Symptom resolution in infiltrating WHO grade II-IV glioma patients undergoing surgical resection

Abstract

Past studies of morbidity in patients with infiltrating gliomas have focused on the impact of surgery on quality of life. Surprisingly, little attention has been given to the rate at which the presenting symptoms improve after surgery, even though this is often the patient’s first concern. This study is an initial effort to provide useful information about symptom resolution and factors predicting persistence of symptoms in glioma patients who undergo surgery. We conducted a retrospective analysis on patients who underwent surgery for World Health Organization (WHO) grade II-IV astrocytoma/oligodendroglioma/oligoastrocytoma at our institution. All patients were seen 2–4 months postoperatively, and asked about the persistence of symptoms they experienced preoperatively. Symptoms reported in clinic were assessed against symptoms reported prior to surgery. Our study includes 56 consecutive patients undergoing surgery for gliomas. Of patients who experienced symptoms initially, headache resolved in 18/27 postoperatively, weakness resolved in 8/14 postoperatively, altered mental status resolved in 8/12 postoperatively, vision problems resolved in 7/11 postoperatively, nausea resolved in 5/7 postoperatively, and ataxia resolved in 4/5 postoperatively. Headache was more likely to resolve in patients with frontal or temporal tumors (p = 0.02). Preoperative Karnofsky Performance Scale (KPS) of 70 or less was associated with longer postsurgical hospital stay (p < 0.01). Younger patients were more likely to experience a resolution of altered mental status (p = 0.04). Our analysis provides data regarding the rate at which surgery alleviates patient symptoms and considers variables predicting likelihood of symptom resolution. Some patients will experience symptom resolution following resection of WHO grade II-IV gliomas in the months following surgery.

1. Introduction

A diverse collection of initial symptoms may herald the diagnosis of a high- or low-grade glioma. Common symptoms encountered by infiltrating glioma patients include headache, weakness, ataxia, vision deficits, and cognitive changes. These tumors often have devastating outcomes for patients and can eventually be fatal. However, survival in glioma patients continues to improve with better medical and surgical therapeutic options.

A number of studies have examined the effect of surgical resection on overall survival and quality of life, but few have explored the rate at which the presenting symptoms improve after surgery. Symptom resolution in glioma patients following tumor resection has not been well described, though this often is the patient’s greatest initial concern. Further, as survival in this population continues to improve, these patients are living longer with their symptoms.

As tumor resection has an important role in treating World Health Organization (WHO) grade II-IV gliomas [1–5], we aim to better understand the frequency of symptom resolution in patients who undergo surgery. To our knowledge, no study has yet examined the frequency of resolution of symptoms most common in glioma patients. This study presents an initial investigation of factors predicting persistence of symptoms despite surgery.

2. Materials and methods

2.1. Patient population

We performed a review of patients who recently underwent surgery at our institution for tumors involving the cerebral hemispheres. Only patients with histopathologic diagnosis of WHO grade II-IV astrocytoma, oligodendroglioma, and mixed oligoastrocytoma were included. We performed a retrospective assessment of hospital charts, medical records, and pathology reports. Only patients with complete medical records were included. This study was performed with approval of our Institutional Review Board (IRB #3199).

2.2. Treatment

Decisions regarding surgical approach to operative goals were by attending physician preference. All patients after 2005 received postoperative temozolomide and radiotherapy fitting with standard practices [6]. Steroids were weaned to the lowest dose possible within 3 weeks of surgery. MRI with gadolinium contrast was performed pre- and postoperatively.

2.3. Outcome assessment

History and physical examinations were performed preoperatively by an attending neurosurgeon. Preoperative Karnofsky Performance Scale (KPS) scores were assessed during this examination. Patients were asked whether symptoms reported preoperatively had changed at follow-up appointments 2–4 months postoperatively. We chose this duration as it is a typical follow-up time for re-imaging in tumor patients. Symptoms reported in clinic at that time were assessed against symptoms reported by patients prior to surgery. New complaints at follow up were also noted.

For initial complaints, including vision changes, nausea, and ataxia, the symptom was considered resolved if the patient reported absence (or near absence) of the symptom at postoperative follow-up. If the symptom, as reported by the patient, worsened or failed to improve completely with surgery, it was considered unresolved. Headache as an initial complaint was assessed by frequency and severity, and was considered resolved if the patient reported headaches occurring less than once per week and not restricting activity. Weakness was recorded as resolved if deficits noted on neurological exam preoperatively were absent postoperatively at clinic. Weakness that failed to improve with surgery or that worsened postoperatively was recorded as unresolved. Mental status was assessed pre- and postoperatively by the attending neurosurgeon as part of the standard neurological exam, and any abnormal exam was documented. Mental status was recorded as resolved if the patient had returned to his or her pre-tumor baseline mental status by the time of the follow-up appointment.

