Emotional Lability as a Symptom of Extra-axial Posterior Fossa Tumors: a Case–Control Review of Neuroanatomy and Patient-Reported Quality of Life

Abstract

Introduction  Emotional lability (EL), the uncontrollable and unmotivated expression of emotion, is a rare and distressing symptom of brainstem compression. In published case reports, EL from an extra-axial posterior fossa tumor was alleviated by tumor resection. The primary aim herein was to radiographically establish the degree of compression from mass lesions onto brainstem structures. Secondarily, we compared changes in patient-reported quality of life (QOL) pre- and postoperatively.

Methods  A retrospective review of posterior fossa tumors treated between 2002 and 2018 at Vancouver General Hospital revealed 11 patients with confirmed EL. Each case was matched to three controls. A lateral brainstem compression scale characterized mass effect at the level of the medulla, pons, and midbrain in preoperative axial T2-weighted fluid-attenuated inversion recovery magnetic resonance imaging (FLAIR MRI) scans. Compression and clinical variables were compared between patient groups. Short Form-36 version 1 health surveys were retrospectively obtained from patient charts to compare pre- versus postoperative changes in survey scores between EL and control patients.

Results  EL symptoms ceased postoperatively for all EL patients. EL tumors exert greater compression onto the pons ( p  = 0.03) and EL patients more commonly have cerebellar findings preoperatively ( p  = 0.003). Patients with EL-causing tumors experienced greater improvement postoperatively in “Health Change” ( p  = 0.05), which was maintained over time.

Conclusion  Findings suggest that compression onto the pons inhibits control over involuntary, stereotyped expression of emotion. This adds to evidence that EL may be attributed to cerebellum deafferentation from cortical and limbic structures through the basis pontis, leading to impaired modulation of emotional response. QOL results augment benefits of offering patients EL-alleviating tumor resection surgery.

Background

Extra-axial brain tumors include a wide range of pathologies originating from surrounding structures of the brain. ¹ Extra-axial tumors in the posterior fossa of the cranial cavity may cause brainstem compression depending on the size and proximity to brainstem structures. Since the brainstem ubiquitously regulates information flow between the cerebrum, the cerebellum, and the rest of the body, ² compression of the brainstem can lead to a variety of nefarious symptoms. ³ Emotional lability (EL), the dysregulation of emotional expression, as a symptom of brainstem compression is a rare phenomenon. There have been fewer than 20 reports of extra-axial tumors causing EL due to brainstem compression.

EL is the uncontrollable and unmotivated expression of emotion, often presenting as laughter or crying. ⁴ There is no perceived mood change associated with these episodes. It is also distinct from mood disorders where there is dysregulation of primary emotion formation, ⁴ which can be localized to the prefrontal cortex and the limbic system. ⁵ EL is also distinct from gelastic seizures, which are paired with abrupt sympathetic system activation. ⁶ The motor behavior of laughter and crying episodes mimics the authentic expression in EL; it is not an issue of generating the motor function. ⁴ EL is common among patients with amyotrophic lateral sclerosis, ⁷ multiple sclerosis, ⁸ severe traumatic brain injury, ⁹ and stroke. ¹⁰ The nosology regarding EL is imprecise, so the following will be included under the umbrella term “emotional lability”: “emotionalism,” “pseudobulbar affect,” “labile affect,” “emotional incontinence,” “pathological laughter,” “pathological crying,” and “pathological laughter and crying.”

Image Analysis Background

The pathophysiological mechanism of EL caused by mass lesions has not yet been determined. In reviewing and synthesizing proposed mechanisms from case reports, overlapping theories were found. While authors cited different structures of the affected brain for causing EL, the general regions remained consistent over reports. To substantiate a stronger foundation for this study, we also extrapolated findings from EL-causing intra-axial tumors and strokes within the brainstem.

Of the published case reports for EL in adult patients with extra-axial tumors ( Table 1 ), meningiomas were found in the following regions: in the petroclival region, ^{11 12 13 14} ventral to the pons, ¹⁵ in the tentorium, ¹⁶ and in the petrous apex region. ¹⁷ Patients with schwannomas ^{6 18 19} and chordomas ^{20 21} have also presented with EL. Similarly, EL has been noted in a subsect of patients with intra-axial tumors, specifically brainstem gliomas, ^{22 23} ependymomas, ²⁴ and cysts. ^{4 20} In all reports, EL was immediately and entirely alleviated with surgical resection of the tumor. In the pediatric population, Hargrave et al reported a cohort of 17 intra-axial pontine glioma patients with pathological laughter that improved or resolved with tumor resection. ²⁵ Of note, EL symptoms resumed at recurrence of the tumor in several patients ²⁵ and EL symptoms often presented earlier than other neurological deficits. ¹³

Table 1. Summary of case reports for emotional lability in adults caused by extra-axial posterior fossa mass lesions.

