8.9.

Nutritional support for GI fistulas

J.Sobocki, B. Szczygieł, T. Naber, S. J. Dudrick

Learning objectives

- To recognize the impact of fistulas at different levels of the gastrointestinal (GI) tract on
digestion, absorption, and metabolism
- To understand the role of nutrition and metabolic care in the nonoperative treatment of
GI fistulas
- To understand the role of enteral nutrition in maintaining intestinal integrity, reducing
infection rates, and decreasing bacterial translocation
- To understand the nutritional goals and means to prepare the patient for elective GI
fistula surgery

8.9.1. Definition and classifications

Definition
A fistula of the gastrointestinal (GI) tract is an abnormal anatomic connection between a
part (or multiple parts) of the digestive tract lumen and the lumen of another epithelialized
structure or surface of the body.

Classifications
Several classifications of fistula exist. They are based on severity, location, or etiology1,2.
According to the location of the tract, fistulas can be divided generally into two groups:
- internal, consisting of an abnormal communication between adjacent hollow viscera (i.e.
gastro-colic, sigmo-vesical), these most often result from complications of disease (i.e
Crohn’s disease, diverticulitis)
- external, consisting of an abnormal communication between the GI tract and the surface
of the body, these most often result from complications of surgery.

Fistulas occur rarely (1%–2.4% of all abdominal operations), but they can present a
number of medical and surgical problems3-5. Wide adoption of damage control surgery,
together with open abdominal procedures in trauma and emergency surgery, have confronted
surgeons with an especially difficult adversary – the exposed complex fistula. External
fistulas can be further divided according to their severity and to their clinical outcome7:
- type I - enterocutaneous fistula (ECF) - connects the intestinal lumen with the skin
surface (i.e. following the tract of the drain after surgery or spontaneous fistula
formation in Crohn’s disease)
- type II - entroincisional fistula (EIF) - connects the intestinal lumen with the surgical
incision (laparotomy)
- type III - enteroatmospheric fistula (EAF) – connects the intestinal lumen with the
granulating surface covering the bowel in patients who have lost part of the anterior
abdominal wall.
- Type IV - laparostomy accompanying fistula (LAF) - similar to EAF but granulation
has not substituted for the lost part of the abdominal wall, the bowel empties into the
peritoneal cavity and outside the body.
 Severity, mortality, hospital stay, and time to definitive surgical repair increase from type I
to IV. Currently, the most challenging fistula for surgeons is the open or exposed enteric
fistula, classified as EAF or LAF. These fistulas, which arise in the midst of an open
abdominal wound, are very difficult to control and there is still much confusion and lack of
consensus on the best approaches to control them6. The intention of treatment is to convert
fistula type IV into type III, III into II, and II into I.

Functional classifications have prognostic value and divide fistulas into:

– Low output <200 ml·day-1
– Medium output 200-500 ml·day-1
– High output >500 ml·day-1

High output fistulas are associated with a greater risk of water-electrolyte and metabolic
disturbances, lead to malnutrition in a shorter time, and have a lower tendency to heal by
nonsurgical treatment. High output fistulas usually arise from the upper GI tract.
Anatomical classifications distinguish fistulas from upper and lower GI tract, selective biliary,
pancreatic, or mixed fistulas.
The following etiological groups can be distinguished: iatrogenic, posttraumatic, foreign
body, inflammatory (i.e. Crohn’s disease), neoplastic, and infective (i.e. tuberculosis,
actinomycosis).

8.9.2. General principles of fistula treatment

As a rule, fistula treatment is nonoperative. Early surgery increases risk of complications