2.4. Statistical analysis

Potential differences in each subgroup of presenting symptom were assessed to identify factors associated with symptom resolution. Categorical variables were compared using Pearson chi-squared test. Fisher exact test was used if more than 80% of values were less than 5. Continuous variables were compared using independent samples Student’s t-test. Confidence limits for proportions were calculated using the Clopper–Pearson Exact method. A p value ≤ 0.05 was considered statistically significant.

3. Results

3.1. Patient characteristics

Patient demographics and tumor characteristics are summarized in Table 1. We identified 56 patients, 24 (43%) women and 32 (57%) men. Patient ages ranged from 20 to 82 years; median age was 50 years. Nine of 56 patients (16%) had WHO grade II astrocytoma/oligodendroglioma/oligoastrocytoma tumors, and 47/56 (84%) had WHO grade III-IV astrocytoma/oligodendroglioma/oligoastrocytoma/glioblastoma tumors. Twenty-three (41%) had left-sided tumors, 26 (46%) had right-sided tumors, and seven (13%) had tumor involvement bilaterally. Twenty-eight (49%) were in the frontal lobe, 10 (18%) were primarily in the temporal lobe, 15 (26%) were in the parietal lobe, and four (7%) were in the occipital lobe.

Table 1.

Patient demographics and tumour characteristics of study group undergoing resection of WHO grade II-IV glioma

Gender	N	Median Age (Range)
Total	56	50 (20–82)
Women	24 (43%)	50 (20–82)
Men	32 (57%)	50 (21–79)
WHO Grade	N
II	9 (16%)
III-IV	47 (84%)
EOR	N
100%	19 (34%)
90%–99%	14 (25%)
70%–89%	23 (41%)
Tumor Location	N
Frontal	28 (49%)
Temporal	10 (18%)
Parietal	15 (26%)
Occipital	4 (7%)
Tumor Side	N
Left	23 (41%)
Right	26 (46%)
Bilateral	7 (13%)
KPS Score	N
Preoperative Median	70

EOR = extent of resection, KPS = Karnofsky Performance Scale, WHO = World Health Organization

3.2. Symptom resolution

Frequencies of presenting symptoms and symptomatic resolution are given in Table 2. Headache was the most commonly reported presenting symptom in 27/56 (48%) patients and improved in 18/27 (67%) patients (95% CI: 46%–84%). Weakness was a presenting symptom in 14/56 (25%) patients, and resolved postoperatively in 8/14 (57%) patients (95% CI: 29% 82%). Altered mental status (AMS) occurred in 12/56 (21%) patients and resolved in 8/12 (67%) patients (95% CI: 35%–90%). Difficulty with vision was a presenting symptom in 11/56 (19%) patients, and resolved in 7/11 (64%) patients (95% CI: 35%–92%). Nausea was a presenting symptom in 7/56 (13%) patients and resolved in 5/7 (71%) patients (95% CI: 29%–96%). Ataxia occurred in 5/56 (9%) patients and resolved in 4/5 (80%) patients following surgery (95% CI: 28%–99%).

Table 2.

Frequency of presenting symptoms and symptom resolution of WHO grade II-IV glioma

Presenting Symptom	Resolution
Headache	18/27 (67%) (95% CI: 46%–84%)
Weakness	8/14 (57%) (95% CI: 29%–82%)
Altered mental status	8/12 (67%) (95% CI: 35%–90%)
Vision Problems	7/11 (64%) (95% CI: 31%–89%)
Nausea	5/7 (71%) (95% CI: 29%–96%)
Ataxia	4/5 (80%) (95% CI: 28%–99%)

3.3. Tumor location and symptomatic outcome

Headache resolution by tumor location is shown in Table 3. Tumor location was associated with symptomatic resolution in patients who presented with headache (frontal: 73%; temporal: 0%; parietal: 90%; occipital: 25%; p = 0.02). Headache resolved postoperatively in 8/11 (73%) and 9/10 (90%) patients presenting with frontal lobe and parietal lobe tumors, respectively. Headache resolution was reported in 0/2 (0%) and 1/4 (25%) patients with temporal lobe and occipital lobe tumors, respectively. Outcomes of other symptoms were not associated with tumor location. Patients with tumors of the left hemisphere were not more likely to have a postoperative KPS of 70 or less, or to have postoperative KPS improvement, than patients with tumors of the right hemisphere (p = 0.07 and 0.24, respectively), as shown in Table 4.

Table 3.

Headache resolution by tumor location after resection of WHO grade II-IV glioma

Location	Resolution
Frontal	8/11 (73%)
Temporal	0/2 (0%)
Parietal	9/10 (90%)
Occipital	1/4 (25%)

Fischer’s exact p = 0.02.

