Spanish expert consensus statement on the use of inhaled levodopa for the treatment of fluctuations in Parkinson’s disease

Abstract

Introduction

The majority of Parkinson’s disease (PD) patients undergoing treatment with levodopa (LD) —the first-line therapy for motor manifestations in PD— experience OFF episodes, characterized by symptom exacerbation. Inhaled LD is an on-demand therapy aimed at managing OFF episodes.

Methods

A scientific committee comprising six PD experts developed statements regarding inhaled LD based on evidence and their clinical practice experience. Using a two-round Delphi study, these statements were evaluated on a Likert scale ranging from 1 (completely disagree) to 9 (completely agree) by a panel of Spanish neurologists specialising in the management of PD patients. Consensus was considered achieved if ≥ 66.67% of responses fell within the same tertile as the median response for each statement ([1–3] disagreement, [4–6] neutral, [7–9] agreement).

Results

Forty-four statements across six thematic areas were developed. A total of 48 experts were invited to participate in the Delphi study, with participation rates of 91.7% in the first round and 83.3% in the second. After two rounds, consensus was reached in agreement on 40 out of 44 statements (91%). There was a high degree of expert consensus on the utility of inhaled LD for the rapid and reliable on-demand treatment of both motor and non-motor fluctuations in PD patients.

Conclusions

Inhaled LD was deemed a useful, effective, and safe on-demand treatment for OFF episodes in PD.

Resumen

Introducción

La mayoría de los pacientes con EP en tratamiento con levodopa (LD) —el tratamiento de primera línea para manifestaciones motoras en enfermedad de Parkinson (EP)— experimenta episodios OFF, con exacerbación de síntomas. La LD inhalada es una terapia a demanda para manejar los episodios OFF.

Métodos

Un comité científico compuesto por seis expertos en EP desarrolló sentencias sobre LD inhalada, basadas en la evidencia y su experiencia en la práctica clínica. Con un estudio Delphi de dos rondas, las sentencias fueron evaluadas en una escala Likert de 1 (completamente en desacuerdo) a 9 (completamente de acuerdo) por un panel de neurólogos españoles expertos en el manejo de pacientes con EP. Se consideró que se alcanzó el consenso si ≥66,67% de las respuestas se situaban en el mismo tercil que la mediana de las respuestas para cada sentencia ([1–3] desacuerdo, [4–6] indeterminado, [7–9] acuerdo).

Resultados

Se desarrollaron 44 sentencias en seis áreas temáticas. Se invitó a participar en el estudio Delphi a 48 expertos, y la tasa de participación fue del 91.7% en la primera ronda y 83.3% en la segunda. Tras las dos rondas, se alcanzó consenso en el acuerdo en 40 de las 44 sentencias (91%). Hubo un alto grado de consenso entre expertos en la utilidad de la LD inhalada para tratar a demanda, de manera rápida y fiable, fluctuaciones motoras y no motoras en pacientes con EP.

Conclusiones

La LD inhalada fue considerada como un tratamiento a demanda útil, eficaz, y seguro para los episodios OFF en EP.

Introduction

Levodopa (LD) is the first-line treatment for motor manifestations of Parkinson’s disease (PD).¹ Over time, the majority of patients under treatment with LD and/or carbidopa (CD) (LD/CD) experience symptom fluctuations between doses, alternating between “on” periods in which symptoms are well controlled and “off” periods in which they are exacerbated.^2,3 “Off” episodes appear as a result of disease progression and the decreased duration of the clinical benefit of LD/CD treatment.^2,3

Approximately 42% of patients experience “off” episodes in the first 2.5 years of treatment; this rate increases to 70% at 5-10 years and 77% after 10 years.⁴ A study using real-life clinical data found that 44% of patients with PD under treatment with LD/CD experienced at least one hour of “off” time per day.⁵ Onset of “off” episodes is sudden and sometimes unpredictable, and may involve both motor and non-motor fluctuations.^2,6 These episodes have a significant effect on patient quality of life, with a negative impact that increases progressively in line with the increase in daily hours of “off” time.⁵

