Introduction

Hypertrophic pyloric stenosis (HPS) is a common gastrointestinal cause of gastric outlet obstruction infants and is one of the most common surgical conditions of infancy. ¹ It is caused by the idiopathic diffuse hypertrophy and hyperplasia of the circular muscle fibres of the pylorus with the proximal extension into the gastric antrum resulting in construction and obstruction of the gastric outlet. In response to outflow obstruction and vigorous peristalsis, stomach musculature becomes uniformly hypertrophied and dilated.

Epidemiology

Pyloric stenosis has an incidence of 2 to 4 per 1000 live births in Western population² and it appears to be less common in infants in African and Asian populations. It is four to five times more common in males.³ The cause of pyloric stenosis is unknown. Genetic, familial, gender and ethnic origin can influence the incidence rates of HPS. Offspring of parents with this condition have a higher risk of developing HPS and in many series first-born males are more frequently than the other siblings.⁴

Clinical presentation

Pyloric stenosis is not present at birth. In the early phase, there may just be regurgitation or occasional non-projectile vomiting. The onset rarely occurs before the first week of life and it commonly present at 2–8 weeks of age, however, it is reported in premature babies prior to this corrected age. The peak usually occurs at four weeks and it is seldom delayed beyond the second to third month. Within a variable time after the onset of symptoms, the vomiting becomes more projectile and it generally occurs during or soon after feeding. However, at times, it may occur up to several hours later. The frequency of vomiting varies, in some infants vomiting occurs after each feed while in others it may be somewhat intermittent.

Shortly after vomiting, the infant is often hungry and will take another feed immediately. The vomitus is non-bilious but due to the frequency and force of vomiting, it may occasionally become blood tinged. The amount of stool may be very small and infrequent depending on the amount of food that reaches the intestine. The degree of dehydration, lethargy and metabolic derangement depends on the time interval between the onset of symptoms and presentation. In some late presentations the infant may be severely malnourished, dehydrated or even collapsed.

It is important to note that it may take some time for the clinical features, metabolic disturbance and ultrasound findings to become established and one should not exclude HPS if symptoms persist / recrudesce after a relatively normal examination of an infant presenting very early. A planned review should features evolve is appropriate. Likewise, it is important to consider HPS in a previously benign ‘posseting’ or ‘refluxy’ baby, whose regurgitation of milk has ‘changed’ in intensity or frequency rather than attributing this to ‘posseting getting worse’. In this situation, the clue to the possibility of HPS is the change in the nature of the previous ‘vomiting’.

Examination findings

With established HPS, the infant fails to thrive due to calorie losses, becomes dehydrated due to fluid losses and appears ‘hungry’ unless significant dehydration or alkalosis has started to affect the infant’s activity level. The infant usually appears ‘bright and active’ unless significantly dehydrated, compared to the infant with a urinary tract infection who may be constitutionally unwell.

On physical examination, gastric distension or visible peristaltic waves may be seen moving from the left upper abdomen toward the epigastrium, and right side in some cases.⁵ The palpable finding of a firm, mobile and non-tender ovoid mass (‘olive’) either to the right of the epigastrium or in the midline, deep to right rectus muscle and under the liver edge is diagnostic. This finding of a palpable mass requires much patience as the success of such a finding is dependent on an empty stomach and a relaxed anterior abdominal wall in a non-crying settled infant. If the stomach is significantly distended during palpation, aspiration of gastric contents using a nasogastric tube may be helpful to increase the likelihood of feeling the hypertrophied muscle. Also, palpation during a test feed may allow a previously non-palpable hypertrophied pylorus to be felt during peristaltic contractions. The best position for palpation is on the infant’s left side. The inability to palpate an olive-shaped mass does not exclude the diagnosis of HPS and often an ultrasound is needed to clarify the diagnosis.⁶

With extensive and protracted vomiting, metabolic derangement will occur. Vomiting of gastric contents leads to depletion of sodium, potassium and hydrochloric acid, which results in the characteristic finding of hypokalaemic, hypochloraemic metabolic alkalosis.⁷ The kidneys conserve sodium at the expense of hydrogen ions, resulting in a paradoxical aciduria. With the increasing degree of dehydration, renal potassium losses are accelerated in an attempt to retain sodium and fluid.

Imaging studies

Some clinicians believe that the palpation of an olive-shaped mass may obviate the need for a confirmatory imaging study, as a positive examination has high specificity.⁸ Plain radiographs will often have been performed, given the history of vomiting, although they are of no diagnostic value. They may show gastric distension.

