients with a posttraumatic chronic subdural hematoma: A prospective study. Journal of Neurosurgery 2004;100: 24–32. 9. Torihashi K, Sadamasa N, Yoshida K, Narumi O, Chin M, Yamagata S. Independent predictors for recurrence of chronic subdural hematoma: A review of 343 consecutive surgical cases. Neurosurgery 2008;63:1125–1129. 10. Kang HL, Shin HS, Kim TH. Clinical analysis of recurrent chronic subdural hematoma. Journal of Korean Neurosurgical Society 2006; 40:262–266. 11. Ko BS, Lee JK, Seo BR, Moon SJ, Kim JH, Kim SH. Clinical analysis of risk factors related to recurrent chronic subdural hematoma. Journal of Korean Neurosurgical Society 2008;43: 11–15. 12. Nakaguchi H, Tanishima T, Yoshimasu N. Factors in the natural history of chronic subdural hematomas that influence their postoperative recurrence. Journal of Neurosurgery 2001;95: 256–262. 13. Nomura S, Kashiwagi S, Fujisawa H, Ito H, Nakamura K. Characterization of local hyperfibrinolysis in chronic subdural hematomas by SDS-PAGE and immunoblot. Journal of Neurosurgery 1994;81:910–913. 14. Kung WM, Hung KS, Chiu WT, Tsai SH, Lin JW, Wang YC, Lin MS. Quantitative assessment of impaired postevacuation
1086
15. 16. 17. 18.
C.-C. Lin et al.
brain re-expansion in bilateral chronic subdural haematoma: Possible mechanism of the higher recurrence rate. Injury 2012; 43:598–602. Kim YI, Lee JH, Park SW, Nam TK, Min BK, Hwang SN. Analysis of prognostic factors for chronic subdural hematoma. Journal of Korean Neurosurgical Society 2008;4:14–18. Oh HJ, Lee KS, Shim JJ, Yoon SM, Yun IG, Bae HG. Postoperative course and recurrence of chronic subdural hematoma. Journal of Korean Neurosurgical Society 2010;48:518–523. Kurokawa Y, Ishizaki E, Inaba K. Bilateral chronic subdural hematoma cases showing rapid and progressive aggravation. Surgical Neurology 2005;64:444–449. Kang MS, Koh HS, Kwon HJ, Choi SW, Kim SH, Youm JY. Factors influencing recurrent chronic subdural hematoma
Brain Inj, 2014; 28(8): 1082–1086
after surgery. Journal of Korean Neurosurgical Society 2007;41: 11–15. 19. Stanisic M, Lund-Johansen M, Mahesparan R. Treatment of chronic subdural hematoma by burr-hole craniostomy in adults: Influence of some factors on postoperative recurrence. Acta Neurochirurgica (Wien) 2005;147:1249–1256. 20. Oishi M, Toyama M, Tamatani S, Kitazawa T, Saito M. Clinical factors of recurrent chronic subdural hematoma. Neurologia Medico Chirurgica (Tokyo) 2001;41:382–386. 21. Nomori H, Ohtsuka T, Naruke T, Suemasu K. Histogram analysis of computed tomography numbers of clinical T1 N0 M0 lung adenocarcinoma, with special reference to lymph node metastasis and tumor invasiveness. Journal of Thoracic & Cardiovascular Surgery 2003;126:1584–1589.