A diverse range of treatments and strategies are currently used to manage “off” periods, including changes in the dose and/or frequency of LD/CD administration, as well as combination treatment with dopamine agonists, monoaminoxidase B (MAO-B) inhibitors, catechol-O-methyltransferase (COMT) inhibitors, amantadine, and zonisamide, among others.^3,7 If fluctuations and/or disabling dyskinesia persist despite rational polytherapy combining different action mechanisms, advanced treatment options are available with demonstrated long- and short-term efficacy and safety; these include deep brain stimulation, apomorphine, or continuous subcutaneous foslevodopa-foscarbidopa infusion or continuous intestinal infusion of LD/CD gel or LD/CD/entacapone.^3,7

Despite optimisation of treatments, patients may continue presenting fluctuations or variations in the control of dopaminergic symptoms, whether motor or non-motor, whose impact may vary greatly between patients in accordance with their lifestyle.³ The management strategy for this type of “off” episodes, which have a considerable impact on daily functioning, consists in on-demand therapies that rapidly alleviate symptoms in the event of unpredictable “off” episodes, particularly when patients are performing activities outside their home.⁸ The available therapies are inhaled LD, subcutaneous apomorphine, and sublingual apomorphine.³ No published clinical practice guidelines prioritise the use of any one on-demand therapy over the others, and their respective benefits and limitations will determine their suitability for each individual patient.^8,9 Inhaled LD has been approved by the European Medicines Agency for the intermittent treatment of episodic motor fluctuations (“off” episodes) in patients with PD under treatment with LD and dopa-decarboxylase (DDC) inhibitors.¹⁰ Despite this, inhaled LD may also improve non-motor symptoms associated with motor fluctuations.

Inhaled LD has been commercially available for 2 years as an on-demand therapy for patients with PD and “off” episodes. The objective of this project was to draft a reference document for the use of inhaled LD in the clinical management of PD, according to the expert opinions of PD specialists in Spain, and to study in greater depth the patient profiles likely to benefit most from this measure.

Methods

The project was conducted in 2 phases. In the first, a scientific committee comprising 6 experts in PD developed a series of statements, based on published evidence and their clinical experience, with recommendations for the use of inhaled LD in patients with PD and “off” episodes. In the second phase, the statements were evaluated by a panel of Spanish neurologists with expertise in the management of PD, following the Delphi methodology.

The Delphi methodology is frequently used in healthcare studies to establish the consensus opinion of a group of experts on a subject through a structured, iterative, anonymised evaluation.^11,12 In this study, the Delphi consensus process was conducted in 2 rounds: one in June-July 2024 and one in September 2024. The statements for which consensus was not reached in the first round were re-evaluated in the second round. Statements were scored using a Likert-type scale from 1 (completely disagree) to 9 (completely agree).

The statements addressed 6 topics: 1) types of “off” episodes and fluctuating symptoms for which inhaled LD is indicated; 2) efficacy of inhaled LD and improvement in activities of daily living; 3) specific clinical profiles, safety, and management of secondary effects; 4) concomitant use with other treatments; 5) the role of inhaled LD in on-demand treatment; and 6) treatment adherence, dosing, and form of administration of inhaled LD.

Data analysis

Descriptive statistics (mean [standard deviation (SD)], median, and range) were used to describe demographic characteristics and the scores obtained for each statement in each round of the Delphi consensus process. The Kolmogorov-Smirnov test was used to analyse normality of data distribution. The internal consistency of the Delphi questionnaire was evaluated using the Cronbach α (reliability was classed as acceptable for α > 0.7; good for α > 0.8; and excellent for α > 0.9).¹³ The intraclass correlation coefficient (r_i) was used to evaluate agreement between members of the expert panel. The Spearman coefficient (r_s) was used to evaluate the correlation between rounds (correlation was classed as little or absent for r_s = 0-0.25; weak for r_s = 0.26-0.50; moderate to strong for r_s = 0.51-0.75; and strong to very strong for r_s = 0.76-1).¹⁴ The kappa index (κ) was used to evaluate qualitative agreement between surveys, with scores divided into tertiles (1-3, 4-6, and 7-9 points) (agreement was classed as poor for κ < 0; slight for κ = 0-0.20; fair for κ = 0.26-0.4; moderate for κ = 0.41-0.60; substantial for κ = 0.61-0.80; and almost perfect for κ = 0.81-1).¹⁵ Consensus was considered to have been established if ≥ 66.67% of responses for each statement were situated within the same tertile as the median. Consensus was classified as disagreement with a statement for scores within the first tertile (1-3), neither agreement nor disagreement for scores within the second tertile (4-6), and agreement for scores within the third tertile (7-9). The coefficient of variation (CV) was calculated for each round, as well as the relative change between rounds ([second round CV – first round CV]/first round CV). A relative change in CV of ≤ 10% between rounds was considered to indicate low variability, meaning that a third round was not necessary. Data analysis was conducted using the SPSS statistics software, version 28.0.

