Predicting clinical outcomes of post-operative focal neurological deficits after glioma resection based on MRI characteristics: A retrospective chart review

doi:10.21203/rs.3.rs-2616875/v1

Download PDF

Research Article

Predicting clinical outcomes of post-operative focal neurological deficits after glioma resection based on MRI characteristics: A retrospective chart review

https://doi.org/10.21203/rs.3.rs-2616875/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background: The standard of care for gliomas includes maximum safe resection of the tumor. This may lead to inadvertent damage to tissue directly or vasculature supplying normal brain tissue. This may result in perilesional brain infarction which is readily seen on early postoperative MRI scans. Their relationship with the presence of a deficit and recovery from said deficit is unclear.

Methods: We did a retrospective chart and radiology review to study this relationship.

Results: Out of 225 included patients, 24% had infarcts on their postoperative MRI while the rest did not have infarcts. The incidence of new deficits in these two groups was not significantly different. The presence of these infarcts did not appear to affect recovery from deficits against patients with deficits and no infarcts. Moreover, the location of the tumor did not significantly correlate with the presence of infarcts, new deficits or recovery from deficits. The extent of resection also failed to show a strong correlation with new infarcts or deficits. Only a small percentage of all patients failed to show some or complete improvement from deficits at 6 months follow up.

Conclusion: Perilesional infarcts are a common finding on postoperative scans after glioma surgery but their presence does not help to predict the presence of a deficit nor the recovery from a deficit.

glioma

surgery

infarct

recovery

outcome

Perilesional infarcts are a common finding after surgery for gliomas but there is no strong correlation with deficits or recovery in our study.

Although a retrospective study, it has a large number of patients included with good quality imaging and standard documentation. The clinical significance of perilesional infarcts is still under investigation. Our work demonstrates that the clinical significance of these infarcts may be smaller than otherwise demonstrated in literature and their role in predicting recovery is less than expected.

Gliomas are the most common (30%) types of brain tumors. [1] Neurosurgeons performing surgery in patients with gliomas are faced with a dilemma, requiring maximum safe resection, the keyword being safe. Patients who undergo extensive resection may acquire new neurological deficits, which may be temporary or permanent, affecting quality of life and reducing survival. On the other hand, a restricted biopsy or debulking exposes the patient to recurrence. [2, 3] A wide variety of radiological modalities and cortical stimulation techniques are available which stimulate functional brain tissue to identify the boundaries between tumor and normal tissue to help resection. [4] Surgery for gliomas has a high incidence of infarctions present in early postoperative scans. The clinical significance of the changes (infarctions) are unclear with no study looking at their relationship with the development of a motor or speech deficit. There is also limited data on how these changes affect subsequent improvement of deficits.

Diffusion-weighted MRI (DWI) can detect infarcts as early as 6 hours of insult, but its usefulness has remained a matter of controversy in post-operative patients. [5] DWI images will demonstrate the restriction of water molecules in brain tissue which has decreased perfusion. These patients show DWI images with a high signal and corresponding low signal on ADC images (Fig. 1). After brain tumor resection, patients may show DWI and ADC changes in post-operative MRI with no apparent correlation to any post-operative neurological deficits. Since gliomas are intrinsic CNS neoplasms, they are often present in proximity to eloquent brain area (EBA), complicating their satisfactory removal. [6]

We planned a retrospective observational study of patients’ post-operative imaging to examine MRI findings associated with deficits that improve with time and those that do not improve in an attempt to better understand these findings and help us in counselling post-operative patients.

