Secondary Logo

Journal Logo

Original Research Article: Pulmonary

Open Gastrostomy for Noninvasive Ventilation Users with Neuromuscular Disease

Bach, John R. MD; Gonzalez, Monica MD; Sharma, Amit MD; Swan, Kenneth MD; Patel, Anuradha MD

Author Information
American Journal of Physical Medicine & Rehabilitation: January 2010 - Volume 89 - Issue 1 - p 1-6
doi: 10.1097/PHM.0b013e3181c55e2c
  • Free


Dysphagic patients with muscular dystrophy, amyotrophic lateral sclerosis (ALS), and other neuromuscular disorders (NMDs) often require gastrostomy tubes for nutritional support. The conventional resort to percutaneous endoscopic gastrostomy (PEG) with esophageal intubation by the gastroscope can sufficiently compromise ventilation to necessitate translaryngeal intubation and general anesthesia. This resulted in extubation failure and tracheostomy for three patients with ALS before they were referred to us. Thus, we sought an approach that would not necessitate intubation.

To minimize the risk of PEG respiratory complications for patients with ALS, the American Academy of Neurology has recommended that patients with ALS undergo PEG before their vital capacities (VCs) have decreased to 50% of predicted normal.1 However, PEG with gastroscope passage by a hole created in an oronasal interface used for noninvasive ventilation was described for patients with <50% of predicted normal VC,2 and several approaches to radiographically inserted gastrostomy (RIG) under local anesthesia have been described. Another approach, a modification of open (Stamm) gastrostomy, was reported by one of our authors in 2001.3 We reported its use for 27 consecutively referred, continuous noninvasive intermittent positive-pressure ventilation (IPPV) or bilevel positive airway pressure–dependent infants with spinal muscular atrophy type 1 without complications.4 The purpose of this study is to report open (Stamm) gastrostomy for adults with severe neuromuscular weakness, and in some cases continuous dependence on noninvasive IPPV, under local anesthesia. This can prolong survival without resort to invasive respiratory management.


Symptoms of chronic alveolar hypoventilation include fatigue, morning headaches, depression, loss of appetite, and hypersomnolence.5 Symptoms, VC in sitting and supine positions, oxyhemoglobin saturation (SpO2), and end-tidal CO2 (EtCO2), were monitored every 2–12 mos for all patients referred to a NMD clinic and having advanced NMD. Nocturnal noninvasive IPPV was instituted to treat symptomatic hypoventilation. Many nocturnal-only noninvasive mechanical ventilation (NIV) users had no ventilator-free breathing ability (VFBA) when supine. Patients using nocturnal noninvasive IPPV gradually extended its use into, and eventually throughout daytime hours as their inspiratory muscle function (VC) diminished and they became continuously ventilator-dependent. All NIV users required continuous ventilatory assistance during the surgical procedure to maintain normal SpO2 and EtCO2.

The IPPV was provided by nasal interface for nocturnal support. It was delivered by 15-mm angled mouthpieces for daytime ventilatory support (Fig. 1) unless the lips were too weak to grab the mouthpiece, in which case nasal IPPV was used around the clock with a variety of interfaces. The need for noninvasive or invasive IPPV for >20 hrs/day was considered full-time ventilatory support. Discontinuation of ventilator use for full-time-supported patients resulted in blood gas derangement and dyspnea leading to asphyxia.

A 47-yr-old man with ALS with 140 ml of VC using IPPV by a 15-mm angled mouthpiece during daytime hours for 19 mos. He also used nasal IPPV for sleep.

Patients whose bulbar-innervated muscle function deteriorated to the point of no longer permitting adequate nutrition by oral intake were referred for open (Stamm) gastrostomy. Ventilator users who were referred to us with indwelling tracheostomy tubes were excluded because they could have undergone PEG without additional airway intubation, and hence, only NIV-dependent patients with NMD were candidates for this study.

