- Email: [email protected]
Predicting Outcomes After Glioma Surgery: Model Behavior
Perspectives Commentary on: Craniotomy for Glioma Resection: A Predictive Model by Missios et al. World Neurosurg 83:957-964, 2015
Predicting Outcomes After Glioma Surgery: Model Behavior Randy L. Jensen
T
he ability to predict outcomes for individual patients undergoing craniotomy for resection of gliomas would be of great value to physicians, patients, families, and thirdparty payers. For instance, having precise knowledge of the personal potential for complications as well as the relative frequency of these unanticipated outcomes could help prepare patients and their loved ones for appropriate postoperative recovery. At present, of course, this is not a reality; however, neurosurgical oncologists are well versed in using molecular biomarkers such as isodehydrogenase-1 or methylguanine methyltransferase methylation status as well as histologic grade to help predict patient overall survival. They also have convincing evidence of the influence of their surgical efforts and patient demographics on overall survival. It is understood that overall survival and progression-free survival can be predicted by age, preoperative Karnofsky Performance Score, and extent of resection (8, 12). Nevertheless, it has not been possible to provide a patient with an individualized expectation of risk for postoperative complications. In a paper recently published in WORLD NEUROSURGERY, Missios et al. have attempted the task of determining preoperative predictive indicators of potential postoperative complications for patients undergoing craniotomy for glioma resection. They have a track record of doing similar analysis for other intracranial tumors: In a study similar to this, they described a predictive model of unfavorable outcomes for patients undergoing resection of benign intracranial tumors (2). Using multivariate analysis of a national database, they identified the effect of various risk factors on the incidence of death (1.3%), unfavorable discharge (22.7%), hydrocephalus (4.2%), cardiac (1.1%) or respiratory
Key words Craniotomy - Glioma - National Inpatient Sample - Risk prediction -
894
Abbreviations and Acronyms DVT: Deep venous thrombosis
www.SCIENCEDIRECT.com
(0.9%) complications, deep venous thrombosis (DVT) (0.5%), pulmonary embolus (2.3%), and acute renal failure (1.5%). In their article in this month’s issue, they evaluated the records of over 20,000 patients with glioma included in a national database who underwent craniotomy and identified similar inpatient postoperative risks, including a 1.6% risk of mortality, and morbidity risk of 25.8% for discharge to rehabilitation, 4.0% for treated hydrocephalus, 0.7% for cardiac complications, 0.5% for respiratory complications, 0.8% for deep wound infection, 0.6% for DVT, 3.1% for pulmonary embolus, and 1.3% for acute renal failure. These numbers proved to be very similar for benign and malignant intracranial disease. Mortality is almost identical in both studies, as is true of most of the other postoperative complications. Discharge to rehabilitation is slightly higher for the glioma group but the difference is small and it is unclear whether it would be of statistical significance. Once again, the authors developed predictive models for individual complications based on a logistic regression analysis and validated in a bootstrapped sample for these same complications. They found that age, minority status, vascular morbidities, other medical conditions, history of coagulopathy, and hyponatremia all affected the risks for negative outcomes. The immediate criticism of both of these studies is that they only describe inpatient complications, and all outcome measures were recorded before the patient was discharged from the acute hospital setting. Although it is true that most complications occur during this period, this does not tell the whole story of the
Department of Neurosurgery, Clinical Neurosciences Center, University of Utah, Salt Lake City, Utah, USA To whom correspondence should be addressed: Randy L. Jensen, M.D., Ph.D. [E-mail: [email protected]] Citation: World Neurosurg. (2015) 84, 4:894-896. http://dx.doi.org/10.1016/j.wneu.2015.05.018
WORLD NEUROSURGERY, http://dx.doi.org/10.1016/j.wneu.2015.05.018
PERSPECTIVES