Author	Year	Title	Country	No. of patients	Diagnosis	Publication details
Cantu and Drew	1966	Pathological laughing and crying associated with a tumor ventral to the pons	United States	1	Meningioma ventral to the pons	Volume 24: issue 6, pp. 1024–1026 (June 1966) in Journal of Neurosurgery
Stevenson et al	1966	A transcervical transclival approach to the ventral surface of the brain stem for removal of a clivus chordoma	United States	1	Clivus chordoma	Volume 24: issue 2, pp. 544–551 (February 1966) in Journal of Neurosurgery
Matsuoka et al	1993	Clival chordoma associated with pathological laughter. Case report	Japan	1	Clival chordoma	Volume 79: issue 3, pp. 428–433 (September 1993) in Journal of Neurosurgery
Shafqat et al	1998	Petroclival meningioma presenting with pathological laughter	United States	1	Petroclival meningioma	Volume 50: issue 6, pp. 1918–1919 (June 1998) in Neurology
Tsutsumi et al	2000	Tentorial meningioma associated with pathological laughter–case report	Japan	1	Tentorial meningioma	Volume 40: issue 5, pp. 272–274 (2000) in Neurologia Medico-Chirurgica
Bhatjiwale et al	2000	Pathological laughter as a presenting symptom of massive trigeminal neuromas: report of four cases	India	4	Trigeminal neuromas	Volume 47: issue 2, pp. 469–472 (August 2000) in Journal of Neurosurgery
Muzumdar et al	2001	Pathological laughter as a presenting symptom of petroclival meningioma	India	1	Petroclival meningioma	Volume 41: issue 10, pp. 505–507 (2001) in Neurologia Medico-Chirurgica
Virani and Jain	2001	Trigeminal schwannoma associated with pathological laughter and crying	India	1	Cerebellopontine angle trigeminal schwannoma	Volume 49: issue 2, pp. 162–165 (June 2001) in Neurology India
McCormick and Lee	2002	Pseudobulbar palsy caused by a large petroclival meningioma report of two cases	United States	2	Petroclival meningiomas	Volume 12: issue, pp. 61–71 (May 2002) in Skull Base
Cheng et al	2003	Petroclival meningioma presenting with pathological laughter: report of a case and review of the literature	Taiwan	1	Petroclival meningioma	Volume 12: issue 4, pp. 187–190 (November 2003) in Acta Neurologica Taiwanica
Nadkarni and Goel	1999	Trochlear nerve neurinoma presenting as pathological laughter	India	1	Trochlear nerve neurinoma	Volume 13: issue 3, pp. 212–213 (July 1999) in British Journal of Neurosurgery
Hassan et al	2018	Pseudobulbar affect due to skull base meningioma resolving after temporal lobectomy for epilepsy	Canada	1	Petrous apex meningioma	Volume 45: issue 4, pp. 485–486 (July 2018) in Canadian Journal of Neurological Sciences
			Total	16

Parvizi et al suggest that the basis pontis is the only region where a discrete lesion can cause EL. ²⁶ The findings from Parvizi et al's seminal papers ^{4 26 27} outline the basis of our hypothesis that increased compression from mass lesions onto the basis pontis structures causes EL symptoms. It is theorized that deafferentation of the cerebellum leads to dysregulation of emotional expression. Projections from cortical and limbic regions of the brain communicate through the pons to the cerebellum. Tumors can cause displacement of brainstem structures from their neutral anatomical position, causing stretching and dislocation of the tracts that communicate through nearby structures. This investigation aims to address a gap in neurophysiology, as there is no current case series to outline the link between EL and brainstem compression from posterior fossa tumors in adults. Studying this disorder furthers our understanding about how the brain, specifically the cerebellum, regulates emotion. There are pertinent clinical applications for this study since EL often presents earlier than other neurological symptoms: an established pathophysiological mechanism for EL could support its inclusion as a useful clinical symptom, for which patients can be screened.

Quality-of-Life Background

The increased psychiatric and psychological encumbrance caused by EL renders the symptom a worthwhile phenomenon to better understand. Stroke patients with EL experience higher rates of depression. ²⁸ When controlled for elevated depression rates, these patients also experience greater reported irritability and delusions of reference compared with the general population. ²⁸ There is a clear social and professional burden associated with this condition as patients are unable to control their outbreaks, leading to socially awkward or inappropriate situations. This can have a detrimental effect of a patient's quality of life (QOL). Patients who suffer from EL after a stroke, due to demyelinating disease, or other irreversible conditions can be treated with antidepressants and dopaminergic agents, which mitigate the symptom at optimal dosing. ²⁹ For patients with EL-causing mass lesions, the symptom can be cured immediately and entirely after the tumor is removed. This provides unique insight into how EL independently affects patient-reported QOL.

