Inuence of Hypertension and Diabetes on Postoperative Outcomes In Cancer Patients Undergoing Moderate/High Risk Surgical Procedures

Background: Systemic hypertension (HTN) and diabetes mellitus (DM) are believed to be risk factors for adverse postoperative outcomes in patients undergoing surgical interventions, but evidence is lacking. This retrospective study evaluated the effects of HTN and DM, alone or in combination, on postoperative outcomes of elective noncardiac surgery in cancer patients. Methods: Patients (n = 844) with malignancies, who underwent elective surgery at a tertiary hospital, were categorised into healthy (group A, n = 339), hypertensive (group B, n = 357), diabetic (group C, n = 21), and hypertensive and diabetic (group D, n = 127) groups. Preoperatively, all patients had systolic blood pressure ≤ 160 mmHg and plasma glucose level ≤ 140 mg/dl. Postoperative in-hospital morbidity and mortality were compared among groups. Results: Postoperative complications occurred in 22 (6.5%), 21 (5.9%), 2 (9.5%), and 11 (8.7%) patients in groups A, B, C, and D, respectively (p = 0.712). HTN (p = 0.538), DM (p = 0.990), and HTN+DM (p = 0.135) did not impact the occurrence of adverse events. Patients with higher surgical risk (ASA III or IV) and those with longer surgical time had higher morbidity and mortality (p = 0.001, p < 0.001, respectively). In multiple logistic regression analysis, ASA status and surgical time were independent risk factors for postoperative complications (both p < 0.001). Conclusion: Cancer patients with preoperative comorbidities, such as HTN and DM, alone or in combination, regardless of other characteristics, do not have an increased risk of adverse postoperative outcomes. Trial registration: Retrospectively registered.


Introduction
Investigations of the occurrence of postoperative complications requires a specialised approach, due to aspects related to both the patient's well-being and quality of life, as well as to the operational and nancial management of the hospital. Although the effect of such complications on the patient is at rst seen a matter of overriding importance, the impact on hospital management also requires special attention, primarily in terms of strategies that seek to reduce the incidence of such complications, and aiming to minimise the costs resulting from the increased hospitalisation duration, to optimise diagnostic and therapeutic resources, and to manage beds. With this concern in mind, a "triple aim" concept for improving health care has been suggested, including improving the individual experience of care, improving the health of populations, and reducing the per capita costs of care for populations [1].
The number of non-cardiac surgeries performed around the world and the cost of these procedures has increased considerably in recent years. This has prompted attempts to prevent the occurrence of adverse postoperative outcomes [2]. In this context, ensuring adequate preoperative evaluation allows the institution of measures to reduce the possible risks arising from the surgical procedure, and consequently the perioperative morbidity and hospital stay [3].
Chronic diseases, such as diabetes mellitus (DM) and systemic hypertension (HTN), have a high prevalence in the general population and, when not adequately controlled, are reported to be related to an increased risk of postoperative complications [4][5][6]. This comorbidity is often associated with a decline in functional reserves and an increasing prevalence of frailty. Thus, patients with comorbidities are more likely to have suboptimal outcomes and experience more treatment-related toxicities. The impact of comorbidity is more evident in cancer patients [7]. A cohort study has demonstrated that patients with comorbidities have an increased risk of all-cause mortality after curative cancer surgery, as compared to patients without comorbidities, but analysed only patients with oesophageal cancer [8].
Although reports are limited, it has been believed that cardiovascular comorbidities can in uence surgical outcomes, but no study was conducted with the primary endpoint of demonstrating the relationship between HTN or DM with postoperative outcomes in cancer patients. Such information is highly relevant for estimating the surgical risk of this population, and for optimizing pre-, peri-, and postoperative care [9,10].However, there is a lack of evidence of the presumed association of HTN and/or DM with the occurrence of postoperative complications.
Therefore, the present study sought to elucidate the in uence of HTN and DM, alone or in combination, on the in-hospital outcomes of cancer patients undergoing non-cardiac surgery that are considered to be of moderate/high risk, at a tertiary cancer center.

