Table 5. Consequences of POPF.
| Consequence | Data from literature |
|---|---|
| Increased mortality | Mortality rates associated with POPF have remained at 1% for 25 years, while overall mortality rates for Grade C POPF is 25% (16,53). The increased mortality in patients with POPF can be attributed to the leak itself, as well as secondary complications. Khobragade et al. retrospectively studied the factors influencing POPF associated mortality following pancreaticoduodenectomy in 592 patients (54). They found that postoperative hemorrhages, BLs, chest complications, positive blood cultures, number of interventions performed on the patient, number of explorations performed, and pre-intervention serum albumin levels influenced mortality rates in patients with Grade C POPF (54) |
| Increased morbidity | Secondary complications can arise following POPF, including IAS and hemorrhage. Related patient morbidity is very high, with rates reported between 30% and 65% (15,27,55-61) |
| Prolonged LOS | For clinically relevant POPF, management usually leads to a prolonged LOS. This can be attributed to treatment of the leak, treatment of secondary POCs, or extended monitoring of patients |
| A study by Veillette et al. retrospectively analyzed the cases of 581 patients undergoing PD. The overall incidence of fistulas was low at 12.9% (75/581). The patients that developed POPF had significantly longer LOS (23.6 vs. 8.7 days, P<0.001) (62) | |
| In another international, multicenter, retrospective study of 1,089 patients that underwent DP, the clinical outcomes of patients were compared based on the 2005 and 2016 ISGPS definition of POPF. In this study, van Hilst et al. found that, according to the 2016 International Study Group in Pancreatic Surgery definition, the median length of stay increased from 7 days in patients with no POPF or Grade A POPF, to 9 days in Grade B POPF. This number increased notably to 29 days for Grade C POPF (63) | |
| Increased reoperation rates | The incidence of Grade C POPF was reported as 1.3% following DP according to a meta-analysis of three studies with 2,635 patients (64). Following PD, this rate was reported as 1.3% in a retrospective systematic review of 5,115 patients at nine high-volume hepato-pancreato-biliary centres in China (65) |
| Additionally, certain secondary conditions may also require surgical treatment, such as peritonitis, hemorrhage, and sepsis | |
| A study by Veillette et al. found that reoperation was required in 6.7% patients that developed POPF, vs. 2.2% patients that did not develop POPF (P=0.04) (62). A study by van Hilst et al. recorded a 13% percent reoperation rate in patients with clinically relevant POPF (CR-POPF), as opposed to 3% in the uncomplicated/Grade A cohort (63) | |
| Increased readmission rates | Readmission to the hospital may be required in cases of late onset POPF, signs of which may not be recognized in the early postoperative period. In the 2008 Veillette et al. study (62), 46 patients suffered from CR-POPF. Of these patients, 33 were classified as having an overt fistula, that is, the signs of POPF were evident by POD 7. The remaining 13 patients were classified as occult fistulas, as the drain outputs did not meet clinical cutoffs to determine the presence of POPF initially, but the patients later experienced abscesses, PPH, or death that was associated with POPF. Among the overt fistula group, 30% of patients required readmission, whereas 77% of patients in the occult fistula groups required readmission (P=0.007). There was no significant difference found between reoperation, bleeding, or mortality among the overt and occult fistula groups. This may be because patients that present with late onset fistula are readmitted just for monitoring, even though they may not necessarily need readmission |
| The study by Veillette et al. also found that readmission was required for 27% patients with POPF, while in the non-POPF cohort, only 11% of patients required readmission (62). Readmission rates varied from 6% in the uncomplicated/Grade A cohort, versus 32% in patients with CR-POPF, according to the van Hilst study (63) | |
| Increased patient cost | Diagnosis of POPF, and treatment of POPF and of secondary complications results in increased patient costs. A 2022 retrospective study conducted by Jajja et al. on a cohort of 997 PD patients between 2010–2017 compared the mean cost for patients with POPF to those without in the United States. Patients with CR-POPF had median cost of $54,727, compared to $23,024 for patients without (P<0.001). They also found that POPF patients incur 2.4 times the overall median cost of PD. The most significant contributor to cost was the postoperative ward stay (median =$39,373). The group also noted a significant increment in cost tied to the grade of POPF with median costs of $32,164 ($13,053) for Grade A, $50,263 ($30,883) for Grade B, and $102,013 ($107,484) for Grade C POPF (P<0.001) (66) |
| Topal et al. conducted a retrospective review of the records of 109 patients that received a curative PD for pancreatic or periampullary tumor. The POPF rate was 12.8%, while overall complication rate was 46.8% (67). The median LOS for the patient categories of POPF, other, and no complications were significantly different [26, 21, and 14 days, respectively (P<0.0001)]. The hospital cost per patient category was significantly correlated with the LOS (P<0.0001). The major factors responsible for increase in inpatient costs were hospitalization and medical staff, but not operating room costs (67) | |
| A similar study by Enestvedt et al. in 2010 found that the median cost for patients with the major POC after PD was $56,224, compared to a cost of $29,038 for uncomplicated recoveries (P<0.001) (68). This was a retrospective study that was conducted on data from 145 patients. A high morbidity rate of 26% resulted in longer LOS (21 vs. 11 days, P<0.001) and length of intensive care unit stay (0.89 vs. 5.3 days, P<0.001) in patients with complications. Multivariate analysis revealed that POPF resulted in 1.3 times the median cost of treatment in patients, compared to non-leak patients. Notably, median pharmacy charges went from $13,306 to $39,640 in patients with major post-operative complications (68) | |
