Epilepsy Surgical Resection Results in Better Seizure Control and Better Long-Term Health Related Quality Outcomes

Commentary

Health-Related Quality of Life and Emotional Well-Being After Epilepsy Surgery: A Prospective, Controlled, Long-Term Follow-Up.

Edelvik A, Taft C, Ekstedt G, Malmgren K. Epilepsia 2017;58:1706–1715.

OBJECTIVE: To evaluate health-related quality of life (HRQOL) and emotional well-being in resective epilepsy surgery and nonoperated patients at long-term follow-up. METHODS: This is a prospective cohort study where patients undergoing presurgical work-up during 1995–1998 completed the Short-Form Health Survey (SF-36) and the Hospital Anxiety and Depression scale (HAD) at baseline, and 2 and 14 years after resective surgery or presurgical evaluation (non-operated patients). SF-36 scores were compared to a normative population. Proportions of patients reaching HRQOL changes of minimum clinically important difference (MCID) were calculated. RESULTS: At 14-year follow-up, operated patients scored equal to or better than the normative sample on all SF-36 domains except Social Functioning and Mental Health. Physical component summary (PCS) was better and mental component summary (MCS) was worse than for the normative sample. Nonoperated patients scored worse than the normative sample on five of eight domains, and on PCS and MCS. Change in seizure status from 2 to 14 years did not affect PCS or MCS means. Improvement reaching MCID from baseline to long-term was seen in 50% (PCS) and 47% (MCS) of operated and in 33% (PCS) and 38% (MCS) of nonoperated patients. Worsening was seen in 18% (PCS) and 22% (MCS) of operated and in 38% (PCS) and 38% (MCS) of nonoperated patients. Differences between groups were nonsignificant. HAD scores did not differ between groups, and the numbers of possible or probable cases were low. Patient satisfaction with surgery was higher in operated seizure-free patients. Only 5% of all operated patients considered surgery not to be overall beneficial. SIGNIFICANCE: At the group level, HRQOL was stable 14 years after surgery compared to after 2 years. Social Functioning and Mental Health were still below, but other domains were similar to the normative sample. Individual patterns did not follow seizure outcome changes, indicating that multiple factors are important for long-term HRQOL.

Understanding the impact of epilepsy treatments directly affects how we treat patients. Outcome measurements include seizure counts, seizure freedom, side effects, and health-related quality of life (HRQOL). Health-related quality of life measurements provide important insight into how patients are functioning mentally and physically. Most studies assess outcomes over short periods, but epilepsy surgical resection clearly has long-term effects. Consideration of how surgical resection impacts HRQOL over many years will guide treatment of future generations of patients with refractory epilepsy.

Quality of life among persons with epilepsy is assessed using different inventories; these scales are not all equal. The Short-Form Health Survey (SF-36) is a generic, widely used, and extensively validated questionnaire that measures eight quality of life domains. Higher domain scores correspond to better health. Domain scores are summarized in two composite scores for Physical Health (physical composite summary [PHS]) and Mental Health (mental health summary [MHS]). The QOL in Epilepsy Inventory-31 (31 items) and QOL in Epilepsy-89 (89 items) are inventories used specifically for epilepsy and have standardized scores. These instruments have demonstrated content, convergent, and construct validity; internal consistency, test–retest reliability; and responsiveness to change (1). Subjects in prospective studies complete these scales at baseline and repeat them at follow-up. Analysis often focuses on statistical significance in change of scores. It is, however, also important to interpret score change in relation to clinical impact on subjects. To address whether statistically significant small changes are clinically relevant, the concept of minimum clinically important difference (MCID) has been developed and used in epilepsy QOL studies. Understanding these scales and how they are analyzed directly impacts relevance of reported findings.

Few prospective longitudinal controlled HRQOL studies have been published, yet none have a follow-up of more than 10 years. Edelvik and colleagues reported 14-year follow-up data from a Swedish cohort of patients who underwent presurgical evaluation and completed both the SF-36 questionnaire as well as the Hospital Anxiety and Depression scale (HAD) at baseline, 2 years, and 14 years after either resective surgery or presurgical evaluation (nonoperated patients).