and mortality and should be avoided. Surgery is limited only to emergency intervention. Even
then, patients must be well resuscitated.
In 1964, Chapman et al. outlined definite priorities, and a time frame for the treatment of
ECF8. More recently the principles of treatment of fistula have been summarised by the
acronym SSNAP (Stabilization, Sepsis control and skin care, Nutrition, Anatomy assessment,
surgical Procedure to deal with fistula) 3,9-11:
- Stabilization refers to the compensation of water-electrolyte imbalance. This should be
achieved within a few hours of diagnosis. In hospitalized patients, severe disturbances
develop rarely, as they are under watchful postoperative medical supervision. Life-
threatening disturbances may develop if the fistula has developed at home after
discharge from hospital.
- Sepsis control is of paramount importance for lifesaving and successful treatment of
fistula. Minimal invasive drainage (USG or CT-guided, minilaparotomy, or damage
control laparotomy) is a strategy of choice and aims to limit metabolic distress and
burden on the immune system resulting in overwhelmed protein loss. Extended bowel
resections and multiple operations are risk factors for fistula and intra-abdominal
sepsis/abscess development12 and creation of short bowel syndrome. Protection of the
skin or wound (damaging fistula secretions contain enzymes and extreme pH) aims to
prevent wound enlargement, additional protein loss, and suffering of the patient. ECF
and EIF are effectively managed with continuous suction drainage and stoma healing
paste administered around the external exit of the fistula. For endoscopically-
accessible fistulas, early endoscopic repair (covered prothesis, clipping etc.) or
endoluminal vacuum therapy may significantly accelerate healing. Endoscopic
procedures are less effective for the treatment of chronic fistulas. In the case of EAF
and LAF, evacuation of the fistula output is challenging. Occlusive, sponge-vacuum
 dressings11 or endoscopic deriving bypass (conversion of EAF to ECF)13 are
procedures of choice.
- Nutritional support is a cornerstone for reduction of mortality and healing of the
fistula.
- Assessment of the anatomy of the fistula is mandatory prior to surgical repair, if the
fistula has not been healed with nonoperative treatment.
- Definitive surgery is dedicated to cases of fistula which have not healed. The operative
procedures are postponed until the general condition, metabolism, and local status
have improved to ensure safe recovery after surgery7,8. The preparation time to surgery
ranges from 3 months for ECF to 12 months for EAF. Patients with LAF are converted
to other types of fistula prior to discharge from hospital (most commonly to EAF).
Surgical procedures may be complex, technically demanding, and consist of
reconstruction of the GI tract, urinary tract, and anterior abdominal wall. Complete
reconstruction may be achieved in one, two, or three stage procedures performed at
intervals over several months. Mortality, morbidity, and fistula recurrence rates are
low in specialized centres and should not exceed 3%, 25%, and 5%, respectively.
An algorithm for management of gastrointestinal fistulas is shown in figure 8.9.1.

8.9.3. Metabolic care and nutrition

The aim of metabolic care and nutrition is to prevent complications and support spontaneous
healing of the fistula.
Major fistula complications include3,9-11:
- hypovolemia and electrolyte imbalance (see chapter 2.6.)
- intra-abdominal abscess and sepsis
- skin/abdominal wall defect progression
- malnutrition (which is present in 55%–90% of fistula patients).

The type and severity of disturbances associated with fistulas reflect the segment of the GI
tract that is affected by the fistula, the volume of fistula output, the severity of inflammation
or sepsis, and comorbidities. Upper GI tract fistula is characterized by high output, and high
concentration of electrolytes, bicarbonate and protein, and, thus, induces metabolic
deterioration much faster. Losses of fluids and bicarbonates by fistulas, accompanied by
sepsis-induced retention of water lead to oliguria, which aggravates water-electrolyte and
acid-base disturbances (typically hypokalemia and acidosis). This cycle, if not properly
treated, leads to acute renal failure. Fluid replacement is mandatory and roughly equals 35
mL·kg-1 body mass plus losses by fistula and plus demand. For individuals older than 70
years of age, the formula is modified to 30 mL·kg-1 body mass. Some patients require
diuretics and temporary catecholamine support. Upper GI fistula output is typically replaced
by 0.9% NaCl + KCl 10mEq·l-1. Duodenal and pancreatic fistulas may require bicarbonate
supplementation.
Characteristics of secretion in different segments of the GI tract are shown in table 8.9.1.
The amount and composition of fistula output is a net result of the secretion and absorption of
the proximal GI segments. Total volume of daily GI secretion reaches 8000 mL in orally fed
patients and a quarter of this amount in fasting individuals. Pharmacological interventions
such as proton pomp inhibitors, somatostatin analogues, or opioid agonists (i.e. loperamide,
racecadotryl) may reduce fistula output but do not improve fistula healing.
 Secretion Volume Na K Cl HCO3-
(mL·d-1) (mmol·l-1) (mmol·l-1) (mmol·l-1) (mmol·l-1)
Salivary gland 1500 10 26 15 50
Gastric 1500 100 10 100 0
Duodenal 2000 130 5 90 10
Ileal 3000 140 5 100 30
Pancreatic 800 140 5 75 115
Biliary 800 150 5 100 35
Colon negligible

Table 8.9.1. Characteristics of secretion in different segments of the gastro-intestinal tract

Nutrition should be introduced when water-electrolyte and acid-base balance have been
reestablished, sepsis is controlled, and other contraindications for nutrition are solved.
Nutrition should not be postponed more than 48 hours, but the quantity and composition of
the mixture may differ and depends on metabolic status. This may start with minimal
macronutrient supply, but supplies of water, electrolytes, micronutrients, and vitamins must
meet full requirements.