Study population

In this retrospective chart review we screened all adult patients (> 17 years) operated for gliomas with surgical resection at our institution by a single surgeon from January 2013 to December 2020. The surgeries were either under general anesthesia or were awake craniotomies with the use of a navigation system. Resections were performed according to the margins on navigation scans but intraoperative brain shift was not systematically adjusted. Surgical morbidity was assessed retrospectively from records. All patients who had a preoperative MRI and early postoperative MRI (within 72 hours) on at least 1.5 Tesla were included. We decided to only study speech and motor deficits for reliable retrospective assessment. All patients were operated upon by the senior author. The presence or absence of new DWI changes (infarcts) was assessed on the postoperative MRI. (Fig. 1) Volume of infarct was determined by the formula ([LxWxH]/2). Gross total resection (GTR) was complete resection of enhancing tumor (HGG) or FLAIR hyperintensity (LGG). Subtotal resection (STR) was between 5–10% residual disease, debulking was < 90% resection and biopsies were only small samples from the bulk of the disease. Patient records were reviewed for new deficits, their onset, progression and clinical course over 6 months.

Variables of interest

Variables to be studied included age, sex, co-morbidities, past or current medications, presence or absence of neurological deficits attributable to the disease before surgery, development of new neurological deficit of any severity, MRI findings on preoperative and post-operative scans (Diffusion-weighted imaging [DWI] and apparent diffusion coefficient [ADC] map), histopathological diagnosis, and the presence or absence of neurological deficits, whether complete or partial, and for recovery. Focal neurologic signs, also known as focal neurological deficits or focal CNS signs, are impairments of nerve, spinal cord, or brain function that affects a specific region of the body, e.g., weakness in the left arm, the right leg, paresis, or plegia. Language deficits may be in the form of dysphasias or aphasia. Only motor and language deficits will be used as these are more reliably documented in retrospective analyses.

Eloquent areas will be defined as per the Sawaya grading system. [7]

Statistics

The descriptive statistics were run to find the different categories within the patients such as gender and age. The relationship between the patient developing a deficit and having post infarct was found through chi square test. We also conducted ANOVA test keeping post op infarct as an independent variable and development of new deficit as dependent variable to test the relationship between them. Pearson’s correlation test was conducted to find the relationship between extent of resection and having a new deficit. The spearman’s Rho test was conducted to test the association between the Sawaya grade of the patient and patients having postoperative infarcts.

Ethics and approvals

The study received ethical review exemption.

We included 255 patients who met the eligibility criteria for the study. There were 155 (69%) males and 70 (31%) females (mean age 41.6 years). By location most common tumor site was insular or some degree of insular involvement (65), followed by parietal (42) and then frontal (38) region. (Fig. 2) MRI done post-operative showed GTR in 124 patients, STR in 61 and debulking while only 8 cases underwent biopsy alone.

Patients with infarct on DWI sequences were compared to patients without infarcts and were followed for neurological deficits. Overall, there were 24% patients with infarcts and 76% without infarcts. Please see Table 1 for detailed results. None of the results showed statistical significance of differences between the two groups.

Table 1

Presence and absence of infarcts in early post-operative MRI scans after glioma resection and their corresponding frequency of new deficits. Most patients recovered either completely or incompletely in the 6 month follow up period, however, there was no significant difference between the two groups.
	Patients with infarcts (24%)	Patients without infarcts (76%)	Significance
New deficit	33%	25%	P > 0.05
No deficit	67%	75%	P > 0.05
Complete deficit recovery	45%	68%	P > 0.05
Incomplete deficit recovery	45%	31%	P > 0.05
No recovery	11%	2.3%	P > 0.05

Regarding the location of tumor based on Sawaya grading, we found no significant correlation between development of infarct and tumor location. (Fig. 3) There was no correlation between Sawaya grade and deficit recovery either. 3 patient records were missing. (Table 2). We also found no significant association between extent of resection in our series and development of a new deficit. (Fig. 4)

Table 2

Relationship between Sawaya grade and recovery. (p-value for all groups was > 0.05)
Grade	Complete recovery	Incomplete recovery	No recovery
1	75%	-	25%
2	62%	31%	6%
3	54%	46%

