doi:10.1093/brain/awu209
Brain 2014: 137; 2759–2772
| 2759
BRAIN
A JOURNAL OF NEUROLOGY
Stimulation of the pedunculopontine nucleus area
in Parkinson’s disease: effects on speech and
intelligibility
Serge Pinto,1 Murielle Ferraye,2,3,4 Robert Espesser,1 Valérie Fraix,3,5 Audrey Maillet,2,3,6
Jennifer Guirchoum,1 Deborah Layani-Zemour,1 Alain Ghio,1 Stéphan Chabardès,5 Pierre Pollak5,7
and Bettina Debû2,3
1
2
3
4
5
6
7
Aix-Marseille Université/CNRS, Laboratoire Parole et Langage (LPL), UMR 7309, 13100, Aix-en-Provence, France
Université Grenoble-Alpes, 38000, Grenoble, France
INSERM U836 – Grenoble Institut des Neurosciences, 38000, Grenoble, France
Donders Centre for Cognitive Neuroimaging, Nijmegen, The Netherlands
Centre Hospitalier Universitaire de Grenoble, 38000, Grenoble, France
Université Lyon I/CNRS, Centre de Neurosciences Cognitives, UMR 5229, Lyon, France
Hôpitaux Universitaires, Genève, Switzerland
Correspondence to: Serge Pinto,
Laboratoire Parole et Langage (LPL), Aix-Marseille Université/CNRS UMR 7309,
5 avenue Pasteur 13100 Aix-en-Provence Cedex, France
E-mail: [email protected]
Improvement of gait disorders following pedunculopontine nucleus area stimulation in patients with Parkinson’s disease has
previously been reported and led us to propose this surgical treatment to patients who progressively developed severe gait
disorders and freezing despite optimal dopaminergic drug treatment and subthalamic nucleus stimulation. The outcome of our
prospective study on the first six patients was somewhat mitigated, as freezing of gait and falls related to freezing were
improved by low frequency electrical stimulation of the pedunculopontine nucleus area in some, but not all, patients. Here,
we report the speech data prospectively collected in these patients with Parkinson’s disease. Indeed, because subthalamic
nucleus surgery may lead to speech impairment and a worsening of dysarthria in some patients with Parkinson’s disease, we
felt it was important to precisely examine any possible modulations of speech for a novel target for deep brain stimulation. Our
results suggested a trend towards speech degradation related to the pedunculopontine nucleus area surgery (off stimulation) for
aero-phonatory control (maximum phonation time), phono-articulatory coordination (oral diadochokinesis) and speech intelligibility. Possibly, the observed speech degradation may also be linked to the clinical characteristics of the group of patients. The
influence of pedunculopontine nucleus area stimulation per se was more complex, depending on the nature of the task: it had a
deleterious effect on maximum phonation time and oral diadochokinesis, and mixed effects on speech intelligibility. Whereas
levodopa intake and subthalamic nucleus stimulation alone had no and positive effects on speech dimensions, respectively, a
negative interaction between the two treatments was observed both before and after pedunculopontine nucleus area surgery.
This combination effect did not seem to be modulated by pedunculopontine nucleus area stimulation. Although limited in our
group of patients, speech impairment following pedunculopontine nucleus area stimulation is a possible outcome that should be
considered before undertaking such surgery. Deleterious effects could be dependent on electrode insertion in this brainstem
structure, more than on current spread to nearby structures involved in speech control. The effect of deep brain stimulation on
speech in patients with Parkinson’s disease remains a challenging and exploratory research area.
Received November 22, 2013. Revised June 17, 2014. Accepted June 26, 2014. Advance Access publication July 30, 2014
ß The Author (2014). Published by Oxford University Press on behalf of the Guarantors of Brain. All rights reserved.
For Permissions, please email: [email protected]
2760
| Brain 2014: 137; 2759–2772
S. Pinto et al.
Keywords: Parkinson’s disease; pedunculopontine nucleus area; subthalamic nucleus; L-DOPA; speech
Abbreviations: DBS = deep brain stimulation; L-DOPA = levodopa; MPT = maximum phonation time; PAG = periaqueductal grey
matter; PPNa = pedunculopontine nucleus area; STN = subthalamic nucleus
Introduction
Initial reports of dramatic improvement of gait disorders following
pedunculopontine nucleus area (PPNa) stimulation (Mazzone
et al., 2005; Plaha and Gill, 2005; Stefani et al., 2007) supported
the idea that the PPNa is involved in the control of locomotion in
humans as had been shown in animal models (Garcia-Rill et al.,
1987; Munro-Davies et al., 1999; Pahapill and Lozano, 2000;
Nandi et al., 2002; Takakusaki et al., 2003; Jenkinson et al.,
2009). The outcome of our prospective study of PPNa stimulation
in six patients with Parkinson’s disease and severe gait disorders
and freezing despite optimal dopaminergic drug treatment and
subthalamic nucleus (STN) stimulation (Ferraye et al., 2010) led
to more mitigated results. Freezing of gait and falls related to
freezing were improved by low frequency electrical stimulation
of the PPNa in some, but not all, patients. Other well-controlled
studies reached similar results, showing a reduction of falls in some
patients with Parkinson’s disease and gait disorders under unilateral PPNa stimulation (Moro et al., 2010). In these studies, no
serious adverse events were observed. However, side-effects
such as oscillopsia, paraesthesia and/or limb myoclonus, although
reversible, could hinder voltage increase (Ferraye et al., 2009,
2010). Other authors did not report such side effects, likely due
to differences in final stereotactic coordinates, leading to a different degree of modulation of structures close to the PPNa (Yelnik,
2007; Zrinzo et al., 2007, 2008; Mazzone et al., 2011, 2013).
Deep brain stimulation (DBS) of other targets, such as the STN,
has proved very efficient in alleviating the cardinal motor symptoms of Parkinson’s disease, but it can lead to stimulation-induced
speech impairment in some patients, sometimes exacerbating a
pre-existent dysarthria. Current diffusion has been proposed as a
possible explanation for such speech impairment side-effects: e.g.
current diffusion towards both the spinal (Pinto et al., 2005) and
the cerebello-thalamo-cortical (Astrom et al., 2010) tracts.
Considering that the PPNa, which belongs to the mesencephalic
reticular formation, lies in somewhat close proximity of passing
fibres and relay nuclei of the cranial nerves (such as nucleus ambiguus of the vagus nerve), as well as the superior cerebellar peduncles, one can hypothesize that current spread towards these
structures could affect speech in patients with PPNa stimulation.
For example, a PET experiment involving healthy volunteers has
shown the involvement of the periaqueductal grey matter (PAG)
in human vocalization, as the PAG is connected with cortical and
subcortical structures involved in voiced speech self-monitoring
(Schulz et al., 2005). These authors described functional connectivity between the PAG and both the premotor/motor and temporal/parietal cortices. They suggested that bilateral activations in
the posterior superior temporal gyri and supramarginal gyri may
support self-monitoring and feedback regulation of human phonation. A recent functional MRI study confirmed such links between
the PAG and structures involved in speech production both in
healthy subjects and patients with Parkinson’s disease (Rektorova
et al., 2012). In addition, strengthening of connectivity between
the PAG and the basal ganglia, posterior superior temporal gyrus,
supramarginal and fusiform gyri, and inferior parietal lobule was
observed in medicated patients with Parkinson’s disease as compared to controls, highlighting functional reorganization in brain
areas involved in feedback control of voiced speech (Rektorova
et al., 2012).
Little is known regarding the influence of PPNa stimulation on
speech production. An improvement in voluntary opening and
closing of the mouth under stimulation of the pedunculopontine
tegmental nucleus was reported (Mazzone et al., 2012), whereas
in a study of grammar processing, no effect of PPNa stimulation
on speech parameters (number of utterances, phonological construction) was found (Zanini et al., 2009). Although these two
reports provided some information regarding a possible role of
the stimulated target in motor or higher-order processes related
to speech, nearly all remains to be known regarding the influence
of PPNa stimulation on speech per se. In the present study, we
analysed speech data prospectively collected in seven patients with
Parkinson’s disease with STN stimulation who underwent PPNa
surgery because of severe freezing of gait. Indeed, STN DBS
may lead to speech impairment and a worsening of dysarthria in
some patients (Pinto et al., 2004, 2005; Tripoliti et al., 2008,
2011, 2014). Therefore, as the PPN is being investigated as a
novel DBS target in movement disorders, it seemed important to
precisely monitor speech outcome following this new surgery to
understand both the eligibility of patients for this surgical procedure and any potential adverse effects. Clinical and gait data of six
of these patients have been reported previously (Piallat et al.,
2009; Ferraye et al., 2010).