Results

Participants in the Delphi consensus process

Forty-eight experts were invited to participate in the Delphi consensus process; a total of 44 (91.7%) responses were received in the first round and 40 (83.3%) in the second. Respondents represented 13 of Spain’s 17 autonomous communities. The participants who completed both rounds of the Delphi process had a mean age of approximately 47 years; 63% were women, 50% had more than 16 years’ clinical experience since residency, 80% practised at public institutions, 63% worked at movement disorders units, and 98% had treated more than 10 patients with inhaled LD (Table 1). No significant differences in participants’ demographic characteristics were observed between the 2 rounds.

Table 1.

Demographic characteristics of the participants.

Characteristic	First round	Second round	P
	(n = 44)	(n = 40)
	n (%)	n (%)
Age in years, mean (SD)	47.0 (8.2)	46.8 (8.3)	.912
Women	28 (63.6)	25 (62.5)	.906
Years of experience since residency			.899
0-5	1 (2.3)	1 (2.5)
6-10	6 (13.6)	5 (12.5)
11-15	14 (31.8)	14 (35.0)
16-20	8 (18.2)	7 (17.5)
> 20	15 (34.1)	13 (32.5)
Work setting			.933
Public	35 (79.5)	32 (80.0)
Private	2 (4.5)	2 (5.0)
Both	7 (15.9)	6 (15.0)
Type of clinic			.762
Specialist clinic	13 (29.5)	13 (32.5)
MDU	29 (65.9)	25 (62.5)
Other	2 (4.5)	2 (5.0)
Patients treated with inhaled LD			.703
0-9	2 (4.5)	1 (2.5)
10-50	32 (72.7)	29 (72.5)
51-100	9 (20.5)	9 (22.5)
> 100	1 (2.3)	1 (2.5)

LD: levodopa; MDU: movement disorders unit; SD: standard deviation.

Results

Consensus was reached for 40 of the 44 statements (91%) (Table 2). The statements for which consensus was not reached after 2 rounds were the following: “Inhaled LD is useful for improving night-time ‘off’ episodes causing arousals and sleep fragmentation (S5)”; “Inhaled LD is useful for treating biphasic dyskinesia” (S6); “The on-demand use of inhaled LD for predictable wearing-off represents the first step in treatment adjustment” (S9); and “Cough is rarely a reason for patients to discontinue treatment with inhaled LD” (S43). It should be noted that education is essential to the correct use of inhaled LD; therefore, a patient education programme has been implemented to improve this area. In the second round, statements S5 and S43 were slightly modified to avoid ambiguity, whereas statements S6 and S9 were left unchanged.

Table 2.

Results of the Delphi consensus process.