Nutritional access
Enteral nutrition is preferred, if possible and effective. Feeding distal to the fistula by naso-
jejunal tube or jejunostomy is preferred for the initial 4-6 weeks. This allows local rest of the
bowel, which facilitates fistula treatment. After 4 weeks, the fistula tract is solid and tight, and
usually shortens and becomes straight, if properly drained. If the fistula has not healed, the
naso-jejunal tube may be replaced by a tube introduced through the tract of the fistula. A
catheter in the tract of the fistula prevents spontaneous healing, thus it is indicated for fistulas
with a poor prognosis for spontaneous healing.
Replacement of the naso-jejunal tube by administration through the fistula tract improves
the patient’s quality of life, facilitates care, reduces the risk of tube impaction (tube is shorter
and larger diameter), and is a good solution for a long-term nutritional preparation for
definitive surgery. Oral or proximal to fistula feeding is acceptable if the prognosis for
spontaneous healing is poor, the feeding meets nutritional needs, is well tolerated, and does
not increase risk of sepsis, and if the fistula is easy to care for and fistula output is not
significantly increased. Enteral nutrition compared to parenteral nutrition is safer and more
effective in achieving nutritional goals.
Parenteral nutrition is the procedure of choice in cases of intestinal insufficiency or
inability to obtain safe and stable enteral access. Central vein catheter is mandatory (CVC or
PICC). Nutrition by peripheral veins is ineffective and may be used only exceptionally (i.e.
temporary CVC loss).

Composition of nutritional mixture

Two periods of nutritional support in fistula patients can be distinguished:
1. Early period (uncontrolled sepsis) nutrition should approach coverage of basic
metabolic rate, fistula-induced losses, and sepsis-induced losses.
 2. Stable period (septic problems are solved) - nutrition should cover basic metabolic
rate, fistula-induced losses, and anabolic needs indispensable for fistula closure.

Losses through the fistula comprise not only water and electrolytes, but protein, energy,
bicarbonates, bile salts, vitamins, and microelements as well. The amount of loss of particular
nutrients corresponds to the volume of fistula output, affected GI segment, and local
inflammation or abscess formation. Protein losses through the fistula are estimated to be 2g of
nitrogen (12.5 g of protein) per one liter of discharge. Additional loses of protein result from
SIRS (cytokine-related proteolysis), immobilization, and irritation of the wound.
Nutritional needs vary from 20 kcal and 1-2 g of protein per kg of body mass (in the early
period) to 35 kcal and 2 g of protein per kg of the body mass (in the stable and recovery
period). No particular formula for calculation can be recommended. Metabolic needs vary
between individuals and in the subsequent periods of disease.
Parenteral nutrition requirements for macronutrients, compared to those for enteral
nutrition, are usually 20%-25% lower. This results from the energetic effects of enteral
feeding, losses of unabsorbed nutrients in the gastrointestinal tract, and lower metabolic limits
for parenterally-administered nutrients.
Evaluation of the impact of nutritional support on the results of management of GI fistulas
is limited by the lack of prospective randomised trials comparing different types of nutritional
intervention. If nutrient requirements can be met, the route of administration of nutrients does
not seem to make any difference to the outcome. Not all patients, however, can be
successfully managed by enteral nutrition, which may increase fistula output. No controlled
trials have shown elemental diets to be more beneficial than standard, less expensive,
polymeric diets.
The risk of the onset of liver function disorder is related to the presence of an
inflammatory process, probable bacterial overgrowth, and an interrupted enterohepatic cycle
of bile acids. Fistuloclysis, defined as reinfusion of fistula output into the distal part of the gut,
may reduce the incidence of liver function disorders12. This technique is, however, complex,
very demanding in the hospital, and usually impossible at home. In many patients with upper
GI fistulas (e.g. oesophageal or duodenal), enteral nutrition can be administered into the gut
distal to the fistula, and parenteral nutrition can be avoided10,13.
Bed rest and inactivity induce catabolism, stimulate protein breakdown, increase risk of
nosocomial infections, and compromise metabolic limits, thus, intensive physical
rehabilitation is mandatory.