Materials and methods
Patients and study design
Seven patients with Parkinson’s disease participated in the study. All
patients presented with severe freezing of gait under chronic L-DOPA
and bilateral STN stimulation treatment. They were included because
gait disorders and freezing were their main complaints. They underwent bilateral electrode implantation within the PPNa according to the
surgical procedure previously published (Ferraye et al., 2010).
Exclusion criteria were surgical contraindications and cognitive impairment (score 5130 on the Mattis dementia rating scale). Patients’
characteristics are summarized in Table 1. Mean ages [ standard
deviation (SD)] at Parkinson’s disease diagnosis, STN surgery and
PPNa surgery were 41.7 9.7 years, 56.8 6.7 years and
63.0 5.8 years, respectively. The mean ( SD) disease duration
was 21.7 6.5 years. Before PPNa surgery, Unified Parkinson’s
Disease Rating Scale (UPDRS, Fahn et al., 1987) motor score improvement induced by L-DOPA when the STN stimulation was turned off,
Effects of PPNa stimulation on PD speech
Brain 2014: 137; 2759–2772
| 2761
Table 1 Demographic characteristics of the patients at the
time of inclusion
Patients
1
2
3
4
5
6
7
Gender
Age at PD diagnosis
Age at STN surgery
Age at PPNa surgery
Disease duration
Daily L-DOPA equivalent
dose (mg)
M
55
64
68
13
1025
F
50
64
68
18
550
M
49
65
72
23
800
M
44
53
57
13
1170
F
31
53
59
28
400
M
27
47
56
29
0*
M
36
52
61
25
880
*Patient 6 was not taking any L-DOPA chronically as medication did not lead to
any improvement beyond that provided by STN stimulation. Hence, he had
stopped taking L-DOPA several years before the PPNa surgery (Ferraye et al.,
2010). PD = Parkinson’s disease.
was 42 18% in average, ranging from 16 to 63% (Ferraye et al.,
2010).
The study was conducted at the Grenoble University Hospital in
accordance with the Declaration of Helsinki (WHO, 2008) and was
approved by the local ethics committee. Before participating, all patients provided written informed consent. Chronic PPNa stimulation
electrical parameters settings were: 1.2–3.8 V for voltage, 60–90 ms
for pulse width and 15–30 Hz for frequency; these stimulation parameters induced improvement in freezing of gait, albeit sometimes moderate, in six of the patients (Ferraye et al., 2010).
The clinical study lasted for 1 year. Speech recordings were carried
out along with the clinical and gait assessments, that is: (i) before
surgery (at baseline, assessment A0); (ii) during a double-blind trial
carried out between four (assessment A1) and 6 months after PPNa
surgery (assessment A3); and (iii) 12 months after PPNa surgery (assessment A12) (Ferraye et al., 2010). For the double-blind trial,
recordings of assessments A1 and A3 were performed after 1 month
of PPNa stimulation either off or on: the first PPNa condition (either
off or on) was set at the beginning of Month 4, and the speech
recording A1 was performed at the end of the month; PPNa stimulation was then turned off during Month 5, and the wash-out speech
recording (A2) was realized at the end of the month (PPNa stimulation
off); finally, the second PPNa stimulation condition (either off or on)
was set at the beginning of Month 6, and the A3 speech recording
was done at the end of the month (Fig. 1). The A12 assessment
allowed speech recordings under all therapeutic combination possibilities. In total, 18 speech recordings were obtained under eight possible
therapeutic conditions over the course of the study.
Speech tasks and temporal
aero-phonatory parameters
Speech was recorded within the Grenoble Neurology ward, in a quiet
non-soundproof room. Patients’ speech was recorded using a dedicated digital and portative device (Microtrack 24/92, M-AudioÕ AvidÕ ), connected with a head-mounted microphone (AKGÕ , model
K440). First, the patients were instructed to take a deep breath, and
then to pronounce and sustain the vowel /a/ comfortably for as long
and as steadily as they could; they were instructed to complete the
task three times and not to strain their voice. Second, an oral diadochokinesis task was performed: the patients had to repeat the pseudoword ‘pataka’ at a fast rate for 30 s. Finally, the patients were asked to
pronounce 10 words and 10 sentences provided by the French
Figure 1 Study protocol: preoperative and postoperative
treatment conditions across the 1-year follow-up of the study
(A1, A2 and A3 were conducted between 4 and 6 months after
surgery).
adaptation (Auzou et al., 1998) of the intelligibility section of the
Frenchay Dysarthria Assessment, version 1 (Enderby, 1980).
Each task corresponded to a specific speech file: four files were thus
acquired for each patient and for each of the therapeutic conditions
(vowel /a/, diadochokinesis, words and sentences). Before analysis,
these audio files were blindly preprocessed (labelling, segmentation)
under a dedicated software environment (Phonédit-Signaix; http://
www.lpl-aix.fr/lpldev/phonedit/). For the vowel /a/ task, an acoustic
analysis was automatically performed based on the calculation of the
maximum phonation time (MPT, in ms) within the signal window
comprised between two cursors delineating the beginning and the
end of the production. The diadochokinesis task allowed for the calculation of breath group durations (in ms), i.e. each period during
which the pseudoword was repeated during a single expiration. The
words and sentences were used in the assessment of speech
intelligibility.
Speech intelligibility
The first step of the assessment procedure was a manual labelling of
stimuli (the words and the sentences produced by the patients) within
the recorded audio files to allow for the automatic extraction of the
files, one file corresponding to one stimulus. Rather than using a single
listener, the words and sentences produced by the patients were evaluated by a group of listeners composing an auditory jury. The auditory
jury fulfilled the following criteria: (i) French-native speakers, without
any history of auditory and/or visual deficit; (ii) unfamiliar with speech
modifications of any neurodegenerative disease; (iii) naı̈ve with regard
to the aim of the experiment. Thirty-three listeners (29 females, four
males; mean age SD = 25.5 12.1 years; mean educational
status SD = 13.1 2.3 years after the fifth grade of French elementary school) were included and participated in the intelligibility assessment. The members of the jury were blind as to the conditions under
which the recording had been made.
Each stimulus (a word or a sentence) was listened to by three different members of the auditory jury. A computerized and random
presentation of auditory stimuli was used (http://www.lpl-aix.fr/
lpldev/perceval/). The stimuli were delivered via headphones. The
listeners, comfortably seated in front of a computer screen, were instructed to write what they had understood. When the stimulus was a
2762
| Brain 2014: 137; 2759–2772
single word, the participants were asked to write down the actual
word. The sentences were all constructed upon the same pattern,
i.e. [noun + verb], the noun being the same across all sentences (in
French ‘L’enfant . . .’, meaning ‘The child . . .’ in English); only the
action verb differed across sentences. Thus, in this sentence context
presentation, only the verb was to be transcribed by the listeners.
When the listeners did not understand the word or the sentence,
they were asked to proceed to the next stimulus by pressing the
‘return’ button of the keyboard. No second listening was possible.
Two sessions were performed under a test-retest protocol. The retest
occurred 1 week after the first session. For the retest, the set of stimuli
used for each listener was the same as for the first run, but the stimuli
were presented in a different random order. The purpose of this retest
session was to increase the reliability of the listeners’ responses. We expected a learning effect, that is, a global increase of correct responses for
the retest session as compared to the test whatever the treatment
condition, as listeners should become more familiar with the task. More
importantly, we expected that the effects induced by the treatment
conditions would remain the same across the two sessions.
S. Pinto et al.
stimulation were the two predictors, with two levels each (off, on).
Interaction between predictors was also considered.
For MPT, duration of the vowel /a/ was the dependent variable and
114 measures were available. There was no significant interaction between the two predictors, L-DOPA and PPNa stimulation treatments
(likelihood ratio test, 2 =1.88, df = 1, P = 0.17): the model without
this interaction was considered for analysis. For the diadochokinesis
task, the number of measures depended on the number of breath
groups produced by the patients: 324 observations were available for
the analysis. There was no significant interaction between the two predictors, L-DOPA and PPN stimulation treatments (likelihood ratio test,
2 = 0.52, df = 1, P = 0.47). The model without this interaction was considered for analysis. For the intelligibility assessment, the actual number
of available responses was 5054. The predictors were L-DOPA (OFF,
ON), PPNa stimulation (off, on), context presentation (single-word, sentence) and session (test, retest). All interactions removed with the backward elimination procedure were clearly non-significant (likelihood ratio
test, 2 5 2, P 4 0.15).
Statistical analyses
Pre-surgery versus 1 year follow-up
comparisons
All analyses were conducted using linear or logit mixed models to
account for the grouping structure (patients, listeners, stimuli) of the
data and for their nature (unbalanced, partially crossed/nested).
Package lme4 (Bates et al., 2013) was used in R software (R
Development Core Team, 2013). All non-significant interactions were
removed using a backward stepwise procedure and using likelihood
ratio tests to compare models. When possible, P-values were computed using a parametric bootstrap procedure and noted as Pboot;
the significance of statistical threshold was set at 0.05.
