Abstract
Background/Aim: With a greater proportion of women with advanced ovarian cancer (AOC) successfully undergoing radical cytoreductive surgery, the demand on peri-operative resources - including intensive care (ICU) beds - is also on the rise. Extended post-operative ICU length of stay (LOS) confers increased patient morbidity and mortality. Several variables associated with prolonged ICU LOS following AOC surgery have been identified. We aimed to evaluate the predictive value of serum lactate levels. Patients and Methods: All patients undergoing ultra-radical surgery for AOC in a large cancer centre over a 34-month period between 2018-2021 were identified via the institution tumour registry. Data were collected retrospectively via electronic care and operating records; biochemistry, radiology, and histopathology databases. Results: In total, 63 patients were identified. Elevated intra-operative serum lactate levels were associated with significantly longer length of ICU post-operative stay. Longer time for hyperlactaemia to normalise following surgery also conferred significantly longer ICU, high dependency and total hospital LOS. Greater blood loss, higher surgical complexity and peritoneal carcinomatosis score, and longer operating time were associated with higher – and persistently elevated – peri-operative lactate levels. Conclusion: Serum lactate in the context of ultra-radical surgery for AOC represents an accessible and inexpensive marker with potential to not only reliably predict LOS, but also to serve as a dynamic prompt for early targeted intervention. Early recognition and correction of hyperlactaemia following AOC may reduce ICU LOS limiting both the resource pressure and patient morbidity/mortality sequelae.
At the time of diagnosis, >60% of patients with ovarian cancer have advanced disease (AOC) (1). Complete surgical cytoreduction, coupled with platinum-based chemotherapy, represents optimal management (2–4). Advances in anaesthesia, surgical technique, and critical care as well as the concentration of ovarian cancer services in specialist high-volume centres have enabled increasing proportions of women with AOC to successfully undergo ultra-radical surgery (5). The boundaries of such surgeries continue to expand evolving from pelvic debulking, to potentially incorporating multi-visceral resection, extensive peritoneal stripping and complete cytoreduction of disease sites previously deemed surgical stopping points such as peri-hepatic, diaphragmatic and cardio-phrenic lymphatic involvement (6–11). Because the incidence of AOC is highest in the post-menopausal population, the majority of patients have, in addition, multiple co-morbidities at the time of diagnosis (12). This combination of disseminated malignancy, surgical radicality, and poor performance status (PS) accounts for 21-36% of all women undergoing surgery for OC requiring post-operative intensive care (ICU) admission (12, 13).
ICU admission following ultra-radical surgery for AOC has implications for both patients and resource demand. Studies collectively report increased mortality in ovarian cancer patients admitted to intensive care post-operatively; further compounded if ICU length of stay (LOS) exceeds 5 days (12, 14–16). The intensive care unit is a specialised clinical area whose optimal running necessitates highly-trained multi-disciplinary staff and 1:1 critical-care nursing, and is provided at significant cost. Therefore, the ability to predict the need for post-operative ICU and identify those at risk of prolonged admission has value; It can be used to improve pre-operative patient counselling, guide peri-operative decision-making and facilitate targeted provision of level 3 care following surgery.
Currently, peri-operative patient risk assessment and ICU demand prediction most often derive from the pre-operative utilisation of validated prognostic scoring systems. Clinical observations, physiological and laboratory measures are combined in advance of surgery to generate and assign individualised patient risk scores as well as to predict morbidity and mortality (17). Such systems can, however, be complex, time-consuming, and may require expensive software to employ. Furthermore, they may fail to take account of the impact of intra-operative events on prognosis (17). To this direction, it has been shown that older patient age, hypoalbuminaemia as well as the intra-operative variables such as prolonged operating time, increased blood loss, higher volume fluid and blood-product resuscitation and concomitant bowel resection, independently associate with increased potential for peri-operative morbidity and significantly longer ICU stay (5, 14, 16, 18–21).
A biomarker of increasing interest in this context is serum lactate, a by-product of glycolysis and measure of end organ perfusion. Hyperlactaemia is well established as a prognostic indicator of poorer outcome in critical illness; with a higher lactate level and longer time to normalise conferring greater risk of death (22–24). In the surgical context – whilst a paucity of evidence exists pertaining to gynaecological oncology – a series of studies on cardiac surgery report elevation of lactate in the peri-operative period as significantly correlating with greater morbidity; increased ventilatory and inotropic requirements and extended length of hospital stay (25, 26). Cardiac surgery patients developing a hyperlactaemia have been shown to experience an 8-10-fold increase in peri-operative mortality (27, 28).
