Diagnosis and recovery patterns in patients with apraxia of speech after stroke

Abstract: Background: Stroke is a leading cause of adult disability. One of its most common consequences is a communication disorder. Beside aphasia (a language disorder), a motor speech disorder may occur, manifested as dysarthria or apraxia of speech (AOS). AOS has been defined as a motor speech disorder that affects an individual's ability to transform a linguistic message into speech motor plans. The most common symptoms associated with AOS include slow rate of speech, impaired articulation with sound errors that are predominately distortions, and disturbed prosody. The effects of AOS vary, from subtle articulation deviations to a complete inability to communicate through speech. AOS is most frequently caused by infarcts in the left middle cerebral artery, which supplies areas involved in both speech, language, and hand motor function. Despite a large amount of research, our knowledge of the exact nature and neurophysiological mechanisms of AOS is limited. Few studies have investigated AOS in an early phase after stroke as well as its resolution longitudinally, and factors predicting recovery are largely unknown. While the effects of focal brain lesions induced by stroke have been frequently studied, less is known about alterations in network connectivity in patients with AOS. Despite a close relationship between speech motor, language, and hand motor function, only few studies have addressed the relation between recovery in these multiple behavioral domains. In clinical settings, there is a lack of valid and reliable assessment instruments for AOS diagnosis that are applicable at all severity levels. Aim: The overall aim of this thesis was to gain more knowledge on the diagnosis and recovery patterns of AOS in individuals in an early phase after stroke. The specific aims for the four studies were: Study I: To study the intra- and interrater reliability of the Apraxia of Speech Rating Scale (ASRS) in assessment of individuals with speech and language impairments in an early phase after stroke. The ASRS was developed for research purposes, and its reliability for clinically active SLPs is unknown. An additional aim was to investigate the applicability of the ASRS in assessment of individuals with severe speech and language impairments. Study II: To describe and evaluate preliminary measures of reliability and validity of a clinical assessment protocol for AOS diagnosis, developed as part of a clinical study with the aims to be applicable in clinical settings and to be valid in the assessment of individuals with speech and language impairments at all severity levels. Study III: To investigate the prevalence of AOS and aphasia in individuals with a hand motor impairment in a subacute phase after stroke, and to compare recovery at six months in speech, language, and hand motor domains. An additional aim was to explore factors predicting recovery from AOS. Study IV: To investigate longitudinal changes in functional connectivity (FC) in speechlanguage networks in individuals with AOS after stroke, from the subacute to the chronic phase, specifically to identify predictors of AOS recovery. Additional aims were to study the relation between FC and degree of severity in AOS and to compare FC strength in patients with AOS after a left hemisphere stroke to that in left hemisphere lesioned stroke patients without speech-language impairment. Methods: For intra- and interrater reliability of the ASRS in study I, five certified speechlanguage pathologists (SLPs) from different hospital departments participated as raters. All worked with neurogenic communication disorders in adults. The ratings were based on video recordings of ten participants in an early phase after stroke showing varying degrees of AOS symptoms, from mild to severe. For measures of intrarater reliability, a rescoring was carried out after a minimum of three weeks. The clinical assessment protocol in study II included ten items, five of which were based on operationalized measures and five were perceptual ratings of AOS characteristics. Interrater reliability for the assessment protocol was based on video recordings of five individuals with varying degrees of AOS symptoms being assessed with the assessment protocol. Eleven certified SLPs participated as raters, all of whom worked with neurogenic communication disorders in adults. For measures of validity, the total scores of the assessment protocol from 39 participants in a subacute phase after stroke were compared against the clinical judgement of an AOS diagnosis. In study III, the prevalence of AOS and aphasia in individuals with a hand motor impairment in a subacute phase after stroke was investigated in a group of 70 participants. Half of the group had a left hemispheric lesion and the other half a right hemispheric lesion. Recovery of AOS, aphasia, and hand motor impairment at six months was investigated in 15 of these participants with a left hemispheric lesion. For measures of functional connectivity in study IV, resting state functional magnetic resonance imaging was applied. Assessments of speech and language impairment and FC in speech-language networks were obtained in nine participants with AOS and concomitant aphasia after a left hemispheric stroke and compared to six left hemispheric lesioned stroke participants without speech-language impairment. Measurements were performed at four weeks and six months after stroke. Functional connectivity was investigated in a network of key regions for speech production: inferior frontal gyrus (IFG), anterior insula (aINS) and ventral premotor cortex (vPMC), all bilaterally to investigate signs of adaptive or maladaptive changes in both hemispheres. Results: In study I, the intrarater reliability for the ASRS total score was moderate on average. The interrater reliability for the total score was poor. The item level values varied between moderate and poor, with lack of agreement on several items. High disagreement was especially noted in ratings of participants with severe speech-language impairments, but varying agreement were also found for participants with milder impairments. In addition, because some of the items on the ASRS require speech output consisting of multisyllabic words and phrases to target the diagnostic marker, limitations when assessing participants with signs of severe AOS with the ASRS were noted. In study II, the interrater reliability for the clinical AOS assessment protocol total score was good, but varied at an item level. The highest reliability was found for items with operationalized measures, while most items with perceptual ratings showed moderate agreement. A high index of validity was found when comparing the total score against the clinical judgement of an AOS diagnosis. In study III, 57% of the participants with a left hemispheric lesion had AOS, while 71% had aphasia. All participants with AOS also had aphasia. Recovery in AOS, aphasia and hand motor impairment at six months correlated positively across speech, language and hand motor domains. The strongest predictor for AOS recovery at six months was the initial aphasia test score. In study IV, recovery of AOS at six months correlated positively with the interhemispheric FC between left and right IFG in the subacute phase. Participants with AOS had a significantly reduced FC between bilateral vPMC in comparison to participants with a left hemispheric lesion without a language impairment, while severity of AOS at six months was related to the FC between bilateral aINS. Conclusion: The results of the two first studies add to the growing body of research that highlights the limitations of diagnosis of AOS solely based on perceptual characteristics, and call for the need to include objective measures in the diagnosis. In addition, if the same set of diagnostic AOS criteria cannot be applied during the course of the disease, it makes it difficult to study its longitudinal course and to identify predictors of recovery. In study III, a high prevalence of AOS with concomitant aphasia was noted in participants with a left hemisphere lesion and a hand motor impairment. Recovery of AOS, aphasia and hand motor followed a parallel trajectory, indicating that shared plasticity mechanisms are driving the recovery. For predictors of AOS recovery, indications that measures of aphasia at the subacute phase may be an important predictor was noticed. In study IV, the degree of AOS recovery at six months was strongly associated with the interhemispheric IFG connectivity strength at the subacute phase, indicating that increased activation in homologous speechlanguage areas in the right hemisphere in the subacute phase is positive for the recovery of AOS at six months. The reduced FC between the interhemispheric vPMC in participants with AOS is in line with earlier findings and confirms the current opinion about the left vPMC as a key region for speech motor programming.