The dependent variables for the linear mixed models were the MPT
duration for the sustained vowel /a/ and the log-transformed breath
group durations for the diadochokinesis task. The log-transformation
of the breath group durations enhanced the normality of the distribution of the residuals of the model. Random slopes were added for the
within-patient variables (L-DOPA, PPNa and STN stimulations) to account for the variability across the seven patients. All these random
effects were allowed to be correlated. For the intelligibility analysis, a
mixed logit model was fitted with the binary response correct/incorrect
as the dependent variable. A supplementary random slope was added
for the within-listener variable (test session) to account for the variability across the 30 listeners. A random intercept was also added to
account for the variability across the stimuli. This model estimated the
correct answer probability, i.e. the intelligibility probability. The phonetic comparison between the reference orthographic transcription of
the word presented to the patients and the free orthographic transcription provided by the listeners allowed for the generation of a
variable reporting incorrect/correct transcriptions (or answers),
mapped to the unintelligibility/intelligibility of the stimulus. When no
response was given, we considered that the stimulus could not be
transcribed. Therefore, it was considered as unintelligible and the
no-response was coded as incorrect.
These statistical analyses were conducted by adjusting models that
used the preoperative (A0) and the 1-year follow-up (A12) postoperative data subsets, and the same dependent variables as for the doubleblind trial. The three predictors were L-DOPA (two levels: OFF, ON),
STN stimulation (two levels: off, on) and PPNa state (three levels:
preoperative, off, on). Interaction between predictors was also
considered.
For the MPT analysis, duration of the vowel /a/ was the dependent
variable and 225 observations were used. The interaction between the
three predictors (L-DOPA, STN and PPNa) was non-significant (likelihood ratio test, 2 = 0.31, df = 2, P = 0.85), as were the 2-way interactions between STN and PPNa (likelihood ratio test, 2 = 0.22, df = 2,
P = 0.89) and between L-DOPA and PPNa (likelihood ratio test,
2 = 1.8, df = 2, P = 0.4). The model without these interactions was
considered for analysis. For the diadochokinesis task analysis, graphical
inspection showed erratic and weak reliable realizations for breath
group ranks superior to eight (10% of the 522 observations).
These data were dropped from the analysis and a total of 470 observations were used. The non-significant interactions (all likelihood ratio
test P values 4 0.26), including the three-way (L-DOPA, STN and
PPNa) and the two-way (STN and PPNa) interactions between predictors, were removed from the final model. For the intelligibility assessment, 9776 productions were available for analysis. The predictors
were L-DOPA (OFF, ON), STN stimulation (off, on), PPNa state (preoperative, off, on), context presentation (single-word, sentence) and
session (test, retest). All non-significant interactions were removed
using a backward elimination procedure (all likelihood ratio test Pvalues 4 0.15). The model without these interactions was considered
for analysis.
Main effect of the pedunculopontine
nucleus area stimulation: outcome of
the double-blind trial
The PPNa main effect was determined by adjusting models that used
postoperative data subsets (A1, A2 and A3), OFF/ON L-DOPA, on
STN stimulation, off/on PPNa stimulation. L-DOPA and PPNa
Results
Double-blind trial: main effect of pedunculopontine nucleus area stimulation
The subset of data for analyses of the PPNa stimulation main effect
was provided by the postoperative assessments (A1, A2, A3).
Effects of PPNa stimulation on PD speech
Brain 2014: 137; 2759–2772
| 2763
Table 2 Stimulation parameter settings of the patients during the double-blind trial (postoperative assessments A1 or A3)
Patient
1
Contacts
Voltage (V)
Pulse width (ms)
Frequency (Hz)
0+1 2
1.8
60
25
2
4+5
1.4
60
25
0 1+
3.2
60
25
3
4 5+
3
60
25
0
2
60
20
4
4
1.6
60
20
STN stimulation remained on during these assessments. The stimulation parameter settings are displayed in Table 2. Bipolar stimulation
was used when the thresholds for side effects were too limiting under
monopolar stimulation.
Maximum phonation time
PPNa stimulation reduced the MPT by 1200 ms, without reaching
statistical significance (b = 1219, t = 1.65, Pboot = 0.09).
L-DOPA administration had a significant deleterious effect
(b = 3,164, t = 2.28, Pboot = 0.026), reducing the MPT by
3100 ms (Fig. 2).
Diadochokinesis breath group duration
PPNa stimulation (b = 0.24, t = 2.0, Pboot = 0.04) and L-DOPA
(b = 0.337, t = 3.3, Pboot = 0.001) both significantly decreased
breath group durations, up to 21% and 30%, respectively (Fig. 2).
We further examined the data by including the breath group rank
in the model, i.e. considering the number and the rank of the
breath groups needed to complete the 30-s repetition of the pseudoword ‘pataka’. The aim was to look for any relationship between the breath group rank and treatment effects (i.e. L-DOPA
and PPNa stimulation). This additional analysis only highlighted a
fatigue effect, constant across all treatment conditions, showing
that the higher the rank, the shorter the duration.
Speech intelligibility
As expected, the effect of the session factor was significant
(b = 0.49, z = 6.1, P 5 0.0001), indicating a better score of
correct responses for the retest session (Fig. 2). There were no
significant interactions between this factor and any other predictor, confirming the reliability of the listeners’ responses. As expected also, the probability of correct responses was globally lower
for the single-word than for the sentence context (b = 1.26,
z = 32.89, P = 0.0037).
PPNa stimulation alone marginally increased the probability of
correct responses (sentence context: b = 0.6, z = 1.86, P = 0.06;
single-word context: b = 0.567; the difference between the two
contexts being non-significant: b = 0.033, z = 0.12, P = 0.9).
L-DOPA alone did not induce any significant decrease of the probability of correct responses (decrease of about two percentage
points in sentence context: b = 0.257, z = 0.78, P = 0.43;
and 15 percentage points in single-word context: b = 0.577;
the difference between the two contexts was non-significant:
b = 0.327, z = 1.46, P = 0.14). The three-way interaction between
0 1+
3
60
20
5
5
2
60
20
0 2+
3.5
60
25
6
4 6+
2.5
60
25
0+1 2
1.8
60
25
7
4+5
1.4
60
25
0
1.6
60
25
4
1.6
60
25
L-DOPA,
PPNa stimulation and context presentation was significant (b = 1.254, z = 3.1, P = 0.0016). Examination of this
interaction revealed that there was no significant interaction between L-DOPA and PPNa stimulation in the sentence context
(b = 0.068, z = 0.23, P = 0.82), while the combined treatment
had a deleterious effect in the single-word context.
In summary, L-DOPA had a significant deleterious effect on
both MPT and diadochokinesis. PPNa stimulation also had somewhat of a deleterious effect, but significance was reached only for
diadochokinesis. For intelligibility, there was no evidence for any
L-DOPA and/or PPNa stimulation deleterious effects, except for
the combined ON L-DOPA/off PPNa stimulation condition in the
single-word context.
Pre-surgery versus 1-year follow-up
comparisons
For these analyses, preoperative (four conditions, combining OFF/
ON L-DOPA and off/on STN stimulation) and A12 postoperative
(eight conditions, combining OFF/ON L-DOPA, off/on STN stimulation and off/on PPNa stimulation) data subsets were considered.
Stimulation parameters at 1-year follow-up are shown in Table 3.
The stimulation parameters were stable for at least 1 month
before the A12 assessment. All patients had cyclic stimulation
with night arrest, except for Patient 6 (who forgot to switch his
stimulation off at night).
Regarding the clinical and perceptual assessment of speech
across the study, individual scores of the UPDRS item 18 are reported in Table 4. The patients in the present study presented with
variable effects of L-DOPA on speech before PPNa surgery. Based
on the global clinical evaluation scoring (item 18 of the UPDRS),
four patients presented with some improvement under L-DOPA
whereas two did not (one patient was not taking any dopaminergic treatment). Individual data of the speech parameters measured are also provided (Supplementary material). No global
impression emerges, as any speech change is highly dependent
on the task and treatment combination state.