In this retrospective chart and radiology review of 225 included patients, we discovered that almost a quarter of them had demonstrable DWI changes consistent with infarcts that were not present on preoperative imaging. (Fig. 1) We have only included patients who had very early postoperative MRI scans to avoid late radiological changes that are associated with surgery. Limiting ourselves to speech and motor deficits that are relatively easily measured and recorded allowed us to improve our data reliability. We had noticed in our practice that patients developed postoperative deficits after glioma resection and even during awake craniotomy guided resection, but with the passage of time we learned that most of our patients improved during their inpatient stay. We also noticed that some patients in our study developed relatively significant volumes of infarcts and showed no demonstrable deficits. Although we did not differentiate whether these infarcts were in eloquent area or not, we looked at the overall association between the presence of these changes, the incidence of deficits and the relationship between DWI changes and prognosis of motor and speech deficits.

Prevalence of infarcts after glioma resection

24% (52) of our patients showed evidence of infarct on MRI. Studies have quoted the incidence of these peritumoral infarctions in the range of 19%-80%. [6, 8, 9] Although different authors differentiated infarcts by varying methods, the most common descriptions were rim shaped around the resection cavity, sector shaped and mixed. A population-based study by Strand et al [8] of 539 included patients showed almost half (44%) had postoperative ischemic findings but a large majority of them (36%) had only rim infarctions. There is no current definition as what is considered a peritumoral infarction. Some authors do not consider DWI abnormalities of less than 3 mm thickness as significant while others classify any abnormalities as significant. Hemostatic agents as well as artifacts can appear as small DWI abnormalities. This makes it difficult to determine their clinical effects. In our study we have attempted to correlate their presence with the presence of deficits. [9] The incidence of infarcts in redo procedures for gliomas has been cited by one paper as 80% vs only 31% in newly diagnosed gliomas. In our study we did not differentiate between initial and recurrent disease but treated every operation as an independent variable. [9]

Volume of infarct

There are studies that show that postoperative infarcts are associated with impaired overall survival. [10] Bette et al have shown in their work that the volume of infarct after surgery for glioblastomas is an independent predictor of overall survival and functional independence but not progression free survival. They have also suggested that this ischemia has a role in tumor progression. Additionally, they have also linked higher infarct volumes with worsened KPS and therefore overall survival [11].

Intraoperative risk factors for infarct development

Sveino Strand et al performed a prospective analysis to determine how intraoperative factors were associated with post-operative infarctions after glioma resections. In their study, the surgeon would immediately fill a questionnaire which included a number of factors including vessel sacrifice, vascularity of the tumor, tumor heterogeneity, brain-tumor interface, consistency and instruments used (suction, ultrasonic aspirator etc). Their study did not find any significant association between surgeon reported factors and peritumoral infarctions in their patient group. They did report different rates amongst different surgeons but these were not significant either. Even though it is clear that vessel damage causes infarctions, self-reporting of perceived important artery or vein sacrifice was not found to be a useful of predictor of infarcts on postoperative scans. [12] The use of ultrasonic aspirators has been purported to cause less peritumoral tissue damage and spares the larger vessels, this is widely contested between neurosurgeons. Strand et al found a higher, non-significant incidence in the use of ultrasonic aspirators for glioma resections. This is not widely studied in literature and we did not adequately document it’s use and vessel damage in our charts. [12] Xenon lamps from modern microscopes have been known to cause damage in recent studies but its significance is unknown at present. [13, 14]

Hou et al performed a prospective study with 75 consecutive insular gliomas resected through a transcortical approach to determine strategies for protecting the numerous important vessels. They had a 58.7% incidence of new infarcts on postop imaging but only 35.5% of their patients developed motor deficits or speech deficits with 41% of these patients showing no improvement at 6 months in motor functions while all speech deficits improved to some degree. They did not clarify what proportion of their patients without infarcts developed deficits. [15]