Their study was limited by loss to follow-up. Significantly more nonoperated patients than operated patients did not complete the 2-year and long-term follow-up surveys. Reported and soon-to-be-discussed differences between operated and nonoperated groups may actually be greater than reflected in findings. It is important to address why 1) these patients were not referred for resective surgery and 2) why they had higher rates of loss to follow up, particularly in a country with national health care. Of the original 114 patients in the nonoperated group, 45 completed the 2-year follow-up and 24 completed the 14-year follow-up. Interestingly, patients who either declined participation at follow-up or could not be reached had significantly higher depression and anxiety scores and lower MHS scores at 2 years than did patients who had long-term follow-up. Perhaps this baseline psychiatric comorbidity influenced both surgical referral and follow-up. Understanding the effects of psychiatric comorbidity on both these aspects highlights the importance of diagnosing and treating these comorbidities.

Consistent with reported randomized epilepsy surgery trials (2, 3) significantly larger proportions of operated than nonoperated patients were seizure free at both 2 years and 14 years. All operated patients with seizures at 14 years had an improved seizure outcome after surgery, that is, reduced seizure frequency or severity. Among nonoperated patients who continued to have seizures at 2 years (87.5%), 67% continued to have seizures at 14 years. How did seizure outcomes translate to HRQOL scores? At 14 years, operated patients had better mean SF-36 domain scores when compared to a normative sample for Physical Functioning and Bodily Pain with resultant better PCS scores. In contrast, Social Functioning and Mental Health scores were worse, leading to significantly lower MCS scores. Operated patients who were seizure free at 2 years had higher MCS scores at 14 years when compared to non-seizure-free patients. Changing seizure status from 2 to 14 years did not, however, affect PCS or MCS means. These findings highlight the importance of early seizure control, as becoming seizure free soon after surgery positively impacted mental functioning. Nonoperated patients scored significantly lower than did the normative sample on both PCS and MCS. Differences between baseline and long-term follow-up reached MCID criteria finding improvement in more operated than nonoperated patients and worsening in more nonoperated patients than in operated patients. Use of the generic SF-36 inventory likely limited the findings because, as discussed, this is not an epilepsy specific questionnaire and not as sensitive to clinically important clinically meaningful changes in HRQOL among epilepsy patients (1). HAD scores did not differ between the groups. Results from a satisfaction survey found that operated patients had higher satisfaction scores and importantly very few operated patients found limited benefit of surgical resection.

A clear finding was decreased social functioning of the operated group when compared to a normative sample, which occurred despite these patients having reported high rates of seizure improvement and overall satisfaction with surgical resection. Social functioning defines individuals' interactions with their environment and the ability to fulfill their roles within such environments as work, social activities, and relationships with partners and family (4). Why did these subjects perform worse on social functioning? Before being referred to surgery, they had seizures for an average of over 18 years. The seizure burden had likely taken its toll, and it was difficult to overcome the long-term impact of seizures, particularly during formative years. In a separate and earlier study, Edelvik et al. reported that no significant changes in employment occurred after surgical resection (5). These findings, combined with the consistent benefit of surgical resection for referred patients, demand that we, as a community, do better and refer refractory patients earlier.

Some important limitations of Edelvik and colleague's current study highlight the difficulty of performing long-term follow-up studies. As discussed, there was significant loss to follow-up, particularly in the nonoperated group. Another interesting point is how patients are defined as either being “seizure free” or “having seizures.” This prospective study defined this based on seizure status in the year prior to inventory completion. Seizures often have a variable presentation with periods of exacerbations and remissions. Their impact, particularly on HRQOL, is ongoing; it is hard to separate out over 14 years how 1 year of either seizure freedom or occurrence is reflected in the HRQOL study outcomes. Nonetheless, the consistency of the findings when compared to the prior 2-year study (6) and other prospective HRQOL studies (1, 7) supports the positive impact of resective epilepsy surgery on long-term QOL outcomes.

Overall, despite limitations of this long-term prospective HRQOL epilepsy investigation by Edelvik and colleagues, the findings support that over a prolonged period, resective epilepsy operated patients had better seizure control than did nonoperated patients and significantly higher physical functioning—but lower mental health functioning—than a normative sample, whereas nonoperated patients had lower physical and mental health functioning than did a normative sample. The combination of persistent difficulties in social functioning after having seizures for many years, even though seizure frequency and severity improved in the operated group and there was a lack of change in scores between 2 and 14 years (despite change in seizure frequency) supports the need for earlier surgical intervention.