Ultrasonography is the now the diagnostic test of choice as it can be performed quickly and without radiation exposure. The accuracy is close to 100% when performed by experienced personnel,⁹ having a sensitivity and specificity of 99.5% and 100% respectively.¹⁰ The sonographic appearance of ‘doughnuts’ or ‘bull’s-eyes’ on cross-section of the pyloric channel is most characteristic. A muscle thickness of the pylorus greater than 4 mm and a pyloric channel length of greater than 17 mm yield a positive predictive value of greater than 90%. For infants less than 30 days of age, these limits may be lower.¹¹

In the absence of ultrasonography, barium upper gastrointestinal study is an effective means of diagnosing HPS. This study may be preferred over ultrasound as the cost-effective initial imaging study when the clinical presentation is atypical for HPS and favours other conditions more amenable to diagnosis by upper gastrointestinal study.¹² Positive findings include an elongated pylorus with antral indentation from the hypertrophied muscle. The pathognomonic finding is the appearance of a ‘railroad track’ sign caused by two thin parallel streams of barium traversing the pylorus. There is also a vigorously peristaltic stomach with delayed or no gastric emptying.

Upper GI endoscopy is performed on the very rare occasions when other imaging modalities are inconclusive and although it would demonstrate pyloric obstruction, it would be difficult to differentiate it from pyloric spasm.

Differential diagnosis

The diagnosis is made by the characteristic clinical manifestations of projectile non-bilious vomiting and the finding of a pathognomic pyloric mass. Other causes of vomiting, especially in early life, include achalasia of the oesophagus and a symptomatic hiatus hernia. Gastro-oesophageal reflux with or without hiatus hernia may have a similar presentation of persistent vomiting after feeding. Other non-pathological causes include feeding technique issues and overfeeding by an inexperienced enthusiastic carer.

Projectile vomiting may occur in some rare diagnoses such as pyloric membrane or pyloric duplication where a palpable mass may be felt.

Metabolic causes may mimic pyloric stenosis. Adrenal insufficiency results in profuse non-bilious vomiting but it is likely to result in metabolic acidosis rather than alkalosis as in pyloric stenosis. Unlike pyloric stenosis, the serum potassium and urinary sodium concentration are elevated in adrenal insufficiency. Recurrent emesis with alkalosis or acidosis may be caused by some inborn errors of metabolism but usually there will be other associated clinical features, such as hypoglycaemia, coma or seizures to suggest a metabolic cause. Any vomiting infant needs to have a clean urine checked to exclude infection.

Management

The definitive treatment is surgical repair by a Ramstedt pyloromyotomy, which is the procedure of choice in which the pyloric mass is split, leaving the mucosal layer intact. This procedure is fairly straightforward, with minimal complications. The pylorus may be accessed by various incision techniques, including laparoscopic means. All methods are considered acceptable practice, with minimal differences in outcomes noted.¹³

The preoperative treatment is directed toward correction of fluid, electrolyte and acid–base imbalance which is important to achieve prior to anaesthesia. The amount of fluid resuscitation is based on the degree of dehydration. Correction of fluid and electrolyte imbalance can usually be achieved within 24-48 hours.

Resuscitation with a bolus dose of intravenous normal saline is required for moderate to severe dehydration and is given at 10–20 mL kg⁻¹. This is followed by rehydration with N/2 + 5% dextrose solution at 1.5 times maintenance rate over 24 hours.

Adequate amounts of both chloride and potassium are necessary to correct metabolic acidosis. The correction of potassium can be achieved by adding 10–20 mEq of KCl per 500 mL of intravenous fluid in patients with normal renal function. Chloride can be adequately replaced by normal saline or N/2 saline with 5% dextrose. During resuscitation urine output and electrolyte determinations should be performed regularly. In general, correction of chloride level to 90 mEq L⁻¹ or greater is adequate for surgery to be performed.

Complications

Pyloromyotomy is associated with a low incidence of morbidity and mortality. A retrospective review of a large number of patients from two centres between 1969 and 1994 showed an overall 19% complication rate.¹⁴ Therefore, complications are minimal when the pyloromyotomy is performed by experienced hands. The mortality associated with this procedure is less than 0.4% in most major centres.¹⁵

Controversies

There is disagreement as to when vomiting becomes significant enough to warrant investigation as infants frequently regurgitate small amounts following a feed especially if they have caregivers with poor feeding technique. An important distinction may be the general appearance of the infant, as infants with regurgitation generally appear relatively well.

It is not easy to palpate a pyloric mass as infants are irritable in the presence of protracted vomiting.

The threshold for investigation with ultrasonography varies considerably.