Improving patient-reported QOL is an important goal of medical and surgical treatment. Skull base tumor patients are treated to mitigate future deterioration and, when possible, to improve current deficits caused by the lesion. Barring intraoperative and postoperative complications, skull base tumor resection surgery can improve patient-reported QOL. ³⁰ Contrastingly, there is a body of knowledge that suggests that QOL improvements are not notable after resection of acoustic neuromas. ³¹ Nevertheless, for patients with EL-causing tumors, postoperative improvements would additionally include cessation of EL symptoms. To assess whether surgically curing EL symptoms leads to a greater improvement in QOL, we compared survey scores between EL patients and control patients. Quantifying patient-reported QOL health scores adds an additional goal of surgery for EL patients: to improve QOL by relieving EL symptoms.

We hypothesized that patient-reported QOL will improve postoperatively for patients with EL. This is in keeping with the expected QOL improvement after skull base tumor resection, barring complications. We also postulated that immediate post- versus preoperative QOL improvements for patients with EL would be greater than those of matched control patients due to postoperative cessation of EL and associated social burdens of the symptom.

Objectives

The primary purpose of this project was to compare preoperative brain imaging between EL and non-EL patients to determine which, and to what extent, brainstem structures were compressed by EL-causing mass lesions. The secondary outcome of this project was to compare, between EL and control patients, the changes in QOL scores pre- versus postoperatively.

Methods

The Vancouver General Hospital (VGH) neurosurgical/neurophysiology database was retrospectively reviewed. This database contains records for all resections of posterior fossa tumors with brainstem compression, since neuromonitoring is essential for cases where tumors are in close proximity to brainstem structures and cranial nerves. The date range for this study was January 1, 2002, to December 31, 2018. This date range was selected because it matches the tenure of the skull base neurosurgeon at VGH. During this time, intraoperative monitoring of cranial nerves became commonplace for complex cases, including those reviewed in this study.

Neurosurgical preoperative clinic notes were reviewed. EL symptoms, when present, were noted in the “Physical Examination” or the “History” section. Patients with explicit mention of having EL symptoms in their consultation report were further screened for exclusionary factors. All eligible patients were included in this investigation.

Control patients were matched to EL cases in a 1:3 ratio (cases:controls). To mitigate confounding factors, controls were assigned in the following manner:

• (a) For each EL patient, a list of possible controls was created. Controls were matched for:

  • – The treating neurosurgeon.

  • – Time of surgical treatment (±2 years).

  • – Tumor pathology.

  • – General tumor location.

• (b) From the list, three patients were randomly selected using a random number generator.

• (c) Scans of the selected controls were reviewed to ensure that the tumor type and general location were consistent with preselection categorization. Any discrepancies in tumors pathology and/or general location were corrected by repeating steps b and c to select another control.

• (d) If the selected patient met the exclusion criteria, steps b and c were repeated to select another control.

Exclusion criteria (for EL and control patients):

• Patients under 18 years of age.

• Patients who had more than one mass lesion visible on their preoperative magnetic resonance imaging (MRI) scan.

• Patients who have had a brain tumor at any other point during their life.

• Patients who were diagnosed with neurofibromatosis type 2.

• Patients who have had any previous psychiatric diagnosis.

• Patients who have had any neurodegenerative or demyelinating disease diagnosis.

• Patients who have had moderate-to-severe brain injury.

• Patients who have had any previous strokes or have evidence of previous infarcts on their preoperative MRI scan.

Clinical Chart Review

A clinical chart review was completed. Patient demographics, clinical presentation, radiological characteristics, surgical features, mortality, clinical outcomes, and radiological outcomes were collected for the cases and matched controls. Clinical variables were compared between EL patients and control patients.

Descriptive statistics are outlined, and clinical characteristics are tabulated and presented in tables. Binomial logistic regression was used to compare clinical characteristics between EL patients and matched controls. For all analysis, the cutoff significance value was set a priori to 0.05.

Image Analysis Review

Preoperative axial T2-weighted fluid-attenuated inversion recovery (FLAIR) MRI scans were reviewed for each case and control patient. If several preoperative scans were available, the one closest in date to the surgery was reviewed. To standardize the process, the following procedure was completed for each scan ( Fig. 1 ). At the level of the medulla, a horizontal line was drawn between the jugular foramina. At the level of the pons, a horizontal line was drawn between the internal acoustic canals. At the level of the midbrain, a horizontal line was drawn between the inferior horns of the lateral ventricle. A vertical line was then drawn at each level from the nasal septum, through the measured center point of the respective aforementioned horizontal line, to the occipital bone. Within the superior-inferior anatomical boundaries of each brainstem structure, the widest point of the tumor was used for the measurement of brainstem compression.