We aimed to consider the value of this parameter in cytoreductive surgery for AOC, evaluating intra-operative lactate as a potential predictor of prolonged ICU stay, as well as determining the impact of time taken for the post-operative resolution of hyperlactaemia on ICU duration of stay. Extrapolating from our hypothesis that raised lactate levels during AOC surgery, and the persistence of such, thereafter, confer an increased ICU LOS, morbidity, mortality, and resources, we searched for peri-operative factors significantly influencing this biochemical marker. Hence, we also explored the potential merit of serial lactate measurements during and after cytoreductive surgery as a reliable guide for both intra-operative decision-making and targeted peri-operative goal-directed therapies – aiming to promptly identify and resolve hyperlactaemia, achieving real-time modification of patient risk and anticipated LOS.
Patients and Methods
Ethics statement. The study was designed and undertaken as a retrospective single-centre investigative data analysis; with no patient identifiers disclosed, nor interventions actioned. Ethical approval was not required.
Patient population. All consecutive patients undergoing ultra-radical cytoreductive surgery for newly-diagnosed histologically-proven FIGO stage III/IV epithelial ovarian/tubal/primary peritoneal cancer within the Derby Gynaecological Cancer Centre (DGCC) University Hospitals of Derby and Burton (UHDB) NHS Foundation Trust over a 34-month period from June 2018-March 2021 were identified using a prospective institutional electronic tumour registry. The DGCC is a large cancer centre, managing all patients with gynaecological malignancy within a defined catchment area of the East Midlands region of the U.K. As all UK NHS patients receive free treatment, with limited private capacity for the management of AOC, our study cohort is likely representative of the wider UK population. Patients undergoing both primary (PDS) and delayed debulking surgery (DDS) following neo-adjuvant chemotherapy (NACT) at DGCC were included; whilst those not admitted directly from operating theatre to ICU and for whom peri-operative serum lactate levels were not available were excluded.
Methods. Integral to management of AOC within our centre is robust multi-disciplinary team (MDT) discussion of all patients to reach consensus on management. Standard approach to ultra-radical surgery entails extensive patient counselling with cancer nurse specialist (CNS) support; pre-operative anaesthetic assessment; surgery led by a minimum of two consultant gynaecological-oncology surgeons with experience of upper abdominal cytoreductive techniques; and anaesthesia performed by a consultant anaesthetic specialist in AOC surgery. Patients are admitted electively to ICU post-operatively, stepping down to ward-level care via high dependency (HDU). Peri-operative lactate levels and supportive care interventions, such as fluid management and inotropes, are undertaken or administered at the discretion of the anaesthetic, critical-care and surgical teams.
Patient baseline characteristics were captured from electronic care records and the UHDB Trust electronic cancer database populating a study proforma with the following demographic and clinical variables: age at surgery; body mass index (BMI); smoking status; and performance status (PS) as assigned at pre-operative anaesthetic review. Additional staging, intra- and post-operative data were retrieved from biochemistry and radiology reports, MDT meeting outcomes and operation notes on tumour type, grade, and stage, estimated intra-operative blood loss (EBL) and operating time. Surgical complexity (SCS), peritoneal carcinomatosis index (PCI) scores and level of cytoreduction achieved, as determined by lead surgeon, were also included in the data collection; with SCS >8 and PCI >20 denoting high complexity surgery and tumour volume (29, 30). Patient tracking records were accessed to calculate length of stay, sub-categorised into ICU, HDU and total in-hospital LOS. All peri-operative serum lactate results, measured in standard units of mmol/l, were retrieved for each patient, identifying both maximum intra-operative and end of surgery levels. A threshold of >2 mmol/l was used to define ‘hyperlactaemia’; and value of >/equal to 4 mmol/l further defined as ‘severe hyperlactaemia’ (23). Time for lactate to normalise post-operatively was calculated from available serial measurements. Post-operative complications were recorded and classified according to the Clavien-Dindo system; and all hospital re-admissions within 30 days of surgery captured (31).
Statistical analysis. Data were collated into frequencies or percentages for categorical variables. Continuous variables were presented as mean, median or range values. To determine the significance of results, odds ratios (ORs) with confidence intervals (CI) were calculated – setting CI at 95% and significance level at p<0.05, and univariate analysis of data performed.
