Neurologists' Attitudes Toward Use and Timing of Deep Brain Stimulation

Discussion

Our study examines the perspectives and attitudes of general neurologists and movement disorder specialists toward the use of DBS and its earlier use in the clinical course of PD. We found that both movement specialists and general neurologists agree on various aspects of DBS candidacy. However, important differences remain in terms of the need for demonstrating levodopa responsiveness as well as the management of medications and dyskinesia before DBS. In terms of timing for DBS, both groups agreed that DBS was more useful when used after the appearance of levodopa complications and that it offers better management than medication alone. The majority of respondents agreed that DBS should be used after a minimum of 3–4 years after diagnosis, suggesting that overall neurologists have a positive attitude toward the earlier use of DBS, particularly after considering an individual's risk/benefit analysis, specific symptoms, and quality of life. We recognize that our results may be influenced by having a sample population with a large majority of movement disorder neurologists.

We compare our results with those from a previous national survey on the attitudes of US neurologists toward DBS conducted by Shih et al. in 2011.¹⁴ This survey found differences between movement disorder specialists (n = 75) and general neurologists (n = 217) in a number of medical strategies used for the management of PD. Our results showed that currently there is more agreement on perspectives toward aspects of DBS candidacy that are well established in the literature. For example, both groups agreed that medical contraindications to brain surgery, an unclear diagnosis, and unclear expectations by patients were reasons to not refer patients to DBS.^{10,11,15,-,17} However, general neurologists and movement disorders specialists continue to show important differences in the recognition that a poor or absent response to levodopa is a poor prognostic consideration for DBS.¹⁴ With levodopa responsiveness considered one of the strongest predictive factors for positive outcomes in DBS,^{11,12,15,-,17} there remains an awareness gap between general neurologists and movement disorder specialists regarding considerations for DBS candidacy. Yet despite this distinction, compared with Shih and Tarsy's¹⁴ survey where 50% of general neurologists did not recognize that poor levodopa response is a poor prognostic indicator, greater than 50% of general neurologists in our survey did recognize this, suggesting improvement in the recognition of DBS prognostic factors over the past decade.

We also saw important differences between general neurologists and movement disorders specialists with regard to perceived risk of DBS. General neurologists considered DBS to have equal risk regardless of whether it is performed earlier or later in the disease progression, whereas specialists perceived greater risk of surgical complications when DBS is offered later. We did not query the age at which respondents discussed or referred patients with PD for DBS. Most respondents discussed DBS 1–6 years after diagnosis, and many (46%) considered a minimum of 2–4 years of disease duration to be appropriate to start considering DBS, but that could be at any age. In the literature, the effect of age on DBS outcomes is a contested topic, with some groups stating that age negatively affects outcomes and increases risk^11,16; others argue that it has no effect.^18,19 The general management of PD medication, dyskinesia, and wearing-off symptoms is another area where differences remain between the 2 groups. This is relevant in terms of considering the earlier use of DBS, as general neurologist might be less likely to refer individuals with PD for DBS earlier than movement disorder neurologists, thus requiring sufficient knowledge of the alternate medication management strategies that are available.

Other areas where we found a wide spectrum of responses, although we did not find relevant differences between clinician types, were connected to perspectives around cost-effectiveness of DBS compared with medication alone and the timing of DBS use in relation to medication options. Both of these perspectives are relevant considerations for the earlier use of DBS in PD. For example, although European cost analysis favors the use of DBS in advanced PD,^20,21 the only cost-effective assessment in the United States focused on early stage is based on preliminary data from a prospective, single-blind, controlled pilot trial.²² Moreover, as a recent systematic review reports: “The infrequent use of randomized, controlled trials to evaluate DBS efficacy, the paucity of data reporting the long-term effectiveness and/or utility of DBS, and the uncertainty surrounding cost data limit our ability to report cost-effectiveness summaries that are robust.” ²³

Similarly, in the past when DBS was considered more of a last resort, patients undergoing the procedure were more likely to be older than 60 years, thus eligible for Medicare. As DBS is considered earlier and earlier in the disease progression, patients might not be covered by Medicare or Medicaid. This raises concerns about social justice, as that would mean that only those patients who can cover the cost of earlier DBS out of pocket would have access to the benefits of the intervention, increasing already known health disparities.

In terms of attitudes toward the earlier use of DBS, although the largest proportion of participants responded that they considered 3–4 years of PD diagnosis to be the minimum amount of time before DBS consideration, many other participants stated that they did not think that there was a minimum duration of PD needed before first considering DBS; others favored a longer period of 5–7 years. Thus, it is clear that the decision on when it is appropriate to consider offering DBS to patients is an issue that remains contested and is likely determined on an individual basis taking into account the quality of life and risk/benefit analysis of each individual patient. The variability in neurologists' decisions regarding who to refer and when reflects the presence of a wide spectrum of attitudes from individual neurologists, rather than a collective consensus on when it is best to use DBS.