Fig. 1.

Example of measuring and calculating brainstem compression at the level of the ( A ) medulla, ( B ) pons, and ( C ) midbrain.

We created a novel lateral brainstem compression scale ( Fig. 2 ) to characterize the degree of mass effect at the level of the medulla, pons, and midbrain. The degree of compression from the mass lesion onto the respective brainstem structure was measured using the demarked lines and categorized according to the lateral brainstem compression scale. Three compression scores, at the medulla, pons, and midbrain, were collected for each patient. Scores at each level were subsequently compared between EL patients and control patients.

Fig. 2.

Lateral brainstem compression scale.

Binomial logistic regression was used to determine whether a statistically significant difference exists for lateral brainstem compression scale scores between EL participants and matched controls. For all analysis, the cutoff significance value was set a priori to 0.05.

Quality-of-Life Review

As standard of care, patients who have a skull base neurosurgery clinic appointment at VGH complete a Short Form-36 version 1 (SF36v1) survey prior to seeing the surgeon. The SF36v1 is a general health survey. The prospectively collected scores were retrospectively obtained from patient charts. Pre- and postoperative SF36v1 survey scores were retrieved for each EL case and control patient.

Only patients with a complete set of at least one pre- and postoperative survey were included.

Several patients had more than one pre- or postoperative survey. The preoperative survey with the date closest to the surgery date was used for all patients in all analyses. Three analyses were completed:

1. First postoperative QOL survey : the first postoperative survey was used for comparison. Per standard of care, the first postoperative follow-up appointment is scheduled for approximately 6 to 8 weeks after surgery.

2. One-year postoperative QOL survey : the last survey within a 1-year postoperative period was used for comparison. Patients are typically scheduled for a 6-month follow-up appointment per standard of care. If a patient has no postoperative survey in the 1 year after surgery, they will be excluded from this analysis.

3. Last available postoperative QOL survey : the last available survey was used for comparison. Long-term follow-up varies greatly based on the patient's ongoing symptoms.

SF36v1 scores were tabulated using the standard scoring protocol for this health survey. ³² Subsection scores and “Health Change” as a separate metric were calculated. In all analyses, each patient served as their own control: the preoperative score was subtracted from the postoperative score to calculate change in QOL score. The changes in scores were compared between EL and control patients for each analysis. A linear mixed-effects model was used to compare the change in SF36v1 subscores and in the health change metric between EL cases and matched controls. For all analysis, the cutoff significance value was set a priori to 0.05.

Results

Clinical Findings

Of the 673 posterior fossa tumor patients treated for brainstem-compressing extra-axial tumors between January 1, 2002, and December 31, 2018, 11 (2%) patients were found to have EL symptoms preoperatively. Of these 11 cases, 7 (64%) were female and the mean age was 50.0 years (range, 32–65 years) at the time of presentation ( Table 2 ). The mean tumor size was 38.1 mm (range, 15–56 mm). Of the 11 EL patients, 7 (64%) had meningiomas, while 4 (36%) had vestibular schwannomas. Of the meningiomas, 5 were petroclival meningiomas, while 2 were cerebellopontine angle meningiomas. EL ceased for all 11 patients postoperatively. The majority of patients in both groups underwent a gross total or a near-total resection of their tumor.

Table 2. Clinical characteristics of EL and control patients.

Variable	EL ( n  = 11)	Control ( n  = 33)	p -Value
Sex
Male	4 (36%)	16 (48%)	0.711
Female	7 (64%)	17 (52%)
Age
Mean (y)	50.0	51.3	0.704
Handedness
Right	9 (82%)	32 (97%)	0.918
Left	2 (18%)	1 (3%)
Tumor laterality
Right	6 (55%)	20 (61%)	0.438
Left	5 (45%)	13 (39%)
Tumor size
Mean (mm)	38.1	37.2	0.695
Cerebellar findings
Present	7 (64%)	1 (3%)	0.003

Note: p -value in bold denotes a statistically significant difference.

A binomial logistic regression was done to assess differences between the EL and control groups. The proportion of EL patients with cerebellar findings, including dysmetria, clumsiness, and slowness, was significantly greater than the proportion in the control population ( p  = 0.003). Cerebellar findings preoperatively were the sole clinical characteristic that demonstrated a statistically significant difference between the two groups.

