Introduction:

Radical cystectomy (RC) is a highly morbid procedure with frequent infectious complications, a
significant risk of anastomotic leaks, and issues with wound infection or dehiscence. Methods for
detecting patients at highest risk of these complications are needed to try and prevent or reduce
morbidity. CRP is a non-specific inflammatory marker often used to guide patient management, with
reported indications in the post-operative period. We hypothesized that a markedly elevated CRP
several days after RC may be a sub-clinical indicator of complication risk due to infection, wound
breakdown, or anastomotic leak.

Methods:

Our prospectively maintained database included CRP measurements starting May 2020. 78 consecutive
patients had a single CRP measured POD#4. Any fever >38°C or leukocytosis (WBC >11.0) after POD#3
through time of discharge was recorded. Complications were graded and classified by the MSKCC
system. Patients who experienced a wound, infection, or anastomotic (I/W/A) complication within 14
days of RC were compared with those who did not. ROC analysis was used to define the cutoff for
elevated CRP in predicting I/W/A complications. Linear association between CRP and maximum WBC
was assessed. Multivariable analysis was performed to find factors associated with I/W/A complications.

Results:

There were not significant differences between patients experiencing 14d I/W/L complications when
comparing sex, age, BMI, tobacco use, Charleston comorbidity index >2, receipt of neoadjuvant
chemotherapy or surgical approach (open vs robotic). CRP performed well on ROC analysis for 14d
I/W/A with an AUC = 0.750 and a CRP of 110 chosen as a cutoff based on Youden’s index.

Multivariable analysis adjusting for elevated CRP, post-operative fever, leukocytosis, demographic
characteristics, and neoadjuvant chemotherapy showed elevated CRP (OR15.3, 95%CI 3.1-75.5, p<0.001)
and WBC >11.0k (10.3, 95%CI 2.2-49.0, p=0.003) to be independently associated with I/W/A complications (table 1). A scatter plot shows there was not a significant linear association between CRP and leukocytosis >11K (R-square = 0.08, Figure 1).

Conclusion:

I/W/A complications, highlighting the need for methods to capture those at increased risk. Our data
suggests a single CRP on POD#4, when elevated, can be a sub-clinical marker of I/W/A complications
independent of leukocytosis. While leukocytosis is a known clinical marker of inflammation, there is not a
strong linear association with CRP in our data set, indicating both may have individual value in assessing
post-RC patients. There should be high suspicion of I/W/A complication for patients with an elevated
CRP and modified discharge pathways or closer follow-up may be strategies to prevent or reduce this morbidity.

Funding: n/a