Prognosis
Data from the literature on the impact of nutritional support on the spontaneous closure of
ECF are controversial. In 1970, Dudrick and colleagues reported data for 78 patients with
gastrointestinal fistulas treated with TPN, in whom the spontaneous closure rate was 70% and
mortality rate 6%14. In three retrospective studies, patients treated before the introduction of
TPN were compared with patients treated with TPN. In each of these series, a significant
increase in the spontaneous closure rate from 27% to 56%, from 34% to 81%, and from 35%
to 65% was noted and a significant decrease in the mortality rate was achieved in patients
treated with TPN.
In the literature, closure rates with nonoperative treatment vary from 19% to 92%.
Considering spontaneously closed fistulas, 90% of them healed in the first month after sepsis
resolution, with an additional 10% closing in the second month, and only a marginal number
after 2 months. With vacuum wound therapies, there are case reports of fistulae closure into
the second and third month11.
 However, not all fistulas can close with nutritional support8. The accurate investigation of
fistulas and the exclusion of possible post-fistula stenosis are important in determining the
possibility of spontaneous healing and in making decisions about nutritional treatment. A low
output fistula originating from the lower part of the GI tract is more likely to close than a
more proximal high output fistula. Fistulas in which gut mucosa can be seen have a minimal
chance of spontaneous closure. The same applies to fistulae with a distal stenosis or in
association with malignancy. The factors responsible for failure of spontaneous fistula closure
have been described by an acronym FRIENDS (Foreign body, Radiation,
Inflammation/Infection, Epithelialization, Neoplasm, Distal intestinal obstruction, and
Steroid-immunosuppression). Local ischemia is another important factor.

8.9.4. Nutritional preparation for elective surgery of the persistent fistula.

During emergency operations, surgical repair of the fistula is not recommended and/or
indicated. Repeated laparotomy and several attempts at primary repair have a small chance for
success, expose the patient to further complications, and risks the creation of short bowel
syndrome. If, after 8 weeks of nutritional support, intensive local treatment, and rehabilitation,
spontaneous closure has not occurred, operative repair of the fistula should be considered15,16.
Elective surgical repair of the persistent fistula should be delayed until the patient’s condition
becomes optimal. Patients must be physically rehabilitated and their nutritional status
improved preoperatively, to enable not only an uneventful postoperative period, but to survive
complications, if they occur. In the case of extended inflammation on a CT-scan or MRI,
and/or a poor condition of the patient, non-emergency surgical procedures should be
postponed17. No particular period of time for patient preparation can be recommended. In
practice, it ranges from 3 months after sepsis resolution in ECF, to 12 or more months in EAF
and LAF patients.
Enteral nutrition through a tube is not recommended for several months, due to low QoL
and higher risk of access-related complications. Enteral access options comprise PEG,
gastrojejunostomy, surgical jejunostomy, or feeding through the fistula tract. Enteral feeding
maintains or reestablishes intestinal integrity, reduces infection rates, decreases bacterial
translocation, reestablishes gut microbiota, and supports the disappearance of multi-drug
resistant bacteria. Stimulation of GI hormone secretion improves metabolic limits (along with
production of apolipoproteins and incretins), reduces hunger, and inhibits gastric secretion.
Trophic effects of enteral nutrition make the bowel wall more solid in consistency and wider
in diameter, which makes surgery technically easier and safer. Diversion enteritis and colitis
resolves and motility of the distal bowel improves. All of these effects reduce anastomotic
leakage risk and enhance rehabilitation after surgery. Patients receiving total parenteral
nutrition can benefit from supplemental enteral nutrition (i.e. 300-500 ml per day of a
standard diet), which should be introduced at least 4 weeks before surgery, if possible.
Long-term home parenteral nutrition is indicated in cases when enteral nutrition is impossible
or unproductive18.
Fistula output should not exceed 1000 mL per day, which corresponds to minimum losses
of 12g of protein, 600 kcal, and significant amounts of electrolytes, microelements, vitamins,
and other crucial compounds. Higher fistula output may make gaining preoperative nutritional
targets impossible and increases the risk of metabolic complications of parenteral nutrition18.
Patients with large defects of the abdominal wall require rehabilitation and body building
techniques to improve remnant muscles and fasciae to make possible abdominal cavity
closure at the time of GI reconstruction procedure.
 In time, the local inflammatory process will subside, which improves the surgical outcome,
and allows avoidance of extended intestinal resection. Multidisciplinary anatomical (CT,
endoscopy, fistulography, etc.) and functional workup must be performed shortly prior to
surgery.
All attempts should be made to save as much gut as possible in order to restore nutritional
autonomy in the future. Postoperatively, full nutritional support, (mainly parenteral in patients
with withdrawn enteral nutrition preoperatively), must be continued, along with antibiotics.
The monitoring of patients must be very conscientious, because the same complications may
develop in the postoperative period as at the onset of the disease. Recurrence rates after fistula
repair surgery reach 5% in the early postoperative period, but three of four of these cases
close successfully on nonoperative treatment.