Imaging Analysis Findings

All 11 EL patients and 33 matched controls were included in the lateral brainstem compression analysis. The mean compression scores for the two groups at each of the three brainstem levels are outlined in Table 3 . A binomial logistic regression found that compared with the control group, EL-causing tumors exert greater compression onto the pons (mean EL compression score = 2.9, mean control compression score = 1.9, two-tailed p  = 0.02). EL tumors compress the pons on average in the 50 to 74% range, while the non-EL tumors compress in the 25 to 49% range. There was no difference in lateral compression found at the level of the midbrain or the medulla ( Fig. 3 ).

Table 3. Lateral brainstem compression scale mean scores.

Brainstem structure	EL ( n  = 11)	Control ( n  = 33)	p -Value
Medulla	1.1	0.9	0.84
Pons	2.9	1.9	0.02
Midbrain	1.5	1.0	0.31

Note: p -value in bold denotes a statistically significant difference.

Fig. 3.

Lateral brainstem compression mean scores for the EL group compared to the control group at the medulla, pons, and midbrain (asterisk denotes a statistically significant difference).

Quality-of-Life Findings

Fig. 4 outlines which patients were included in each of the three QOL analyses. To effectively detect QOL change, only those with a complete set of pre- and postoperative SF36v1 health surveys were compared with their matched controls. Similarly, only controls with a complete set of SF36v1 health surveys were included. Eight EL patients and 20 control patients fit this requirement. For the “1-year postoperative QOL survey” analysis, patients without a postoperative survey in the 1 year after surgery were additionally excluded. For all three analyses, the median was used to describe the time between two events because outlier data points skew the mean values, rendering the mean an ineffective tool for analyzing this dataset.

Fig. 4.

Patient inclusion flowchart for the QOL analyses.

First Postoperative QOL Survey (EL, n  = 8; Controls, n  = 20)

This analysis compared patients' preoperative survey to their first postoperative survey. The median time between the preoperative survey and surgery was 5.6 weeks (range, 0.9–37.1 weeks) for EL patients and 9.9 weeks (range, 2.0–68.6 weeks) for control patients. The median time between the surgery and the first postoperative survey was 9.8 weeks (range, 5.9–72.9 weeks) for EL patients and 8.6 weeks (range, 5.9–35.6 weeks) for control patients.

The change in QOL scores are outlined in Table 4 . The mean total SF36v1 improvement experienced by EL patients (167-point improvement) was greater than that experienced by control patients (143-point improvement). “Health Change” is not included in the sum of SF36v1 subscores. A linear mixed-effects model found that patients with EL-causing tumors experience greater improvement postoperatively in the “Health Change” category (two-tailed p  = 0.05). “Health Change” demonstrates patient perception of health and function. Of note, control patients were found to experience greater improvement than EL patients postoperatively in the “Role Limitations Due to Physical Health” (two-tailed p  = 0.03). Changes in the remaining SF36v1 subcategories did not demonstrate a statistically significant difference ( Fig. 5 ).

Table 4. Mean change in quality-of-life scores for each SF36v1 subsection and health change as a separate metric using the first postoperative survey.

Patient group	SF36v1 subsection
	∆PF	∆RP	∆RE	∆EF	∆EW	∆SF	∆BP	∆GH	∆HC
EL	−2.0	−3.1	29.3	35.6	38.5	29.9	17.8	20.6	53.1
Control	8.3	32.5	31.8	16.9	13.2	14.4	14.2	12.3	8.8
p -Value	0.166	0.023	0.359	0.462	0.573	0.559	0.663	0.719	0.047

Abbreviations: BP, bodily pain; EF, energy/fatigue; EL, emotional lability; EW, emotional well-being; GH, general health; HC, health change; PF, physical functioning; RE, role limitations due to emotional problems; RP, role limitations due to physical health; SF, social functioning; SF36v1, Short Form-36 version 1.

Note: p -values in bold denote a statistically significant difference.

Fig. 5.

Change in quality-of-life score for each SF36v1 subsection, the sum of subsections, and health change as a separate metric using the first postoperative survey. BP: bodily pain; EF: energy/fatigue; EW: emotional well-being; GH: general health; HC: health change; PF: physical functioning; RE: role limitations due to emotional problems; RP: role limitations due to physical health; SF: social functioning. Asterisk denotes a statistically significant difference.

One-Year Postoperative QOL Survey (EL, n  = 6; Controls, n  = 14)

This analysis compared patients' preoperative survey to their last survey within a 1-year postoperative period. There were 2 EL patients removed from this analysis because their postoperative survey(s) did not fall within the 1-year window after surgery. Matched controls for these 2 EL patients were also removed from this analysis. Ultimately, 6 EL patients and 14 controls were included in this analysis. For patients who had only one postoperative survey that happened to be completed within the first year after surgery, the QOL score change for this analysis would be congruent with those stated in the “first postoperative QOL survey” section.

