Parkinson’s Disease and Gastrointestinal Non Motor Symptoms: Diagnostic and Therapeutic Options – A Practise Guide

Abstract

Abstract

Gastrointestinal (GI) disturbances in Parkinson’s disease (PD) are varied, involve the upper and lower GI tract and are evident in all stages of the disease. Recognition and re-evaluation of these non motor symptoms (NMS) due to the course of PD is important. They have a major impact on the efficacy of oral anti­parkinsonian medication and health related quality of life. Treatment needs to be tailored to the specific patient case with evaluation of PD stage, the specific GI NMS and comorbidities. This article provides an overview of the pharmacological and non-pharmacological therapeutic options for GI NMS in PD.

INTRODUCTION

Gastrointestinal (GI) disturbances are common non motor symptoms (NMS) in patients with Parkinson’s disease (PD). They are various, involve the whole GI tract and are evident in all stages of PD. Furthermore, constipation serves as a prodromal marker of the early non motor disease phase and is estimated as a premotor symptom like hyposmia [1]. GI NMS often get worse with advanced PD [2] and have a negative influence on reliable therapy with oral anti­parkinsonian medication. Malabsorption caused by delayed gastric emptying and/or prolonged transport times, leads to loss of medication effect, cumulates of doses or erratic absorption causing motor fluctuations [3, 4]. GI NMS combined with the negative impact on motor control results in major discomfort and a reduced health related quality of life (HRQoL) in patients with PD [5]. Furthermore, GI disturbances are a main cause of emergency admission of PD patients [6] and can lead to life-threatening complications such as aspiration in dysphagia, weight loss due to malabsorption and intestinal obstruction due to low transit in the intestinal tract with the risk of perforation. Regular assessment of GI NMS and a specific treatment are of major importance in the long term care of PD patients.

The evidence base for most therapeutic options is not yet substantial and lacks level 1 evidence as randomized, double-blind, placebo-controlled trials for treatment of GI NMS are missing. The American Academy of Neurology (AAN) and the Movement Disorders Society (MDS) have issued recent guidance on the evidence based management of NMS in PD including gastrointestinal symptoms [7, 8]. Pharmacological and non-pharmacological treatment strategies are summarised in Table 1.

METHODS

This paper is based on a broad literature review in pubmed for English language human studies under the following search terms: gastrointestinal non motor symptoms AND Parkinson’s disease; gastrointestinal NMS AND Parkinson’s disease; treatment AND non motor symptoms AND Parkinson’s disease; therapies AND non motor symptoms AND Parkinson’s disease. Articles presenting primary research results, review articles and case reports have been considered and if suitable referenced in this article.

GASTROINTESTINAL NON MOTOR SYMPTOMS – DIAGNOSTIC PROCEDURES AND THERAPEUTIC OPTIONS

For most of the diagnostic procedures and both the non-pharmacological and the pharmacological therapies randomized, double-blind, placebo-controlled trials are missing. Hence there is insufficient evidence to make adequate conclusions on efficacy but recommendations based on the published trials, reports, widespread clinical use and practise experience are reasonable. The practice implications for these treatments may be useful and worth trying on an individual base. As most of the studies are observational with a duration of less than six months, practise recommendations for the long-term management of the different GI NMS can not be drawn. There is a huge gap for research opportunities.

The most important step in the diagnosis of GI NMS is thorough history taking and neurological and physical examination. Stool diaries, structured interviews using questionnaires and patient self-completed NMS questionnaires may be helpful [9, 10] to provide a first assumption of the affected anatomical side, a PD related cause or possible medication side effects or comorbidities, an impression of the severity of the gastrointestinal NMS and the need for further diagnostic investigations.

We have chosen the most commonly reported GI NMS in PD to provide a more detailed presentation of possible therapeutic options. Diagnostic and therapeutic options for the different GI NMS in PD patients are summarised in Table 1.

Sialorrhea with drooling of saliva

Sialorrhea with chronic drooling is common in PD patients and can cause a range of physical and psychosocial complications, including perioral skin maceration and dermatitis, increased risk of aspiration-related lung infections and may be socially disabling due to stigmatization [11]. Management of sialorrhea is important and options include pharmacological and non-pharmacological methods for saliva control: practicing oral motor exercises, using chewing gum to encourage regular swallowing, using intraoral devices, biofeedback or rarely surgical interventions on the salivary glands [12].

There are four Class II studies for treatment of sialorrhea with drooling of saliva in PD patients [13–16] showing that botulinum toxin (BoNT) type A and B significantly reduce the amount of saliva production after injection of the parotid and/or submandibular glands. Ondo et al. [16] injected 1000 units of BoNT type B in each parotid gland and 250 units in each submandibular gland. The authors observed significant reductions in the Drooling Rating Scale, the Drooling Severity and Frequency Scale and the visual analogue scale assessing disability incurred by sialorrhea in the BoNT group compared to the placebo group. In comparison, Lagalla et al. [17] used much higher doses with 4000 units of BoNT type B in each parotid gland and received a significant reduction in the score on sialorrhea rating scales and in objective measurements for the BoNT treated group compared to the placebo group. Although the amount of BoNT units vary clearly in the different studies, BoNT is probably safe as side effects after BoNT type A and B injection were reported as mild. Side effects included dry mouth, mild transient swallowing difficulties and rarely, severe dysphagia, diarrhoea and neck pain [16–19]. Therefore, the AAN [7] and the MDS [8] recommend BoNT to be used in PD patients but with caution as dysphagia or worsening weakness can occur. Injection should be given by well-trained physicians. It is still in discussion whether ultrasound-guided injections are superior to blind injections [18, 20–22] and whether saliva reduction is superior improved by injection of the submandibular instead of the parotid glands or both [18, 23]. Therefore, larger placebo controlled studies are needed. Guidubaldi et al. [24] showed similar effectiveness and safety of BoNT type A and B injection in controlling sialorrhea in twelve PD patients.

Sublingual atropine drops as a simple and cost effective treatment for sialorrhea was investigated in six PD patients. This open-label pilot study showed statistically significant declines in saliva production, both objectively and subjectively [25]. Systemic adverse effects, including delirium and worsening of hallucinations, were described. Therefore sublingual atropine should only be tried in selected PD patients.

The AAN [7] and the MDS [8] recommend glycopyrrolate (an anticholinergic drug) as possibly useful for the short term treatment of sialorrhea in PD patients but state that there is insufficient evidence for treatment exceeding one week. Arbouw et al. [26] presented a four week, randomized, double-blind, placebo-controlled, crossover trial demonstrating that oral glycopyrrolate with a dose of 1 mg administered twice daily improved significantly the mean sialorrhea score in comparison to placebo (p = 0.011). Furthermore, this led to a clinically relevant improvement in 39% of PD patients of at least 30% in comparison to one patient (4.3% ) with placebo (p = 0.021). Glycopyrrolate was well tolerated with no significant differences in side effects to placebo treatment. Other studies [27, 28] using glycopyrrolate for the treatment of drooling of saliva in general medical conditions observed side effects that include dry mouth, difficult urinary retention, constipation, drowsiness and blurred vision. As it can not cross the blood-brain barrier due to its synthetic structure, centrally-mediated side effects, e.g. alteration of cognition, are not anticipated in comparison to other anticholinergic medication. Glycopyrrolate is available as oral tablet and as solution.

Sublingually ipratropium bromide spray compared to placebo was investigated in a five week, double-blind, crossover study for the treatment of sialorrhea in PD patients by Thomsen et al. [29]. The study showed no significant difference in weight of saliva measured at baseline or at the end of two weeks treatment with ipratropium bromide compared to placebo. Post hoc analysis did not reveal any significant change on UPDRS part II item 6 between ipratropium bromide and placebo, no differences in the number of doses administered by patients and no change in the subjective assessment of drooling severity after ipratropium bromide use. The MDS [8] concluded that there is insufficient evidence for the efficacy of ipratropium bromide spray as the sample size was low although the results appear to be negative. As ipratropium bromide spray is regularly used in pulmonary diseases there are no safety concerns of local application concerning general side effects of anticholinergic medication, e.g. cognitive, urinary, intraocular pressure changes [30]. Only dryness of the mouth can be expected inclinical use.

Lloret et al. [31] explored the safety and anti-sialorrhea efficacy of three different (0.3, 1, 3 mg) single doses of intra-oral slow dissolving thin films containing tropicamide versus placebo in 19 PD patients. Tropicamide 1 mg resulted in a significant decrease of patient’s subjective feelings of saliva levels. Objectively saliva volume was more reduced after tropicamide in comparison to placebo but significance was not reached. As no adverse events were detected a phase II study is currently underway (NCT01844648).

To our knowledge there have not been any recent studies on radiation and surgical management of drooling in PD patients. Radiation of the salivary glands has been investigated in patients with Motor Neuron Disease but studies did not reveal a statistically significant reduction of drooling afterradiation [32].

In conclusion, the symptom sialorrhea is best managed by a multidisciplinary team with a conservative approach initially and can progress to more invasive procedures if necessary. Pharmacological intervention can be effective but often the benefits are short lived because of side effects or high treatment costs. Main therapeutic option is the injection of botulinum toxin. More evidence is available but consensus guidelines defining which toxin type, which amount of mouse units and which glands to target are most effective, are lacking.

Surgery is usually a last resort in general management of drooling but to our knowledge there are no reports in PD patients.

Following the GI system in aboral direction, diagnostic procedures and treatment options for these NMS are summarised in Table 1. Dysphagia as a NMS originates of the pharynx, larynx or oesophagus whereas nausea and vomiting due to delayed gastric emptying and gastroparesis originate of the stomach. Constipation due to prolonged colonic transit time as a NMS of the lower GI tract is discussed in more detail.

Constipation

As one proposed mechanism is “slow-transit” constipation resulting in prolonged intestinal passage and reduced stool frequency [33, 34], treatment of constipation in PD patients should start with non-pharmacologic approaches like dietary advice and increase in daily activity levels. Parkinson medication should be adapted if necessary and medication known to exacerbate constipation should be discontinued if possible.

It was shown that increased fibre uptake leads to subjective and objective improvements in constipation related to PD [35, 36]. Ashraf et al. [35] studied the effect of psyllium (5.1 g twice a day) versus placebo on symptoms and colorectal function in 12 PD patients with constipation. PD patients treated with psyllium presented with increased stool frequency and weight but colonic transit or anorectal function was not altered. Psyllium as dietary fibre is not absorbed by the small intestine and absorbs excess water while stimulating normal bowel elimination.

In clinical practise osmotic laxatives and stool softener such as lactulose, polyethylene glycol, magnesium hydroxide, magnesium citrate, magnesium sulfate, sodium phosphate and docusate sodium are common in use and may be beneficial in conjunctionwith increased fibre and fluid intake. As these medication contain poorly absorbed ions or molecules that retain water in the intestinal lumen the effects are mainly mechanical and do not change the activity of the gastrointestinal and pelvic muscles. Therefore efficacy should be evaluated on a regular base as symptoms may actually worsen in outlet constipation and increase the problem in these patients over time [37, 38]. The AAN [7] and the MDS [8] recommend isosmotic macrogol (polyethylene glycol electrolyte solution) as likely efficacious for treatment of constipation in PD patients based on one Class II study which found a significant improvement of bowel movement frequency (p <  0.002) and stool consistency (p <  0.006) [39]. Macrogol is contained in products such as Movicol ^® or Miralax ^® and works on an osmotic basis. It can be used with an acceptable risk without specialized monitoring due to its low toxicity.

Doi et al. [40] investigated the effects of the extract of the dietary herb Rikkunshi-to (RKT) on gastroparesis in 20 PD patients, using objective parameters given by the (13)C-sodium acetate expiration breath test (gastric emptying study). A significant reduction of the peak time of the (13)C-dose-excess curve representing a significant shortening of the gastric emptying time was observed after administration of RKT. RKT was well tolerated and no patient experienced adverse effects, except for a bitter taste.

The use of prokinetic agents (5-HT agonists), which increase the acetylcholine release from the enteric cholinergic neurons and augment the peristaltic reflex to accelerate gastrointestinal transit, have been evaluated in small studies in PD patients. Possible adverse cardiovascular effects, especially under Tegaserod, have been described initially but have not emerged as a significant issue in further studies [41, 42].

Cisapride [43] significantly accelerated the colonic transit without overshoot effects, such as diarrhoea or worsening of PD motor symptoms. As cisapride also blocks potassium channels fatal cardiac arrhythmias have been observed.

Mosapride [44] led to subjective improvements in bowel frequency and difficult defecation; furthermore to objective improvements by significant shortening of colonic transit time and improvement of outlet constipation by augmentation of lower gastrointestinal tract motility. Serious adverse effects have not been reported and mosapride was well tolerated by all patients except for one with epigastric discomfort.

Tegaserod, a serotonin subtype 4 receptor partial agonist which is approved for the treatment of chronic idiopathic constipation, was investigated in a randomized, double-blind, placebo-controlled study of 15 PD patients [45]. Sullivan et al. reported a trend for decreased constipation by subjective improvement of satisfaction with bowel habits and GI symptoms in the tegaserod treatment group compared to the placebo group over a four week period. An observational study of five PD patients reported an improvement of both, bowel movement frequency and stool consistency in most of the patients. These effects lasted up to 1.5 years [46]. In both studies neither major side effects nor a worsening of PD symptoms have been described.

Lubiprostone was effective in a placebo controlled trial for the short-term treatment of constipation in PD patients as there was a high significant clinical global improvement of constipation by 64% of PD patients receiving the drug versus 18.5% PD patients receiving placebo [47]. Side effects were mild, most commonly intermittent loose stools.

Nizatidine, a selective histamine H2-receptor antagonist and a cholinomimetic, was found to shorten gastric emptying time significantly by reduction of the peak time of the (13) C-sodium acetate expiration breath test in 20 PD patients with constipation [48]. Nizatidine was well tolerated.

There is high quality of evidence that Linaclotide is effective and generally safe in chronic idiopathic constipation with comparable side effects of treatment and placebo group with diarrhea as the main adverse event [49, 50]. To our knowledge there are not yet any studies in PD patients.

A systematic review and meta-analysis of the efficacy of prebiotics, probiotics and synbiotics in chronic idiopathic constipation [51] revealed that probiotics appeared to have beneficial effects as a mean increase in the number of stools per week could be found based on two randomized controlled trials [52, 53]. Synbiotics also appeared beneficial. Ford et al. [51] concluded that there is insufficient evidence to recommend any of the three therapies in chronic idiopathic constipation as there were few open trials each with unclear risk of bias. Cassani et al. [54] studied the effect of regular intake of probiotics (Lactobacillus casei Shirota) on constipation in 40 PD patients for short-term and found a significant improvement of stool consistency and bowel habits (feeling of bloated, abdominal pain, sensation of incomplete emptying) after probiotic intake.

The injection of BoNT can be used for treatment of constipation in PD patients as it may be caused by slow transit, pelvic floor dysfunction or a combination of both. Pelvic floor dysfunction is characterized by functional obstruction at the pelvic outlet due to a failure of the puborectalis muscle to relax during defecation or a paradoxical puborectalis contraction as a kind of focal dystonia [55–58]. There are two studies, Albanese et al. [59] and Cadeddu et al. [60] in PD patients with outlet-obstruction constipation evaluating the effectiveness of BoNT injection. BoNT was injected under transrectal ultrasound guidance in the puborectalis muscle (100 MU in total in two sites on either side of the muscle). Both studies reported a decrease in anal tone during straining on assessment with anorectal manometry. Furthermore, defecography two months after BoNT injection showed significant improvement in anorectal angle during straining so that evacuation is possible. Neither studies reported systemic side effects nor permanent sphincteric damage as the resting pressure and maximum voluntary contraction were unchanged in comparison with the values before and after injection of BoNT. Both studies indicate that outlet obstruction is an important cause for constipation in some PD patients and that injection of BoNT provides an early symptomatic improvement for short-term use which is safe and simple. However placebo-controlled studies with long-term follow up are a need as the duration of efficacy of BoNT, the effect of repeated treatments and the optimal dose of BoNT remains unclear in the treatment of constipation in PD patients.

Non pharmacological concepts like biofeedback training can be a successful approach to pelvic floor dysfunction or dyssynergia in patients with constipation. Techniques vary and a skilled and experienced therapist is necessary but overall success rates of 67% can be revealed [61–63]. To our knowledge there are no studies investigating these methods in PDpatients.

Effects of oral Parkinson medication on gastrointestinal non motor symptoms

GI NMS in PD patients are associated with the disease itself but dopaminergic medication such as dopamine agonists and levodopa may impact both positively and negatively on these symptoms. Dopaminergic medication can worsen or even cause GI symptoms as side effects due to stimulation of receptors in the area postrema cause nausea and D2 receptor-related stimulation of intestinal dopaminergic neurons causes inhibition of intestinalmotility [64, 65].

Nausea and vomiting are common side effects of all dopaminergic medication. Therefore, a slow or stepwise increase of dopaminergic medication, a pre-treatment with domperidone, initial medication intake together with food (with the exception of levodopa) and informing of patients about these possible side effects may prevent medication induced nausea and increase patients’ compliance. Dopaminergic medication can worsen PD associated gastric emptying and gastroparesis due to an additional inhibitory effect on the upper GI motility contributing to nausea [65]. However dopaminergic medication improves gastrointestinal symptoms connected to motor off state such as sialorrhea, dysphagia, and constipation in PD patients [66, 67]. Woitalla et al. [68] hypothesized that transdermal drug administration may result in fewer GI symptoms. The authors investigated the effect of switching from oral anti-PD medication to rotigotine transdermal patch in 58 PD patients with GI symptoms in a prospective study. The intensity and the number of gastrointestinal complaints improved after switch to rotigotine patch after around six weeks. As this was a small open label study, a larger, randomized clinical trial of rotigotine transdermal patch versus oral applied dopamine agonists should be conducted to evaluate the transdermal effect of dopaminergic medication on GI NMS in PD patients. Furthermore, in the RECOVER study, there was a numerical, but not significant difference in favour of rotigotine for GI NMS in the Non Motor Symptom Scale suggesting that there was at least no apparent worsening of GI symptoms under the treatment with rotigotine [69].