This lack of consensus on DBS candidacy and timing is reflected in the literature. For example, this is seen in critiques of the EARLYSTIM trial, which changed perspectives on treatment paradigms,⁵ and resulted in the FDA's updating 1 device manufacturer approval by lowering the minimum duration of PD diagnosis before DBS consideration to only 4 years.⁴ Although the EARLYSTIM trial was deemed revolutionary by many, other groups criticized that the demographics of the participants in the study made it impossible to extrapolate the results to a larger population and that the placebo effect of patients expecting benefit from DBS was enough to affect the results.^24,-,26 Regardless, the study met its primary end point of improved quality of life when DBS is implemented earlier in the course of motor complications, a factor that was cited by the majority of our survey respondents as an important factor determining the reason for DBS referral. Future work may be needed to define the components of quality of life that are most predictive of benefit from DBS, which may have direct implications on timing of DBS referral. For example, an analysis of EARLYSTIM results shows that changes in depression, disability, and pain influenced the health-related quality-of-life improvement in the DBS group.²⁷

There is an absence of official guidelines from national and international professional societies that provide a recommended framework in which to consider DBS timing. Current guidelines inform clinicians as to the presence or absence of PD symptoms that may benefit from surgery or pose additional risk of adverse outcomes from DBS. To add to the problem, when guidelines do exist, they are not updated regularly. For example, the last update to the AAN medical and surgical treatment for PD guidelines was in 2006.²⁸ Even clinical guidelines across the globe that were recently updated still fail to provide a practical framework as to when it is appropriate to start considering referral for DBS surgery.^15,29,-,32 Considering that several recent studies have brought new evidence to light regarding DBS outcomes in motor and not motor symptoms of PD,^5,33,34 it seems reasonable to reflect the current knowledge landscape and provide more informed recommendations to both neurologists and movement disorder neurologists.

The need for a consensus on parameters around timing of DBS is compounded by the rapid expansion of new therapies for PD. In the past few years, several new medications for PD have come to market, which may affect the decision making about when to recommend DBS because there are more medications that can be tried. Inhaled levodopa was approved by the FDA in December 2018 for treatment of off episodes in PD.³⁵ Istradefylline, a selective adenosine A2 antagonist, was approved in August of 2019 for reduction of off time.³⁶ Opicapone, a novel COMT inhibitor administered once daily, was approved in April of 2020.³⁷ Sublingual apomorphine was approved by the FDA as recently as May of 2020, as another rescue therapy for off time.³⁸

Increased surgical options are also under study. MRI-guided focused ultrasound thalamotomy for medically refractory PD tremor was FDA approved in 2018,³⁹ and preliminary positive results from a phase I study of noninvasive pallidotomy⁴⁰ spurred a large, multinational Phase II study of MRI-guided focused ultrasound pallidotomy for medically refractory motor fluctuations and/or dyskinesia. Gene therapy trials such as the AADC gene (Voyager therapeutics)⁴¹ and a study of GBA gene therapy are currently ongoing.^42,43 Last, cell-based therapies are again being researched with increased vigor as a potential treatment for motor fluctuations.⁴⁴ This changing landscape of PD therapies therefore raises a key question: Should these interventions be tried before recommending DBS (specifically the nonsurgical interventions)? The consequences of delaying DBS to try yet another medication are not well known, especially given that emerging medications could be seen as newer versions of past medications. Similarly, the long-term effects of earlier DBS are not fully understood, despite recent 5-year data from a single center showing promise in delaying disease progression and reducing polypharmacy,⁷ thus making the decision for earlier and earlier surgical intervention more challenging.

Addition of each novel therapy increases the complexity of treatment decisions. Although several tools were developed to help clinicians make more informed decisions about referring patients for DBS consultation, including the Stimulus patient-specific referral criteria,¹⁶ the FLASQ-PD questionnaire,⁴⁵ and the MANAGE PD tool,⁴⁶ they are not in widespread use and do not facilitate decision making about DBS timing. If a more standardized approach to DBS referral that balances the timing of the intervention with the need for individualized decision making can be defined through published guidelines, additional tools, or widespread use of existing tools, this would ensure that patients are being referred to the therapeutic modality that offers them the greatest clinical benefit at the optimal time. This consistency in the DBS referral process, together with registry information about DBS outcomes, would also enable better comparison of the effect of DBS across different populations. Further research on patients' preferences is also needed to better understand the factors shaping the decision to undergo or not to undergo DBS earlier than later in the disease progression.

Our study has a number of limitations. Although the response rate of the survey was 18%, the sample size was still small, with 164 participants. In addition, because of important differences in age category, with respondents to the survey tending toward younger populations, and differences in practice setting, generalizability of the results is limited. Furthermore, we recognize that the majority of our respondents self-identified as movement disorders neurologists (84%), which may have resulted in self-selection bias based on the type of participants who chose to respond to the survey. This number is also higher than the estimated proportion of movement disorder neurologists in this country, which could make our results difficult to extrapolate to the general population. Similarly, our survey demographic has significant differences in age category and practice setting compared with the active AAN US membership. However, AAN member profile data may differ from survey data due to changes in demographic factors, like practice setting, that members have not recently updated in their member profile. Finally, the importance placed on cost-effectiveness of DBS may infer a cost-related bias of our surveyed population. Our survey of AAN members represents largely US-based respondents; a complimentary survey of European neurologists could have provided a broader view of DBS perceptions, perhaps favoring earlier use of DBS and greater use of polypharmacy with differing concern for cost-effectiveness.

This survey sought to explore current perspectives and attitudes toward the use of DBS for PD, especially pertaining to earlier usage of DBS. Overall, the majority of respondents felt that DBS could be considered at a minimum of 3–4 years after diagnosis, in line with recent changes to FDA recommendations. However, the changing landscape of PD medication and device-assisted treatments and the emerging evidence base for the use of DBS earlier in the disease course both highlight the importance of developing professional guidelines that could help neurologists navigate when to discuss this intervention with their patients as the next viable option for treatment. Although we recognize that an individual risk/benefit analysis still supersedes any particular guideline, updating current guidelines to reflect the present landscape of medical and device-assisted treatments that includes recommendations about their relative timing may assist in providing more consistency in treatment pathways. This can help improve assessments of outcomes and help shape public policy to ensure that DBS is offered to patients in a responsible and timely manner.