Study population
We conducted a single-centre, retrospective cohort study involving all patients who underwent elective non-cardiac surgery of moderate/high risk screened at a preoperative assessment clinic. Data were extracted from electronic clinical records. Surgical time was de ned as the time from the moment of the incision until skin closure, excluding the anaesthesia time. Adverse outcomes included any type of complication (clinical and/or surgical) that increased the length of hospital stay for each procedure performed, or death. Postoperative outcomes were studied individually and jointly as "morbimortality".

Ethical aspects
The IBCC Oncologia Research Ethics Board in Sao Paulo, Brazil, reviewed and approved this study. Data con dentiality was maintained at all times, in accordance with Brazilian ethical regulation.

Statistical analyses
Qualitative characteristics were summarised using absolute and relative frequency, according to adverse intra-hospital outcomes. Pearson's chi-square test was employed as appropriate. Quantitative characteristics were described according to adverse in-hospital outcomes, using summary measures (mean, standard deviation, median, minimum, and maximum) and compared using Student's t-tests or Mann-Whitney tests.
Bivariate logistic regression analyses were used to adjust for multiple predictive factors and their interactions. Odds ratios (ORs) and 95% con dence interval (95%CIs) were used to quantify the relationship between the outcomes of interest and each independent factor, while multiple logistic regression analysis was used to analyse the in uence of parameters on the occurrence of adverse outcomes.

Statistical analyses were conducted in Statistical Package for the Social Sciences (SPSS) for Windows
Software Version 20.0 (IBM SPSS Inc., Armonk, NY, USA). All statistical assessments were two-sided. Probability (p)-values below 0.05 were considered statistically signi cant.

Results
The cohort included 844 patients undergoing elective surgeries. Figure 1 shows the number of patients excluded from the study at each stage of analysis.
Baseline patient characteristics are summarised in Table 1. The cohort included 116 men and 728 women. Patients in group D were signi cantly older than those in group A (p < 0.001) and B (p < 0.001), and those in group B were older than those in group A (p < 0.001). Patients with comorbidities had more ECG (p = 0.002), radiological (p < 0.001), and laboratory (p < 0.001) abnormalities. There were no signi cant differences among groups in terms of ASA surgical risk (p = 0.554) and the duration of surgery (p = 0.396).   (Table 3).  To determine whether the presence of comorbidities has a direct in uence on preoperative surgical risk (ASA status), we analysed the effects of the presence of these comorbidities on in-hospital outcomes (morbidity and mortality), excluding the surgical risk. The presence of HTN and DM alone or in combination did not have a signi cant effect on adverse events when surgical risk was excluded (Table 4). Four deaths (0.47%) were observed in the studied population. Three deaths resulted from postoperative clinical complications and one was died due to unsuccessful surgery in three sequential approaches. One death occurred in a patient from the healthy group, two in patients from the hypertensive group, and one in a patient from the hypertensive and diabetic group. There were no deaths in the diabetic group. The mortality rate was not signi cantly different among groups (p = 0.864). Due to the small number of deaths, they are presented together with adverse events under the name of mortality and morbidity in Table 2.
In patients with normal ECG, there were 42 adverse events, as compared to 14 adverse events among patients with at least one signi cant ECG change (p = 0.254; Table 2). Abnormal chest radiography was noted in 13.4% patients, of which 10.4% had adverse outcomes, as compared to 5.8% of the group with normal X-ray (p = 0.075). One hundred sixty-six patients (19.7%) had at least one altered laboratory test, according to the pre-established criteria. Of these 11.4% had some postoperative complication, as compared to 5.6% of those who had completely normal preoperative examinations (Table 2; p = 0.007).
Patients who were classi ed as surgical risk ASA III/IV had a higher risk of postoperative complications than those classi ed as ASA I/II (Table 2; p = 0.001). Patients with operative time exceeding 5 hours had signi cantly more in hospital postoperative complications (Table 2; p < 0.001). Multiple logistic regression analysis also con rmed that surgical risk and operative time (Table 3; both p < 0.001) were independent predictors for postoperative adverse events.