		Round 1			Round 2
No.	Statement	Median range	Participants in agreement/disagreement range	Consensus	Median range	Participants in agreement/disagreement range	Consensus
	Section 1. Types of “off” episodes and fluctuating symptoms for which inhaled LD is indicated
S1	Inhaled LD is efficacious for treating “off” episodes in patients with wearing-off.	7-9	42 (95.5)/0	Consensus/agreement
S2	Inhaled LD is efficacious for treating “off” episodes in patients with more complex or unpredictable fluctuations than wearing-off.	7-9	40 (90.9)/1 (2.3)	Consensus/agreement
S3	Inhaled LD is efficacious for treating morning akinesia.	7-9	43 (97.7)/0	Consensus/agreement
S4	Inhaled LD enables better response to LD in patients with delayed-on for some LD doses.	7-9	37 (84.1)/1 (2.3)	Consensus/agreement
S5^a	Inhaled LD is useful for improving night-time “off” episodes causing arousals and sleep fragmentation, according to clinical experience. (Round 1 questionnaire).	7-9	25 (56.8)/3 (6.8)	No consensus/agreement	4-6	17 (42.5)/5 (12.5)	No consensus/unclear
	Inhaled LD is useful for improving night-time “off” episodes causing arousals and sleep fragmentation. (Round 2 questionnaire).
S6^a	Inhaled LD is useful for treating biphasic dyskinesia.	4-6	15 (34.1)/9 (20.5)	No consensus/unclear	4-6	14 (35.0)/7 (17.5)	No consensus/unclear
S7	Patients with non-motor fluctuations and responding well to LD benefit from inhaled LD.	7-9	41 (93.2)/0	Consensus/agreement
S8	Patients with episodes of “off” painful dystonia benefit from the use of inhaled LD.	7-9	40 (90.9)/1 (2.3)	Consensus/agreement
S9^a	The on-demand use of inhaled LD for predictable wearing-off represents the first step in treatment adjustment.	4-6	9 (20.5)/20 (45.5)	No consensus/unclear	1-3	13 (32.5)/20 (50.0)	No consensus/disagreement
S10	The on-demand use of inhaled LD represents a therapeutic tool for the treatment of predictable wearing-off that is not controlled regardless of the number of doses of oral medication.	7-9	38 (86.4)/0	Consensus/agreement
S11	The on-demand use of inhaled LD for predictable wearing-off should not delay the use of second-line therapies when symptoms are not controlled with the best optimised oral therapy.	7-9	36 (81.8)/1 (2.3)	Consensus/agreement
	Section 2. Efficacy of inhaled LD and improvement in activities of daily living
S12	The effect of inhaled LD causes a clinically relevant change in UPDRS-III score.	7-9	38 (86.4)/2 (4.5)	Consensus/agreement
S13	Inhaled LD can rapidly, significantly, and predictably improve the patient’s motor status.	7-9	40 (90.9)/0	Consensus/agreement
S14	The effect of inhaled LD on symptoms persists for at least 60 min after administration.	7-9	35 (79.5)/1 (2.3)	Consensus/agreement
S15	The efficacy of inhaled LD for treating “off” episodes persists in the long term (longer than 1 year).	7-9	37 (84.1)/0	Consensus/agreement
S16	On-demand treatment of unpredictable or sudden-onset “off” episodes with inhaled LD is beneficial for activities of daily living.	7-9	38 (86.4)/0	Consensus/agreement
S17	Using inhaled LD may be an effective strategy in advance of situations with greater work activity or social repercussions.	7-9	36 (81.8)/2 (4.5)	Consensus/agreement
S18	The use of inhaled LD in patients with infrequent but severe “off” episodes gives them confidence to attend family events at any time of day.	7-9	41 (93.2)/0	Consensus/agreement
S19	The use of inhaled LD in patients with prolonged morning akinesia enables them to comfortably perform their daily routines to fulfil their workplace duties.	7-9	38 (86.4)/0	Consensus/agreement
S20	The use of inhaled LD in patients with delayed “on” after eating enables them to continue attending their evening leisure activities.	7-9	36 (81.8)/1 (2.3)	Consensus/agreement
S21	Inhaled LD enables very socially active individuals not to miss their favourite activities.	7-9	39 (88.6)/0	Consensus/agreement
	Section 3. Specific clinical profiles, safety, and management of secondary effects
S22	Inhaled LD is indicated in elderly patients, with no need for dose adjustment.	7-9	35 (79.5)/1 (2.3)	Consensus/agreement
S23	Caution should be exercised when using inhaled LD in patients with chronic respiratory diseases, such as COPD or other chronic underlying lung diseases.	7-9	40 (90.9)/0	Consensus/agreement
