Blood loss during three column osteotomies: influence on outcomes and mitigation strategies

Introduction

Three column osteotomies (3CO) are a powerful tool for the correction of rigid sagittal spinal deformities (1). The most frequently employed 3CO techniques include pedicle subtraction osteotomies (PSO) and vertebral column resection (VCR) (2). Unlike posterior column osteotomies which rely on mobile anterior segments to achieve correction, 3CO can achieve sagittal correction of up to 25–35 degrees per level even in patients with fixed deformities (Figure 1).

Figure 1 Pre- (left) and post-operative (right) standing films following L3 pedicle subtraction osteotomy in a patient with flat back syndrome.

However, 3CO surgeries are technically demanding procedures that are associated with prolonged operative times, large amounts of intraoperative blood loss which frequently require significant volumes of blood product transfusion, and high rates of post-operative neurological and medical complications (3). The large blood loss associated with these surgeries is of particular concern as there are well established correlations between increased blood loss and adverse outcomes following spine surgery (4,5). With this in mind, significant efforts have been made to develop effective strategies for minimizing blood loss during 3CO surgeries. One commonly employed strategy to achieve this is the perioperative dosing of prothrombotic agents, with the most commonly used agent being tranexamic acid (TXA) (6,7). Despite its widespread use, debate persists regarding the overall risk-to-benefit ratio of TXA due to its associations with adverse thrombotic events such as deep vein thrombosis (DVT) and pulmonary embolism (PE) (8,9).

While the spinal deformity literature has described the effects of these various approaches with respect to minimizing blood loss, little data exists quantifying the effect of these efforts on longer-term postoperative outcomes in this population of patients undergoing high risk surgery (5). This study aimed to investigate the associations between various patient and perioperative factors on intraoperative blood loss as well as the relationship between intraoperative blood loss on postoperative outcomes, specifically length of stay (LOS), readmission, and reoperation at 30 days, 90 days, and one year in patients undergoing 3CO for deformity correction. In addition, we investigated the effects of intraoperative TXA administration on these outcomes as well as rates of DVT and PE. We present this article in accordance with the STROBE reporting checklist (available at https://jss.amegroups.com/article/view/10.21037/jss-23-143/rc).

Methods

Patients undergoing PSO or VCR between 2014 and 2021 at Pennsylvania Hospital were identified using current procedural terminology (CPT) codes. Authors A.K.O. and V.A. were the primary surgeons for all patients. Indications for surgery were reviewed and patients who underwent 3CO for infectious, oncologic, or acute traumatic indications were excluded. The final cohort contained only patients undergoing 3CO for degenerative spine disease. Only patients with at least 2 years of clinic follow-up were included. Patient records were further reviewed to extract demographic variables including gender, race, ethnicity, age at surgery, height, weight, and body mass index (BMI) at the time of surgery. Additional surgical data, including history of prior spinal surgery, number of levels fused, estimated blood loss (EBL), and intraoperative transfusion records were determined from operative notes and anesthesia records. In addition to EBL, estimated blood volume (EBV) was calculated for each patient using the Nadler equation to calculate percent EBV lost (% EBV lost = EBL/EBV × 100%). Blood loss was also quantified by determining changes in hemoglobin (Hgb) concentration from pre-operative baseline compared to both immediate post-op and post-op day one values another study has shown that percentage change in Hgb may be associated with adverse outcomes in patients undergoing spinal surgery (10). Post-operative outcomes including intensive care unit (ICU) LOS, total hospital LOS, readmissions, reoperations, and incidence of DVT/PE were also determined from chart review.

Statistical analysis

Univariable analyses, including Pearson’s correlations and T-tests for continuous characteristics and c² testing for categorical characteristics, were used to investigate relationships between variables of interest and LOS, reoperation, and readmission. To further investigate the influence of blood loss and TXA use on hospital LOS, linear regression models, including TXA use, age, operative length, and percent EBV loss with either ICU or Hospital LOS as the dependent variable were performed. All analyses were performed with SPSS version 28 (IBM, Armonk, NY, USA).

Ethical statement

The study was conducted in accordance with the Declaration of Helsinki (as revised in 2013). Given the minimal risk nature of this retrospective study, the ethical approval was deemed exempt from full review by the University of Pennsylvania Institutional Review Board. Informed consent requirements were waived due to the retrospective nature of the study.

Results

Patient demographics and surgical characteristics

Forty-two patients met inclusion criteria. Patient demographics are shown in Table 1. The average age at surgery was 61.1±11.0 years and average BMI was 29.8±6.1 kg/m²; 45% of patients were male; 84% of patients had a prior spinal fusion before undergoing 3CO. Average operative length was 581±108 minutes; 83% of 3CO were performed in the lumbar spine and 17% in the thoracic spine. L3 was the most commonly treated level (36%). The average number of levels fused was 12.8±2.9; 35.7% of patients underwent additional posterior column osteotomies, and 9.5% underwent anterior lumbar interbody fusion (ALIF) in addition to posterior fusion; 7.1% of patients underwent staged surgery. All patients undergoing staged surgery had an ALIF performed prior to undertaking PSO.