Besides failure to sustain nutritional needs because of dysfunctional deglutition and bolus manipulation and possibly aspiration, as inspiratory musculature deteriorates many patients become undernourished because they are too tachypneic to swallow safely. Tachypneic candidates for gastrostomy must fail to gain weight despite being provided with 1000–1500 ml IPPVs by mouthpiece during meals. Some patients can maintain minute ventilation by grabbing the mouthpiece about every 15 secs, giving them more than sufficient time to manipulate food boluses and trigger swallowing. All patients' modified barium swallow examinations indicated severe dysphagia, and further peroral intake was contraindicated or the patient himself or herself noted that meals took too long to assuage hunger. Additional inclusion criteria included dependence on NIV during the procedure to maintain normal SpO2 and EtCO2, VC in the sitting or supine position <40% of predicted normal, and the ability to cooperate and give consent. Although 41 of the patients were dependent on <20 hrs/day of NIV at the time of the surgery, 21 of the 41 had no VFBA when supine, the position in which the gastrostomy tube needed to be inserted, hence they required full NIV support to survive the procedure. Exclusion criteria were presence of a tracheostomy tube because this would have permitted PEG under general anesthesia without resort to translaryngeal intubation and the presence of intrinsic lung or airways disease that resulted in SpO2 baseline <95% despite eucapnia and optimal use of mechanically assisted coughing (MAC). Thus, a protocol using NIV and MAC as needed to maintain normal SpO2 (>94%) in ambient air for effectiveness and safety during and after the procedure necessitates that the patients have sufficiently healthy lungs.6

The Stamm modification passes the tube through the omentum en route to the abdominal wall. The omentum ensures a seal to the site and eliminates the need for tacking sutures between the stomach and peritoneum.3 Because this simplification enables the use of a smaller midline incision, less intra-abdominal manipulation, and no peritoneal sutures, it can be performed comfortably under local anesthesia. Potential complications include tube dislocation, hemorrhage, peritonitis, and wound infection. Potential respiratory complications include hypoventilation and airway mucus plugging. All patients had outpatient evaluations by the surgeon and physiatrist managing the pulmonary care within 2 wks before and after the procedure.

Along with local anesthesia, the patients were judiciously administered intravenous sedation while receiving full noninvasive ventilatory support by oronasal interface. Midazolam of 0.5 mg was titrated in slowly to a maximum of 4 mg. Ketamine was diluted to achieve a concentration of 10 mg/ml and was titrated in aliquots of 1 ml until the patients closed their eyes but were responsive to voice commands. The maximum dose of ketamine used was 100 mg. Thus, we used only agents that cause minimal respiratory depression, and all patients were comfortable.

Intravenous sedation is often poorly tolerated by patients with chronic ventilatory impairment. Surgery requiring general anesthesia or intravenous sedation or both is avoided for patients with VCs <40% of predicted normal.7–9 However, we used closed systems (oronasal interface) of full noninvasive ventilatory support (Fig. 2). Either assist-control mode ventilation at volumes of 700–1500 ml or pressure-control mode at pressures of 18–25 cm H2O with backup rates of 10–12/min were used. During the procedure, EtCO2 was monitored in the interface, and pulse oximetry was monitored. EtCO2 correlates with PaCO2 (r = 0.94–0.98) except in the presence of severe concomitant lung or vascular disease with congestive heart failure, which was not present in our patient population.10–12 We did not perform preprocedure arterial blood gases because all patients' SpO2 and EtCO2 had been monitored in the outpatient clinic and SpO2 at home. The SpO2, EtCO2, and serum bicarbonate baselines were stable and within normal limits both before and at the time of the procedure for all patients. For patients using oronasal, nasal, and mouthpiece IPPV, EtCO2 was measured by the hole normally used to monitor airway pressures in the 15-mm tube adapters connecting the interface and the ventilator tubing. The pickup can also be slipped under nasal and oronasal interfaces for EtCO2 measurements.13

A 19-yr-old patient with Duchenne muscular dystrophy with VC of 360 ml and requiring continuous noninvasive ventilatory support, undergoing open gastrostomy under local anesthesia with normal alveolar ventilation provided by closed system NIV (Hybrid interface, Teleflex Medicine, Research Triangle Park, NC).

Because the patients were managed in ambient air, any SpO2 <95% indicated artifact, hypercapnia, the presence of airway debris, or intrinsic lung disease. Inadequate ventilation either necessitated interface adjustment to decrease air leakage or adjustment of ventilator settings. When ventilation was normal (eucapnia), the desaturation was as a result of airway secretions that were cleared by suctioning or MAC using mechanical insufflation-exsufflation (CoughAssist, Respironics International, Inc., Murrysville, PA). Exsufflation-timed abdominal thrusts, normally routine, were avoided during the procedure and for 5-days postoperative. Thus, SpO2 was maintained at ≥95% in ambient air to maintain adequate ventilation by NIV, guide the airway secretion expulsion by MAC, and preclude any possibility of ventilatory or respiratory failure. Postoperative complications and survival outcomes were monitored. Data are presented as means ± standard deviations and ranges.


