The predictive value of preoperative frailty screening for postoperative outcomes in older patients undergoing surgery for non-metastatic colorectal cancer

Abstract

In patients with colorectal cancer (CRC), surgical resection is the mainstay of treatment. Older patients are at an increased risk for postoperative complications due to a higher prevalence of frailty, which can be defined as the aging-associated decline in functioning across organ systems that leads to a higher vulnerability towards stressors (such as surgery) [[1]Xue Q.L. The frailty syndrome: definition and natural history.Clin Geriatr Med. 2011; 27: 1-15https://doi.org/10.1016/j.cger.2010.08.009Abstract Full Text Full Text PDF PubMed Scopus (936) Google Scholar]. Measuring frailty next to standard risk screening may aid in risk prediction and guide the implementation of preoperative interventions such as prehabilitation (preoperative physical optimization). There is currently no universally implemented frailty screening tool in colorectal surgery. One of the screening instruments that has been proposed is the Groningen Frailty Indicator (GFI) [[2]Stevernik N. Slaets J. Schuurmans H. van Lis M. Steverink N. Slaets J.P.J. et al.Measuring frailty: development and testing the GFI (Groningen frailty indicator).Gerontologist. 2001; 41: 236-237Google Scholar], which is a questionnaire that addresses impairments in daily activities, health problems and psychosocial functioning. A previous study in patients with stage I-IV CRC showed that GFI was associated with postoperative sepsis [[3]Reisinger K.W. van Vugt J.L.A. Tegels J.J.W. Snijders C. Hulsewe K.W.E. Hoofwijk A.G.M. et al.Functional compromise reflected by sarcopenia, frailty, and nutritional depletion predicts adverse postoperative outcome after colorectal cancer surgery.Ann Surg. 2015; 261: 345-352https://doi.org/10.1097/SLA.0000000000000628Crossref PubMed Scopus (319) Google Scholar]. The aim of this study was to assess the predictive value of the GFI next to standard preoperative variables for postoperative adverse outcomes in older patients undergoing surgery for CRC. For this retrospective cohort study, patients ≥70 years were eligible for inclusion if they had undergone abdominal surgery for CRC between 2014 and 2017 in Isala Hospital, Zwolle, The Netherlands. Patients who had metastatic cancer, two or more colorectal tumors, concurrent malignancy for which they had received treatment within the previous five years, or who did not have preoperative frailty data available were excluded. The local Medical Research Ethics Committee confirmed that the study was not subject to the Dutch Medical Research Involving Human Subjects Act and requirement for informed consent was waived. The Institutional Review Board approved of this study. The following data were collected from electronic medical records (EMR): age, sex, American Society of Anesthesiologists (ASA) -classification [[4]Owens W.D. Felts J.A. Spitznagel E.L. ASA physical status classifications: a study of consistency of ratings.Anesthesiology. 1978; https://doi.org/10.1097/00000542-197810000-00003Crossref PubMed Scopus (1567) Google Scholar], body mass index (BMI), Charlson Comorbidity Index [[5]Charlson M.E. Pompei P. Ales K.L. MacKenzie C.R. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation.J Chronic Dis. 1987; 40: 373-383Abstract Full Text PDF PubMed Scopus (33755) Google Scholar], preoperative anemia (hemoglobin <8.5 for males and < 7.5 for females), location of tumor (colon vs rectum), tumor stage (American Joint Committee on Cancer [AJCC] classification [[6]Edge S. Byrd D. Compton C. Fritz A. Greene F. Trotti A. AJCC cancer staging manual.7th ed. Springer, France2010Google Scholar]), urgency of surgery (elective vs urgent/acute), surgical approach (laparoscopic vs open/conversion), whether a stoma was constructed, and whether (neo)adjuvant therapy was administered. The GFI-questionnaire was administered by a nurse during a preoperative outpatient visit and entered into the EMR. The GFI consist of questions relating to impairments in daily activities including needing help with shopping, walking around, (un)dressing or going to toilet (four questions), health issues including self-reported poor physical fitness, weight loss, polypharmacy, memory problems, or vision or hearing problems (six questions), and psychosocial functioning including feelings of emptiness, loneliness, abandonment, depression or anxiety (five questions) (Supplementary Table S1). Every item can provide one point, resulting in a score ranging from 0 to 15. A score of ≥4 was considered indicative of frailty [[2]Stevernik N. Slaets J. Schuurmans H. van Lis M. Steverink N. Slaets J.P.J. et al.Measuring frailty: development and testing the GFI (Groningen frailty indicator).Gerontologist. 