ultimate outcome of a given patient. One could ask what happens to the patient as an outpatient? Do the functional outcomes of unfavorable discharges improve over time after receiving physical and occupational therapy? This would certainly decrease the rates of unfavorable outcomes after craniotomy for tumor, as has been reported in other publications (10, 14). This is even more evident when considering cognitive deficits, which have been shown to take longer than 3 months for recovery, especially when the operation is located within the language domain (13). On the other hand, the approach of recording complications at the time of discharge misses delayed unexpected outcomes. For instance, a DVT that is discovered or occurs in an outpatient, a frequent occurrence, would be underestimated in this study. In fact, both of these studies have very low rates of pulmonary embolus and DVT compared with many contemporary series. The most intriguing portion of this report is the rate of discharge to rehabilitation/unfavorable discharge. Bekelis et al. report that the risk of this outcome was approximately 25% and found it was slightly higher for the glioma group. It would be really helpful to know more about patients with this outcome. What were the reasons for this need for further inpatient care? Presumably, it was related to new or existing neurologic deficits but the authors are unable to make this distinction because of the limitations of an administrative database. It would be interesting to separate out patients who were admitted with existing neurologic deficits preoperatively that would have necessitated inpatient rehabilitation regardless of the surgical intervention. One potential takehome message from these studies is that a quarter of patients undergoing craniotomy for tumor will require inpatient rehabilitation after their surgery. This is a useful fact to give patients and families in preoperative counseling. Even more useful would be the ability to determine exactly which patient characteristics make a given patient more at risk of a given new neurologic deficit. The ultimate goal would be to predict the chance of a specific neurologic deficit for a given lesion in a particular location of the brain. For instance, what is the relative risk for a patient to experience a language deficit after resection of an insular oligodendroglioma World Health Organization grade II with isodehydrogenase-1 mutation and sparing of the medial lenticulostriate arteries? At present, that is a “pie in
REFERENCES 1. Abu Hamdeh S, Lytsy B, Ronne-Engstrom E: Surgical site infections in standard neurosurgery procedures—a study of incidence, impact and potential risk factors. Br J Neurosurg 28:270-275, 2014. 2. Bekelis K, Kalakoti P, Nanda A, Missios S: A predictive model of unfavorable outcomes after benign intracranial tumor resection. World Neurosurg 2015: In press.
the sky” wish but this study sets the groundwork to move in this direction. In the past, most studies on craniotomy outcomes focused on infection rates (1, 4-7). One early exception was a study examining the complication rate after craniotomy for tumor (3). Brell et al. found a similar overall major complication rate of 27.5%, with neurologic complications being the most frequently encountered. They did not find a statistically significant relationship between complication rate and eloquence of the tumoral area. Infratentorial tumor location, previous radiotherapy, and reoperations were strongly related to the incidence of regional complications. Age over 60 years and severe concomitant disease were risk factors for systemic complications. In a similar fashion, others have demonstrated the difficulty in predicting postoperative functional outcome even for gliomas in eloquent areas (14). A more recent attempt to identify preoperative risk factors in a single-institution study found a similar in-hospital mortality rate of 1.0% (11). In-hospital systemic or infectious complications and permanent or transient central nervous system deficits occurred in 6.7% and 11.2% of the patients, respectively. Preoperative factors of age over 60 years, increased C-reactive protein level (>3 mg/L), and American Society of Anesthesia Class 4 were associated with in-hospital systemic and infectious complications. There are convincing data that surgeons and hospitals with high case volumes for craniotomy for tumor have lower complication rates overall (15). Trinh et al. found that there were racial and thirdparty payer influences on which patients used the high-volume hospitals and surgeons. This was supported by a study of patients undergoing craniotomy for brain metastasis that also found that race and sex were highly significant in overall patient outcome including the complication rate (9). Nuno et al. also pointed out that the complication rate was dependent on the histologic type of the primary cancer. These studies further confound the idea of finding generalizable complication rates for patients undergoing craniotomy for tumor. Nevertheless, any attempts to define predictive measures to help patients understand the types of complications and the relative frequency of those unfavorable outcomes should be commended. By working together to collect, analyze, and disseminate outcome data for patients undergoing craniotomy for tumor, the neurosurgical community may make this possible.