The median time between the preoperative survey and surgery was 6.7 weeks (range, 2.1–37.1 weeks) for EL patients and 10.6 weeks (range, 2.6–45.1 weeks) for control patients. The median time between the surgery and the postoperative survey was 23.8 weeks (range, 10.3–43.9 weeks) for EL patients and 24.5 weeks (range, 7.3–43.9 weeks) for control patients.

Change in QOL scores are outlined in Table 5 . The mean total SF36v1 improvement experienced at this time point by control patients (156-point improvement) was greater than that experienced by EL patients (126-point improvement), as demonstrated in Fig. 6 . “Health Change” is not included in the sum of SF36v1 subscores. A linear mixed-effects model found that patients with EL-causing tumors experience greater improvement at this postoperative time point in the “Health Change” category (two-tailed p  = 0.03). Changes in the remaining SF36v1 subcategories did not demonstrate a statistically significant difference.

Table 5. Mean change in quality-of-life scores for each SF36v1 subsection and health change as a separate metric using the last survey within 1 year after surgery.

Patient group	SF36v1 subsection
	∆PF	∆RP	∆RE	∆EF	∆EW	∆SF	∆BP	∆GH	∆HC
EL	10.7	12.5	16.7	32.5	20.7	29.2	−5.7	9.2	62.5
Control	8.9	50.0	26.1	14.6	20.3	21.4	8.4	6.4	12.5
p -Value	0.692	0.193	0.639	0.246	0.590	0.280	0.076	0.510	0.026

Abbreviations: BP, bodily pain; EF, energy/fatigue; EL, emotional lability; EW, emotional well-being; GH, general health; HC, health change; PF, physical functioning; RE, role limitations due to emotional problems; RP, role limitations due to physical health; SF, social functioning; SF36v1, Short Form-36 version 1.

Note: p -value in bold denote a statistically significant difference.

Fig. 6.

Change in quality-of-life score for each SF36v1 subsection, the sum of subsections, and health change as a separate metric using the last survey within 1 year after surgery. BP: bodily pain; EF: energy/fatigue; EW: emotional well-being; GH: general health; HC: health change; PF: physical functioning; RE: role limitations due to emotional problems; RP: role limitations due to physical health; SF: social functioning. Asterisk denotes a statistically significant difference.

Last Available Postoperative QOL Survey (EL, n  = 8; Controls, n  = 20)

This analysis compared patients' preoperative survey to their last available postoperative survey. If patients had only one postoperative survey, the QOL score change for this analysis would be congruent with those stated in the “first postoperative QOL survey” section. If the last available survey was within the 1-year postoperative window, the QOL score change for this analysis would be congruent with those stated in the “1-year postoperative QOL survey” section. For patients with several postoperative surveys, the relative long-term change in QOL can be noted in this analysis.

The median time between the preoperative survey and surgery was 5.6 weeks (range, 0.9–37.1 weeks) for EL patients and 9.9 weeks (range, 2.0–68.6 weeks) for control patients. The median time between surgery and the last available postoperative survey was 71.9 weeks (range, 27.9–426.9 weeks) for EL patients and 35.6 weeks (range, 7.4–262.1 weeks) for control patients.

The change in QOL scores are outlined in Table 6 . The mean total SF36v1 improvement experienced by EL patients (141-point improvement) was greater than that experienced by control patients (132-point improvement), as demonstrated in Fig. 7 . “Health Change” is not included in the sum of SF36v1 subscores. A linear mixed-effects model found that there was no statistically significant difference between EL patients and controls patients in any SF36v1 subscore or in the “Health Change” metric in this analysis.

Table 6. Mean change in quality-of-life score for each SF36v1 subsection and health change as a separate metric using the last available postoperative survey.

Patient group	SF36v1 subsection
	∆PF	∆RP	∆RE	∆EF	∆EW	∆SF	∆BP	∆GH	∆HC
EL	13.6	9.4	20.9	35.6	28.5	20.4	5.5	7.5	43.8
Control	8.5	36.3	28.3	12.4	17.4	18.1	3.1	7.5	18.8
p -Value	0.855	0.190	0.530	0.212	0.753	0.847	0.849	0.381	0.498

Abbreviations: BP, bodily pain; EF, energy/fatigue; EL, emotional lability; EW, emotional well-being; GH, general health; HC, health change; PF, physical functioning; RE, role limitations due to emotional problems; RP, role limitations due to physical health; SF, social functioning; SF36v1, Short Form-36 version 1.