Table 4.

Univariate analysis of factors influencing return of function after resection of WHO grade II-IV glioma

	Headache	Weakness	Vision	Altered mental status	Nausea	Postoperative KPS improvement	Postoperative KPS ⩽70
Sex	1.2 (0.25–6.24)	1.7 (0.12–24.26)	1.3 (0.11–15.70)	2.0 (0.99–4.00)	0.5 (0.19–1.33)	0.8 (0.19–2.95)	1.5 (0.50–4.84)
Male vs. Female	p = 0.78†	p = 0.70†	p = 0.65*	p = 0.14*	p = 0.29*	p = 0.68†	p = 0.51†
Preoperative KPS	3.4 (0.60–19.64)	0.5 (0.32–0.94)	1.3 (0.11–15.70)	0.8 (0.50–1.12)	1.5 (0.67–3.33)	2.0 (0.37–11.00)	17.4 (3.97–76.20)
⩽70 vs >70	p = 0.16†	p = 0.22†	p = 0.65*	p = 0.78*	p = 0.50*	p = 0.41†	p < 0.01†
Tumor Side**	0.7 (0.14–3.61)	8.3 (0.63–110.02)	0.8 (0.05–13.63)	0.38 (0.02–6.35)	0.7 (0.02–18.06)	2.3 (0.57–9.22)	0.3 (0.08–1.15)
Left vs. Right	p = 0.68†	p = 0.09†	p = 0.72*	p = 0.50*	p = 0.71*	p = 0.24†	p = 0.07†
EOR	2.0 (0.38–10.58)	0.3 (0.33–2.76)	0.5 (0.02–11.08)	1.0 (0.06–15.99)	0.5 (0.19–1.33)	1.0 (0.26–3.77)	1.2 (0.40–4.09)
STR vs. GTR/NTR	p = 0.41†	p = 0.30*	p = 0.62*	p = 0.76*	p = 0.29*	p = 0.99†	p = 0.67†
WHO Grade	1.6 (0.14–18.00)	1.8 (0.99–3.07)	1.2 (0.07–19.63)	1.6 (0.99–2.46)	1.5 (0.85–2.64)	0.4 (0.05–4.04)	0.5 (0.11–2.13)
II vs. III/IV	p = 0.70†	p = 0.37†	p = 0.72*	p = 0.67*	p = 0.71*	p = 0.47†	p = 0.34†

EOR = extent of resection, GTR = gross total resection, KPS = Karnofsky Performance Status, NTR = near total resection, STR = subtotal resection, WHO = World Health Organization.

^†Denotes chi-squared p value.

^*Denotes Fischer exact p value.

^**Excluding tumors involving both hemispheres.

3.4. Importance of KPS

Low preoperative KPS was associated with greater length of hospital stay (LOS) following surgery, as shown in Table 5. Patients with a preoperative KPS of 70 or less spent an average of 6.1 days in the hospital following surgery, while those with a preoperative KPS greater than 70 spent an average of 3.2 days in the hospital following surgery (p < 0.01). Three patients (1 preoperative KPS > 70; 2 preoperative KPS ≤ 70) with an extreme duration of hospital stay (greater than 21 days) were not included in this statistic. Postoperative KPS was not more likely to improve in patients with a preoperative KPS of 70 or less (OR: 2.0; 95% CI: 0.37–11.00; p = 0.41). Preoperative KPS of 70 or less predicted a postoperative KPS of 70 or less (OR: 17.4; 95% CI: 3.97–76.20; p < 0.01). This is shown in Table 4.

Table 5.

Factors influencing outcome after resection of WHO grade II-IV glioma

Effect of preoperative KPS on length of hospital stay*
Variable	N	Hospital Days (mean)	SEM	p value
Preoperative KPS ⩽ 70	33 (65%)	6.1	0.5	<0.01*
Preoperative KPS > 70	18 (35%)	3.2	0.3
Effect of age on recovery from altered mental status
Variable	N	Age (mean)	SEM	p value
Return to baseline	8 (66%)	47	5.6	0.04*
No return to baseline	4 (33%)	64	3.9

KPS = Karnofsky Performance Scale; SEM = standard error of the mean.

^*p value < 0.05 is considered significant.

^*Excluding patients with hospital stays longer than 21 days.

3.5. Effects of age and gender

Age appears to affect resolution of preoperative AMS after surgery, as is also shown in Table 4. A return to baseline mental status was seen in 8/12 (66%) patients with an average age of 47 years, and 4/12 (33%) patients with an average age of 64 years did not have a return to baseline mental status (p = 0.04). Age did not appear to affect resolution of any other symptoms. Male versus female sex had no influence on symptom resolution.