Herwaldt LA: Risk factors and outcomes associated with surgical site infections after craniotomy or craniectomy. J Neurosurg 120:509-521, 2014. 5. Federico G, Tumbarello M, Spanu T, Rosell R, Iacoangeli M, Scerrati M, Tacconelli E: Risk factors and prognostic indicators of bacterial meningitis in a cohort of 3580 postneurosurgical patients. Scand J Infect Dis 33:533-537, 2001.
18-month prospective survey. J Neurosurg 109: 729-734, 2008. 8. Marko NF, Weil RJ, Schroeder JL, Lang FF, Suki D, Sawaya RE: Extent of resection of glioblastoma revisited: personalized survival modeling facilitates more accurate survival prediction and supports a maximum-safe-resection approach to surgery. J Clin Oncol 32:774-782, 2014.
3. Brell M, Ibanez J, Caral L, Ferrer E: Factors influencing surgical complications of intra-axial brain tumours. Acta Neurochir (Wien) 142: 739-750, 2000.
6. Korinek AM: Risk factors for neurosurgical site infections after craniotomy: a prospective multicenter study of 2944 patients. The French Study Group of Neurosurgical Infections, the SEHP, and the C-CLIN Paris-Nord. Service Epidemiologie Hygiene et Prevention. Neurosurgery 41: 1073-1079; discussion 1079-1081, 1997.
9. Nuno M, Mukherjee D, Elramsisy A, Nosova K, Lad SP, Boakye M, Yu JS, Black KL, Patil CG: Racial and gender disparities and the role of primary tumor type on inpatient outcomes following craniotomy for brain metastases. Ann Surg Oncol 19:2657-2663, 2012.
4. Chiang HY, Kamath AS, Pottinger JM, Greenlee JD, Howard MA 3rd, Cavanaugh JE,
7. Lietard C, Thebaud V, Besson G, Lejeune B: Risk factors for neurosurgical site infections: an
10. Oppenlander ME, Wolf AB, Snyder LA, Bina R, Wilson JR, Coons SW, Ashby LS, Brachman D,
WORLD NEUROSURGERY 84 [4]: 894-896, OCTOBER 2015
www.WORLDNEUROSURGERY.org
895
PERSPECTIVES
Nakaji P, Porter RW, Smith KA, Spetzler RF, Sanai N: An extent of resection threshold for recurrent glioblastoma and its risk for neurological morbidity. J Neurosurg 120:846-853, 2014. 11. Reponen E, Korja M, Niemi T, Silvasti-Lundell M, Hernesniemi J, Tuominen H: Preoperative identification of neurosurgery patients with a high risk of in-hospital complications: a prospective cohort of 418 consecutive elective craniotomy patients. J Neurosurg 123:594-604, 2015. 12. Sanai N, Polley MY, McDermott MW, Parsa AT, Berger MS: An extent of resection threshold for
newly diagnosed glioblastomas. J Neurosurg 115: 3-8, 2011. 13. Satoer D, Visch-Brink E, Smits M, Kloet A, Looman C, Dirven C, Vincent A: Long-term evaluation of cognition after glioma surgery in eloquent areas. J Neurooncol 116:153-160, 2014. 14. Spena G, D’Agata F, Panciani PP: Buglione di Monale M, Fontanella MM: Supratentorial gliomas in eloquent areas: which parameters can predict functional outcome and extent of resection? PLoS One 8:e80916, 2013.
15. Trinh VT, Davies JM, Berger MS: Surgery for primary supratentorial brain tumors in the United States, 2000-2009: effect of provider and hospital caseload on complication rates. J Neurosurg 122: 280-296, 2015. Citation: World Neurosurg. (2015) 84, 4:894-896. http://dx.doi.org/10.1016/j.wneu.2015.05.018 Journal homepage: www.WORLDNEUROSURGERY.org Available online: www.sciencedirect.com 1878-8750/$ - see front matter ª 2015 Elsevier Inc. All rights reserved.
Photo by Jeffrey A. Brown, MD. “The iconic Basilica Santa Maria della Salute presides over a fleet of covered Venetian gondolas docked on the Grand Canal at dusk.”