Fig. 7.

Change in quality-of-life score for each SF36v1 subsection, the sum of subsections, and health change as a separate metric using the last available postoperative survey. BP: bodily pain; EF: energy/fatigue; EW: emotional well-being; GH: general health; HC: health change; PF: physical functioning; RE: role limitations due to emotional problems; RP: role limitations due to physical health; SF: social functioning.

Discussion

The main findings of this project were that (1) patients with EL preoperatively have greater brainstem compression at the level of the pons, (2) patients with EL preoperatively more often have cerebellar findings on neurological exam, and (3) patient perception of health improves in the period immediately after surgery and continues for at least 1 year when resection cures EL symptoms. This is the largest case series to date on adult extra-axial posterior fossa tumors that cause EL.

The homogeneity in surgical treatment paradigm and postoperative care received by patients strengthens our foundational understanding of this rare clinical phenomenon. All data was collected from a single institution and all patients were treated by the same fellowship-trained skull base neurosurgeon, meaning that all patients were treated according to a uniform surgical treatment paradigm. This homogeneity limits discrepancies that may arise when comparing surgeons who subscribe to a “wait-and-watch” approach versus those who prefer aggressive resection. The uniformity of postoperative care and rehabilitation resources available to patients at VGH strengthens the QOL findings herein. Cases were matched to controls that had a similar timeframe for surgery (±2 years) to ensure that these consistencies were maintained.

Brainstem Compression

Our results demonstrate that EL tumors compress the pons laterally to a greater degree than non-EL tumors. Specifically, EL tumors were found to compress the pons on average in the 50 to 74% range, while the non-EL tumors compress in the 25 to 49% range. Compression scores at the level of the medulla and midbrain did not demonstrate a statistically significance difference between groups. Additionally, there was no discrepancy in mean tumor size between groups. These findings support our hypothesis that the basis pontis may be implicated in the EL pathway.

The cerebellum is essential for movement coordination. The current body of evidence in the literature suggests that the cerebellum also modulates stereotyped expressions of emotion according to specific contextual information. ⁴ Depending on social context, expressions of laughter and crying to the same stimulus can be scaled up, suppressed marginally, or inhibited completely by input from the cerebellum. This theory was introduced by Parvizi and colleagues in 2001 as an alternative theory to the “disinhibition model,” which was proposed by Wilson in 1924. ³³ Parvizi's model emphasizes the cerebellum as a center for coordination of preprogrammed stereotyped expressions of emotion through the corticopontocerebellar pathway. Through this pathway, key cognitive and social context provided by telencephalic structures is relayed through the basis pontis to the cerebellum, where the information is used for regulation. The modulation of expressions by the cerebellum to social context is a learned pattern of behavior. Subsequently, this information is relayed back to telencephalic inductor and effector sites in cortical and subcortical regions of the brain, leading to context-appropriate laughter and crying.

To contextualize the brainstem compression findings herein, we propose that disruption of the pons, in this case due to compression from mass lesion, leads to deafferentation of the cerebellum from cortical and limbic structures through the basis pontis in a manner consistent with the model proposed by Parvizi and colleagues. Ultimately, this impairs modulation of emotional responses to contextual information, rendering patients unable to control their laughter and/or crying. The study results suggest that compression onto the pons inhibits control over involuntary, stereotyped expression of emotion.

Cerebellar Findings

Compared to control patients, EL patients more commonly have motor cerebellar findings preoperatively during a standard neurological exam. Greater disruption of EL tracts is correlated with greater disruption of motor tracts to the cerebellum. EL patients more frequently present with dysmetria, clumsiness, and slowness when asked to complete coordinated movements. This is likely because motor tracts communicating to the cerebellum may also be affected by the increasingly compressed basis pontis. While EL is a motor response, it is stipulated that the EL-related circuitry involves different tracts from those engaged in “non-EL” motor responses. This may explain why patients with cerebellar findings, like dyscoordination, commonly do not exhibit EL symptoms and, inversely, why not all EL patients in our cohort had cerebellar findings during their neurological exam.

Quality-of-Life Findings

The 6- to 8-week postoperative survey scores and the 1-year postoperative survey scores demonstrate the acute change in QOL. This is reflective of the immediate cessation of EL symptoms postoperatively. In support of our hypothesis, it was found that EL patients experience greater acute improvement postoperatively in the “Health Change” metric, which ascertains perception of health. In six of the eight SF36v1 subsections, EL patients experienced improvement postoperatively. This finding on perceived health change augments the benefits of obtaining EL-alleviating surgery for patients.