| Increased need for intensive care | Patients that develop CR-POPF usually require intensive care due to the high morbidity that includes life-threatening conditions such as septic shock, peritonitis, and multiorgan failure |
| In the study of PD patients by Veillette et al., it was found that development of POPF led to significantly increased time in the ICU [20 (SD 26.7) days vs. 17 (SD 3.4) days, P<0.001] (62). Similarly, van Hilst et al. found that percentage of patients that required ICU admission ranged from 2% in uncomplicated patients, or those with Grade A POPF, to 4% in Grade B POPF, to 59% in patients with Grade C POPF (63) | |
| It should be noted that admission to the ICU can vary by center based on the facilities available in the surgical ward. While some centers are better equipped to handle complications, such that only organ failure warrants an ICU admission, other centers, other centers may have a lower clinical threshold for this (62) | |
| DGE | DGE is a common complication following pancreatic surgery, characterized by prolonged retention of food within the stomach without mechanical obstruction. This complication is divided into three grades (A, B, and C) based on nasogastric intubation, type of diet a patient can tolerate, a patient’s general health, whether a prokinetic drug is used, and the need for further tests (69). DGE can lead to symptoms such as nausea, vomiting, bloating, and early satiety, significantly impacting postoperative recovery and quality of life. Beyond impacts on patients alone, DGE places significant burden on healthcare systems, whereby patients with DGE may require additional medical interventions, prolonged hospital stays, nutritional support, and medications, leading to higher healthcare costs (69) |
| In a prospective study of 267 patients undergoing pancreaticoduodenectomy (>80% pylorus-preserving, antecolic-reconstruction), 49 patients (17.8%) developed DGE, with 5.1% classified as Grade B and 3.6% as Grade C (70). DGE patients were more likely to present with multiple complications (32.6% vs. 4.4%, ≥3 complications, P<0.001), including POPF (42.9% vs. 18.9%, P=0.001) and IAA (16.3% vs. 4.0%, P=0.012). They also had a longer hospital stay (median, 12 vs. 7 days, P<0.001) and were more likely to require transitional care upon discharge (24.5% vs. 6.6%, P<0.001). Multivariate analysis revealed that predictors for DGE included POPF [OR =3.39 (1.35–8.52), P=0.009] and IAA [OR =1.51 (1.03–2.22), P=0.035] (70). A separate study by Futagawa et al. found that DGE (particularly Grade C) negatively affects cancer-specific survival, using data from 383 patients who underwent pancreaticoduodenectomy (140 with DGE) (71). Five-year overall survival rates were 32.7% and 41%, respectively, for the DGE versus non-DGE group (P=0.02), highlighting the seriousness of this complication (71) | |
| PPH | Hemorrhages are most commonly caused due to a pseudoaneurysm of a large visceral artery that develops due to the artery being in contact with pancreatic fluid, which is high in proteolytic digestive enzyme. Some commonly afflicted arteries include the common hepatic, splenic, gastroduodenal and superior mesenteric artery. If these are left untreated, the pseudoaneurysm may expand and rupture, which is associated with severe hemorrhage and hemodynamic instability. In these scenarios, early intervention is critical (16). Generally, the rate of PPH varies from 3% to 16.8% (72-75) |
| A retrospective study of 347 patients that underwent PD was conducted by Khuri et al. 18 (5.18%) of which suffered from PPH (72). Of these, 5.6% (1/18) suffered early PPH, while the remaining 94.4% (17/18) suffered late PPH. Pseudoaneurysm was recorded as the cause for 6/17 late PPH. Significance was shown in the severity of late PPH and vascular pseudoaneurysms (P=0.001) (72) | |
| In another retrospective study of 1,122 patients, the rate was reported as 3% for both pancreatectomies and PD (74). Rate of early PPH was 21% and tended to extraluminal. Late PPH was intraluminal in 18/26 patients (69%). The study found that PPH significantly increased LOS (P<0.01), but not mortality rates. Additionally, PPH occurred following discharge in 39% of patients (74) | |
| Postoperative sepsis | Sepsis can result from bacterial infection due to contamination of the operative field by biliary and enteric contents, which may explain why it is more common following Whipple than DP (16). The complication presents itself clinically as the development of a fever, DGE, abdominal pain, rising inflammatory markers such as C-reactive proteins, and turbidity of drain fluid (16,76). Drain fluid can be assessed through Gram stain, bacterial cultures, and microscopy. Patients presenting these symptoms must be assessed for peripancreatic fluid collections with CT scans (16). These collections can usually be managed through image guided percutaneous or endoscopic drainage (16) |
| In a retrospective analysis, Behrman et al. studied 192 patients that underwent elective pancreatectomies (77). 16.3% (32/192) patients developed IAS. The risk factors for IAS included an early onset POPF, and soft pancreatic texture (77). The study found prolonged LOS (28.5 vs. 15.2 days), and a higher mortality rate (15.8% vs. 1.8%) in patients with IAS compared to patients without IAS |
POPF, postoperative pancreatic fistula; BL, bile leak; LOS, length of stay; POCs, post-operative complications; PD, pancreaticoduodenectomy; DP, distal pancreatectomy; ISGPS, International Study Group in Pancreatic Surgery; POD, post-operative day; PPH, postoperative pancreatic hemorrhage; ICU, intensive care unit; SD, standard deviation; DGE, delayed gastric emptying; IAA, intra-abdominal abscess; OR, odds ratio; PPH, postoperative pancreatic hemorrhage; CT, computed tomography; IAS, intra-abdominal sepsis.