Characterization of pre-operative anemia in patients undergoing surgery by a gynecologic oncologist and association with post-operative complications

Safe to save blood in ovarian cancer surgery - time to change transfusion habits

BACKGROUND

Patients with advanced ovarian cancer (AOC) undergoing surgery are often subjected to red blood cell (RBC) transfusions. Both anemia and RBC transfusion are associated with increased morbidity. The aim was to evaluate patient recovery after the implementation of patient blood management (PBM) strategies.

METHODS

This retrospective cohort study included 354 patients with AOC undergoing surgery at Skane University Hospital Lund, Sweden, between January 2016 and December 2021. The gradual implementation of PBM strategies included restrictive RBC transfusion, tranexamic acid as standard medication before laparotomies and intravenous iron administered to patients with iron deficiency. Severe complications were defined as Clavien-Dindo (CD) grade ≥ 3a. Logistic and linear regression analyses were used to evaluate the differences between three consecutive periods.

RESULTS

After the implementation of new strategies, 52% of the patients had at least one transfusion compared to 83% at baseline (p < 0.001). There was no difference in the rate of severe complications (CD ≥ 3a) between the groups, adjusted odds ratio 0.55 (95% CI 0.26-1.17). The mean difference in hemoglobin before chemotherapy was -1.32 g/L (95% CI -3.04 to -0.22) when adjusted for blood loss and days from surgery to chemotherapy. The length of stay (LOS) decreased from 8.5 days to 7.5 days (p 0.002).

INTERPRETATION

The number of patients transfused were reduced by 31%. Despite a slight increase in anemia rate, severe complications (CD ≥ 3a) remained stable. The LOS was reduced, and chemotherapy was given without delay, indicating that PBM is feasible and without causing major severe effects on short-term recovery.

The impact of perioperative transfusions on the oncologic outcomes of patients with ovarian cancer: A population-based study

Perioperative blood transfusion in ovarian cancer patients was associated with a 28% increase in all-cause mortality. The negative impact of perioperative blood transfusion extends beyond the immediate postoperative period.

OBJECTIVES

The effect of perioperative blood transfusions on long-term oncologic outcomes of patients with advanced ovarian cancer undergoing cytoreductive surgery remains uncertain. Our study aims to determine the association between perioperative blood transfusion and all-cause mortality in this population.

METHODS

Using province-wide administrative databases, patients with advanced ovarian cancer who underwent surgery between 2007 and 2021 as part of first-line treatment were identified. Perioperative transfusion was defined as any transfusion from date of surgery to discharge from hospital. Multivariable Cox proportional hazards regression models were used to determine if there was an independent association of transfusion with all-cause mortality, accounting significant confounders.

RESULTS

A total of 5891 patients had cytoreductive surgery for advanced ovarian cancer between 2007 and 2021, of which 2898 (49.2%) had interval cytoreductive surgery (ICS) and 2993 (50.8%) had primary cytoreductive surgery (PCS). Perioperative blood transfusion was given to 37.3% of patients (40.5% ICS and 34.2% PCS). On multivariable analysis, there was an increased hazard of all-cause mortality for patients receiving perioperative transfusion compared to those who did not (hazard ratio: 1.28; 95% CI: 1.20-1.37). The association of increased all-cause mortality was observed starting 1 year after surgery, was sustained thereafter, and seen in both ICS and PCS groups.

CONCLUSION

Perioperative blood transfusion after cytoreductive surgery for ovarian cancer is common in Ontario, Canada and was significantly associated with an increase in all-cause mortality. Blood transfusion is a poor prognostic factor, and the negative impact of blood transfusion persists beyond the immediate postoperative period.

Low serum creatinine levels are associated with major post-operative complications in patients undergoing surgery with gynecologic oncologists

OBJECTIVE

Serum creatinine is a byproduct of muscle metabolism, and low creatinine is postulated to be associated with diminished muscle mass. This study examined the association between low pre-operative serum creatinine and post-operative outcomes.