Infarcts in patients with low grade gliomas

Berger et al’s retrospective study of 82 LGG patients showed a 23% incidence of infarcts with 68% of these being associated with a deficit. Infarcts were found to be significantly more common in recurrent and insular gliomas. There was a statistically significant higher chance of having a postoperative deficit with an infarct and a lower chance of improvement with the presence of an infarct. They did not quantify what they considered to be an infarct with no indication of whether rim restriction was considered an infarct. [16] Intraoperative MAP reduction to 20–30% below baseline in such patients is considered a risk for infarct development as well. [17]

Infarcts in patients with high grade gliomas

Berger at al performed a retrospective review of 239 patients with HGG and determined a 12.5% incidence of infarcts of which 43% developed deficits. 35% of patients without infarcts and 57% of those with infarcts had motor deficits immediately after surgery which improved to 25% and 37% respectively at 6 months. However, their study does not clarify how many patients developed a new deficit as in the two groups, 37% and 47% patients already had preoperative motor deficits. They also performed a subgroup analysis for speech deficits in which the presence of an infarct did not affect speech deficits while those without speech deficits showed a small improvement. [18]

Limitations

There are several limitations including the retrospective chart review nature of the study and limitation to a single surgeon’s experience.

Our data failed to show any significant association with the presence of infarcts after glioma resection and presence or absence of deficits. Infarcts also did not appear to affect recovery from said deficits with a large number of patients recovering partially or completely.