Study outcome measures. The primary outcome of the study was to evaluate the relationship between intra-operative serum lactate and intensive care, high dependency and overall hospital length of stay. The secondary outcome measure was to evaluate the relationship between time for serum lactate to normalise post-operatively and intensive care, high dependency and overall hospital length of stay. The tertiary outcome was to identify peri-operative factors associated with significantly higher intra-operative serum lactate levels.
Results
Patient population. A total of 63 patients were identified for inclusion in the study – see Table I. Median age was 66.3 years (range=47.1-84.1 years). Among all patients, 20.6% (n=13) were aged >/equal to 75 years. Median body mass index (BMI) was 25.0 (18.8-42) kg/m2; with 14.3% (n=9) of the cohort having BMI >/equal to 35 kg/m2. Smokers accounted for 9.5% (n=6) of patients. A pre-operative performance status (PS) of >/equal to 2 was assigned to 17.5% (n=11). Serous carcinoma was the most common underlying histopathological diagnosis (n=60, 95.2%) and the majority of patients had radiologically defined stage IVb disease (n=33, 52.4%). Primary and delayed debulking surgery was performed in 31.7% (n=20) and 68.3% (n=43) of women, respectively. Mean SCS and PCI scores were 7 (2-16) and 17 (3-38), respectively. Average blood loss at ultra-radical surgery was 985 ml (range=200-3,000 ml). Rates of complete cytoreduction to R0 were 95.0% (n=19) for PDS and 97.7% (n=42) for DDS. Mean LOS for ICU, HDU and total hospital stay were 3104 (731-13,730) min, 6,330 (1,048-22,029) min and 16.0 (6-78) days respectively. A total of 39 patients (61.9%) experienced post-operative complications varying from grade I-V; the most prevalent being grade II (n=16, 25.4%). The rate of patients re-admitted within 30 days of surgery was 14.3% (n=9).
Patient population baseline characteristics.
Impact of maximum intra-operative serum lactate on length of stay, complication, and re-admission rates. All patients (n=63) developed an intra-operative hyperlactaemia of >2 mmol/l during surgery for AOC. The maximum serum lactate recorded for each patient was categorised according to severity of hyperlactaemia – either as remaining <4 mmol/l (n=41) or reaching >/equal 4 mmol/L (n=22). Comparative analysis demonstrated a maximum intra-operative lactate level equal to or exceeding 4 mmol/l – that is, development of severe hyperlactaemia – as conferring a significantly longer post-operative ICU LOS (p=0.01). No significant difference was demonstrated, however, between cohorts when comparing duration of both HDU and total hospital admission (p>0.05). Maximum intra-operative lactate had no impact on the likelihood of developing complications, or post-operative re-admission (p>0.05; Table II).
Impact of maximum intra-operative serum lactate on length of stay, complication and re-admission rates.
Impact of end of surgery serum lactate on length of stay, complication and re-admission rates. End of surgery serum lactate levels were also categorised as being either <4 mmol/l (n=43) or >/equal 4 mmol/l (n=20). Data demonstrated the persistence of severe hyperlactaemia on closing was associated with a significantly longer post-operative ICU stay (p=0.04). A higher end of surgery lactate had no, however, significant impact on HDU, total hospital LOS or rates of post-surgical complications, including re-admission (Table III).
Impact of end of surgery serum lactate on length of stay, complication and re-admission rates.
Impact of time for serum lactate to normalise post-operatively on length of stay, complication, and re-admission rates. Time for hyperlactaemia to normalise post-operatively was compared as taking either <1,500 min (n=35) or >/equal 1,500 min (n=28). A longer time for elevated lactate levels to resolve following surgery was associated with significantly longer LOS in both ICU (p<0.00001) and HDU (p=0.005), as well as increased duration of total inpatient admission (p=0.03). No significant difference according to this variable was demonstrated, however, regarding complication and re-admission rates (p>0.05; Table IV).
Impact of time taken for serum lactate to normalise post-operatively on length of stay, complication and re-admission rates.
Impact of peri-operative factors on intra-operative serum lactate levels. Age, body mass index, smoking status and pre-operative patient performance status. The patient age, BMI, smoking status and pre-operative PS had no significant effect on maximal intra-operative, end of surgery lactate level or time taken for hyperlactaemia to resolve post-surgery (Table V).