896
www.SCIENCEDIRECT.com
WORLD NEUROSURGERY, http://dx.doi.org/10.1016/j.wneu.2015.05.018
Predicting Outcomes After Glioma Surgery: Model Behavior Randy L. Jensen
T
he ability to predict outcomes for individual patients undergoing craniotomy for resection of gliomas would be of great value to physicians, patients, families, and thirdparty payers. For instance, having precise knowledge of the personal potential for complications as well as the relative frequency of these unanticipated outcomes could help prepare patients and their loved ones for appropriate postoperative recovery. At present, of course, this is not a reality; however, neurosurgical oncologists are well versed in using molecular biomarkers such as isodehydrogenase-1 or methylguanine methyltransferase methylation status as well as histologic grade to help predict patient overall survival. They also have convincing evidence of the influence of their surgical efforts and patient demographics on overall survival. It is understood that overall survival and progression-free survival can be predicted by age, preoperative Karnofsky Performance Score, and extent of resection (8, 12). Nevertheless, it has not been possible to provide a patient with an individualized expectation of risk for postoperative complications. In a paper recently published in WORLD NEUROSURGERY, Missios et al. have attempted the task of determining preoperative predictive indicators of potential postoperative complications for patients undergoing craniotomy for glioma resection. They have a track record of doing similar analysis for other intracranial tumors: In a study similar to this, they described a predictive model of unfavorable outcomes for patients undergoing resection of benign intracranial tumors (2). Using multivariate analysis of a national database, they identified the effect of various risk factors on the incidence of death (1.3%), unfavorable discharge (22.7%), hydrocephalus (4.2%), cardiac (1.1%) or respiratory
Key words Craniotomy - Glioma - National Inpatient Sample - Risk prediction -
894
Abbreviations and Acronyms DVT: Deep venous thrombosis
www.SCIENCEDIRECT.com
(0.9%) complications, deep venous thrombosis (DVT) (0.5%), pulmonary embolus (2.3%), and acute renal failure (1.5%). In their article in this month’s issue, they evaluated the records of over 20,000 patients with glioma included in a national database who underwent craniotomy and identified similar inpatient postoperative risks, including a 1.6% risk of mortality, and morbidity risk of 25.8% for discharge to rehabilitation, 4.0% for treated hydrocephalus, 0.7% for cardiac complications, 0.5% for respiratory complications, 0.8% for deep wound infection, 0.6% for DVT, 3.1% for pulmonary embolus, and 1.3% for acute renal failure. These numbers proved to be very similar for benign and malignant intracranial disease. Mortality is almost identical in both studies, as is true of most of the other postoperative complications. Discharge to rehabilitation is slightly higher for the glioma group but the difference is small and it is unclear whether it would be of statistical significance. Once again, the authors developed predictive models for individual complications based on a logistic regression analysis and validated in a bootstrapped sample for these same complications. They found that age, minority status, vascular morbidities, other medical conditions, history of coagulopathy, and hyponatremia all affected the risks for negative outcomes. The immediate criticism of both of these studies is that they only describe inpatient complications, and all outcome measures were recorded before the patient was discharged from the acute hospital setting. Although it is true that most complications occur during this period, this does not tell the whole story of the
Department of Neurosurgery, Clinical Neurosciences Center, University of Utah, Salt Lake City, Utah, USA To whom correspondence should be addressed: Randy L. Jensen, M.D., Ph.D. [E-mail: [email protected]] Citation: World Neurosurg. (2015) 84, 4:894-896. http://dx.doi.org/10.1016/j.wneu.2015.05.018
WORLD NEUROSURGERY, http://dx.doi.org/10.1016/j.wneu.2015.05.018
PERSPECTIVES
ultimate outcome of a given patient. One could ask what happens to the patient as an outpatient? Do the functional outcomes of unfavorable discharges improve over time after receiving physical and occupational therapy? This would certainly decrease the rates of unfavorable outcomes after craniotomy for tumor, as has been reported in other publications (10, 14). This is even more evident when considering cognitive deficits, which have been shown to take longer than 3 months for recovery, especially when the operation is located within the language domain (13). On the other hand, the approach of recording complications at the time of discharge misses delayed unexpected outcomes. For instance, a DVT that is discovered or occurs in an outpatient, a frequent occurrence, would be underestimated