3.6. New postoperative symptoms

Eighteen of 56 (32%) patients developed one or more new symptoms following surgery. Five of 29 (17%) patients developed headache who did not have headache prior to surgery, and 5/42 (12%) patients developed weakness who did not have weakness prior to surgery. Other new postoperative symptoms were less common, and are shown in Table 6.

Table 6.

New postoperative symptoms after resection of WHO grade II-IV glioma^*

New Symptom**	N
Headache	5/29 (17%)
Weakness	5/42 (12%)
Altered mental status	3/44 (7%)
Vision problems	2/45 (4%)
Nausea	4/49 (8%)
Ataxia	2/51 (4%)

^*Patients may have developed more than one symptom.

^**Excluding patients who presented with the symptom initially.

4. Discussion

Glioma patients may present with a range of different symptoms. With the exception of early efforts by Cushing [7], and later Bucy [8], relatively little attention has been given to understanding the persistence of symptoms in these patients following surgery. This study is a preliminary analysis of symptom resolution, and factors predicting persistence of symptoms, in WHO Grade II-IV patients who undergo surgery. We found headache to be the most common presenting symptom, and noted that patients often experience symptom resolution following surgery.

Past studies of postoperative morbidity in glioma patients have relied on quality of life (QoL) assessments, and these studies in glioma patients have yielded mixed results. One study has suggested improved QoL following surgical resection owing to improved cognitive function and alleviation of neurologic symptoms [9], while another found modern glioma surgery does not affect QoL [10]. Various QoL metrics have been validated for use in glioma patients [11–15]. However, our study differs from QoL studies that account for many variables, as the variables reported in this work are clean and direct.

In our study, headache was the most common presenting symptom, occurring in 27/56 (48%) patients. This is consistent with past reports. Chang et al. previously reported headache to be the most common presenting symptom in patients with WHO grade III-IV gliomas [16]. Headache has been previously reported as a presenting symptom in 37% to 56% of patients [16–20]. We observed an overall resolution of headache in 18/27 patients (67%, 95% CI: 49%–84%).

Forsyth and colleagues reported that frontal headaches are most common in brain tumor patients [17], suggesting frontal meninges may be more sensitive to irritation by intracranial pathology. However, the mechanism behind this is unclear. This may be linked to the association we found between headache resolution and tumor location (p = 0.02). Patients with frontal and parietal tumors were more likely to have improvement of their headaches following surgery (73% and 90%, respectively) than patients with occipital or temporal tumors (25% and 0%, respectively). Interestingly, previous studies have noted tumors located occipitally to be a predictor of disability [10].

Patients with motor deficits were likely to see improvement postoperatively. Of 14 patients noted to present with weakness prior to surgery, eight had resolution of their symptoms (57%; 95% CI: 29%–82%). Incidence of motor deficits was 25%, which is within the range given in previous reports [20,21]. Of note, only one patient presenting with weakness in this study had a tumor involving the primary motor cortex. In those whose weakness did not resolve, one tumor was located in the basal ganglia, one tumor was located in the corpus callosum, and the remaining tumors involved the frontal lobe. Otten and colleagues have noted that weakness may be due to decreased connectivity within motor functional networks in patients with tumors not involving primary motor areas [22].

Our study supports that preoperative functioning is important to patient outcomes, as low KPS scores predicted longer hospital stays. Further, postoperative KPS score was not significantly more likely to improve in patients with low preoperative KPS scores (p = 0.41). Fitting with earlier reports [23], we saw no significant difference in symptom resolution with respect to tumor hemisphere. Finally, our KPS analysis seems representative of larger pools of glioma patients, reflected by only slightly lower preoperative KPS scores compared to other, larger trials [16].

Improvement of AMS postoperatively is of particular importance as cognitive changes have been shown to be an independent prognostic factor in survival of glioma patients [24]. We demonstrated that older patients were less likely to have a return to baseline cognitive functioning. Our rate of AMS as a presenting symptom is similar to rates cited in other studies [10].

One previous study has shown the impact of symptomatic worsening on patient prognosis [25], which can help guide therapeutic decisions. However, our study lacked the size to assess effects of symptom resolution on patient prognosis, and further study is needed to understand how symptomatic improvement postoperatively is associated with overall survival. Additionally, seizure data was not included in this study, as management and symptoms related to epilepsy were beyond our scope. Our study represents an initial effort to better understand symptomatic resolution in glioma patients following surgical resection.

5. Conclusion

This analysis provides data regarding the rate at which surgery alleviates patient symptoms. Some patients will experience symptomatic resolution following resection of WHO grade II-IV glioma in the months following surgery.