2001; 41: 236-237Google Scholar]. The primary outcome was overall 30-day postoperative complications. Secondary outcomes were Comprehensive Complication Index [[7]Slankamenac K. Graf R. Barkun J. Puhan M.A. Clavien P.A. The comprehensive complication index: a novel continuous scale to measure surgical morbidity.Ann Surg. 2013; 258: 1-7https://doi.org/10.1097/SLA.0b013e318296c732Crossref PubMed Scopus (777) Google Scholar], delirium, (prolonged) length of stay (LOS) (≥15 days), new post-discharge institutionalization (nursing home/rehabilitation) if living at home before surgery, and 30-day mortality. Delirium was considered present if an explicit diagnosis was made by the treating physician or a geriatrician or if haloperidol was prescribed. Baseline characteristics, treatments and outcomes between frail and non-frail patients were compared with chi squared test, Fisher's exact test, t-test, or Mann-Whitney U test. To assess the relationship between frailty and study outcomes, multivariable logistic regression models were constructed with frailty and potential confounder variables that were associated with frailty and the outcome in univariable analysis (p < 0.20). The variables that were considered were age, sex (male vs female), BMI (continuous), ASA-score (III-IV vs I-II), Charlson Comorbidity Index (score ≥ 2 vs 0–1), preoperative anemia, tumor location (colon vs rectum), and tumor stage (AJCC III vs I-II). Potential confounders were added to the multivariable model in a forward stepwise fashion starting with the variable that had the strongest association with the outcome. The variable was retained in the multivariable model if the beta coefficient of the central determinant (frailty) changed by more than 10%. The results of the logistic regression analyses were reported with odds ratios (OR) with 95% confidence intervals (CI). A two-tailed p-value <0.05 was considered statistically significant. All analyses were performed with SPSS version 27.0. The flow chart of patient inclusion is depicted in Fig. 1. A total of 425 patients ≥70 years underwent abdominal surgery for CRC during the inclusion period. Of these, 194 were excluded according to the prespecified criteria. The final study sample consisted of 231 patients. Forty-four (19%) patients were frail. For these patients, the median GFI score was 5 (interquartile range [IQR] 4–7). For non-frail patients, the median GFI score was 1 (IQR 1–2). Baseline characteristics and study outcomes between frail and non-frail patients are summarized in Table 1. Frail patients were older, had higher Charlson Comorbidity Index scores, and higher ASA-scores. Frail patients more often had a colon tumor and more often underwent open surgery.Table 1Comparison of baseline and treatment characteristics and study outcomes between frail and non-frail patients according to Groningen Frailty Indicator.TotalN = 231Non-frailN = 187FrailN = 44p-valueBaseline characteristicsFemale sex128 (55)98 (52)30 (68)0.06Age, median (IQR)76 (73–81)75 (72–80)81 (77–84)<0.001BMI, median (IQR) (missing: 1)25.8 (23.8–28.9)26.1 (24.1–29.0)24.7 (23.2–28.5)0.13ASA-score (missing: 33)I-II154 (78)136 (83)18 (53)<0.001III-IV44 (22)28 (17)16 (47)Charlson Comorbidity Index0–1167 (72)144 (77)23 (52)<0.001≥264 (28)43 (23)21 (48)Preoperative anemia135 (58)104 (56)31 (71)0.07Pathological tumor stage (AJCC)0.18I-II166 (72)138 (74)28 (64)III65 (28)49 (26)16 (36)Tumor locationColon168 (73)128 (68)40 (91)0.003Rectum63 (27)59 (32)4 (9)Living alone (missing: 4)86 (38)61 (33)25 (57)0.004Living situationIndependent225 (97)186 (99)39 (89)0.001Nursing home6 (3)1 (1)5 (11)Treatment characteristicsUrgent surgery13 (6)10 (5)3 (7)0.70Surgical approachOpen83 (36)61 (33)22 (50)0.03Laparoscopic148 (64)126 (67)22 (50)Primary anastomosis200 (87)162 (87)38 (86)0.96Ostomy54 (23)47 (25)7 (16)0.19Neoadjuvant therapy38 (17)34 (18)4 (9)0.14Adjuvant therapy29 (13)24 (13)5 (11)0.79Postoperative outcomesOverall complications134 (58)101 (54)33 (75)0.01Comprehensive Complication Index8.7 (0.0–22.6)8.7 (0.0–20.9)17.9 (2.2–29.0)0.008Delirium10 (4)4 (2)6 (14)0.004LOS, days, median (IQR)6 (4–8)5 (4–8)8 (6–12)<0.001LOS >14 days21 (9)13 (7)8 (18)0.04New post-discharge institutionalization *Based on 225 patients living independently before surgery.18 (8)6 (3)12 (27)<0.00130-day mortality3 (1)3 (2)0 (0)1.00Results are shown as number and percentage unless otherwise indicated. Significant results (p < 0.05) are shown in boldface.Abbreviations: AJCC American Joint Committee on Cancer; ASA American Society of Anesthesiologists; BMI body mass index; IQR interquartile range; LOS length of stay. Based on 225 patients living independently before surgery. Open table in a new tab Results are shown as number and percentage unless otherwise indicated. Significant results (p < 0.05) are shown in boldface. Abbreviations: AJCC American Joint Committee on Cancer; ASA American Society of Anesthesiologists; BMI body mass index; IQR interquartile range; LOS length of stay. Complications occurred in 134 (58%) patients. Frail patients were more likely to have complications (75% vs 54%, p = 0.01). In multivariable analysis adjusted for age and sex, frailty remained an independent predictor of complications (OR 2.6, 95% CI 1.2–5.6, p = 0.02). Regarding the overall complication burden, the median Comprehensive Complication Index was 8.7 (IQR 0.0–22.6). Frail patients had a higher burden of complications than non-frail patients (median 17.9 vs 8.7, p = 0.008). Ten (4%) patients were diagnosed with postoperative delirium. Frail patients had a higher risk for developing delirium (14% vs 2%, p = 0.004; OR 7.2, 95% CI 1.9–26.8, p = 0.003). Of the other preoperative predictors, only increasing age was also associated with a higher risk of delirium (OR 1.2, 95% CI 1.1–1.4, p = 0.007). No multivariable analysis was performed due to the low number of delirium cases. The median LOS was six days (IQR 4–8). Frail patients were admitted for three days longer than non-frail patients (median 8 vs 5 days, p < 0.001). Frail patients were at a higher risk for LOS ≥15 days (18% vs 7%, p = 0.04). In multivariable analysis adjusted for Charlson Comorbidity Index, the association did not remain statistically significant (OR 2.5, 95% CI 0.9–6.6, p = 0.07). Of the 225 patients living independently before surgery, eighteen (8%) were discharged to nursing home or rehabilitation center. Frail patients were more likely to have this outcome (27% vs 3%, p = 0.006). In multivariable analysis adjusted for age, frailty remained independently associated with post-discharge institutionalization (OR 9.3, 95% CI 3.0–28.2, p < 0.001). In this cohort study in patients ≥70 years undergoing surgery for CRC, preoperative frailty as measured with the GFI questionnaire was an independent predictor of overall complications and post-discharge institutionalization. The higher rate of complications in frail patients may be a reflection of their lower physical capacity to recover from major surgery. Complications can arise in all patients regardless of physical resilience, but frail patients may be more prone for developing subsequent complications (e.g., pneumonia followed by acute renal failure or anastomotic leak complicated by subsequent delirium). This may explain the higher complication burden in the frail group. Furthermore, frail patients were seven times more likely to develop postoperative delirium. Frailty and age were the only preoperative variables associated with delirium, underscoring the importance of frailty screening as part of risk prediction. Being able to remain functionally independent after surgery is a major goal for older patients [[8]Mohile S.G. Hurria A. Cohen H.J. Rowland J.H. Leach C.R. Arora N.L. et al.Improving the quality of survivorship for older adults with cancer.Cancer. 