Maximum phonation time
The model revealed that preoperative MPT did not differ significantly
from the postoperative PPNa off stimulation (b = 1,013,
t = 0.99, Pboot = 0.29), neither from the postoperative PPNa on
stimulation (b = 2,027, t = 1.47, Pboot = 0.096) (Fig. 3). There
was no significant effect of L-DOPA administration alone
(b = 422, t = 0.4, Pboot = 0.7). There was a significant beneficial
effect of STN stimulation alone, increasing MPT by up to 2600 ms
2764
| Brain 2014: 137; 2759–2772
S. Pinto et al.
Figure 2 Postoperative double-blind trial: effects of the PPNa stimulation (off, on) and L-DOPA (OFF, ON). Top left: MPT (in ms); top
right: breath group duration (in ms), measured during the diadochokinesis (DDK) task; bottom: speech intelligibility, as measured by the
correct answer probability (results of the test assessments; upper part under the sentence presentation, lower part under the single word
presentation). STN stimulation remained on during these assessments. Values correspond to either means or medians, as specified in the yaxis labels. For the diadochokinesis task, there were missing data (Patient 3 OFF L-DOPA, on PPNa stimulation; Patient 5, ON L-DOPA, on
PPNa stimulation). Bold black lines indicate measured values; red lines indicate values predicted by the statistical model (from the fixed
effects only); dotted lines indicate individual measured values (one line per patient).
(b = 2639, t = 2.2, Pboot = 0.02). This effect was lost under L-DOPA,
as indicated by the marginal interaction between L-DOPA and STN
stimulation (b = 1719, t = 1.87, Pboot = 0.06).
Diadochokinesis breath group duration
Preoperative breath group durations did not differ significantly
from both the postoperative PPNa off (b = 0.23, t = 2.1,
Pboot = 0.07) and on stimulation states (b = 0.23, t = 1.99,
Pboot = 0.09) (Fig. 3). There was no significant effect of L-DOPA
alone (b = 0.06, t = 0.66, Pboot = 0.5). There was a significant
beneficial effect of STN stimulation alone (b = 0.21, t = 2.58,
Pboot = 0.01), increasing breath group durations by 22%. This
beneficial effect disappeared under L-DOPA, as revealed by the
significant interaction between L-DOPA and STN stimulation
(b = 0.36, t = 3.26, Pboot = 0.001): breath group durations
Effects of PPNa stimulation on PD speech
Brain 2014: 137; 2759–2772
| 2765
Table 3 Stimulation parameter settings of the patients at 1-year follow-up (postoperative assessment A12)
Patients
1
2
Contacts
Voltage (V)
Pulse width (ms)
Frequency (Hz)
0+1 2
2.7
60
25
4 5 6+
3.3
60
25
0 1 2+
2.5
90
20
3
4 5 6+
2.5
90
20
4
0 1 2+
2.8
60
30
4 5 6+
2.4
60
30
0 1+
2.4
60
15
5
5
2.3
60
15
1 2+
2.5
60
20
6
5 6+
2
60
20
7
0 1 2+
3.8
60
15
4 5 6+
1.2
60
15
0
1.4
60
25
4
1.6
60
25
Table 4 UPDRS item 18 (speech) scores for the seven patients with Parkinson’s disease: pre- versus post-operative data
Patients
Assessment
Therapeutic condition
1
2
3
4
5
6
7
A0
OFF L-DOPA, off STN
ON L-DOPA, off STN
OFF L-DOPA, on STN
ON L-DOPA, on STN
OFF L-DOPA, off STN, off PPNa
OFF L-DOPA, off STN, on PPNa
OFF L-DOPA, on STN, off PPNa
OFF L-DOPA, on STN, on PPNa
ON L-DOPA, off STN, off PPNa
ON L-DOPA, off STN, on PPNa
ON L-DOPA, on STN, off PPNa
ON L-DOPA, on STN, on PPNa
4
2
3
4
3
3
3
3
3
3
3
3
2
2
2
3
3
3
3
3
3
3
3
3
3
3
3
3
4
3
2
3
3
2
1
1
1
2
2
1
1
1
1
1
1
2
1
1
1
2
2
1
1
1
1
1
1
2
NA
1
NA
3
1
1
2
NA
NA
NA
NA
3
3
3
3
2
4
3
3
4
3
3
3
A12
NA = Patient 6 was not taking any L-DOPA; - = missing data.
decreased by 10%, reaching values below those observed under
the ‘all treatments off’ condition. The additional analysis including
the breath group rank demonstrated again the fatigue effect already seen in the double-blind trial.
Speech intelligibility
This analysis needed the estimation of two variants of the model,
and consequently the significance threshold was Bonferroni-corrected at P 5 0.025 (Fig. 3). As expected, probability of correct
response was significantly smaller for the test than for the retest
session (b = 0.55, z = 10, P 5 0.0001). No interaction was
found between this factor and any other predictor, confirming
again the reliability of the listeners’ responses. As expected the
sentence context globally increased the probability of correct responses; an interaction between the stimuli context presentation
and PPNa state revealed different changes between preoperative
and postoperative states under the two contexts.
In the sentence context, both OFF and ON L-DOPA, the probability of correct responses was significantly greater preoperatively
than under both postoperative states [OFF L-DOPA: off PPNa
(b = 0.6, z = 2.26, P = 0.02); on PPNa (b = 1.36, z = 5.1,
P 5 0.0001); ON L-DOPA: off PPNa (b = 0.8, z = 3,
P = 0.025); on PPNa (b = 1.13, z = 4.25, P 5 0.0001)]. This
pattern was observed whatever the STN stimulation condition,
as the interaction between STN stimulation and PPNa state was
not significant (likelihood ratio test, 2 = 0.4, df = 2, P = 0.8). In
single word context, the probability of correct responses did not
differ significantly between the preoperative and postoperative
states whether OFF L-DOPA (on PPNa: b = 0.26, z = 1.1,
P = 0.28; on PPNa: b = 0.38, z = 1.5, P = 0.13) or ON
L-DOPA (off PPNa: b =
0.47, z = 1.86, P = 0.06; on PPNa:
b = 0.15, z = 0.6, P = 0.55). This was true under both STN
stimulation conditions, the interaction between STN stimulation
and PPNa state being non-significant (likelihood ratio test,
2 = 0.4, df = 2, P = 0.8).
No significant main effect of L-DOPA alone, i.e. off STN stimulation, was found. Conversely, there was a significant beneficial
effect of STN stimulation alone, i.e. OFF L-DOPA (b = 0.736,
z = 3.8, P = 0.0001). Examination of the significant interaction between L-DOPA and STN stimulation (b = 0.81, z = 6.97,
P 5 0.0001) revealed a deleterious effect of these combined treatments on speech intelligibility.
In summary, for the MPT and diadochokinesis tasks, analyses
did not reveal any significant change between preoperative and
postoperative PPNa stimulation states. For speech intelligibility, in
the sentence context only, the postoperative off and on
PPNa states were significantly degraded as compared with the
preoperative state. For all speech tasks, there was a positive
effect of STN stimulation that was inhibited by the administration
of L-DOPA.
Discussion
We examined the effects of PPNa stimulation, L-DOPA administration, and STN stimulation alone and combined on speech production, comparing PPNa preoperative and postoperative
2766
| Brain 2014: 137; 2759–2772
S. Pinto et al.
Figure 3 Pre- versus post-PPNa surgery: effects of the PPNa state (pre-op, off, on), L-DOPA (OFF, ON) and STN stimulation (off, on).
Top left: MPT (in ms); top right: breath group duration (in ms), measured during the diadochokinesis (DDK) task; bottom: speech
intelligibility, as measured by the correct answer probability (results of the test assessments). The displayed mean values are those
predicted by the statistical models (from the fixed effects only).
treatment states. The results of the postoperative double-blind
trial, conducted under STN stimulation, showed a significant deleterious effect of L-DOPA on temporal aero-phonatory speech variables (MPT and diadochokinesis). PPNa stimulation also had a
deleterious effect on these speech measures, but the effect was
only significant for diadochokinesis. Regarding intelligibility, there
was no evidence for any L-DOPA and/or PPNa stimulation deleterious effects (except for the combined ON L-DOPA/on PPNa
stimulation condition in the single-word context). The analysis of
the preoperative and 1-year postoperative assessments enabled
comparisons of all treatment conditions. Overall, for the MPT
and diadochokinesis tasks, we did not find any significant
change between the preoperative and postoperative PPNa stimulation states. Similarly, comparison of PPNa stimulation off versus
on conditions did not reveal any large influence on speech.
However, speech intelligibility in the sentence context was significantly impaired under the postoperative off and on PPNa states, as
compared to the preoperative state. For all speech tasks, there was
a positive effect of STN stimulation that was inhibited by L-DOPA:
whether before or after PPNa surgery, there was a significant
Effects of PPNa stimulation on PD speech
deleterious interaction between L-DOPA and STN stimulation so
that the beneficial effect of STN stimulation observed when the
patients were OFF medication was blocked by administration of
L-DOPA. Yet administration of L-DOPA had little effect when the
patients were off STN stimulation. The significance of these results
will be discussed in reference to the brainstem structures known to
be involved in vocal behaviour (observed in animal studies) and
speech motor control (demonstrated in humans), before considering antagonist interactions mechanisms between treatments as an
explanation for speech modulations.