Of the 62 NIV users undergoing open gastrostomy, 41 were part-time (<20 hrs/day), and 18 were continuously NIV-dependent, with most having no significant VFBA, after being part-time dependent for a period of time. Twenty-one of the 41 part-time users had no VFBA supine. The ages, duration of illness, VCs, and noninvasive IPPV use requirements before gastrostomy are noted in Table 1. Open gastrostomy was successful for 62 consecutive times without the need to resort to general anesthesia and PEG. There were no surgical, anesthetic, or postoperative respiratory complications. Of the 62 patients, 3 did not use NIV before the procedure but needed it during the procedure to maintain alveolar ventilation and normal SpO2 and used it nocturnally after the procedure to relieve symptoms. The other 59 required NIV before, at, and after gastrostomy. Thirty-three patients who were less than full-time NIV-dependent preoperatively became continuously NIV-dependent eventually for a total of 51 patients who were spared respiratory failure despite continuous NIV support facilitated by the success of this procedure.

Patient demographic, pulmonary function, and ventilator use

All patients, without exception, gained weight after the institution of enteral nutritional support. Several accidentally dislodged tubes were replaced by Foley catheters of the same caliber (size, 24F). In 15 cases, VC temporarily increased by 125 ± 80 ml despite advancing NMD. Thirteen patients with advanced bulbar ALS eventually underwent tracheotomy when their SpO2 decreased and remained <95% despite NIV and MAC in accordance with reported indications for tracheotomy.13 None of the patients with Duchenne muscular dystrophy or without ALS underwent tracheotomy despite, in 16 cases, continuous ventilator dependence. Mean postgastrostomy survival was 38.8 ± 6.2 mos with 26 of the patients still alive.


This work presents gastrostomy under local anesthesia for patients who are up to being continuously ventilator-dependent without tracheostomy tubes and with VCs <40% of predicted normal. Thus, counter to American Academy of Neurology recommendations, gastrostomy does not need to be performed while deglutition is functional, the VC >50% of normal, and the patient capable of autonomous breathing.1

When symptomatic or when supine VC is much less than VC in the sitting position, patients often require nocturnal ventilatory assistance to sleep reclining. With advancing weakness and need for continuous noninvasive IPPV, snuggly fitting oronasal interfaces at full-support settings provide a closed system of ventilatory support (Fig. 2) similar to that provided by translaryngeal or tracheostomy tubes with inflated cuffs. Even under these circumstances, open gastrostomy can be performed safely provided that it is performed under local anesthesia and noninvasive IPPV and MAC used as needed to maintain normal SpO2 in ambient air. It is impossible to develop respiratory failure with normal SpO2 in ambient air. Low spans of bilevel positive airway pressure, a commonly used form of ventilatory assistance, are inadequate for advanced patients. With normal alveolar ventilation, SpO2 <95% is most often a result of aspiration of saliva or food in this patient population.13 MAC expels airway mucus and reverses associated oxyhemoglobin desaturations <95%.6,13,14 Because an oronasal interface does not protect a patient from aspiration, continuous oximetry monitoring in ambient air was vital during the procedure as feedback to use MAC and airway suctioning as needed. Airway suctioning and MAC reversed all oxyhemoglobin desaturations <95% and, therefore, clinically significant aspiration.

Four of our six tachypneic patients with Duchenne muscular dystrophy and decreased appetites as a result of hypercapnia and inadequate time to trigger the swallowing reflex gained weight after the institution of mouthpiece IPPV during meals.

RIG also permits gastrostomy under local anesthesia. There have been previous reports of the use of NIV for patients with ALS during PEG2,15–18 as well as patients with ALS undergoing RIG with VC <50% of normal.16–21 In all cases, supplemental oxygen was used, but none had VCs <10% of normal as did most of our patients with <300 ml of VC, and none was continuously NIV-dependent with little or no VFBA before or after the procedure. Thus, the patients seemed to be strong enough to avoid hypercapnic ventilatory failure despite oxygen delivery and sedatives. The studies reported equal or fewer complications and difficulties using RIG rather than PEG.16–21 However, there were patients for whom RIG failed and PEG was performed under general anesthesia. The overall major complication rates of RIG were 5%–10% when compared with that in this study (0). Thus, gastrostomy should be performed by approaches that permit local anesthesia with the use of noninvasive IPPV and MAC as needed. Further comparison of RIG and modified Stamm approaches is warranted. Open gastrostomy under local anesthesia facilitates the continuation of noninvasive respiratory management for patients including up to continuous noninvasive ventilatory support for patients with chronic ventilatory failure.