in this study. In fact, both of these studies have very low rates of pulmonary embolus and DVT compared with many contemporary series. The most intriguing portion of this report is the rate of discharge to rehabilitation/unfavorable discharge. Bekelis et al. report that the risk of this outcome was approximately 25% and found it was slightly higher for the glioma group. It would be really helpful to know more about patients with this outcome. What were the reasons for this need for further inpatient care? Presumably, it was related to new or existing neurologic deficits but the authors are unable to make this distinction because of the limitations of an administrative database. It would be interesting to separate out patients who were admitted with existing neurologic deficits preoperatively that would have necessitated inpatient rehabilitation regardless of the surgical intervention. One potential takehome message from these studies is that a quarter of patients undergoing craniotomy for tumor will require inpatient rehabilitation after their surgery. This is a useful fact to give patients and families in preoperative counseling. Even more useful would be the ability to determine exactly which patient characteristics make a given patient more at risk of a given new neurologic deficit. The ultimate goal would be to predict the chance of a specific neurologic deficit for a given lesion in a particular location of the brain. For instance, what is the relative risk for a patient to experience a language deficit after resection of an insular oligodendroglioma World Health Organization grade II with isodehydrogenase-1 mutation and sparing of the medial lenticulostriate arteries? At present, that is a “pie in
REFERENCES 1. Abu Hamdeh S, Lytsy B, Ronne-Engstrom E: Surgical site infections in standard neurosurgery procedures—a study of incidence, impact and potential risk factors. Br J Neurosurg 28:270-275, 2014. 2. Bekelis K, Kalakoti P, Nanda A, Missios S: A predictive model of unfavorable outcomes after benign intracranial tumor resection. World Neurosurg 2015: In press.
the sky” wish but this study sets the groundwork to move in this direction. In the past, most studies on craniotomy outcomes focused on infection rates (1, 4-7). One early exception was a study examining the complication rate after craniotomy for tumor (3). Brell et al. found a similar overall major complication rate of 27.5%, with neurologic complications being the most frequently encountered. They did not find a statistically significant relationship between complication rate and eloquence of the tumoral area. Infratentorial tumor location, previous radiotherapy, and reoperations were strongly related to the incidence of regional complications. Age over 60 years and severe concomitant disease were risk factors for systemic complications. In a similar fashion, others have demonstrated the difficulty in predicting postoperative functional outcome even for gliomas in eloquent areas (14). A more recent attempt to identify preoperative risk factors in a single-institution study found a similar in-hospital mortality rate of 1.0% (11). In-hospital systemic or infectious complications and permanent or transient central nervous system deficits occurred in 6.7% and 11.2% of the patients, respectively. Preoperative factors of age over 60 years, increased C-reactive protein level (>3 mg/L), and American Society of Anesthesia Class 4 were associated with in-hospital systemic and infectious complications. There are convincing data that surgeons and hospitals with high case volumes for craniotomy for tumor have lower complication rates overall (15). Trinh et al. found that there were racial and thirdparty payer influences on which patients used the high-volume hospitals and surgeons. This was supported by a study of patients undergoing craniotomy for brain metastasis that also found that race and sex were highly significant in overall patient outcome including the complication rate (9). Nuno et al. also pointed out that the complication rate was dependent on the histologic type of the primary cancer. These studies further confound the idea of finding generalizable complication rates for patients undergoing craniotomy for tumor. Nevertheless, any attempts to define predictive measures to help patients understand the types of complications and the relative frequency of those unfavorable outcomes should be commended. By working together to collect, analyze, and disseminate outcome data for patients undergoing craniotomy for tumor, the neurosurgical community may make this possible.
Herwaldt LA: Risk factors and outcomes associated with surgical site infections after craniotomy or craniectomy. J Neurosurg 120:509-521, 2014. 5. Federico G, Tumbarello M, Spanu T, Rosell R, Iacoangeli M, Scerrati M, Tacconelli E: Risk factors and prognostic indicators of bacterial meningitis in a cohort of 3580 postneurosurgical patients. Scand J Infect Dis 33:533-537, 2001.