We will first examine the apparent contradiction in the outcome
of the two analyses regarding the influence of PPNa stimulation
on speech. Indeed, while comparison of the PPNa stimulation off
and on data collected during the double-blind A1 to A3 assessments showed a deleterious effect on breath group duration and,
to a lesser extent, on MPT, no such degradation was found when
comparing the PPNa off and on conditions 1 year after surgery.
Nevertheless, although the effects did not reach statistical significance but for speech intelligibility, all speech variables were
degraded 1 year after surgery, as compared to the preoperative
state: indeed, the relative magnitude of the changes, i.e. the reduction of performance between the pre- and postoperative states
as expressed in per cent of baseline level, were noteworthy (10
to 20% of the initial MPT or diadochokinesis durations). The lack
of statistical significance of these deleterious effects on speech
aero-phonatory measures was likely due to the high variability
between patients as well as the small number of patients. Thus,
although the final outcome at 1 year resembles a lesion effect,
rather than a stimulation effect, this is likely not the whole story.
Indeed, our own as well as other published studies examining the
influence of PPNa stimulation on gait have reported long-lasting
effects (up to 2 weeks) after switching off the generator (Ferraye
et al., 2010; Moro et al., 2010; Ostrem et al., 2010). In the present study, the wash-out between the two stimulation conditions
at 1-year follow-up was 24 h, a retrospectively insufficient duration to wash-out the after-effect of 6 months of chronic stimulation. In contrast, during the double-blind study, patients were
evaluated at least 1 month after the stimulator arrest (2 months
for the A2 assessment for the patients whose first randomized
condition, A1, was off). Thus, only the data from the doubleblind study actually reflect a true PPNa stimulation effect, and
our data clearly demonstrate a deleterious effect of PPNa stimulation on the aero-phonatory variables. Yet, our results also
strongly suggest that PPNa surgery in itself may have somewhat
deleterious effects on speech production. This may not come as a
total surprise, provided the role of ponto-mesencephalic structures
in voice control and speech production.
Brainstem structures involved in
vocalization
Mainly, two cerebral pathways of sound generation in humans can
be pointed out (Ackermann, 2008): (i) one subserving non-verbal
vocalizations and involving projections from the anterior cingulate
gyrus and adjacent mesiofrontal areas via midbrain structures,
such as the PAG and adjacent tegmentum, to central pattern
Brain 2014: 137; 2759–2772
| 2767
generators and cranial nerve motor nuclei located within the
lower brainstem; (ii) and a second one allowing for proper
human speech production (i.e. verbal utterances), associated
with a more extensive brain network (cf. Price, 2012, for a
review). Thus, apart from the relay nuclei involved in the pyramidal and extra-pyramidal tracts, including the ponto-cerebellar
pathways, other midbrain regions are known to be involved in
vocal production.
As described in animal studies, the brainstem tegmentum, and
particularly the PAG, is an important convergence zone for orofacial motor neurons (Fardin et al., 1984; Zhang et al., 1995; Davis
et al., 1996; Jurgens, 2002). The function of the PAG is not fully
understood, as it is known to be involved in several quite distinct
functions, e.g. analgesia or defensive behaviour (Behbehani,
1995). Segregation of these functions within the PAG has been
suggested upon the recent understanding of its anatomical subdivisions (Linnman et al., 2012). Regarding vocal behaviour, it has
been proposed that
‘the PAG serves as a link between sensory and motivation-controlling
structures on the one hand, and the periambigual reticular formation
coordinating the activity of the different phonatory muscles on the
other’ (Jurgens, 1994).
Indeed, limbic structures such as the anterior cingulate, the thalamus and the hypothalamus project fibres into the PAG. Animal
studies also showed that enhancement of vocal production following injections of GABAergic and glutamatergic agonists was not
due to the activation of passing fibres but to the periaqueductal
neurons themselves (Lu and Jurgens, 1993; Siebert and Jurgens,
2003). Rather than being the site of vocal pattern generation, the
PAG could be serving gating functions (Siebert and Jurgens,
2003). In addition, the PAG controls the synchronous activity of
neurons of the lower brainstem that are known to control vocal
fold tension and respiration (Davis et al., 1996). This latter fact
argues for a possible dependence of our experimental tasks on
PAG activity: in the double-blind trial, which elicited the off
versus on PPNa stimulation comparison, a relative negative
impact of PPNa stimulation was observed for the two more
‘vocal’ tasks, namely the maximal phonation time and the diadochokinesis task. However, for the more ‘verbal tasks’, namely the
word and sentence production, this effect was not observed. For
this latter task, it was as if the motor cortical vocal control pathway bypassed the PAG (Jurgens, 2009).
Yet, vocal production cannot be restricted to the PAG and one
should also consider the dorsolateral tegmentum and the prelemniscal area located between the superior colliculus and the superior
olive. The ventrolateral parabrachial region is also involved in
speech motor control, as demonstrated in animal studies of respiration (Chamberlin and Saper, 1994). Finally, animal studies have
demonstrated an almost complete lack of direct connections between the various phonatory motor neurons along the neuraxis,
from the level of the pontine brainstem down to the lumbar spinal
cord (Jurgens, 2009). This suggested a role for the reticular formation of the lower brainstem in vocal motor coordination, as it is
extensively and concomitantly connected with all phonatory motor
nuclei. Thus, neurophysiological examination of brainstem
2768
| Brain 2014: 137; 2759–2772
involvement in vocal behaviour supports the idea that degeneration of brainstem structures may contribute to voice deficits in
Parkinson’s disease.
Should any effect on speech be expected
when stimulating midbrain structures?
Regarding the PPNa itself, low-frequency stimulation of the pedunculopontine tegmental nucleus significantly improved openingand-closing jaw movements in 14 patients with Parkinson’s disease
without medication (Mazzone et al., 2012). This finding underlined the role of the brainstem nuclei in oromotor control, which is
very relevant in speech motor control. In our own patients, the
stimulation sites were the dorsal PPN and the ventral part of the
cuneiform nucleus (Ferraye et al., 2010). This area is more medial,
superior, and posterior to the one targeted by Mazzone et al.
(2012). Our results at 1 year demonstrated a trend towards significance for the deterioration of all assessed speech dimensions,
whether the PPNa stimulation was turned off or on, when compared to the preoperative state. As stated previously, a long-lasting effect of the stimulation was a likely reason for the lack of
differences between the PPNa off and on states. Although we
cannot exclude a deleterious effect of the surgery per se on
speech, the data from the double-blind trial strongly suggest
that turning on PPNa stimulation, that is, delivering current, may
exacerbate speech changes.
Neuromodulation of midbrain regions in humans, namely the
PAG and periventricular grey matter, has been proposed since
the 1970s for the treatment of chronic neuropathic pain
(Hosobuchi et al., 1977; Richardson and Akil, 1977). This surgical
procedure induced long-term and effective pain alleviation in selected patients (Kumar et al., 1997), the outcome depending on
the nature and topography of the chronic pain (Pereira et al.,
2013). Few side effects or complications have been reported
(Bittar et al., 2005). Mainly, ventral PAG stimulation may induce
various percepts, ranging from pleasant sensations, warmth and
well-being at stimulation frequencies 550 Hz to perception of
diffuse burning or anxiety at higher frequencies; dorsal PAG stimulation typically induces unpleasant sensations of fear, doom, anxiety and agitation (Kaplitt et al., 2003). Higher stimulation settings
may induce gaze abnormalities via current spreading. In these
studies on PAG and/or periventricular grey matter stimulation
for intractable pain, none of the side-effects referred to any
speech/voice deficits (Kumar et al., 1997; Kaplitt et al., 2003;
Bittar et al., 2005; Pereira et al., 2013). Yet, electrical stimulation
of the PAG can also provoke laughing (Sem-Jacobsen and
Torkildsen, 1960) and its destruction, mutism (Esposito et al.,
1999). More recently, neuroimaging studies in humans have confirmed the connection between the PAG and cortical and subcortical structures involved either in speech production (Rektorova
et al., 2012) or vocalization (Schulz et al., 2005). It is usually
accepted that the maximal current spread diameter does not
exceed 2 mm at the voltages we used, questioning the hypothesis
that the current reached the PAG in our patients.