2016; 122: 2459-2568https://doi.org/10.1016/j.physbeh.2017.03.040Crossref PubMed Scopus (454) Google Scholar]. In this study, frail patients had a nine-fold higher risk for post-discharge institutionalization. A more complicated postoperative course likely contributed to the higher number of admissions to a care facility in the frail group. The high risk of functional decline as a consequence of surgery should be discussed with frail patients during the preoperative outpatient visit. Frail patients might benefit from preoperative interventions aimed at improving their resilience before surgery [9Bruns E.R.J. Argillander T.E. Schuijt H.J. van Duijvendijk P. van der Zaag E.S. Wassenaar E.B. et al.Fit4SurgeryTV at-home prehabilitation for frail older patients planned for colorectal cancer surgery: a pilot study.Am J Phys Med Rehabil. 2019; 98: 399-406https://doi.org/10.1097/PHM.0000000000001108Crossref PubMed Scopus (22) Google Scholar, 10Carli F. Baldini G. From preoperative assessment to preoperative optimization of frail older patiens.Eur J Surg Oncol. 2021; 47: 519-523https://doi.org/10.1016/j.ejso.2020.06.011Abstract Full Text Full Text PDF PubMed Scopus (10) Google Scholar, 11Minnella E.M. Awasthi R. Gillis C. Fiore Jr., J.F. Liberman A.S.S. Charlebois P. et al.Patients with poor baseline walking capacity are most likely to improve their functional status with multimodal prehabilitation.Surgery. 2016; 160: 1070-1079https://doi.org/10.1016/j.surg.2016.05.036Abstract Full Text Full Text PDF PubMed Scopus (103) Google Scholar]. These interventions can range from preoperative adjustments arising from the Comprehensive Geriatric Assessment [[12]Ellis G. Gardner M. Tsiachristas A. Langhorne P. Burke O. Harwood R.H. et al.Comprehensive geriatric assessment for older adults admitted to hospital.Cochrane Database Syst Rev. 2017; 2017https://doi.org/10.1002/14651858.CD006211.pub3Crossref Scopus (89) Google Scholar] to prehabilitation programs. Previous studies have suggested that prehabilitation involving physical exercise and nutritional enhancement may prevent complications in frail patients undergoing CRC surgery [[13]van der Hulst H.C. Bastiaannet E. Portielje J.E.A. van der Bol J.M. Dekker J.W.T. Can physical prehabilitation prevent complications after colorectal cancer surgery in frail older patients?.Eur J Surg Oncol. 2021; https://doi.org/10.1016/j.ejso.2021.05.044Abstract Full Text Full Text PDF Scopus (6) Google Scholar,[14]De Klerk M. Van Dalen D.H. Nahar-Van Venrooij L.M.W. Meijerink W.J.H.J. Verdaasdonk E.G.G. A multimodal prehabilitation program in high-risk patients undergoing elective resection for colorectal cancer: a retrospective cohort study.Eur J Surg Oncol. 2021; https://doi.org/10.1016/j.ejso.2021.05.033Abstract Full Text Full Text PDF PubMed Scopus (9) Google Scholar]. This study has some limitations. We performed a relatively large number of (exploratory) analyses, and the conclusions should therefore be regarded as hypothesis-generating. Future (prospective) studies are needed to confirm the findings of the present study. In addition, the predictive value of GFI should be compared with other frailty instruments for adverse postoperative outcomes in colorectal surgery. It should also be noted that the study cohort likely has a selection bias towards the fittest individuals as frailest patients are more often offered non-surgical therapy (especially in the setting of rectal cancer [[15]Montroni I. Ugolini G. Saur N.M. Spinelli A. Rostoft S. Millan M. et al.Personalized management of elderly patients with rectal cancer: expert recommendations of the European Society of Surgical Oncology, European Society of Coloproctology, International Society of Geriatric Oncology, and American College of Surgeons Commission on Cancer.Eur J Surg Oncol. 2018; 44: 1685-1702https://doi.org/10.1016/j.ejso.2018.08.003Abstract Full Text Full Text PDF PubMed Scopus (79) Google Scholar]). Furthermore, GFI was a part of preoperative work-up, meaning that frail patients were referred to a geriatrician more often and likely received additional preoperative interventions. This may have led to an underestimation of the association between frailty and postoperative outcomes. In older patients undergoing surgery for CRC, GFI-frailty was associated with postoperative complications and new institutionalization. Frailty screening with GFI can potentially aid in shared decision-making in the preoperative phase and guide the implementation of preoperative interventions such as prehabilitation.