1. Miller RG, Rosenberg JA, Gelinas DF, et al: Percutaneous endoscopic gastrostomy and enteral nutrition in amyotrophic lateral sclerosis. J Neurol 1995;242:695–8
2. Boitano LJ, Jordan T, Bendit JO: Noninvasive ventilation allows gastrostomy tube placement in patients with advanced ALS. Neurology 2001;56:413–4
3. Zickler RW, Barbagiovanni JT, Swan KG: A simplified open gastrostomy under local anesthesia. Am Surg 2001;67:806–8
4. Bach JR: Medical considerations of long-term survival of Werdnig-Hoffmann disease. Am J Phys Med Rehabil 2007;86:349–55
5. Bach JR, Alba AS: Management of chronic alveolar hypoventilation by nasal ventilation. Chest 1990;97:52–7
6. Gomez-Merino E, Bach JR: Duchenne muscular dystrophy: Prolongation of life by noninvasive respiratory muscle aids. Am J Phys Med Rehabil 2002;81:411–5
7. Aprin H, Bowen JR, MacEwen DG, et al: Spine fusion in patients with spinal muscular atrophy. J Bone Joint Surg Am 1982;64:1179–87
8. Sakai DN, Hsu JD, Bonnett CA, et al: Stabilization of the collapsing spine in Duchenne muscular dystrophy. Clin Orthop Relat Res 1977;128:256–60
9. Smith AD, Koreska J, Eng P, et al: Progression of scoliosis in Duchenne muscular dystrophy. J Bone Joint Surg Am 1989;71:1066–74
10. Lindahl SG, Yates AP, Hatch DJ: Relationship between invasive and noninvasive measurements of gas exchange in anesthetized infants and children. Anesthesiology 1987;66:168–75
11. Morley TF, Giaimo J, Maroszan E, et al: Use of capnography for assessment of the adequacy of alveolar ventilation during weaning from mechanical ventilation. Am Rev Respir Dis 1993;148:339–44
12. Wiedemann HP, McCarthy K: Noninvasive monitoring of oxygen and carbon dioxide. Clin Chest Med 1989;10:239–54
13. Bach JR, Bianchi C, Aufiero E: Oximetry and indications for tracheotomy in amyotrophic lateral sclerosis. Chest 2004;126:1502–7
14. Bach JR: Respiratory muscle aids: Patient evaluation and the oximetry, respiratory aid protocol, in Bach JR (ed): Noninvasive Mechanical Ventilation. Philadelphia, Hanley and Belfus, 2002, pp 165–88
15. Gregory S, Siderowf A, Golaszewski AL, et al: Gastrostomy insertion in ALS patients with low vital capacity: Respiratory support and survival. Neurology 2002;58:485–7
16. Chio A, Galletti R, Finocchiaro C, et al: Percutaneous radiological gastrostomy: A safe and effective method of nutritional tube placement in advanced ALS. J Neurol Neurosurg Psychiatry 2004;75:645–7
17. Thorton FJ, Fotheringham J, Alexander M, et al: Amyotrophic lateral sclerosis: Enteral nutrition provision—Endoscopic or radiologic gastrostomy? Radiology 2002;224:713–7
18. Desport JC, Mabrouk T, Bouillet P, et al: Complications and survival following radiologically and endoscopically guided gastrostomy in patients with amyotrophic lateral sclerosis. Amyotroph Lateral Scler Other Motor Neuron Disord 2005;6:88–103
19. Gottschalk A, Strotzer M, Feuerbach S, et al: CT-guided percutaneous gastrostomy: Success rate, early and late complications [in German]. Rofo 2007;179:387–95
20. Shaw AS, Ampong MA, Rio A, et al: Entristar skin-level gastrostomy tube: Primary placement with radiologic guidance in patients with amyotrophic lateral sclerosis. Radiology 2004;233:392–9
21. Lewis D, Ampong MA, Rio A, et al: Mushroom-cage gastrostomy tube placement in patients with amyotrophic lateral sclerosis: A 5-year experience in 104 patients in a single institution. Eur Radiol 2009;19:1763–71

Open Gastrostomy; Noninvasive Mechanical Ventilation; Amyotrophic Lateral Sclerosis; Duchenne Muscular Dystrophy; Mechanical Insufflation-Exsufflation

© 2010 Lippincott Williams & Wilkins, Inc.