Hence, other midbrain structures being close to and located
posterior to the targeted structure (Alam et al., 2011) might
S. Pinto et al.
also be possible sites for current spread inducing speech modulation. The PPNa lies in somewhat close proximity with passing
fibres and relay nuclei of the vagus nerve, so that PPNa stimulation could affect speech via current diffusion towards the corticobulbar tract. Indeed, extracerebral vagus nerve stimulation
(Bergey, 2013) leads to side effects such as hoarseness and
cough, even if transient (The Vagus Nerve Stimulation Study
Group, 1995; Handforth et al., 1998) and reducing with time
(Morris and Mueller, 1999). However, no functional signs of
vagal stimulation were observed in our patients, suggesting that
such possibility would be unlikely. Stimulation of the superior cerebellar peduncles neighbourhood may also provoke speech deficits,
in case of current diffusion towards cortico-ponto-cerebellar fibres
(Astrom et al., 2010). Shall we consider this possibility? The question lies open. Finally, the fact that STN and PPN surgeries may
per se have exerted the deleterious effect on speech we report
here cannot be excluded. Indeed, the trajectory and final target in
the PPNa differs from those reported by other authors, who did
not report speech degradation following sole PPN stimulation at
therapeutic stimulating parameters. An interaction between STN
and PPNa stimulations remains possible.
A deleterious treatment interaction on
speech?
Pedunculopontine nucleus area and subthalamic nucleus
stimulation
Our results failed to identify any effect resulting from the combination of both PPNa and STN stimulations. This is in keeping
with previously published results. Indeed, improvements provided
by STN and PPN stimulations, alone or combined, failed to affect
speech motor dimensions explored in the sole study of grammar
processing available so far (Zanini et al., 2009), where only morphological and syntactic errors seemed to be positively modulated
by PPN and/or STN stimulations. In the Zanini et al. (2009) study,
it is not possible to know whether there was any lesion effect, as
no pre-surgery assessments were available. In addition, while in
the Italian study the patients received bilateral STN and PPN electrodes in a single surgical session, in the present study PPNa surgery occurred several years after STN surgery. Hence, disease had
further progressed between the two surgeries, likely contributing
to speech worsening. Such differences impede direct comparison
between the two studies. Possibly, the combination of PPN and
STN stimulations can lead to a relative further improvement as
compared with STN stimulation alone, despite the fact that
PPNa stimulation alone, contrary to STN stimulation alone,
hardly has any effect on the classical Parkinson’s disease motor
triad (Ferraye et al., 2011). Such a synergistic effect between the
two stimulations does not seem to hold regarding our speech
evaluation. Also, the effects induced by PPNa stimulation might
be ascribed to different levels of degeneration occurring in the
PPNa area, which may involve acetylcholine and/or non-acetylcholine neurons (Braak and Del Tredici, 2004a; Pienaar et al.,
2013) and fibres en passant rather than neuronal cell bodies
(Mazzone et al., 2013). These possibilities may further differentiate the mechanism of action of DBS in the STN, in which a
Effects of PPNa stimulation on PD speech
relatively homogeneous population of neurons is present, from the
mechanism underlying the DBS of PPNa, where the population of
neurons is heterogeneous, with a consistent loss of cells in
Parkinson’s disease (Braak et al., 2004b).
L-DOPA
and pedunculopontine nucleus area stimulation
It is important to note that disease evolution was quite advanced
in the group of patients included in this study (mean disease duration = 21.7 6.5 years). This must be kept in mind when trying
to interpret the treatment and interaction effects, as both the
progressive nature of Parkinson’s disease and possible additional
pathological processes, namely non-dopaminergic, may have
played a role in speech decline, which was already present
before PPNa surgery. One cannot exclude that speech impairment
could have developed as the disease progressed and L-DOPA
became less effective. We did not find any interaction between
dopaminergic therapy and PPNa stimulation (except for the sole
three-ways interaction between L-DOPA, PPNa stimulation and
context presentation in the double-blind trial). This is in-line with
another study that failed to observe either a positive or negative
interaction effect between dopaminergic therapy and PPN stimulation on oromotor jaw movements (Mazzone et al., 2012).
Oromotor movements have been studied after unilateral PPN
stimulation. The comparison of this study with our present work
and Zanini et al. (2009)’s study that also involved bilateral PPNa
stimulation, may be hampered by the different configuration of
current spreading, which was reasonably more confined under
unilateral than bilateral stimulation However, in the report by
Mazzone et al. (2012), patients were studied in the acute
phase, a few days after receiving a unilateral PPN electrode. In
addition, recordings were performed after a few minutes or hours
of stimulation only, whereas our patients were examined after
several months of chronic bilateral stimulation. It therefore
seems quite difficult to try and compare the results of the two
studies, as the methodology differs widely.
It is known that worsening of speech impairment with
Parkinson’s disease progression parallels an increasing severity of
non-dopaminergic cerebral lesions (Agid et al., 1990; Braak et al.,
1995; Halliday et al., 2011). Although some studies pointed out
beneficial effects of L-DOPA on Parkinson’s disease speech
(Sanabria et al., 2001; Goberman et al., 2002; De Letter et al.,
2007), others have reported a lack of significant change between
the two medication conditions (De Letter et al., 2003; Skodda
et al., 2011). The patients in the present study presented with
variable effects of L-DOPA on speech before PPNa surgery.
Based on the global clinical evaluation scoring (item 18 of the
UPDRS), three patients presented with mild improvement under
L-DOPA whereas three did not (one last patient was not taking
any dopaminergic treatment). Results of the double-blind trial
clearly demonstrated a deleterious effect of L-DOPA on speech
parameters, without any significant negative interaction with
PPNa stimulation.
L-DOPA
and subthalamic nucleus stimulation
When comparing preoperative and 12-month postoperative states,
whereas L-DOPA intake and STN stimulation alone had no and
positive effects on speech dimensions, respectively, a negative
Brain 2014: 137; 2759–2772
| 2769
interaction between these two treatments was observed both
before and after PPNa surgery. Except for the one patient who
did not take any L-DOPA, the remaining six patients had been
receiving dopaminergic therapy for 410 years. In addition, STN
stimulation was also a chronic treatment for at least 4 years at the
time of PPN surgery. No consensus is available regarding the long
term effects of the combination of dopaminergic therapy and STN
stimulation on speech impairment. It is generally agreed that STN
stimulation does not have any major effect on speech intelligibility.
Yet, degradation at 1-year follow-up was reported (Tripoliti et al.,
2011, 2014). In our study, we observed an improvement, rather
than degradation, of speech intelligibility under STN stimulation
alone. This effect existed before the PPNa surgery and did not
seem to be modulated by PPNa stimulation. Thus, our results
open the intriguing possibility that some observed decrease in
speech intelligibility could be brought about by a negative interaction between dopaminergic and STN stimulation treatments: this
antagonist combination suggests that the beneficial STN stimulation effect was blocked by the L-DOPA administration.
This effect could also reflect a specific pathological mechanism
associated with a negative interaction between L-DOPA intake and
STN stimulation. Such a mechanism would depend on crossmodulations between basal ganglia components during speech,
which are to date still not fully understood even under healthy
conditions. When trying to understand the mechanisms of DBS,
recent resting state functional MRI findings in Parkinson’s disease
have provided evidence that STN-DBS modulates all components
of the motor cortico-striato-thalamo-cortical loop, including the
direct and indirect basal ganglia pathways (Kahan et al., 2014):
specifically, the effective connectivity of the STN afferents and
efferents were reduced, whereas cortico-striatal, thalamo-cortical
and direct pathways were strengthened. Our results suggest that
the patients’ speech benefited from STN stimulation per se (Fig.
3). This effect could be related to such basal ganglia modulations.
However, decreased preoperative speech intelligibility ON medication and disease duration have been shown to rank among the
most significant predictors of deterioration of speech intelligibility
following STN surgery in the OFF medication/on STN stimulation
state (Tripoliti et al., 2014). These authors reported that unlike
motor outcome, the effect on speech was not related either to
the patients’ age at the time of surgery or to the preoperative
motor scores. Also, the only predictor of speech deterioration
ON medication/on STN stimulation was the preoperative OFF
medication speech score. Altogether, these findings would suggest
that the presence of significant non-dopaminergic lesions preoperatively should predict a poor speech outcome after the surgery, whether with or without L-DOPA. Considering the fact that
the patients of the present study had rather long disease duration,
one could imagine that the loss of the beneficial effect of STN
stimulation under L-DOPA could depend upon: (i) a strengthened
cortico-striatal pathway by STN stimulation; which is (ii) exacerbated by the administration of exogenous L-DOPA. At the level of
the striatum, both in its associative and sensorimotor parts, endogenous dopamine is released during speech (Simonyan et al.,
2013). If the loss of dopaminergic neurons contributed to the
speech deterioration over Parkinson’s disease progression, the inability of the patients to compensate for the combined effects of
2770
| Brain 2014: 137; 2759–2772
L-DOPA
and STN stimulation (Tripoliti et al., 2014) might result
from a difficulty in managing the effect of exogenous L-DOPA in
the context of a basal ganglia system modulated by STN stimulation. One possible explanation could then be the modification
induced by L-DOPA of the striatal neuron firing patterns ‘restored’
by STN stimulation.