Copyright of Brain Injury is the property of Taylor & Francis Ltd and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
Quantitative assessment of post-operative recurrence of chronic subdural haematoma using mean haematoma density Chai-Ching Lin1, Yu-Mei Lu2,3, Tzu-Hsuan Chen4, Shin-Ping Wang5, Sheng-Huang Hsiao6,7*, & Muh-Shi Lin1,8,9* 1
Department of Biotechnology and Animal Science, National Ilan University, Ilan, Taiwan, 2Department of Health Promotion and Health Education, National Taiwan Normal University, Taipei, Taiwan, 3Department of Public Health, Tzu Chi University, Hualien, Taiwan, 4Department of Physical Medicine and Rehabilitation, Mackay Memorial Hospital, Hsinchu, Taiwan, 5Office of Research and development, National Yang-Ming University, Taipei, Taiwan, 6Department of Neurosurgery, Taipei City Hospital, Renai Branch, Taipei, Taiwan, 7Institute of Traditional Medicine, School of Medicine, National Yang-Ming University, Taipei, Taiwan, 8Department of Surgery, Faculty of Medicine, School of Medicine, National Yang-Ming University, Taipei, Taiwan, and 9Department of Neurosurgery, Taipei City Hospital, Zhong Xiao Branch, Taipei, Taiwan
Abstract
Keywords
Purpose: The purpose of this study was to establish a quantitative method with which to assess the post-operative recurrence of chronic subdural haematoma (CSDH). Methods: CT scans were reviewed from 44 consecutive patients with CSDHs who underwent burr hole drainage between July 2008 and January 2012. The area of the haematoma was quantified according to the mean haematoma density (MHD) using computer-based image analysis of pre-operative brain CT scans. MHD as well as other variables of patients with and without post-operative recurrences was statistically compared. Results: Post-operative recurrence was noted in six of the 44 patients that underwent surgical procedures. Among these variables, high MHD, separated type and bilateral and skull base involvement of CSDHs were shown to be significantly related to post-operative recurrence (p50.05). Controlling for separated type in logistic regression analysis revealed the OR of MHD as statistically significant indicators with a p value of less than 0.05 (OR ¼ 1.243; 95% CI ¼ 1.003–1.54). Conclusion: This study provides statistical proof that MHD is a significant, independent, prognostic factor for the post-operative recurrence of CSDH. As such, consideration of MHD could aid in the prediction of post-operative prognosis of CSDHs.
Computer-based image analyses, chronic subdural haematoma, mean haematoma density, post-operative recurrence
Introduction Since Wepfer’s [1] report of chronic subdural haematoma (CSDH) in 1656, considerable advances have been made in characterizing the pathophysiology, diagnostic methods and management of this condition. CSDH has a high recurrence rate even after surgical treatment, ranging from 3–20% [2–5]. Unfortunately, the reasons for this high rate of recurrence have yet to be fully elucidated. The discovery of a new predictive factor could significantly benefit the diagnosis and treatment of this condition. Although there is little consensus regarding the underlying mechanism of CSDH recurrence [6, 7], a number of studies investigating the pathogenesis of recurrent CSDH have reported risk factors, including advanced age, bilateral CSDH, CSDH with cranial base extension and a tendency for bleeding such as that observed in leukaemia, liver disease and chronic renal failure [8–11].
*These authors contributed equally to this work. Correspondence: Muh-Shi Lin, MD, PhD, and Sheng-Huang Hsiao, MD, PhD, Department of Neurosurgery, Number 87, Tongde Road, Nangang District, Taipei City 115, Taiwan. Tel: +866-2-27861288. Fax: +866-227888492. E-mail: [email protected]
History Received 6 August 2013 Accepted 24 February 2014 Published online 3 April 2014
Various neuroimaging features of CSDH on computed tomography (CT) scans and magnetic resonance imaging are regarded as predictive factors of post-operative recurrence. Cases of high and mixed haematoma densities on CT scans (separated type of CSDH) have demonstrated a higher tendency toward re-bleeding [11–13]. However, these studies provide only qualitative data. No previous quantitative studies have analysed the relevant neuroimaging characteristics or compared the differences between recurrent and non-recurrent CSDHs. This research team previously reported the quantitative image analysis of the brain re-expansion rate in unilateral and bilateral CSDHs [14]. This computerassisted quantitative method has proven practical and simple to use. The current study developed a simple, computer-based method for the analysis of CT images to facilitate the quantification of density characteristics in CSDHs in terms of mean haematoma density (MHD). Then a statistical examination was conducted of its practicability in predicting postoperative recurrence. The results demonstrate the effectiveness of the proposed method of MHD calculation for the prediction of patient outcomes. To the best of the authors’ knowledge, no such method has been previously reported in the literature.