Funding

None

Goodenberger ML, Jenkins RB (2012) Genetics of adult glioma. Cancer Genet 205(12):613–621. 10.1016/j.cancergen.2012.10.009
Smith JS, Chang EF, Lamborn KR, Chang SM, Prados MD, Cha S, Tihan T, Vandenberg S, McDermott MW, Berger MS Role of extent of resection in the long-term outcome of low-grade hemispheric gliomas. J Clin Oncol. 2008 Mar 10;26(8):1338-45. doi: 10.1200/JCO.2007.13.9337. PMID: 18323558
Gil-Robles S, Duffau H (2010) Feb;28(2):E8 Surgical management of World Health Organization Grade II gliomas in eloquent areas: the necessity of preserving a margin around functional structures. Neurosurg Focus. doi: 10.3171/2009.12.FOCUS09236. PMID: 20121443
Szelényi A, Bello L, Duffau H, Fava E, Feigl GC, Galanda M, Neuloh G, Signorelli F, Sala F (2010) Feb;28(2):E7 ; Workgroup for Intraoperative Management in Low-Grade Glioma Surgery within the European Low-Grade Glioma Network. Intraoperative electrical stimulation in awake craniotomy: methodological aspects of current practice. Neurosurg Focus. doi: 10.3171/2009.12.FOCUS09237. PMID: 20121442
Girot M, Leclerc X, Gauvrit JY, Verdelho A, Pruvo JP, Leys D (2003) Cerebral magnetic resonance imaging within 6 hours of stroke onset: inter- and intra-observer reproducibility. Cerebrovasc Dis. ;16(2):122-7. doi: 10.1159/000070591. PMID: 12792169
Jakola AS, Berntsen EM, Christensen P, Gulati S, Unsgård G, Kvistad KA, Solheim O (2014) Surgically acquired deficits and diffusion weighted MRI changes after glioma resection-A matched case-control study with blinded neuroradiological assessment. PloS one. Jul 3;9(7):e101805
Sawaya R, Hammoud M, Schoppa D, Hess KR, Wu SZ et al (1998) Neurosurgical Outcomes in a Modern Series of 400 Craniotomies for Treatment of Parenchymal Tumors. Neurosurgery 42:1044–1055
Strand PS, Berntsen EM, Fyllingen EH, Sagberg LM, Reinertsen I, Gulati S, Bouget D, Solheim O (2021 Nov) Brain infarctions after glioma surgery: prevalence, radiological characteristics and risk factors. Acta Neurochir (Wien) 163(11):3097–3108. 10.1007/s00701-021-04914-zEpub 2021 Sep 1. PMID: 34468884; PMCID: PMC8520515
Gempt J, Förschler A, Buchmann N, Pape H, Ryang YM, Krieg SM, Zimmer C, Meyer B, Ringel F (2013) Apr;118(4):801-8 Postoperative ischemic changes following resection of newly diagnosed and recurrent gliomas and their clinical relevance. J Neurosurg. doi: 10.3171/2012.12.JNS12125. Epub 2013 Feb 1. PMID: 23373806
Chambless LB, Kistka HM, Parker SL, Hassam-Malani L, McGirt MJ, Thompson RC (2015) The relative value of postoperative versus preoperative Karnofsky Performance Scale scores as a predictor of survival after surgical resection of glioblastoma multiforme. J Neurooncol 121:359–364
Sveino Strand P, Gulati S, Millgård Sagberg L, Solheim O Intraoperative risk factors for peritumoral infarctions following glioma surgery.Brain Spine. 2022 Jun7;2:100903. doi: 10.1016/j.bas.2022.100903. PMID: 36248115; PMCID: PMC9559966.
Bette S, Wiestler B, Kaesmacher J, Huber T, Gerhardt J, Barz M, Delbridge C, Ryang YM, Ringel F, Zimmer C, Meyer B, Boeckh-Behrens T, Kirschke JS, Gempt J Infarct volume after glioblastoma surgery as an independent prognostic factor.Oncotarget. 2016 Sep20;7(38):61945–61954. doi: 10.18632/oncotarget.11482. PMID: 27566556; PMCID: PMC5308702.
Choudhry IK, Kyriakedes J, Foad MB (2013 Mar) Iatrogenic burn caused by an operating microscope: case report. J Hand Surg Am 38(3):545–547. 10.1016/j.jhsa.2012.11.027Epub 2013 Jan 20. PMID: 23337461
Gayatri P, Menon GG, Suneel PR (2013) Jul;25(3):267 – 70 Effect of operating microscope light on brain temperature during craniotomy. J Neurosurg Anesthesiol. doi: 10.1097/ANA.0b013e3182894a01. PMID: 23459259
Hou Z, Huang Z, Li Z, Deng Z, Li G, Xu Y, Wang M, Sun S, Zhang Y, Qiao H, Xie J Incidence of ischemic complications and technical nuances of arteries preservation for insular gliomas resection.Front Surg. 2022 Oct14;9:956872. doi: 10.3389/fsurg.2022.956872. PMID: 36311934; PMCID: PMC9614341.
Berger A, Tzarfati G, Costa M, Serafimova M, Korn A, Vendrov I, Alfasi T, Krill D, Aviram D, Ben Moshe S, Kashanian A, Ram Z, Grossman R Incidence and impact of stroke following surgery for low-grade gliomas. J Neurosurg. 2019 Dec 27:1–9. doi: 10.3171/2019.10.JNS192301. Epub ahead of print. PMID: 31881532
Vlisides P, Mashour GA (2016) Perioperative stroke. Can J Anaesth 63:193–204
Berger A, Tzarfati G, Serafimova M et al (2022) Risk factors and prognostic implications of surgery-related strokes following resection of high-grade glioma. Sci Rep 12:22594. https://doi.org/10.1038/s41598-022-27127-5

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

Predicting clinical outcomes of post-operative focal neurological deficits after glioma resection based on MRI characteristics: A retrospective chart review

Status:

Version 1

Abstract

Figures

Key Points

Importance of the study

Introduction

Materials And Methods

Results

Discussion

Prevalence of infarcts after glioma resection

Volume of infarct

Intraoperative risk factors for infarct development

Infarcts in patients with low grade gliomas

Infarcts in patients with high grade gliomas

Limitations

Conclusion

Declarations

References

Additional Declarations

Status:

Version 1