Conclusion
Stimulation of the PPNa cannot be considered as an additional and
a late compensatory resource in the pathological process of
Parkinson’s disease. On the contrary, it could represent a selective
choice in relation to specific symptom patterns, which require a
careful selection of patients and a new conceptual interpretation
of the disease and surgery. Stimulation of the PPNa, and possibly
of surrounding structures as in the case of a spread of current to
adjacent pathways, may contribute to the induction of stimulation-related speech deterioration as observed in the present
study. However, little is known as regards the specific role of
subcortical and brainstem structures, such as the STN and PPNa,
on speech function. Thus, the effect of deep brain stimulation of
these structures on speech in patients with Parkinson’s disease
remains a challenging research area. Speech impairment after
PPNa stimulation is a possible outcome that should be taken
into account when considering such surgery, particularly as interactions between treatments rather than stimulation per se seem to
determine speech outcome.
Acknowledgements
The authors thank Mr Ludovic Jankowski and Mr Alain Purson for
their help for the speech data preprocessing.
Funding
The authors would like to thank their financial supports: Michael J
Fox Fondation (project number: R06098CC), Fondation de France
(project number: R07044CS), Medtronic (for the implantable material). This study also benefited from the support of the French government, through the French National Agency for Research (ANR)
which financed the ‘Brain and Language Research Institute’ (BLRI)
Labex framework (ANR-11-LABX-0036) and the ‘Investments of the
Future’ A*Midex project (ANR-11-IDEX-0001-02).
Supplementary material
Supplementary material is available at Brain online.
References
Ackermann H. Cerebellar contributions to speech production and speech
perception: psycholinguistic and neurobiological perspectives. Trends
Neurosci 2008; 31: 265–72.
S. Pinto et al.
Agid Y, Graybiel AM, Ruberg M, Hirsch E, Blin J, Dubois B, et al. The
efficacy of levodopa treatment declines in the course of Parkinson’s
disease: do nondopaminergic lesions play a role? Adv Neurol 1990; 53:
83–100.
Alam M, Schwabe K, Krauss JK. The pedunculopontine nucleus
area: critical evaluation of interspecies differences relevant for its
use as a target for deep brain stimulation. Brain 2011; 134 (Pt 1):
11–23.
Astrom M, Tripoliti E, Hariz MI, Zrinzo LU, Martinez-Torres I, Limousin P,
et al. Patient-specific model-based investigation of speech intelligibility
and movement during deep brain stimulation. Stereotact Funct
Neurosurg 2010; 88: 224–33.
Auzou P, Ozsancak C, Jan M, Leonardon S, Menard JF, Gaillard MJ,
et al. Clinical assessment of dysarthria: presentation and validation of
a method. Rev Neurol 1998; 154: 523–30.
Bates D, Maechler M, Bolker B. lme4: linear mixed-effects models using
S4 classes. . R package version 1.0-5. 2013. Available from: http://
CRAN.R-project.org/package=lme4.
Behbehani MM. Functional characteristics of the midbrain periaqueductal
gray. Prog Neurobiol 1995; 46: 575–605.
Bergey GK. Neurostimulation in the treatment of epilepsy. Exp Neurol
2013; 244: 87–95.
Bittar RG, Kar-Purkayastha I, Owen SL, Bear RE, Green A, Wang S, et al.
Deep brain stimulation for pain relief: a meta-analysis. J Clin Neurosci
2005; 12: 515–9.
Braak H, Del Tredici K. Poor and protracted myelination as a contributory
factor to neurodegenerative disorders. Neurobiol Aging 2004a; 25:
19–23.
Braak H, Braak E, Yilmazer D, Schultz C, de Vos RA, Jansen EN. Nigral
and extranigral pathology in Parkinson’s disease. J Neural Transm
Suppl 1995; 46: 15–31.
Braak H, Ghebremedhin E, Rub U, Bratzke H, Del Tredici K. Stages in the
development of Parkinson’s disease-related pathology. Cell Tissue Res
2004b; 318: 121–34.
Chamberlin NL, Saper CB. Topographic organization of respiratory responses to glutamate microstimulation of the parabrachial nucleus in
the rat. J Neurosci 1994; 14 (11 Pt 1): 6500–10.
Davis PJ, Zhang SP, Bandler R. Midbrain and medullary regulation of
respiration and vocalization. Prog Brain Res 1996; 107: 315–25.
De Letter M, Santens P, De Bodt M, Van Maele G, Van Borsel J, Boon P.
The effect of levodopa on respiration and word intelligibility in people
with advanced Parkinson’s disease. Clin Neurol Neurosurg 2007; 109:
495–500.
De Letter M, Santens P, Van Borsel J. The effects of levodopa on tongue
strength and endurance in patients with Parkinson’s disease. Acta
Neurol Belg 2003; 103: 35–8.
Enderby P. Frenchay dysarthria assessment. Br J Disord Commun 1980;
15: 165–73.
Esposito A, Demeurisse G, Alberti B, Fabbro F. Complete mutism after
midbrain periaqueductal gray lesion. Neuroreport 1999; 10: 681–5.
Fahn S, Elton RL. Members of the UPDRS Development Committee.
Unified Parkinson’s Disease Rating Scale. In: Fahn S, Marsden CD,
Calne DB, Goldstein M, editors. Recent developments in Parkinson’s
disease. Florham Park: MacMillan; 1987. p. 153–64.
Fardin V, Oliveras JL, Besson JM. A reinvestigation of the analgesic effects induced by stimulation of the periaqueductal gray matter in the
rat. I. The production of behavioral side effects together with analgesia. Brain Res 1984; 306: 105–23.
Ferraye MU, Debu B, Fraix V, Goetz L, Ardouin C, Yelnik J, et al. Effects
of pedunculopontine nucleus area stimulation on gait disorders in
Parkinson’s disease. Brain 2010; 133 (Pt 1): 205–14.
Ferraye MU, Debu B, Fraix V, Krack P, Charbardes S, Seigneuret E, et al.
Subthalamic nucleus versus pedunculopontine nucleus stimulation in
Parkinson disease: synergy or antagonism? J Neural Transm 2011;
118: 1469–75.
Ferraye MU, Gerardin P, Debû B, Chabardès S, Fraix V, Seigneuret E,
et al. Pedunculopontine nucleus stimulation induces monocular oscillopsia. J Neurol Neurosurg Psychiatry 2009; 80: 228–31.
Effects of PPNa stimulation on PD speech
Garcia-Rill E, Houser CR, Skinner RD, Smith W, Woodward DJ.
Locomotion-inducing sites in the vicinity of the pedunculopontine nucleus. Brain Res Bull 1987; 18: 731–8.
Goberman A, Coelho C, Robb M. Phonatory characteristics of parkinsonian speech before and after morning medication: the ON and OFF
states. J Commun Disord 2002; 35: 217–39.
Halliday G, Lees A, Stern M. Milestones in Parkinson’s disease–clinical
and pathologic features. Mov Disord 2011; 26: 1015–21.
Handforth A, DeGiorgio CM, Schachter SC, Uthman BM, Naritoku DK,
Tecoma ES, et al. Vagus nerve stimulation therapy for partial-onset
seizures: a randomized active-control trial. Neurology 1998; 51:
48–55.
Hosobuchi Y, Adams JE, Linchitz R. Pain relief by electrical stimulation of
the central gray matter in humans and its reversal by naloxone.
Science 1977; 197: 183–6.
Jenkinson N, Nandi D, Muthusamy K, Ray NJ, Gregory R, Stein JF, et al.
Anatomy, physiology, and pathophysiology of the pedunculopontine
nucleus. Mov Disord 2009; 24: 319–28.
Jurgens U. The role of the periaqueductal grey in vocal behaviour. Behav
Brain Res 1994; 62: 107–17.
Jurgens U. Neural pathways underlying vocal control. Neurosci Biobehav
Rev 2002; 26: 235–58.
Jurgens U. The neural control of vocalization in mammals: a review.
J Voice 2009; 23: 1–10.
Kahan J, Urner M, Moran R, Flandin G, Marreiros A, Mancini L, et al.
Resting state functional MRI in Parkinson’s disease: the impact of deep
brain stimulation on ‘effective’ connectivity. Brain 2014; 137 (Pt 4):
1130–44.
Kaplitt MG, Rezai AR, Lozano AM, Tasker R. Deep brain stimulation for
chronic pain. In: Winn HR, editor. Youmans neurological surgery.
Philadelphia, PA: Saunders; 2003. p. 3119–31.
Kumar K, Toth C, Nath RK. Deep brain stimulation for intractable pain: a
15-year experience. Neurosurgery 1997; 40: 736–46; discussion 46–7.
Linnman C, Moulton EA, Barmettler G, Becerra L, Borsook D.
Neuroimaging of the periaqueductal gray: state of the field.
Neuroimage 2012; 60: 505–22.