Mean haematoma density for CSDH
DOI: 10.3109/02699052.2014.901559
Materials and methods Patients This study reviewed brain CT scans and the medical reports of 46 consecutive patients with CSDHs who underwent surgery in the institute during the period between July 2008 and January 2012. Patients who had undergone craniotomy or presented with coagulopathy due to liver disease or chronic renal failure were excluded from this study. These patients underwent burr hole craniostomy drainage on one side (unilateral CSDH, n ¼ 34) or both sides (bilateral CSDHs, n ¼ 12). Two of these patients, with homogenous isodense haematoma (two unilateral), were excluded from this study due to difficulties in distinguishing the contours of the haematomas from the brain parenchyma using CT scans. This resulted in a total of 44 patients. This study was approved by the ethics committee of the institute. Follow-up CT scans of every patient were performed 24 hours after surgery to assess the extent of CSDH evacuation and subsequently at monthly intervals (up to 6 months) or until the total disappearance of the subdural collection and clinical recovery to neurologically stable status. In the event of persistent neurological deficit or clinical deterioration, CT scanning was performed earlier. A second operation was indicated for patients who were both radiologically and symptomatically proven to have had a recurrence. Specifically, surgery was recommended when persistent neurological symptoms were present and follow-up CT scans revealed an increase in subdural collection with cerebral compression on the operated side, compared to CT findings obtained 24 hours after surgery. Computer-assisted quantitative analyses of CSDHs The procedures used to trace the margins of the haematoma on brain CT scans are outlined in Figure 1 of a previous report [14]. In brief, the boundary of the haematoma was traced and defined using image analysis software (GE PACS Web System). The density of the traced haematoma was calculated and presented in Hounsfield units (HU) for each axial slice. Assessment of CSDHs in terms of MHDs This study defined MHD as the average density expressed in HU over the entire volume of CSDH. MHD in unilateral CSDH was calculated using the following formula: the mean of Ai, where Ai ¼ HU of the traced haematoma quantified for each axial CT slice (i ¼ serial CT slice number). MHD in bilateral CSDH was calculated using the following formula: the mean of Bi, where Bi ¼ the average HU of the traced haematoma on both sides of each axial CT slice (i ¼ serial CT slice number). Statistical methods Data are presented as the mean for continuous variables and number (percentage) for categorical variables. This study used the chi-square test and Mann-Whitney U-test to assess differences between recurrence and non-recurrence
1083
groups. The relationship between variables and the rate of recurrence was calculated using logistic regression analysis. Variables with an odds ratio (OR) exceeding 1 as well as a 95% confidence interval (CI) other than 1 were considered independent prognostic factors [15]. A p value of 50.05 was considered statistically significant. All analysis was performed using SPSS software (version 15.0; SPSS Inc., Chicago, IL).
Results Patient characteristics Between July 2008 and January 2012, 44 consecutive patients presenting with CSDHs underwent burr hole drainage. This included 30 males and 14 females with a mean age of 45.14 ± 5.90 years (range ¼ 34–56 years). Among these, 32 patients required unilateral burr hole drainage (72.7%) and 12 required bilateral operations (27.3%). Among the 44 patients, six patients (13.6%) presented post-operative recurrence in the presence of neurological deficit and CT scan findings [12, 16]. The average time to recurrence after burr hole drainage was 36.8 days. Patient-related data are presented in Table I. Univariate analysis for post-operative recurrence Univariate analysis was performed to examine factors associated with the post-operative recurrence of CSDH. Several factors were shown to be significantly related to the risk of recurrence. These included separated type, bilateral, skull base involvement of CSDH and MHD (Table II). The impacts of bilateral CSDHs, separated type CSDH and CSDH with skull base involvement on recurrence have previously been reported as important prognostic factors for CSDHs [9, 12, 13, 17, 18], thereby corroborating previous findings [14]. The current study, however, is unique in its identification of MHD as a significant factor closely associated with the recurrence of CSDH.
Table I. Patient demographic and clinical data (n ¼ 44). Age, years Gender, n (%) Male Female Burr hole drainage of CSDH, n (%) Unilateral burr hole drainage Bilateral burr hole drainage Pre-operative characteristics of CSDH Laterality, n (%) Unilateral Bilateral Definitive CT number (HU) Skull base involvement, n (%) Present Absent Separate type,* n (%) Present Absent Post-operative recurrence, n (%) Present Absent Duration of follow-up (months) *Nakaguchi’s classification of CSDH.
45.14 ± 5.901 (34–56) 30 (68.2%) 14 (31.8%) 32 (72.7%) 12 (27.3%) 32 (72.7%) 12 (27.3%) 21.23 ± 7.917 (9.5–38.82) 11 (25%) 33 (75%) 5 (11.4%) 39 (88.6%) 6 (13.6%) 38 (86.4%) 25.73 ± 7.934 (10–39)
1084
C.-C. Lin et al.