18-month prospective survey. J Neurosurg 109: 729-734, 2008. 8. Marko NF, Weil RJ, Schroeder JL, Lang FF, Suki D, Sawaya RE: Extent of resection of glioblastoma revisited: personalized survival modeling facilitates more accurate survival prediction and supports a maximum-safe-resection approach to surgery. J Clin Oncol 32:774-782, 2014.
3. Brell M, Ibanez J, Caral L, Ferrer E: Factors influencing surgical complications of intra-axial brain tumours. Acta Neurochir (Wien) 142: 739-750, 2000.
6. Korinek AM: Risk factors for neurosurgical site infections after craniotomy: a prospective multicenter study of 2944 patients. The French Study Group of Neurosurgical Infections, the SEHP, and the C-CLIN Paris-Nord. Service Epidemiologie Hygiene et Prevention. Neurosurgery 41: 1073-1079; discussion 1079-1081, 1997.
9. Nuno M, Mukherjee D, Elramsisy A, Nosova K, Lad SP, Boakye M, Yu JS, Black KL, Patil CG: Racial and gender disparities and the role of primary tumor type on inpatient outcomes following craniotomy for brain metastases. Ann Surg Oncol 19:2657-2663, 2012.
4. Chiang HY, Kamath AS, Pottinger JM, Greenlee JD, Howard MA 3rd, Cavanaugh JE,
7. Lietard C, Thebaud V, Besson G, Lejeune B: Risk factors for neurosurgical site infections: an
10. Oppenlander ME, Wolf AB, Snyder LA, Bina R, Wilson JR, Coons SW, Ashby LS, Brachman D,
WORLD NEUROSURGERY 84 [4]: 894-896, OCTOBER 2015
www.WORLDNEUROSURGERY.org
895
PERSPECTIVES
Nakaji P, Porter RW, Smith KA, Spetzler RF, Sanai N: An extent of resection threshold for recurrent glioblastoma and its risk for neurological morbidity. J Neurosurg 120:846-853, 2014. 11. Reponen E, Korja M, Niemi T, Silvasti-Lundell M, Hernesniemi J, Tuominen H: Preoperative identification of neurosurgery patients with a high risk of in-hospital complications: a prospective cohort of 418 consecutive elective craniotomy patients. J Neurosurg 123:594-604, 2015. 12. Sanai N, Polley MY, McDermott MW, Parsa AT, Berger MS: An extent of resection threshold for
newly diagnosed glioblastomas. J Neurosurg 115: 3-8, 2011. 13. Satoer D, Visch-Brink E, Smits M, Kloet A, Looman C, Dirven C, Vincent A: Long-term evaluation of cognition after glioma surgery in eloquent areas. J Neurooncol 116:153-160, 2014. 14. Spena G, D’Agata F, Panciani PP: Buglione di Monale M, Fontanella MM: Supratentorial gliomas in eloquent areas: which parameters can predict functional outcome and extent of resection? PLoS One 8:e80916, 2013.
15. Trinh VT, Davies JM, Berger MS: Surgery for primary supratentorial brain tumors in the United States, 2000-2009: effect of provider and hospital caseload on complication rates. J Neurosurg 122: 280-296, 2015. Citation: World Neurosurg. (2015) 84, 4:894-896. http://dx.doi.org/10.1016/j.wneu.2015.05.018 Journal homepage: www.WORLDNEUROSURGERY.org Available online: www.sciencedirect.com 1878-8750/$ - see front matter ª 2015 Elsevier Inc. All rights reserved.
Photo by Jeffrey A. Brown, MD. “The iconic Basilica Santa Maria della Salute presides over a fleet of covered Venetian gondolas docked on the Grand Canal at dusk.”
896
www.SCIENCEDIRECT.com
WORLD NEUROSURGERY, http://dx.doi.org/10.1016/j.wneu.2015.05.018