Lu CL, Jurgens U. Effects of chemical stimulation in the periaqueductal
gray on vocalization in the squirrel monkey. Brain Res Bull 1993; 32:
143–51.
Mazzone P, Lozano A, Stanzione P, Galati S, Scarnati E, Peppe A, et al.
Implantation of human pedunculopontine nucleus: a safe and clinically
relevant target in Parkinson’s disease. Neuroreport 2005; 16: 1877–81.
Mazzone P, Padua L, Falisi G, Insola A, Florio TM, Scarnati E. Unilateral
deep brain stimulation of the pedunculopontine tegmental nucleus
improves oromotor movements in Parkinson’s disease. Brain Stimul
2012; 5: 634–41.
Mazzone P, Sposato S, Insola A, Scarnati E. The deep brain stimulation of
the pedunculopontine tegmental nucleus: towards a new stereotactic
neurosurgery. J Neural Transm 2011; 118: 1431–51.
Mazzone P, Sposato S, Insola A, Scarnati E. The clinical effects of deep
brain stimulation of the pedunculopontine tegmental nucleus in movement disorders may not be related to the anatomical target, leads
location, and setup of electrical stimulation. Neurosurgery 2013; 73:
894–906.
Moro E, Hamani C, Poon YY, Al-Khairallah T, Dostrovsky JO,
Hutchison WD, et al. Unilateral pedunculopontine stimulation improves falls in Parkinson’s disease. Brain 2010; 133 (Pt 1): 215–24.
Morris GL III, Mueller WM. Long-term treatment with vagus nerve
stimulation in patients with refractory epilepsy. The Vagus Nerve
Stimulation Study Group E01-E05. Neurology 1999; 53: 1731–5.
Munro-Davies LE, Winter J, Aziz TZ, Stein JF. The role of the pedunculopontine region in basal-ganglia mechanisms of akinesia. Exp Brain
Res 1999; 129: 511–7.
Nandi D, Liu X, Winter JL, Aziz TZ, Stein JF. Deep brain stimulation of
the pedunculopontine region in the normal non-human primate. J Clin
Neurosci 2002; 9: 170–4.
Ostrem JL, Christine CW, Glass GA, Schrock LE, Starr PA.
Pedunculopontine nucleus deep brain stimulation in a patient with
Brain 2014: 137; 2759–2772
| 2771
primary progressive freezing gait disorder. Stereotact Funct
Neurosurg 2010; 88: 51–5.
Pahapill PA, Lozano AM. The pedunculopontine nucleus and Parkinson’s
disease. Brain 2000; 123 (Pt 9): 1767–83.
Pereira EA, Wang S, Peachey T, Lu G, Shlugman D, Stein JF, et al.
Elevated gamma band power in humans receiving naloxone suggests
dorsal periaqueductal and periventricular gray deep brain stimulation
produced analgesia is opioid mediated. Exp Neurol 2013; 239: 248–55.
Piallat B, Chabardes S, Torres N, Fraix V, Goetz L, Seigneuret E, et al.
Gait is associated with an increase in tonic firing of the sub-cuneiform
nucleus neurons. Neuroscience 2009; 158: 1201–5.
Pienaar IS, van de Berg W. A non-cholinergic neuronal loss in the pedunculopontine nucleus of toxin-evoked parkinsonian rats. Exp Neurol
2013; 248: 213–23.
Pinto S, Gentil M, Krack P, Sauleau P, Fraix V, Benabid AL, et al.
Changes induced by levodopa and subthalamic nucleus stimulation
on parkinsonian speech. Mov Disord 2005; 20: 1507–15.
Pinto S, Ozsancak C, Tripoliti E, Thobois S, Limousin-Dowsey P,
Auzou P. Treatments for dysarthria in Parkinson’s disease. Lancet
Neurol 2004; 3: 547–56.
Plaha P, Gill SS. Bilateral deep brain stimulation of the pedunculopontine
nucleus for Parkinson’s disease. Neuroreport 2005; 16: 1883–7.
Price CJ. A review and synthesis of the first 20 years of PET and fMRI
studies of heard speech, spoken language and reading. Neuroimage
2012; 62: 816–47.
R Development Core Team. R: A language and environment for statistical computing. 2013, version 3.0.2. Available from: http://www.Rproject.org/.
Rektorova I, Mikl M, Barrett J, Marecek R, Rektor I, Paus T. Functional
neuroanatomy of vocalization in patients with Parkinson’s disease.
J Neurol Sci 2012; 313: 7–12.
Richardson DE, Akil H. Pain reduction by electrical brain stimulation in
man. Part 1: acute administration in periaqueductal and periventricular
sites. J Neurosurg 1977; 47: 178–83.
Sanabria J, Ruiz PG, Gutierrez R, Marquez F, Escobar P, Gentil M, et al.
The effect of levodopa on vocal function in Parkinson’s disease. Clin
Neuropharmacol 2001; 24: 99–102.
Schulz GM, Varga M, Jeffires K, Ludlow CL, Braun AR. Functional neuroanatomy of human vocalization: an H215O PET study. Cereb Cortex
2005; 15: 1835–47.
Sem-Jacobsen CW, Torkildsen A. Depth recording and electrical stimulation in the human brain. In: Ramey ER, O’Doherty DS, editors.
Electrical studies on the unanesthetized brain. New-York: Haper and
Row (Hoeber); 1960. p. 275–90.
Siebert S, Jurgens U. Vocalization after periaqueductal grey inactivation
with the GABA agonist muscimol in the squirrel monkey. Neurosci Lett
2003; 340: 111–4.
Simonyan K, Herscovitch P, Horwitz B. Speech-induced striatal dopamine
release is left lateralized and coupled to functional striatal circuits in
healthy humans: a combined PET, fMRI and DTI study. Neuroimage
2013; 70: 21–32.
Skodda S, Flasskamp A, Schlegel U. Instability of syllable repetition in
Parkinson’s disease-influence of levodopa and deep brain stimulation.
Mov Disord 2011; 26: 728–30.
Stefani A, Lozano AM, Peppe A, Stanzione P, Galati S, Tropepi D, et al.
Bilateral deep brain stimulation of the pedunculopontine and subthalamic nuclei in severe Parkinson’s disease. Brain 2007; 130 (Pt 6):
1596–607.
Takakusaki K, Habaguchi T, Ohtinata-Sugimoto J, Saitoh K, Sakamoto T.
Basal ganglia efferents to the brainstem centers controlling postural
muscle tone and locomotion: a new concept for understanding
motor disorders in basal ganglia dysfunction. Neuroscience 2003;
119: 293–308.
The Vagus Nerve Stimulation Study Group. A randomized controlled trial
of chronic vagus nerve stimulation for treatment of medically intractable seizures. Neurology 1995; 45: 224–30.
Tripoliti E, Limousin P, Foltynie T, Candelario J, Aviles-Olmos I, Hariz MI,
et al. Predictive factors of speech intelligibility following subthalamic
2772
| Brain 2014: 137; 2759–2772
nucleus stimulation in consecutive patients with Parkinson’s disease.
Mov Disord 2014; 29: 532–8.
Tripoliti E, Zrinzo L, Martinez-Torres I, Frost E, Pinto S, Foltynie T, et al.
Effects of subthalamic stimulation on speech of consecutive patients
with Parkinson disease. Neurology 2011; 76: 80–6.
Tripoliti E, Zrinzo L, Martinez-Torres I, Tisch S, Frost E, Borrell E, et al.
Effects of contact location and voltage amplitude on speech and
movement in bilateral subthalamic nucleus deep brain stimulation.
Mov Disord 2008; 23: 2377–83.
WHO. 59th General Assembly. In: Association WM, editor. Seoul, Korea:
WHO; 2008.
Yelnik J. PPN or PPD, what is the target for deep brain stimulation in
Parkinson’s disease? Brain 2007; 130 (Pt 9): e79.
S. Pinto et al.
Zanini S, Moschella V, Stefani A, Peppe A, Pierantozzi M, Galati S, et al.
Grammar improvement following deep brain stimulation of the subthalamic and the pedunculopontine nuclei in advanced Parkinson’s
disease: a pilot study. Parkinsonism Relat Disord 2009; 15: 606–9.
Zhang SP, Bandler R, Davis PJ. Brain stem integration of vocalization:
role of the nucleus retroambigualis. J Neurophysiol 1995; 74: 2500–12.
Zrinzo L, Zrinzo LV, Hariz M. The pedunculopontine and peripeduncular
nuclei: a tale of two structures. Brain 2007; 130 (Pt 6): e73.
Zrinzo L, Zrinzo LV, Tisch S, Limousin PD, Yousry TA, Afshar F, et al.
Stereotactic localization of the human pedunculopontine nucleus:
atlas-based coordinates and validation of a magnetic resonance
imaging protocol for direct localization. Brain 2008; 131 (Pt 6):
1588–98.