Blood loss

Average EBL, determined from anesthetic records was 2,095±750 mL. Percent EBV lost was determined to control for patient size. Average percent EBV lost was 45.1%±20.3%. Baseline Hgb was 13.7±1.7 g/dL and the average post-operative Hgb drop was 2.2±2.0 g/dL. With regards to intraoperative transfusion, the average number of packed red blood cell (pRBC) transfusions was 3.2±2.4. Total units transfused (pRBC, plasma, and platelets) averaged 4.8±3.8.

Percent EBV lost was negatively correlated with BMI (Pearson correlation: −0.369, P=0.02) and preoperative Hgb (Pearson correlation: −0.410, P=0.006) (i.e., as BMI or Hgb increased, EBV lost decreased). There was a strong positive correlation between percent EBV lost and units of pRBC transfused (Pearson correlation: 0.506, P<0.001). Although there were no statistically significant differences in EBL, % EBV lost, or ∆Hgb between patients with prior fusion compared to those without, postop Hgb was significantly higher in those with a prior fusion (11.3 vs. 96 g/dL, P=0.003).

TXA

Twenty-six patients (61.9%) received TXA intraoperatively, results are reported in Table 2; 88% received bolus plus continuous infusion, typically 5 mg/kg/h, 8% received continuous infusion only, 4% received bolus only. There were no significant differences in BMI (29.2 vs. 30.7 kg/m², P=0.45), age (59.8 vs. 63.0 years, P=0.34), length of surgery (588 vs. 571 minutes, P=0.61), or percent EBV lost (47.8% vs. 41.0%, P=0.28) between patients who received TXA and those who did not. There were no significant associations between TXA administration and readmission or reoperation at any time points. There was a trend towards fewer blood transfusions intraoperatively in patients receiving TXA that did not reach statistical significance (3.7 vs. 2.4, P=0.09). On univariate analysis, TXA administration was associated with significantly longer ICU (3.2 vs. 1.7 days, P=0.02) and hospital LOS (10.1 vs. 6.1 days, P=0.003). Linear regression showed TXA use was associated with 4.1 days increase in hospital LOS (P=0.006) while controlling for age, gender, and American Society of Anesthesiologists (ASA) score. TXA use was associated with a 2.2 day increase in ICU LOS, but this was no longer statistically significant (P=0.08) when controlling for the same variables. To further investigate this relationship between TXA use and increased LOS, rates of postoperative medical complications were compared between the TXA and no TXA groups. The overall DVT/PE rate was 15.4%. Upon stratification by intraoperative TXA use, however, no DVT/PE were seen in patients who did not receive TXA, but the DVT/PE rate was 32% in the TXA group.

Reoperation and readmission

The 30-day readmission rate was 14%. With respect to patients who were readmitted within 30 days and those who were not, there were no significant differences in EBL, percent EBV lost, preop Hgb, ∆Hgb. Similar findings were found for readmissions at 90 days and 2 years after surgery. Including patients who underwent reoperation during the index admission, 16.7% of patients underwent reoperation within 30 days. Indications for reoperation were postoperative radiculopathy requiring hardware repositioning or foraminotomy at level of PSO (n=3), mechanical wound disruption requiring revision (n=2), epidural hematoma (n=1), and abdominal exploration for chyle leak related to anterior exposure for anterior interbody fusion (n=1). There were no associations seen between metrics of intraoperative blood loss, including both EBL and % EBV lost, or blood transfusion and reoperation. These same findings were seen at 90 days and 2 years post-operatively.

Hospital and ICU LOS

Increased ICU LOS was positively correlated with increased operative length (Pearson correlation: 0.359, P=0.02), percent EBV lost (Pearson correlation: 0.356, P=0.02), and number of units of red blood cell transfusion (Pearson correlation: 0.565, P<0.001). Increased hospital LOS was associated with increased number of intraoperative blood transfusions (Pearson correlation: 0.415, P=0.006). Preoperative Hgb showed a significant negative correlation with hospital LOS (i.e., as preop Hgb increased, LOS decreased, Pearson correlation: −0.331, P=0.03).

Linear regression analyses controlling for age, gender, and ASA class showed that each 21% increase in EBV loss was associated with a one day increase in ICU LOS (B =0.047, P=0.01). To quantify the impact of blood transfusion on LOS, a similar regression analyses was performed with hospital LOS as the dependent variable and including age, gender, ASA score, and number of units of pRBC transfused as predictors. While controlling for these variables, each pRBC transfusion was associated with a 0.728 day increase in hospital LOS (Table 3).

Discussion

There is a large body of literature devoted to investigating the drivers of intraoperative blood loss as well as methods to minimize blood loss in patients undergoing spine surgery (4,5). Less data is available, however, quantifying the actual effects of excessive blood loss on perioperative outcomes such as LOS, readmission, and reoperation, especially for less commonly performed surgeries such as 3CO (11,12). In the present study we found that while there were no associations between various markers of blood loss (EBL, percent EBV lost, ∆Hgb, transfusion requirements) and readmission and reoperation, there were significant associations with blood loss and increases in both ICU and hospital LOS in patients undergoing 3CO. Increasing number of red blood cell transfusions were also associated with increased hospital and ICU LOS. In univariable analyses, increasing preoperative Hgb was associated with decreasing LOS. Furthermore, the data showed no significant differences in blood loss or transfusion requirements in patients who received TXA compared to those who did not. However, patients who received TXA had significantly greater LOS even while controlling for select confounding variables, specifically age, gender, and ASA score. Investigation of postoperative medical complications revealed that this may be driven, at least in part, by the significantly higher rates of DVT and PE in the TXA group.