METHODS

This retrospective cohort study utilized the 2014-2021 National Surgical Quality Improvement Program to identify patients undergoing surgery with gynecologic oncologists. Patients with missing pre-operative creatinine, end-stage renal disease, sepsis, septic shock, dialysis, or pregnancy were excluded. Pre-operative creatinine was categorized into markedly low (≤0.44 mg/dL), mildly low (0.45-0.64 mg/dL), normal (0.65-0.84 mg/dL), and four categories of elevated levels (0.85-1.04, 1.05-1.24, 1.25-1.44, and ≥1.45 mg/dL). Outcomes included major (≥Grade 3) 30-day complications, categorized into any complications, wound, cardiovascular and pulmonary, renal, infectious, and thromboembolic complications. Also examined were 30-day readmissions, reoperations, and mortality. Logistic regressions assessed the association between creatinine and complications, with stratification by albumin and sensitivity analysis with propensity score matching.

RESULTS

Among 84 786 patients, 0.8% had markedly low, 19.6% mildly low, and 50.2% normal creatinine; the remainder had elevated creatinine. As creatinine decreased, the risks of major complications increased in a dose-dependent manner on univariable and multivariable analyses. A total of 9.6% (n=63) markedly low patients experienced major complications, second to creatinine ≥1.45 mg/dL (9.9%, n=141). On multivariable models, both markedly and mildly low creatinine were associated with higher odds of major complications (OR 1.715, 95% CI 1.299 to 2.264 and OR 1.093, 95% CI 1.001 to 1.193) and infections (OR 1.575, 95% CI 1.118 to 2.218 and OR 1.165, 95% CI 1.048 to 1.296) versus normal. Markedly low creatinine had similar ORs to creatinine ≥1.45 mg/dL and was further associated with higher odds of cardiovascular and pulmonary complications (OR 2.301, 95% CI 1.300 to 4.071), readmissions (OR 1.403, 95% CI 1.045 to 1.884), and mortality (OR 2.718, 95% CI 1.050 to 7.031). After albumin stratification, associations persisted for markedly low creatinine. Propensity-weighted analyses demonstrated congruent findings.

CONCLUSIONS

Low creatinine levels are associated with major post-operative complications in gynecologic oncology in a dose-dependent manner. Low creatinine can offer useful information for pre-operative risk stratification, surgical counseling, and peri-operative management.

Shifting trends and sicker patients: Reassessing hysterectomy performed for benign indications by gynecologic oncologists

OBJECTIVE

To assess trends and differences in patient characteristics, complications, and distributions of hysterectomy for benign indications by benign gynecologists (BG) and gynecologic oncologists (GO).

METHODS

This retrospective cohort study identified patients undergoing hysterectomy for benign indications using the National Surgical Quality Improvement Program data from 2014 to 2021. Exclusions were made for gynecologic or disseminated cancers, ascites, non-gynecologic surgeons, and cesarean hysterectomies. Primary outcome was major (≥Grade 3) 30-day complications, categorized into any complications, wound, cardiovascular and pulmonary, renal, infectious, andthromboembolic complications. Thirty-day readmissions, reoperations, and mortality were also analyzed. Propensity score matching was performed in a 1:1 match of GO to BG patients and was compared. Linear regressions assessed trends.

RESULTS

Among 198,767 patients, 18% (n = 37,707) underwent hysterectomy for benign indications with GO. GO patients exhibited more risk factors for complications and differed significantly from BG patients in comorbidities and perioperative characteristics. Overall, GO patients had higher major complication rates (3.1% vs 2.2%, p < 0.001) and for several other composite complications. After matching, compared to BG, GO-performed hysterectomies had similar rates of major complications (3.0% vs 3.0%, p = 0.55) and no differences in other composite complications, except fewer reoperations (1.2 % vs 1.5%, p < 0.01) and wound complications (0.4% vs 0.5%, p = 0.02) in GO patients. Over the eight years, the percentage of GO-performed hysterectomy (β = 0.41, R2 = 0.71,p < 0.01) increased significantly whereas BG-performed surgeries decreased by the same magnitude. BG had a significant decrease in frail patients (β = -0.47, R2 = 0.90, p < 0.01), but GO did not (β = -0.36, R2 = 0.38, p = 0.10).

CONCLUSIONS

GO are performing more hysterectomies for benign indications on higher-risk patients. However, on a matched cohort, risks of major complications were similar between GO and BG.