Impact of peri-operative factors on intra-operative serum lactates and time to normalise post-operatively.
Surgical complexity score. When comparing patients assigned a low/intermediate SCS (n=39) versus high SCS (n=24), both the maximum intra-operative and end of surgery serum lactate levels were significantly higher if SCS was equal to, or exceeded, 8 (p=0.0005 and p=0.002 respectively). The time taken for lactate to return within normal limits following surgery was also significantly longer for the high surgical complexity cohort (p=0.0001; Table V).
Peritoneal carcinomatosis index score. The impact of PCI score on severity of intra-operative hyperlactaemia was compared, considering serum lactate values for those with PCI <20 (n=47) versus >/equal 20 (n=16). In the context of a higher PCI, both the maximum lactate reached during surgery and value at end of surgery were significantly higher (p=0.001); as was the time taken for such to correct in the post-operative period (p=0.006; Table V).
Surgical blood loss. The impact of estimated blood loss on intra-operative lactate was evaluated, comparing patients with EBL of <1,000 ml (n=37) with >/equal 1,000 ml (n=26). Greater surgical blood loss was associated with significantly higher maximum intra-operative serum lactate (p=0.009), end of surgery lactate (p=0.019) and longer time for hyperlactaemia to resolve (p=0.002; Table V).
Operating time. Peri-operative lactate results were compared according to operating time – categorising duration of surgery as either <8 h (n=44) or >/equal to 8 h (n=19). Prolonged ultra-radical surgery, exceeding 8 h, generated significantly higher intra-operative serum lactate levels – and was associated with an extended time for elevated levels to normalise (p=0.001; Table V).
Discussion
Our results have demonstrated that, in the context of ultra-radical debulking surgery for AOC, intra-operative serum lactate measurements reaching >/equal 4 mmol/l, persistence of severe hyperlactaemia to and beyond end of surgery and a time interval of >/equal 1,500 min for elevated lactate levels to return to normal range are all associated with significantly longer duration of ICU stay. Delayed resolution of a post-operative hyperlactaemia was also shown to correlate with both a significantly increased HDU and total inpatient LOS. Variables identified as influential on peri-operative lactate trend were SCS >/equal 8, PCI score >/equal 20, EBL >/equal 1,000 ml and operating time >/equal 8 h. Each parameter was associated with significantly higher lactate levels and longer time for serum values to return to within normal limits.
Attributed to a combination of patient age, a typically co-morbid state often exacerbated by the pathogenesis of AOC and radicality of abdomino-pelvic surgery required to achieve complete cytoreduction in advanced disease, patients with disseminated ovarian cancer are more likely to require post-operative ICU than those undergoing surgery for other gynaecological malignancies of equivalent stage (19). ICU admission following debulking surgery for AOC is associated with increased mortality; an increased 30-day mortality of 11-27% for ovarian cancer patients admitted to ICU post-operatively has been reported (14, 15, 21, 32, 33). A further negative impact on survival has been reported in association with ICU LOS >5 days or in the context of unplanned admission (12, 16). Evidence to date has identified a number of individual factors as reliable predictors of need for post-operative ICU support following surgery for AOC (5, 14, 16, 18–20, 23, 34). One such study of 255 patients reported age >60 years, serum albumin <3.5 g/dl and CA 125 >500 U/ml as all associated with significant greater likelihood of requiring ICU post-surgery; with hypoalbuminaemia remaining independently significant on multivariate analysis (p=0.005) (19). Intra-operative variables conferring increased ICU demand in AOC include tumour volume >6 cm, concomitant lymphadenectomy or bowel resection, fluid resuscitation in excess of 5,000 ml, sub-optimal debulking, as well as both duration of surgery exceeding 200 min and end of surgery later than 1,900 h (16, 19).