There is data available that suggest that delayed gastric emptying can be caused or worsened by levodopa itself. In a study by Roberston et al. [70] eight healthy subjects underwent scintigraphic evaluation of gastric emptying with and without prior administration of levodopa. After administration of levodopa, gastric emptying was significantly prolonged. These results suggest a direct local effect on upper gastric motility by levodopa. It is unknown whether levodopa in PD patients causes a further delay in gastric emptying or whether it might correct delay to some extent by ensuring more coordinated gastric contractions [71]. Motility of the lower intestinal tract causing constipation does not seem to be worsened by levodopa [72, 73]. There is evidence that anorectal abnormalities change in parallel with the motor fluctuations related to on and off periods in PD patients [74].

The importance of the effectiveness of Parkinson medication was demonstrated as a greater delay in gastric emptying was found following a solid meal compared with a liquid meal in advanced, but not mild, PD patients [75]. Therefore, PD patients should avoid solid meals pre- and post-levodopa intake to maximize duodenal delivery of levodopa from the stomach. It has been shown that motor fluctuations in PD may be at least partly due to erratic delivery of levodopa from the stomach to the duodenal and jejunal absorption sites. This is supported by a correlation between plasma levels of dopamine with improved mobility following oral administration versus directly administration to the duodenum via a nasoduodenaltube [76, 77].

To improve levodopa absorption in the GI tract, eradication of Helicobacter pylori (H. pylori) should also be considered in PD patients as it has been shown that H. pylori binds levodopa. This leads to a decrease in levodopa absorption from the GI tract. In correspondence, eradication of H. pylori improved levodopa absorption and resultant clinical state [78–81].

Advanced therapies like apomorphine or duodopa pump are alternative options in PD patients with gastrointestinal dysfunction as both therapies are independent of swallowing, gastric motility and in case of apomorphine additionally of duodenal absorption. Therefore, medication effect is more independent of gastrointestinal dysfunction and more reliable.

Furthermore promising improvements of gastrointestinal NMS have been reported under both therapies [67, 83].

There are only few studies investigating the effect of subthalamic nucleus (STN) deep brain stimulation (DBS) on gastrointestinal symptoms in PD patients. It is especially difficult to distinguish between the direct effect of the stimulation and possible positive indirect effects on GI NMS, e.g. due to an improvement in mobility or reduction of dopaminergic medication with disappearance of side effects.

Improvements in dysphagia have been reported after STN DBS by Ciucci et al. [84]. They observed a significant improvement of the pharyngeal transit time under DBS stimulation compared to without DBS stimulation. The positive effect of the stimulation was explained by an overcome of the PD related bradykinesia and return of the pharyngeal stage motor pattern to performance levels approximating those of “normal” deglutition. This observation is consistant with the immediate positive effect of DBS stimulation on dysphagia described by Wolz et al. [85]. Improvement in bowel function after DBS has been reported [86, 87]. Halim et al. [88] observed an improvement in bowel function especially in early onset PD patients in comparison to late onset PD patients in a small case series and Zibetti et al. [89] reported improvement up to 24 months. Arai et al. [90] showed that gastric emptying was improved by STN DBS but not by treatment with levodopa using the (13)C-acetate breath test.

CONCLUSION

Gastrointestinal non motor symptoms in Parkinson’s disease are highly prevalent and involve all levels of the GI tract. GI NMS cause significant impairment both as symptoms and by disturbing of anti-parkinsonian medication effect with a negative influence on motor control of PD. These make it important to recognise GI NMS, to be aware of basic approaches to assess them and to offer specific treatment. To tailor specific need of treatment for the individual PD patient, clinicians need to be aware how PD medication can impact these symptoms and should be familiar with common treatment options. More research and clinical trials are needed to provide evidence based recommendations.

FINANCIAL DISCLOSURE / CONFLICT OF INTEREST

Lisa Klingelhoefer: No conflict of interest and no financial disclosures concerning this article.

Heinz Reichmann: No conflict of interest and no financial disclosures concerning this article.