Brain Inj, 2014; 28(8): 1082–1086
Table II. Variable factors and clinical outcomes.
Variable Age, years Gender Female Male Characteristics of CSDH Mean haematoma density (HU) Skull base involvement Present Absent Bilateral Present Absent Separate type Present Absent
Recurrence (%) (n ¼ 6)
No recurrence (%) (n ¼ 38)
48.17 ± 6.676
44.66 ± 5.720
2 (33.3) 4 (66.7)
12 (31.6) 26 (68.4)
33 ± 5.98
19.37 ± 6.48
6 (100) 0 (0)
5 (13.2) 33 (86.8)
5 (83.3) 1 (16.7)
7 (18.4) 31 (81.6)
4 (66.7) 2 (33.3)
1 (2.6) 37 (97.4)
p Value 0.17 0.93 ( 2 ¼ 0.007)
0.0037* 0.000* ( 2 ¼ 20.84) 0.001* ( 2 ¼ 11.01) 0.000* ( 2 ¼ 21.10)
*Significant at p50.05.
Table III. Logistic analyses of factors with impact on recurrence. Variable Mean haematoma density (HU) Separate type CSDH
Odds ratio
95% CI
p Value
1.243 14.966
1.00–1.54 0.61–367.48
0.047* 0.098
*Significant at p50.05. Reference group is non-recurrence patients.
Logistic analysis of risk of post-operative recurrence in CSDH Following univariate analysis, a forward stepwise logistic model was assembled to identify independent prognostic factors. Separated type, bilateral, skull base involvement of CSDH and high MHD were closely related to recurrence. These factors were further evaluated using logistic regression analysis. Following analysis of the two variants (separated type and MHD) in a logistic regression model, it was found that, after controlling for separated type, the OR of MHD was statistically significant with a p value of less than 0.05 (OR ¼ 1.243; 95% CI ¼ 1.003–1.54). It is postulated that a one-unit increase in the MHD of CSDH could be expected to increase the odds of recurrence by a factor of 1.243 (Table III). These findings provide statistical evidence to support the idea that MHD generated using the proposed computer-assisted quantitative method can be used as a significant independent prognostic factor for the assessment of post-operative recurrence of CSDH.
Discussion The high rate of post-operative recurrence in CSDHs is a critical issue. To overcome this clinical problem, a simple model was established for image processing to facilitate the quantification of density characteristics in CSDHs using MHD. This study retrospectively reviewed the results of 44 CSDH patients who had undergone burr hole drainage. The results show that MHD confers significant benefits in predicting post-operative recurrence. The findings indicate the following: (1) MHD is a significant factor associated with the recurrence of CSDH as well as separated type, bilateral
and skull base involvement in CSDH; and (2) when controlling for separated type CSDH, MHD is independently correlated with recurrence. It is postulated that a one-unit increase in the MHD of CSDH would increase the odds of recurrence by a factor of 1.243. These findings corroborate the findings obtained in a previous study [14] and provide statistical proof that the calculated MHD is independently associated with recurrence. The data demonstrate that this quantitative technique is suitable for the prediction of outcome in patients with CSDHs. Nakaguchi et al. [12] described the respective stages in the pathogenesis of CSDHs. During the initial formation (homogenous type), the tendency toward re-bleeding is low, due to the balance between fibrinolysis and coagulation. The following stage showed greater vascularity than the homogenous stage and brain CT scans revealed a high density layer along the inner membrane (laminar type). Following the maturation of CSDHs, the tendency for re-bleeding from the neomembrane of the outer membrane was high and the blood components were separated by gravity (separated type). As the outer membrane underwent fibrosis, CSDHs resolved in the trabecular stage and the tendency for bleeding was relatively low [11, 12]. The tendency toward re-bleeding depends on the balance between coagulative and fibrinolytic activities. Furthermore, according to Kurokawa et al. [17], mixed high and low intensity signals in T2WI or low intensity signals in T1WI are the most reliable predictive factors in revealing the rapid aggravation of bilateral CSDHs. These clinical findings and qualitative analysis prompt analytical quantification. Despite the high rate of post-operative CSDH recurrence, the underlying reasons have yet to be elucidated. Many researchers have investigated the risk factors related to postoperative CSDH recurrence. Stanisic et al. [19] reported a number of risk factors including separated type CSDH, midline displacement of more than 5 mm and the appearance of acute subdural clots in CT scans. The current study demonstrates that patients with bilateral, separated type or CSDH with skull base involvement are more prone to recurrence. These findings are in agreement with those of a previous report [14] as well as other clinical reports
Mean haematoma density for CSDH
DOI: 10.3109/02699052.2014.901559
[12, 13, 17, 20]. This study moved beyond qualitative analysis by quantifying the area of pre-operative haematomas using computer-based image analysis of CT scans in terms of MHD, which was shown to be a significant independent prognostic factor associated with recurrence. Patients with CSDHs may differ with regard to the volumes of the hyperdense and hypodense components of haematoma observed in CT imaging. In addition, these differences in the density of haematomas may present different clinical manifestations. CSDHs of higher density were not fully matured where repeated micro-haemorrhages from the neocapillary network in the outer membrane took place. The active growth of blood vessels into the membrane of CSDH was shown to be proportional to the amount of re-bleeding in the haematoma cavity, resulting in CSDHs of higher density. In stages associated with a higher rate of recurrence (i.e. separated type or high density homogenous type) [11–13], the ratio of hyperdense to hypodense components within a haematoma tends to be higher. The proposed quantitative method determines the MHD of an entire subdural haematoma, providing an overall average value based on HU as a standard. Thus, patients with a majority of hyperdense components within a haematoma (as determined via quantitative methods) tend to have higher MHD. When the two variants (separated type and MHD) were analysed in a logistic regression model, it was found that, after controlling for separated type, the OR of MHD was statistically significant with a p value of less than 0.05 (OR ¼ 1.243; 95% CI ¼ 1.003–1.54). It is postulated that a one-unit increase in the MHD of CSDH would increase the odds of recurrence by a factor of 1.243. The results indicate that this simple, quantitative method could be effectively implemented in daily clinical practice. Furthermore, histogram analysis using HU is currently used to predict lymph node metastasis and tumour invasiveness in lung adenocarcinoma [21]. This study was subject to a number of limitations. First, this method is not applicable to patients with homogeneous isodense CSDHs, because the contours of the haematoma cannot be clearly distinguished from brain parenchyma in CT scans, which can affect the accuracy of MHD calculation [14]. This study excluded patients with homogeneous isodense CSDHs (the most common). One previous study reported that the incidence of CSDH recurrence in high and mixed density groups was significantly higher than that observed in low and isodensity groups (p50.001) [11]. A key objective in this study was the development of a simple, practical method with which to quantify the density of CSDHs in terms of MHD. Specifically, the level of MHD tends to be lower in low/ isodensity CSDHs and higher in high/mixed density CSDHs. Thus, patients with high and mixed density CSDHs should be aware of the high incidence of CSDH recurrence and be cautious with regard to high MHD levels. Although the data obtained is statistically significant, that study included a relatively small number of patients. Further quantitative studies of patients with CSDHs should be undertaken. The use of the proposed model would enable clinicians to alert physicians to the possibility of CSDHs prior to surgery. For patients without neurological deficits but high MHD, the operation could be delayed until MHD levels have subsided. However, in these situations, neurological observation should
1085
be strictly followed because high MHD is significantly related to the likelihood of post-operative recurrence. Moreover, patients with high MHD who undergo surgical treatments should be monitored particularly closely during peri-operative haemostasis, post-operative critical care and long term followup. It is believed that this pilot study deserves application and assessment in clinical practice. Further longitudinal studies are warranted to examine its practicability in the prediction of post-operative recurrence.
Conclusion This study provided quantitative evidence related to the efficacy of using MHD for the evaluation of CSDH. Patients with higher MHD are more likely to undergo post-operative recurrence; therefore, these patients deserve greater clinical attention. To that end, the proposed image processing method enables the rapid screening of patients with CSDHs.
Declaration of interest The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.