Our study expands on this current evidence base in addition identifying serum lactate and lactate trend as reliable predictors of postoperative ICU, HDU and total hospital LOS following AOC surgery. Physiological lactate homeostasis is achieved by balancing production from multiple sources, including skin, red blood cells, muscle, and gastro-intestinal tract, with hepato-renal clearance, maintaining normal serum levels below 2 mmol/l (23). Hyperlactaemia occurs, however, when this balance is tipped by either increased production, decreased clearance or both (23). Two distinct mechanisms of excess lactate production have been described. The first characterised by anaerobic glycolysis, triggered in all shock states and the second, perhaps less well understood, a non-hypoxic metabolic process insidiously occurring in defined patho-physiological states, including malignancy (23). A peri-operative hyperlactaemia driven by the above processes is common and may be further compounded by the impairment of hepatic clearance induced by intra-operative hypovolaemia, hypotension, and hypothermia (23–26). Common medications, including anti-epileptics, biguanides such as metformin, salicylates including aspirin, as well the anaesthetic agent propofol can also potentially exacerbate a hyperlactaemia (23). High circulating levels of serum lactate have been shown to reduce cardiac contractility as well as limit vascular responsiveness to inotropic intervention. Hence, a high peri-operative serum lactate represents both an indirect marker of underlying pathophysiological instability or stress and a direct potentiator of an increasingly refractory cycle of cardiovascular compromise explaining the significant increase in morbidity, need for cardio-respiratory support, LOS and overall mortality risk reported in surgical patients developing hyperlactaemia (17, 23–25, 27, 28). This understanding serves to support the value of serum lactate as a guide for peri-operative decision-making and targeted goal-directed therapies aiming to both identify and promptly correct hyperlactaemia, minimising the potential for a persistent hyperlactaemia state to establish and hence the additional patient morbidity, mortality, and resource demand.
To develop this concept of modifying peri-operative risk and ICU requirement through early and pro-active management of hyperlactaemia further, we identified other peri-operative variables, in addition to the patho-physiological drivers above, appearing to directly influence lactate levels and trend specific to the context of AOC cytoreduction. Our study demonstrated high SCS, PCI score, EBL, and prolonged operating time as all associated with a significantly increased likelihood of developing a prolonged hyperlactaemia during and after surgery. Whilst the surgical complexity and PCI scores correlating with achieving R0 are non-modifiable at the time of surgery, close communication with the anaesthetic team ensures surgeon awareness of intra-operative lactate pattern. A prolonged, refractory hyperlactaemia should prompt re-evaluation of appropriateness of ongoing pursuit of complete cytoreduction versus potential need to limit radicality – particularly in cases demanding high complexity surgery for large volume, multi-site disease; and in the context of co-morbid, frail patient. EBL and operating time represent potentially more readily modifiable factors. Appropriate triaging of surgeries to specialist high volume centres, performance by surgeons with experience of cytoreductive techniques, consideration of pre-operative single-bolus tranexamic acid, intra-operative attention to meticulous haemostasis, and consideration of ‘buddy-operating’ system to counter surgeon fatigue are all measures associated with minimising blood loss, maximising surgical efficiency and hence limiting potential for surgical stress to trigger a difficult to correct hyperlactaemia and the resultant sequelae (35).
Our paper has both strengths and limitations. We report on a representative cohort of patients with AOC undergoing standard surgical techniques and approach to post-operative management, enabling reliable extrapolation of study findings to other centres. Our study represents an element of novelty, contributing to a currently limited evidence-base pertaining to the value of peri-operative serum lactate levels in the ultra-radical setting. Our population size was, however, small; and data collected retrospectively risking underpowering, as well as potential for selection, information, and recall bias. Whilst univariate analysis usefully identified significant associations, we acknowledge that it fails to account for the impact of additional confounding variables on results. Hence, further multi-centre evaluation is warranted incorporating larger patient populations, multivariate analysis, as well undertaking both survival and cost analyses to generate more robust conclusions on the clinical merit of serum lactate as a predictor of both short and long-term risk, LOS and as a useful therapeutic target in the context of AOC cytoreductive surgery.
Conclusion
Our study demonstrated serum lactate as an accessible, inexpensive, and reliable predictor of increased ICU, HDU and total hospital LOS following cytoreductive surgery for AOC. Representing a potentially modifiable marker, we suggest that serial peri-operative lactate levels are of value in this setting, guiding targeted goal-directed therapy and informing intra-operative decision-making, with the potential to optimise patient outcomes and limit service demand through the early recognition and correction of hyperlactaemia.
Footnotes
Authors’ Contributions
SA, AP: concept, literature review, data collection, statistical analysis, writing, editing. McMG, VC, HC: data collection. VA, AB, SA: editing.
Conflicts of Interest
The Authors have no conflicts of interest to declare in relation to this study.
- Received February 18, 2022.
- Revision received March 7, 2022.
- Accepted March 8, 2022.
- Copyright © 2022 International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved.