The Treatment of Posterior Drooling by Botulinum Toxin in a Child with Cerebral Palsy

INTRODUCTION Drooling is the unintentional loss of saliva from the mouth, known as anterior drooling (1). However, anterior drooling has to be distinguished from posterior drooling which refers to saliva that is spilled over the tongue through the faucial isthmus. Under physiological conditions this initiates the pharyngeal phase of swallowing, during which the larynx closes to protect the airways followed by relaxation of the upper esophageal sphincter. Whenever the trigger to swallow is impaired or missing, pooled saliva may lead to posterior drooling mostly apparent from an alarming congested breathing, coughing, gagging, vomiting, and at times aspiration into the trachea. Unrecognized and silent pneumonia can occur (2). The risk of posterior drooling can be enhanced by the fact that many disabled children are taken care of in a supine position for a substantial part of the day. Many children with cerebral palsy (CP) suffer from gastroesophageal reflux (3). In healthy subjects, exposure of the distal esophagus to acid results in an immediate increase of saliva secretion. The possible function of this is that swallowed saliva plays a role in the defense of esophageal mucosa to acid-induced injuries. Reflux in children with CP causes stimulation of pH-sensitive receptors in the mucosa of the distal esophagus which activates the esophageal-salivary reflex leading to an increase of salivary flow rate (3,4). This may exacerbate anterior and posterior drooling. Botulinum neurotoxin (BoNT) injections in the salivary glands in the treatment of anterior drooling demonstrate promising efficacy (5-8). This case report summarizes the results of repeated bilateral single-dose BoNT injections into the submandibular glands in a patient with CP with severe drooling, aspiration, and recurrent pneumonia. The primary treatment intention was to decrease salivary flow in an effort to reduce anterior as well as posterior drooling. PATIENT AND METHODS A male (9 years and 4 months of age, diagnosis CP, quadriplegic spastic athetosis, wheelchair dependent) was referred to our outpatient clinic because he constantly drooled in a manner that clothes, furniture, and objects became wet. He used to withdraw from peer activities because of his drooling. In the two years prior to consultation episodes of pneumonia occurred averaging seven times a year. He had gastroesophageal reflux and was mainly fed by a percutaneous endoscopic gastrostomy (PEG). Stasis of saliva in the oral cavity and hypopharynx was always present. Congested breathing resulted in restless sleep during which he choked on his own saliva. He intermittently used an anticholinergic drug that effectively reduced salivary flow reducing the discomfort of drooling. However, a diversity of side effects made long-term use impossible. Written informed consent was obtained from the parents. To measure the salivary flow rate absorbent dental cotton rolls (Salivette®; Sarstedt B.V., Etten-Leur, The Netherlands) were placed directly at the orifices of the larger salivary glands for a 5-minute period: salivary flow rate [mL/min] = (weight increase of dental rolls/time of collection). The assessments were conducted by the same speech pathologist, always in the morning following 1 hour of starvation. The procedure was repeated 30 minutes later. Because of the close anatomical relation between the sublingual and submandibular glands, separate flow rates could not be obtained. The compound salivary flow from these two glands is referred to as submandibular flow (SubFl). The drooling quotient (DQ), visual analogue scales, and the Teacher Drooling Scale (TDS) were used to assess the degree of drooling (9). The TDS rates anterior drooling on a 5-point scale. A score 5 indicates constantly wet and saliva leaking on clothes and furniture, 3 means occasionally drooling, and 1 indicates no drooling. The DQ was scored during two separated periods of 10 minutes. Every 15 seconds (40 observations in 10 minutes) the presence or absence of drooling was assessed. The DQ was expressed as a percentage estimated from the ratio of observed drooling episodes and the total number of observations (DQ [%] = 100 × number of drooling episodes/40) (10). Visual analog scales of exactly 10 cm without visible subdivisions were presented to the parents. The average degree of drooling severity had to be indicated concerning the 2 weeks before assessment. The patient suffered persistent drooling. Bilateral BoTN injections in the submandibular glands were recommended. It is generally accepted that the submandibular glands produce 60-70% of resting saliva when the individual neither eats nor drinks. The parotids respond primarily during eating and drinking with which we did not want to interfere. The patient was followed in our clinic for several years and repeatedly received BoNT injections. Two weeks before the third injection and 4 weeks after this intervention, video fluoroscopic swallow studies (VFSS) were performed to assess the efficacy of swallowing. Intervention Under general anesthesia and ultrasound guidance a single dose of 50 units BoTN (Botox®; Allergan, Nieuwegein, The Netherlands) was divided over the left and right submandibular gland. The patient returned for follow-up measurements (Flow, DQ, VAS) at 2, 8, 16, and 24 weeks after the first BoNT injection and at 4, 8, 18, and 28 weeks after the second injection. RESULTS The first BoNT treatment was given at the age of 9 years and 6 months. In the months following injection the salivary flow was investigated by the swab method. Baseline submandibular follow (0.59 mL/min) was reduced by 47% two weeks after injection and 36% 24 weeks