References 1. Sambasivan M. An overview of chronic subdural hematoma: Experience with 2300 cases. Surgical Neurology 1997;47:418–422. 2. Laumer R. Implantation of a reservoir for refractory chronic subdural hematoma. Neurosurgery 2002;50:672. 3. Arbit E, Patterson Jr RH, Fraser RA. An implantable subdural drain for treatment of chronic subdural hematoma. Surgical Neurology 1981;15:175–177. 4. Ernestus RI, Beldzinski P, Lanfermann H, Klug N. Chronic subdural hematoma: Surgical treatment and outcome in 104 patients. Surgical Neurology 1997;48:220–225. 5. Mellergard P, Wisten O. Operations and re-operations for chronic subdural haematomas during a 25-year period in a well defined population. Acta Neurochirurgica (Wien) 1996;138: 708–713. 6. Lee KS, Bae WK, Doh JW, Bae HG, Yun IG. Origin of chronic subdural haematoma and relation to traumatic subdural lesions. Brain Injury 1998;12:901–910. 7. Markwalder TM, Seiler RW. Chronic subdural hematomas: To drain or not to drain? Neurosurgery 1985;16:185–188. 8. Frati A, Salvati M, Mainiero F, Ippoliti F, Rocchi G, Raco A, Caroli E, Cantore G, Delfini R. Inflammation markers and risk factors for recurrence in 35 patients with a posttraumatic chronic subdural hematoma: A prospective study. Journal of Neurosurgery 2004;100: 24–32. 9. Torihashi K, Sadamasa N, Yoshida K, Narumi O, Chin M, Yamagata S. Independent predictors for recurrence of chronic subdural hematoma: A review of 343 consecutive surgical cases. Neurosurgery 2008;63:1125–1129. 10. Kang HL, Shin HS, Kim TH. Clinical analysis of recurrent chronic subdural hematoma. Journal of Korean Neurosurgical Society 2006; 40:262–266. 11. Ko BS, Lee JK, Seo BR, Moon SJ, Kim JH, Kim SH. Clinical analysis of risk factors related to recurrent chronic subdural hematoma. Journal of Korean Neurosurgical Society 2008;43: 11–15. 12. Nakaguchi H, Tanishima T, Yoshimasu N. Factors in the natural history of chronic subdural hematomas that influence their postoperative recurrence. Journal of Neurosurgery 2001;95: 256–262. 13. Nomura S, Kashiwagi S, Fujisawa H, Ito H, Nakamura K. Characterization of local hyperfibrinolysis in chronic subdural hematomas by SDS-PAGE and immunoblot. Journal of Neurosurgery 1994;81:910–913. 14. Kung WM, Hung KS, Chiu WT, Tsai SH, Lin JW, Wang YC, Lin MS. Quantitative assessment of impaired postevacuation
1086
15. 16. 17. 18.
C.-C. Lin et al.
brain re-expansion in bilateral chronic subdural haematoma: Possible mechanism of the higher recurrence rate. Injury 2012; 43:598–602. Kim YI, Lee JH, Park SW, Nam TK, Min BK, Hwang SN. Analysis of prognostic factors for chronic subdural hematoma. Journal of Korean Neurosurgical Society 2008;4:14–18. Oh HJ, Lee KS, Shim JJ, Yoon SM, Yun IG, Bae HG. Postoperative course and recurrence of chronic subdural hematoma. Journal of Korean Neurosurgical Society 2010;48:518–523. Kurokawa Y, Ishizaki E, Inaba K. Bilateral chronic subdural hematoma cases showing rapid and progressive aggravation. Surgical Neurology 2005;64:444–449. Kang MS, Koh HS, Kwon HJ, Choi SW, Kim SH, Youm JY. Factors influencing recurrent chronic subdural hematoma
Brain Inj, 2014; 28(8): 1082–1086
after surgery. Journal of Korean Neurosurgical Society 2007;41: 11–15. 19. Stanisic M, Lund-Johansen M, Mahesparan R. Treatment of chronic subdural hematoma by burr-hole craniostomy in adults: Influence of some factors on postoperative recurrence. Acta Neurochirurgica (Wien) 2005;147:1249–1256. 20. Oishi M, Toyama M, Tamatani S, Kitazawa T, Saito M. Clinical factors of recurrent chronic subdural hematoma. Neurologia Medico Chirurgica (Tokyo) 2001;41:382–386. 21. Nomori H, Ohtsuka T, Naruke T, Suemasu K. Histogram analysis of computed tomography numbers of clinical T1 N0 M0 lung adenocarcinoma, with special reference to lymph node metastasis and tumor invasiveness. Journal of Thoracic & Cardiovascular Surgery 2003;126:1584–1589.
Copyright of Brain Injury is the property of Taylor & Francis Ltd and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
