Whipple's pancreaticoduodenectomy

This topic had appeared in Question 25 from the first paper of 2020, where the college asked for "the complications of this procedure that are of relevance to its ICU management". The impetus for this SAQ may well have been the simple fact that this procedure is not without its complications, and many of these complications end up becoming the ICU's problem. Or, perhaps it was the concern, raised by many a surgical staff, that the ICU trainees do not appear to have a solid grasp of what exactly was done during one of these eponymous procedures (Whipples, Ivor-Lewis, Benthal, etc), or what complications to expect. As assessment drives learning, this SAQ would have an intelligent response to that concern.

The time-poor candidate has not time for idle gibberish, and needs one solid reference to read before the test date. The best resource for this purpose is probably the 2020 ERAS society guidelines,  which is a manageable 19-page document with a nice summary table.

Pancreaticoduodenectomy

A Whipple procedures (or Whipple's procedure) is a pancreaticoduodenectomy. It's known as a "Whipples" after Allen Oldfather Whipple, who in the 1930s modified a much older procedure originally described by Alessandro Codivilla (in 1898). Codivilla pioneered the technique of pancreaticoduodenectomy, resecting the head of pancreas distal stomach and proximal duodenum en bloc with the hospital's only surgical assitant (a monsieur Bartolo dal Monte), with the patient under chloroform. He then attached the bile duct to the remaining duodenal stump, and reattached the rest of the duodenum to the stomach in a manner reminscent of a roux-en-Y. What he did with the stump of the pancreatic duct is unclear, but he probably just ligated it, and it ultimately formed some sort of fistula.  Walther Kausch was the next to refine the procedure in 1912, but his article is too German for the author to determine how he entfernte das Carcinom; all that can be established is that it was a staged procedure and not all of the duodenum was removed. Whipple basically performed the exact same procedure in 1935 (and the patient died within 30 hours) but in contrast to the others he removed all of the duodenum. Somehow, the procedure ended up being named after him (probably because Kausch was not around to contest it, having died in 1928).

Anyway. None of this would have earned any marks.  The college wanted a list of anastomoses:

• Bile duct anastomosed to jejunum
• Stomach stump anastomosed to jejunim
• Pancreatic duct anastomosed to jejunum
  
  Like so:
  
  

Complications of pancreaticoduodenectomy

As one might imagine of a 6-8 hour procedure which ends with three anastomoses in the abdomen, "the complications of this procedure that are of relevance to its ICU management" are numerous, and making a coherent answer to this sort of question really relies on the candidate's abilities to categorise them quickly. What is offered below was constructed mainly from Butturini et al (2006), McEvoy (et al (2014) and Karim et al (2018), of which the latter offer some percentages from their case series of 98 patients. None of them offer any sort of structure, so a simple alphabetic systems-based one has been produced to help non-surgical people.

For a good summary, Aoki et al (2017) have this excellent bar graph based on Japanese data:

Now, to put an ICU spin on this:

Surgical complications

• Specific to the Whipple procedure:
  • Anastomotic leak: the most worrying anastomosis here being the pancreatic anastomosis. The rate of leak is estimated between 2 and 50% (Crippa et al, 2007)
  • Pancreatic fistula:  defined as drain output of amylase-rich fluid;  in 20%. Fuks et al (2009) report that of these, the most severe require emergency reoperation. The fistula is actually the least disturbing complication, as the digestive juices have a mechanism for their egress if a fistula is forming; a collection which is not draining is therefore just sitting there, eroding into the walls of the gut and mesenteric arteries. Most sane people would agree that this is not something you can manage conservatively. 
• Common to all major abdominal surgery
  • Abdominal wound infection (in 23.5%)
  • Wound dehiscence

Airway-related concerns

• These patients may return from theatre intubated
• The decision to extubate them is more delicate than usual, given their predisposition to pain-associated atelectasis
• This is complicated further by the relative contraindication for any positive airway pressure (eg. high flow nasal prongs and NIV). Given the relative fragility of the anastomotic sites and the extremely high risks of morbidiy should anoy of them leak, one might be forgiven for being extra careful. There is no data regarding the increased risk of anastomotic breakdown in this group, and equipoise is absent which makes collecting evidence more difficult. All that can be safely concluded is that the aerophagy and gastric distension can't possibly help the anastomotic healing.

Respiratory complications

• Atelectasis
• Hospital-acquired pneumonia
• The use of NIV and HFNP being streng verboten, there is quite a susbtantial risk of respiratory failure. In one prospective cohort (Junejo et al, 2014) the risk of post-op respiratory failure was 38%, and that's in a group of people who appear to have passed some sort of cardiopulmonary pre-op assessment.

Vascular complications

• Intra-abdominal haemorrhage occurs in <2%, but accounts for 38% of the deaths following Whipples- mainly from the gastroduodenal artery. Pancreatic fistula is one of the most important determinants of this. 
• Hepatic ischaemia (due to compression or inadvertent damage to the hepatic artery) appears to be extremely rare
• Splenic or portal vein thrombosis are somewhat more common
• The persistent need for vasopressor support is common, as these patients often suffer from a prolonged period of low-grade inflammatory vasodilated shock, and have prolonged reliance on epirdural analgesia. The use of vasopressors does not appear to increase the risk of fistulae.

Pain

• The transverse subcostal incision which is usually required is a significant source of pain
• The operation takes place high enough that even a high thoracic epidural may not cover all of the involved structures. But, it is still recommended, as it is far superior to all the other methods.

Electrolyte derangement

• Normal anion gap metabolic acidosis can be produced by a high-volume pancreaticojejunal fistula
• Pancreatectomy can give rise to insulin dependence

Fluid balance

• High output pancreatic fistula can produce a lot of fluid loss
• High volume NG drainage due to delayed gastric emptying can also cause you to become rapidly very volume-depleted
• The 2012 ERAS guidelines recommend a neutral fluid balance for optimal outcomes, which is hard with an oliguric hypotensive patient

Nutritional problems

• Delayed gastric emptying (20%)
• Prolonged period of fasting post operatively
• These people have pancreatic cancer, so they are not coming off a particularly high nutritional baseline

Intra-abdominal collections (12.2%)

• Bilious collection ("biloma") in <2%
• Hepatic abscess (due to the reflux of enteric contents up the bile duct)

Thromboembolic disease

• Apart from the major surgery they've just had, these people are also prone to clots because of their underlying malignancy
• The 2012 ERAS guidelines recommend 4 weeks of low molecular weight heparin.