Discussion
This study evaluated the effects of HTN and DM, alone or in combination, on the postoperative outcomes of elective noncardiac surgery in cancer patients. Our ndings suggest that patients with preoperative comorbidities, such as HTN and DM, alone or in combination, do not have an increased risk of adverse postoperative outcomes. Rather preoperative risk (ASA status) and operative time affected the risk of postoperative adverse events.
Preoperative clinical evaluation of patients undergoing elective surgery can be performed for multiple purposes, including for detecting possible disorders that may compromise a satisfactory postoperative course [11]. Through medical history and physical examination, pre-existing or unknown diseases that may increase the risk of complications during the perioperative period can be detected. HTN and DM have a high prevalence among the population. In patients with neoplasms, the prevalence of HTN is around 40%, reaching 74% in the group aged over 65 years [6,[12][13][14][15]. Similarly, in this population, DM has a prevalence of 21% [16].
It has been considered that the presence of HTN or DM alone, or both pathologies together, may increase the likelihood of postoperative complications after non-cardiac surgery, as these are risk factors for cardiovascular disease [17][18][19]; Our results indicated that patients with HTN and DM, alone or in combination, that is well-controlled presurgically, does not impact the postoperative outcomes of cancer patients undergoing non-cardiac intermediate/high-risk surgical interventions, as compared to patients without DM or HTN. A recent study [20] that evaluated 203 patients in a mixed population (with and without cancer), but with risk factors for cardiovascular complications, who underwent major non-cardiac thoracic surgery also found no association between the presence of HTN or DM with adverse outcomes in the postoperative period.
Abnormalities on chest radiography and ECG were not associated with an increase in morbidity and mortality in our study, corroborating ndings by previous studies that showed that requesting these preoperative tests for asymptomatic patients offers little or no bene t [21][22][23]. In elderly individuals, chest X-ray may be abnormal, but without necessarily increasing the surgical risk [21]. There are also doubts in the literature regarding the bene t of using ECG in the preoperative period in asymptomatic patients, considering that the incidence of abnormal results increases with age and may not be related to increased adverse outcomes [22,23].
In the composition of surgical risk, the presence of comorbidities is relevant. In our study, preoperative risk assessment was categorised into ASA I/II and ASA III/IV risk status. We noted that the presence of a higher surgical risk was an independent predictor of postoperative complications, which has been established previously. This indicates that we had suitably adjusted for confounding factors. In addition, we performed multivariate logistic regression analysis with and without including this surgical risk variable; the presence of these comorbidities did not correlate with an increase in the number of adverse events, irrespective of inclusion of ASA status.
Most procedures (70%) had an operative time < 3 hours; patients with a prolonged surgical time (> 5 hours) had signi cantly more adverse outcomes, as corroborated by previous studies [24,25]. Early recognition and intervention for any perioperative complications may help these patients and also reduce health care costs.
Our study had some limitations that were mainly inherent to the retrospective analysis of administrative databases. The study was conducted retrospectively at a single centre, which might limit the generalisation of its results. Additional limitations include that the vast majority of procedures had an effective duration of surgery of no more than 3 hours, and knowing that the surgical time is a predictor of postoperative complications, it is not clear whether these comorbidities would become signi cant in longer surgeries. In addition, most patients in this study were female. Given the scarcity of studies seeking to correlate the characteristics inherent to patients in the preoperative period with the outcomes in the postoperative period, further research is needed improve understanding of the impact of these comorbidities in different populations of individuals submitted to more diverse surgical procedures.
In summary, with the use of a representative database, we found that, in a population of cancer patients undergoing elective moderate/high risk surgeries, a previous diagnosis of HTN and DM, alone or combined (properly controlled), did not correlate with an increase in in-hospital morbidity and mortality.