after injection (Table 1). Anterior drooling severity was scored by the Teacher Drooling Scale. Baseline TDS was 5. Eight weeks after the first injections TDS decreased to 2, indicating sometimes drooling. In accordance with this, the DQ was reduced by 91.3% (from 46 to 4). The outcome of the VAS assessments showed a similar trend. According to the parents' opinion drooling was severe at baseline (VAS: 13) whereas it improved substantially 8 weeks after BoNT injections to mild drooling (VAS: 78). Approximately 6 months after the BoNT injections (control visit 24 weeks) parents reported that, in their opinion, drooling had started to increase (VAS: 34). The SubFL and DQ had increased slightly although a substantial difference still existed compared to baseline (Table 1). Based on the registrations of SubFL, DQ, and VAS a good impression was obtained about anterior drooling and the results after BoNT administration. In addition, it became clear that his posterior drooling, as was present before the BoNT injections, had improved. Two weeks after treatment parents reported his breathing to be improved because of less pooling of saliva in the oral cavity. During the entire control period he slept better, his condition improved, and he had no pneumonia. Although the controls of anterior drooling remained fairly good, parents reported congested breathing again after 7 months, which had been absent after BoNT treatment and for which no objective test or observational study existed. Based on the clinical picture, suggesting the recurrence of posterior drooling, a second BoNT treatment was executed. The results of the follow-up measurements are presented in Table 1.TABLE 1: Assessment of salivary flow rate and drooling before and after treatmentThe symptoms of posterior drooling disappeared again after the second injection and the pulmonary condition improved. During an outpatient visit, 10 months after the second treatment, pneumonia was reported. A third BoNT treatment was executed 12 months after the second injection. VFSS's were done before and after the third injection in order to visualize the probable false route of saliva into the trachea and to gain better insight in the mechanism of swallowing. The first VFSS was made some weeks before the third BoNT injection when he was recovering from an episode of pneumonia. To conduct the study the patient received 3 mL thickened barium contrast. The oral stage of swallowing showed poor and insufficient bolus formation. The tongue demonstrated inadequate peristaltic movements with limited transport of the bolus. The onset of the pharyngeal phase of swallowing was delayed and pooling occurred in the valleculae and pyriform sinuses. Following three attempts to swallow, the material was partially spilled over into the trachea. Residual contrast medium was seen after the first swallowing actions and extended efforts were needed to clear the pharynx. There was no immediate cough or gag response to the aspiration which resulted in silent aspiration. The second VFSS was taken 12 weeks after the third BoNT injection. His tongue movements were stronger, and he showed no refusal to swallow. In contrast to the first study, the onset of the pharyngeal phase was not delayed. Pooling in the valleculae still occurred but was followed by an efficient swallow act, during which no aspiration was observed and less residual contrast medium was seen. DISCUSSION Our patient was treated repeatedly with Botox® injected in the submandibular glands requiring ultrasound guidance for safe injections (11). The interval between treatments ranged 7 to 12 months. BoNT intraglandular injections give an anticholinergic effect because it temporary causes a presynaptic blockage of acetylcholine secretion. A marked decrease in submandibular salivary flow and anterior drooling after the first injection was achieved. Posterior drooling was evaluated indirectly via a change in clinical symptoms like congested breathing, swallowing difficulties, and aspiration with associated pneumonia. A striking finding was the long lasting effect after the second BoNT injection. It is hypothesized that hypotrophy of the injected salivary glands may be induced due to long lasting denervation. From the questionnaires it was concluded that fear to swallow decreased, oral intake increased, and he gained more body weight. Reduced salivary flow rate caused less episodes of aspiration. The patient trusted himself more to swallow. In addition, there might have been less congestion of saliva at the faucial isthmus resulting in a decrease of defensive coughing relieving pharyngeal irritation. The vicious circle of irritation with avoidance to swallow tended to return when the salivary flow from the submandibular glands increased again. Video fluoroscopic swallowing studies were executed before and after the third BoNT injection which indicated that swallowing and deglutition were negatively influenced during episodes of diminished general health (e.g. during recurrent pneumonia). The incidence of pneumonia decreased impressively after BoNT injections improving his physical condition. The VTSS 12 weeks after treatment showed an efficient swallow act without aspiration. In conclusion, posterior drooling can lead to (silent) aspiration. In the case described BoNT injections in the submandibular glands resulted in a substantial reduction of submandibular salivary flow rate. From the clinical course, after injections a reduction of posterior drooling was clear with the absence of recurrent pneumonia as the most important finding. Acknowledgment This work has been supported by a grant from the Johanna KinderFonds, a fund raising consortium in the field of child rehabilitation, Arnhem, The Netherlands.