S24	Inhaled LD is particularly attractive in patients with disorders of gastrointestinal motility or absorption.	7-9	42 (95.5)/0	Consensus/agreement
S25	Inhaled LD does not require dose adjustment according to sex.	7-9	40 (90.9)/0	Consensus/agreement
S26	Inhaled LD does not require dose adjustment in smokers.	7-9	42 (95.5)/0	Consensus/agreement
S27	Inhaled LD is safe at the pulmonary level in patients without relevant respiratory comorbidities.	7-9	41 (93.2)/1 (2.3)	Consensus/agreement
S28	In patients with PD and presenting cough after the use of inhaled LD, administration may be divided into several inhalations, rather than just one.	7-9	35 (79.5)/3 (6.8)	Consensus/agreement
S29	Inhaled LD may be used in patients with PD and presenting dyskinesia.	7-9	41 (93.2)/0	Consensus/agreement
S30	Inhaled LD is safe in patients with PD, with cough being the most frequent adverse reaction; however, it is mild and transient in the majority of cases, and may be caused by local irritation.	7-9	44 (100)/0	Consensus/agreement
	Section 4. Concomitant use with other treatments
S31	Inhaled LD is indicated in patients with Parkinson disease, provided that they are using LD/DDC inhibitors (carbidopa/benserazide).	7-9	40 (90.9)/3 (6.8)	Consensus/agreement
S32	Treatment with inhaled LD is appropriate in patients receiving such second-line treatments as LD and apomorphine infusion pumps (provided that they are under treatment with LD/DDC inhibitors) if they present “off” episodes that interfere in activities of daily living.	7-9	40 (90.9)/0	Consensus/agreement
S33	Treatment with inhaled LD is appropriate in patients with deep brain stimulation systems if they present “off” episodes that interfere in activities of daily living, provided that they are under treatment with LD/DDC inhibitors.	7-9	43 (97.7)/0	Consensus/agreement
S34	Inhaled LD presents the same drug interactions as oral LD; therefore, concomitant treatment with non-selective MAO inhibitors (phenelzine, tranylcypromine, moclobemide) is not recommended, and tricyclic antidepressants must be used with caution.	7-9	43 (97.7)/0	Consensus/agreement
	Section 5. The role of inhaled LD in on-demand treatment
S35	Inhaled LD is easier to administer than subcutaneous and sublingual apomorphine because there is no need for testing to identify the optimal dose (effectiveness/tolerability).	7-9	37 (84.1)/0	Consensus/agreement
S36	Inhaled LD has a better tolerability profile than subcutaneous and sublingual apomorphine, with less risk of secondary effects.	7-9	40 (90.9)/0	Consensus/agreement
S37	Insufficient evidence is available to establish which of the 3 on-demand treatments for “off” episodes (subcutaneous apomorphine, sublingual apomorphine, or inhaled LD) has the greatest efficacy.	7-9	42 (95.5)/0	Consensus/agreement
S38	An advantage of inhaled LD is that it does not require premedication with domperidone, unlike other on-demand dopaminergic therapies (subcutaneous or sublingual apomorphine).	7-9	42 (95.5)/1 (2.3)	Consensus/agreement
S39	The most appropriate on-demand therapy for “off” episodes should be selected on an individual basis, accounting for each patient’s symptoms and comorbidities, potential secondary effects, and the patient's preferences about the route and form of administration.	7-9	44 (100)/0	Consensus/agreement
	Section 6. Treatment adherence, dosing, and form of administration
S40	Recognition of “off” periods requires training and education of the neurologist/healthcare professional and the patient to identify motor and non-motor fluctuations at onset.	7-9	44 (100)/0	Consensus/agreement
S41	The indicated dose of inhaled LD is 2 capsules, administered in 2 consecutive inhalations at an interval no longer than 10 min.	7-9	44 (100)/0	Consensus/agreement
S42	LD should be inhaled when motor or non-motor symptoms of an “off” period begin to reappear.	7-9	42 (95.5)/1 (2.3)	Consensus/agreement
S43^a	Cough is rarely a reason for patients to discontinue treatment with inhaled LD. (Round 1 questionnaire).	4-6	18 (40.9)/12 (27.3)	No consensus/unclear	4-6	11 (27.5)/13 (32.5)	No consensus/unclear
	Cough is frequently a reason for patients to discontinue treatment with inhaled LD. (Round 2 questionnaire).
S44	Patient support programmes assist them in detecting and better recognising the disease (identifying fluctuations) and improving use of the inhaler.	7-9	43 (97.7)/0	Consensus/agreement