Figure 8.9.1. Algorithm for management of gastrointestinal fistulas

Summary

The therapeutic goals in the management of postoperative GI fistulas are: prevention of

complications and closure of the fistula, or preparation for definitive surgery.
Achieving these goals is not easy, and often impossible, in the short term, especially in the
malnourished patient. Treatment of fistulas is complex and based on bowel rest, and enteral
nutrition, if possible, but parenteral nutrition if not, pharmacological suppression of secretion,
suction drainage, physical rehabilitation, and careful monitoring of all vital functions.
Whenever possible, some enteral nutrition should be introduced early, at least as a part of
nutritional support. Elective surgery should be considered if, after 3–5 weeks of nutritional
support, spontaneous closure has not occurred. Timing of elective surgery should meet criteria
of optimal metabolic improvement, functional rehabilitation, and regression of local
inflammation. Emergency surgery is indicated in patients who develop uncontrolled sepsis, or
severe haemorrhage, but minimally-invasive procedures (such as radiologically-guided
percutaneous drainage) are preferred. Endoscopic procedures (clipping, stenting, endoluminal
vacuum, PEDS, etc.) may significantly accelerate healing of the fistula.

References:

1. Berry SM, Fischer JE. Classification and pathophysiology of enterocutaneous fistulas.

Surg Clin North Am 1996;76:1009-18.
2. Sitges-Serra A, Jaurrieta E, Sitges-Creus A. Management of postoperative
enterocutaneous fistulas: the roles of parenteral nutrition and surgery. Br J Surg 1982;69:147-
50.
3. Soeters PB, Ebeid AM, Fischer JE. Review of 404 patients with gastrointestinal
fistulas. Impact of parenteral nutrition. Ann Surg 1979;190:189-202.
4. Dudrick SJ, Maharaj AR, McKelvey AA. Artificial nutritional support in patients with
gastrointestinal fistulas. World J Surg 1999;23:570-6.
5. Visschers RG, Olde Damink SW, Winkens B, Soeters PB, van Gemert WG. Treatment
strategies in 135 consecutive patients with enterocutaneous fistulas. World J Surg
2008;32:445-53.
6. Schecter WP, Hirshberg A, Chang DS, et al. Enteric fistulas: principles of
management. J Am Coll Surg 2009;209:484-91.
7. Ravindran P, Ansari N, Young CJ, Solomon MJ. Definitive surgical closure of
enterocutaneous fistula: outcome and factors predictive of increased postoperative morbidity.
Colorectal Dis 2014;16:209-18.
8. Lee MJ, Heywood N, Sagar PM, Brown SR, Fearnhead NS, p CDC. Surgical
management of fistulating perianal Crohn's disease: a UK survey. Colorectal Dis
2017;19:266-73.
9. Polk TM, Schwab CW. Metabolic and nutritional support of the enterocutaneous
fistula patient: a three-phase approach. World J Surg 2012;36:524-33.
10. Falconi M, Pederzoli P. The relevance of gastrointestinal fistulae in clinical practice: a
review. Gut 2001;49 Suppl 4:iv2-10.
11. Meguid MM, Campos AC. Nutritional management of patients with gastrointestinal
fistulas. Surg Clin North Am 1996;76:1035-80.
12. Rinsema W, Gouma DJ, von Meyenfeldt MF, Soeters PB. Reinfusion of secretions
from high-output proximal stomas or fistulas. Surg Gynecol Obstet 1988;167:372-6.
13. Levy E, Frileux P, Cugnenc PH, Honiger J, Ollivier JM, Parc R. High-output external
fistulae of the small bowel: management with continuous enteral nutrition. Br J Surg
1989;76:676-9.
14. Dudrick SJ, Wilmore DW, Steiger E, Mackie JA, Fitts WT, Jr. Spontaneous closure of
traumatic pancreatoduodenal fistulas with total intravenous nutrition. J Trauma 1970;10:542-
53.
15. Lloyd DA, Gabe SM, Windsor AC. Nutrition and management of enterocutaneous
fistula. Br J Surg 2006;93:1045-55.
16. Foster CE, 3rd, Lefor AT. General management of gastrointestinal fistulas.
Recognition, stabilization, and correction of fluid and electrolyte imbalances. Surg Clin North
Am 1996;76:1019-33.
17. Visschers RG, van Gemert WG, Winkens B, Soeters PB, Olde Damink SW. Guided
treatment improves outcome of patients with enterocutaneous fistulas. World J Surg
2012;36:2341-8.
18. Compher C, Winkler MF, Guenter P, Steiger E. Nutrition Management of Home
Parenteral Nutrition Among Patients With Enterocutaneous Fistula in the Sustain Registry.
JPEN J Parenter Enteral Nutr 2017:148607117695246.