The 6- to 8-week postoperative surveys unexpectedly demonstrated a greater improvement postoperatively in the “Role Limitations Due to Physical Health” subsection for the control group. This can be attributed to one outlier patient in the EL group whose deterioration postoperatively skewed the small dataset as he suffered from prolonged debilitating diplopia during his first postoperative follow-up. Notably, the diplopia resolved completely by his second follow-up appointment and his QOL survey scores improved concordantly. Even with the outlier EL patient, the sum of all subscores demonstrates that EL patients experience greater average improvement compared with the control group.

Notably, at the 1-year postoperative time point, control patients experience greater mean improvement in total SF36v1 score; this difference does not reach statistical significance. EL patients have a greater mean improvement in SF36v1 total score at the other two time points. One of two phenomena is thought to be responsible for this outcome: (1) this is a function of the reduced sample size for this analysis (EL, n  = 6; controls, n  = 14) or (2) at this time point, the control patients truly have a greater improvement than EL patients. The evidence toward the former rationale is superior; there was a reduced sample size in this specific analysis due to the nature of the 1-year timeline. As sample size decreases, the effect of outlier data points on the mean total score increases. Since no specific SF36v1 subsection demonstrated a statistically significant difference between the control population and the EL population, the greater total score change for the control group can be attributed mostly to the reduced sample size.

Although there was no statistically significant difference in QOL subscores between the EL and control groups in the long-term follow-up surveys, EL patients were found to have greater mean improvement in the total survey score (141-point improvement) compared with the control patients (132-point improvement). Similarly, the change in “emotional well-being” subscore were greater in the EL group than in the control group on average. These findings were present in spite of the diminished outcomes for the EL group in the “Role Limitations Due to Physical Health” subsection during the immediate postoperative phase, which ultimately stabilized over time.

It can be concluded that in the long term, metrics scored by the SF36v1 survey stabilize between both groups as patients generally experience long-term improvements postoperatively. Despite the challenging anatomy of brainstem-compressing tumor resections, lasting and disabling postoperative complications were very rare in our patient population.

Limitations and Future Direction

This series was retrospective and thus has all the limitations inherent to this study design. The presence of EL symptoms noted in patients was dependent on the accuracy of medical records. Since the link between EL and brainstem-compressing posterior fossa tumors is not widely known, there may be underreporting of the symptom. Since there is no expected frequency for EL symptoms, it is possible that the outbreaks go unnoticed during a standard 1-hour neurosurgery consultation. Patients and their family members may not associate the inappropriate outbreaks to a brain tumor, thus failing to report EL when prompted for the patient's history and symptoms.

The study was limited in its small sample size, leading to the potential for data skew. We were unable to obtain QOL scores for all 11 EL and 33 control patients largely because this study is based on data collected for standard of care, so there was no regimented research timeline. While uncommon, some patients did not complete the QOL surveys due to frank refusal or forgetfulness.

Future imaging research should aim to perform comparative analysis with prospectively collected diffusion tensor imaging studies so that affected brainstem regions can be mapped pre- and postoperatively for a better understanding of the tracts involved in the regulation of emotional expression. In this investigation, we did not analyze tumor compression of the brainstem in the anterior-posterior direction. Tumors generally have a sidedness, and the lateral scale was sufficient to capture the degree of compression at each brainstem level. An anterior-posterior compression analysis would be useful for truly midline clival tumors, which are rare. Furthermore, none of the tumors in our study were pure midline clival tumors.

Future studies reporting QOL changes should consider utilizing prospective data collection methods to follow patients over an extended period of time to maximize and standardize data points collected for each patient. To improve generalizability and shorten accrual timelines, a multicenter study is suggested. Lastly, the SF36v1 health survey is a broad QOL assessment; a brain tumor–specific QOL survey (e.g., Functional Assessment of Cancer Therapy-Meningioma survey) may be considered for future prospective studies to ascertain QOL as it relates to neurological status in addition to metrics measured by general health surveys.

Conclusion

This retrospective study is the largest case series to date that investigates adult extra-axial posterior fossa tumors that cause EL. This study strengthens the body of evidence that EL may be attributed to deafferentation of the cerebellum from cortical and limbic structures through the basis pontis, leading to impaired modulation of emotional response. The clinical and image analysis results suggest that compression onto the pons inhibits control over involuntary, stereotyped expression of emotion. The QOL results indicate that increased improvement in health perception postoperatively for EL patients augment the benefits of offering patients EL-alleviating surgery. With greater understanding of how mass lesions can cause EL, general practitioners and neurosurgeons can specifically ask patients with brainstem-compressing tumors whether they have experienced this symptom. More accurate screening and reporting of EL symptoms would be helpful for both clinical purposes and future research aims.