COPD: chronic obstructive pulmonary disease; DDC: dopa-decarboxylase; LD: levodopa; MAO: monoaminoxidase; UPDRS-III: Unified Parkinson's Disease Rating Scale, part III.

a

Consensus was not reached for these statements after 2 rounds.

The questionnaires used for the 2 rounds presented good internal consistency (Cronbach α: 0.880 [P <  .001] and 0.859 [P <  .001] in round 2) and high reliability (r_i: 0.842 [P <  .001] in round 1 and 0.822 [P <  .001] in round 2) (Table 3). Spearman correlation coefficients showed strong to very strong agreement between rounds for the total questionnaire (r_s = 0.957). The kappa index also showed a high degree of qualitative agreement. CVs for the first and second rounds were 0.1900 ± 0.0575 and 0.1933 ± 0.0520, respectively. Variability between rounds was low, with a relative increase in CV of 1.75%; therefore, a third round of the consensus process was ruled out.

Table 3.

Internal consistency of the questionnaire.

Section	Round 1		Round 2
	Cronbach α (P)	ri (P)	Cronbach α (P)	ri (P)
Total score (44 items)	0.880 (< .001)	0.842 (< .001)	0.859 (< .001)	0.822 (< .001)
Section 1 (11 items). Types of “off” episodes and fluctuating symptoms for which inhaled LD is indicated.	0.728 (< .001)	0.696 (< .001)	0.675 (< .001)	0.609 (< .001)
Section 2 (10 items). Efficacy of inhaled LD and improvement in activities of daily living.	0.790 (< .001)	0.755 (< .001)	0.789 (< .001)	0.754 (.001)
Section 3 (9 items). Specific clinical profiles, safety, and management of secondary effects.	0.498 (< .001)	0.499 (< .001)	0.597 (< .001)	0.541 (< .001)
Section 4 (4 items). Concomitant use with other treatments.	0.497 (< .001)	0.498 (< .002)	0.497 (< .001)	0.498 (.002)
Section 5 (5 items). The role of inhaled LD in on-demand treatment.	0.516 (< .001)	0.576 (< .001)	0.492 (.002)	0.462 (.002)
Section 6 (4 items). Treatment adherence, dosing, and form of administration.	0.248 (< .104)	0.162 (< .104)	0.257 (.105)	0.165 (.105)

LD: levodopa; r_i: intraclass correlation coefficient.

Discussion

In this study, a scientific committee developed 44 statements on the use of inhaled LD in patients with PD and “off” episodes, which were subsequently evaluated by an expert panel of 44 neurologists specialising in PD, following the Delphi methodology. The experts reached a high level of consensus (90.1% of statements) on the value of inhaled LD for rapid, reliable, on-demand treatment of motor and non-motor fluctuations in patients with PD. The majority of statements were scored within the third tertile (7-9 points) by over 80% of respondents, far exceeding the minimum of 66.7% required to establish consensus; this underscores the high level of agreement.

The panel failed to reach consensus for only 4 statements (S5, S6, S9, and S43). Regarding the use of inhaled LD to treat night-time “off” episodes (statement S5) and biphasic dyskinesia (S6), the lack of consensus is reasonable, and reflects the need for research to establish the use of the treatment in these situations, as no evidence is currently available. Currently, selection of an on-demand therapy depends on the preferences of the neurologist, as no published guidelines indicate specific uses for the different options available. Although the experts considered inhaled LD to be easier to administer (S35) and safer (S36) than the 2 apomorphine formulations, there was no consensus on statement S9 (the use of inhaled LD as the first step in treatment adjustment for predictable wearing-off). The lack of consensus on statements S6 and S9 persisted in the second round, with the same level of disagreement. The lack of consensus on statement S6 is due to the lack of data on whether inhaled LD might be efficacious in different manifestations of biphasic dyskinesia, due to the heterogeneity in the time of day at which this complication may appear, as well as its severity. The lack of consensus on statement S9 demonstrates the difficulty in PD management of establishing a treatment algorithm applicable in different healthcare settings. It should be noted that the panel agreed that insufficient evidence was available for selecting one treatment option over another (S37); given the different characteristics of these therapies, no single option can be identified as the treatment of choice for the first step of treatment adjustment.^8,9

In a recent Delphi study, an expert group agreed that certain clinical scenarios justify the use of on-demand therapies in PD; however, they did not distinguish between the 3 available options.¹⁶ Furthermore, that study did not address the issues for which our study did not establish consensus.¹⁶

The lack of consensus on statement S43 (cough as a reason for discontinuation of treatment with inhaled LD) may be due to ambiguous interpretation of the statement: though coughing is a frequent secondary effect of inhaled LD, there was no consensus on its role as a common reason for patients to discontinue treatment (S43). In clinical trials, the rate of treatment discontinuation due to cough was 15.6%.¹⁰ No specific guidelines are available on how inhaled LD should be used in patients experiencing cough; however, consensus was reached on the approach in clinical practice, with the dose being divided into several inhalations (S28).

For some statements, some respondents gave scores outside the tertile for which consensus was achieved, which may indicate a lack of knowledge about certain issues. For instance, the indications for inhaled LD include elderly patients, and dose does not need to be adjusted according to age or smoking¹²; however, some of the scores for statements S22 and S26 suggest that some experts were unaware of this. In other cases, it is possible that the statement was misinterpreted. For example, some scores for statement S31 (“inhaled LD is indicated in patients with PD provided that they are using LD/DDC inhibitors”) conflicted with the consensus in favour of the statement; this is striking, as the statement refers to the approved indication for inhaled LD.¹⁰

Regarding the type of “off” episodes and symptom fluctuations for which inhaled LD is indicated, we observed a high level of agreement with the statement that the treatment is efficacious for improving non-motor fluctuations (S7) and “off” painful dystonia (S8), 2 of the types of fluctuations that have been associated with the greatest impact on everyday functioning in patients with PD.¹⁷ When patients perform activities outside the home, non-motor fluctuations more frequently go unnoticed than motor fluctuations; furthermore, they have been associated with detection of earlier loss of sustained response to dopaminergic medication.¹⁸ Our expert panel did not reach consensus on whether inhaled LD should be considered the first step in the management of fluctuations in PD (S9). However, in patients with less predictable non-motor fluctuations or wearing-off episodes, which may appear in specific or demanding day-to-day situations, or on days when oral drugs were not optimally absorbed, inhaled LD is considered a good therapeutic tool, regardless of the number of doses of oral medication taken (S10); this is consistent with evidence from the literature.¹⁹

In accordance with the demonstrated efficacy of inhaled LD in randomised clinical trials,^20–23 a high level of consensus was reached for all statements related to the efficacy of the treatment and the improvement in activities of daily living (S12-S21); this suggests that experts considered inhaled LD to positively impact patient quality of life.

Similarly, and once more in accordance with the published evidence,^10,24–26 consensus was achieved for all statements related to specific patient profiles that may benefit from inhaled LD (elderly patients [S22], patients with chronic respiratory disease [S23], and patients with disorders of gastrointestinal motility or absorption [S24]) and to safety and management of secondary effects (S25-S30). Very high levels of consensus were reached for the majority of these statements.

In accordance with the available evidence and the summary of product characteristics for inhaled LD,^10,20,26 consensus was reached on all statements related to the use of concomitant treatments with inhaled LD. The treatments considered appropriate were LD/DDC inhibitors (S31) and such other antiparkinsonian drugs as MAO-B inhibitors, COMT inhibitors, dopamine agonists, amantadine, and LD and apomorphine infusion pumps (S32); the treatment was also considered appropriate in patients with deep brain stimulation systems and “off” episodes interfering in activities of daily living (provided that they were under treatment with LD/DDC inhibitors) (S33). As with oral LD, experts agreed that inhaled LD was not recommended in patients receiving non-selective MAO inhibitors (phenelzine, tranylcypromine, moclobemide), and that tricyclic antidepressants must be used with caution (S34). However, inhaled LD can be combined with MAO-B inhibitors (selegiline, rasagiline, safinamide).

Of the 3 on-demand treatments currently used for PD, experts considered inhaled LD to be the easiest to administer (S35), the safest and best tolerated (S36), and to present the advantage that antiemetic premedication with domperidone is not required (S38). However, it is important to distinguish between ease of administration and ease of implementation. Sublingual apomorphine is simpler to administer than inhaled LD, but its implementation is more complex on account of the need for testing at a healthcare centre to establish the optimal dose and safety. Further research is needed to provide evidence on when a given on-demand therapy should be selected over another, and the experts agreed that insufficient evidence was available to determine the most efficacious option for “off” episodes (S37).

It should be noted that, regarding the comparison against other rescue therapies, sublingual apomorphine was not commercially available when the consensus process was conducted; therefore, although the participants were aware of the published studies, they were unable to base their opinions on clinical experience with the treatment.

Experts agreed that healthcare professionals and patients require training and education to recognise “off” periods (S40). Similarly, we observed a very high level of agreement regarding the dosing of LD (2 inhalations) when the symptoms of an “off” period are detected (S41, S42).

The main strength of this study is its systematic approach, employing the Delphi methodology to establish consensus. The main limitation is that the data is based on expert opinions, which may not always be directly linked to the scientific evidence. However, respondents to the questionnaire were demonstrated experts in the management of patients with Parkinson’s disease and in on-demand therapies, an essential characteristic in consensus studies.

The Delphi consensus methodology presents inherent, significant limitations. Among these, we may note bias in the selection of experts, which may compromise the representativeness of the consensus; the possibility that consensus may have been influenced by peer pressure; the subjectivity of responses, as they are based as much on opinions as on empirical data; the lack of representativeness and critical evaluation of individual reasoning; and the difficulty of maintaining up-to-date information, as it is not possible to continually integrate new evidence.

Conclusions

Following the Delphi methodology, we established a high level of consensus among Spanish neurologists on the value of inhaled LD for rapid, reliable, on-demand treatment of motor and non-motor fluctuations in patients with PD. Inhaled LD represents an efficacious tool for the management of different types of “off” episodes, particularly in the case of predictable wearing-off, complex or unpredictable motor fluctuations, morning akinesia, and delayed-on with oral LD. However, its usefulness in treating night-time “off” episodes and biphasic dyskinesia is unclear.

Although inhaled LD is not considered the first line of treatment when patients develop fluctuations, it is known to be valuable in specific situations, such as unpredictable fluctuations or problems with the absorption of oral medication, and even in patients who have already received such advanced therapies as deep brain stimulation or continuous infusion of levodopa or apomorphine. With respect to its impact on daily life, inhaled LD enables patients to maintain their work, social, and family activities, and is particularly helpful for treating morning akinesia, postprandial delays in drug effect, and oscillations in the effect of dopaminergic drugs affecting essential activities of daily living.

Inhaled LD has also shown significant benefits in the treatment of non-motor fluctuations and “off” painful dystonia, which severely affect everyday functioning and may indicate early loss of sustained response to dopaminergic treatment.

Inhaled LD has a favourable safety profile, even in elderly patients or patients with dyskinesia, although caution should be exercised in patients with chronic respiratory diseases. Compared to other on-demand therapies, such as sublingual or subcutaneous apomorphine, respondents considered inhaled LD to be more tolerable and easier to implement, as it does not require premedication. Although participants agreed that inhaled LD was the easiest treatment to administer, this question should also be posed to patients to verify whether they also consider it to be the easiest to use. Mild, transient cough is the most frequent secondary effect.

Future studies are needed to evaluate all 3 on-demand therapies to compare their efficacy and clearly establish their usefulness in specific situations.

CRediT authorship contribution statement

JPM contributed to the design and development of the study, interpretation of the results, writing of the first draft, and subsequent revision of the manuscript. MBE, EFA, LLM, NLA, ASF, and DSG contributed to the design and development of the study, interpretation of the results, and revision of the manuscript.

Funding

This study is sponsored by Esteve.

All Delphi consensus statements were written by the scientific committee, and the analysis of the results, conclusions, and drafting of the article were performed by the authors named in the article.

Acknowledgements

The authors would like to thank the members of the expert panel for their participation in this Delphi study: Javier Abril, Raquel Alarcia, Fernando Alonso, Iciar Avilés-Olmos, Helena Bejr-Kasem, Teresa Buongiorno, Iria Cabo, Nuria Caballol, Teresa Cáceres, Mar Carmona, María Cerdán, Beatriz de la Casa, Tania Delgado, Julio Dotor, Francisco Escamilla, Raúl Espinosa, Javier Gutiérrez, Carolina Hernández, Lidia Hernández, Ion Infante, Silvia Jesús, Inés Legarda, Idaira Martín, Raúl Martínez, Marina Mata, Carlos Ordás, Rafael Pérez, María Carmen Riveira, Ana Rodríguez, Javier Ruiz, Juan José Soria, Isabel Sastre, Gabriel Salazar, Juan Salas, Almudena Sánchez, Pilar Sánchez, Pilar Sanz, Beatriz Tijero, Javier del Val, María Rosa Velicia, Lidia Vela, Dolores Vilas, and Rosa Yáñez. We are also grateful to Juan José de la Cruz Troca for performing the statistical analysis; and Angela Rynne Vidal, PhD, and Anna Nualart, PhD, of VML health, for their editorial services.