Our reported metrics of blood loss during 3CO are in line with existing reports in the literature. In a multicenter study of patients undergoing 3CO for deformity correction, Bianco et al. reported average percent EBV loss of 55% which is comparable to the finding of 45% in the present study (3). Our average EBL of 2,095 mL also matches other reports in the literature of patients undergoing 3CO (6). Various studies have shown that blood transfusions are associated with increased risk of post-operative infection, thrombotic events, intraoperative complications, and, as in the present study, increased LOS, as summarized in a recent review from Zhou et al. (13). While the mechanisms behind these associations are not fully delineated, efforts to develop more restrictive transfusion practices have resulted in beneficial effects on patient outcomes and LOS (14). In the present study, it is demonstrated that not only is any pRBC transfusion correlated with worsened to outcomes, but the study also showed a linear association between increasing transfusion volume and poor outcomes. Thus, even in surgeries where transfusions are likely to be required, interventions that reduce blood loss and minimize transfusion requirements can still have beneficial effects on outcomes even without eliminating the need for transfusion. Rather than accepting the need for large transfusion volumes during 3CO, this data shows that strategies that produce even modest reductions in blood loss and transfusion requirements have the potential to provide incremental benefits in perioperative outcomes.

However, with respect to TXA use, our finding corroborates recent evidence showing a lack of effect of TXA on blood loss in patients undergoing 3CO (6). Others have shown that blood loss during 3CO is not uniformly distributed across the various stages of surgery, with the greatest proportion of blood loss occurring during the osteotomy itself (15). Given the difficulties in obtaining hemostasis due to the substantial surface area of osteotomized bone, traditional methods to facilitate hemostasis, such as the use of TXA, may prove insufficient in reducing blood loss during this stage of 3CO. These two factors may explain the lack of substantial differences in total EBL between those who received TXA and those who did not (15). In addition, although there was a trend towards decreasing transfusion requirements, this finding did not reach statistical significance. Furthermore, despite this trend, TXA use was still associated with increased LOS, indicating that it did not ameliorate the effect of blood transfusion on LOS.

In contrast with the present study, Haddad et al. reported no significant differences in LOS or hypercoagulable complications based on TXA exposure (6). Similarly, prior meta-analyses in the spine literature have not shown an association with TXA use and thrombotic complications (16), it is possible that the risk profile for patients undergoing 3CO is different from the general spine surgery population due to increased length of surgery, blood loss, and decreased level of mobility following surgery. The combination of the prothrombotic effect of TXA with the prolonged operative times, transfusion requirements, and limited postop mobility of 3CO patients may have a synergistic effect on the risk of thrombotic complications and prolonged LOS seen after 3CO.

The association between preoperative Hgb and LOS also highlights the importance of pre-operative optimization in improving post-operative outcomes. Given the retrospective and non-randomized nature of the present study, it remains to be seen if preoperative Hgb serves only as a marker for overall health status, or if the lower starting Hgb itself results in an increased need for transfusion and therefore increased LOS. Future studies may help delineate acceptable thresholds for preoperative Hgb and the role of interventions to raise this value to improve perioperative outcomes.

Limitations

This study is limited by its retrospective nature which introduces potential sources of bias. The single center nature also limited the size of the study population and has implications for the generalizability of the present findings. The small size also prevents more robust controls for confounding factors and other sources of unmeasured bias in the sample. While some reported outcomes were in-line with the existing literature, the discrepancy between the present study and the existing literature with respect to post-operative thrombotic complications should also be interpreted carefully. Finally, the single-center, retrospective nature may have prevented the capture of additional reoperation or readmission events that occurred outside of our health system. Given the challenges in studying less frequently performed surgeries like 3CO, future multi-institutional studies may be performed to address concerns about generalizability while also providing sufficient statistical power to delineate the complex interrelationships between blood loss, blood transfusion, and TXA use on perioperative outcomes after 3CO.

Conclusions

In a retrospective study of patients undergoing 3CO, we found that increasing blood loss and increasing transfusion requirements were associated with prolonged LOS. TXA use did not result in decreased blood loss, but was associated with increased rates of DVT, PE, and LOS. 3CO are powerful procedures to achieve significant correction, but significant work remains to be done to reduce the high rates of complications associated with these procedures.

Acknowledgments

Funding: None.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://jss.amegroups.com/article/view/10.21037/jss-23-143/rc

Data Sharing Statement: Available at https://jss.amegroups.com/article/view/10.21037/jss-23-143/dss

Peer Review File: Available at https://jss.amegroups.com/article/view/10.21037/jss-23-143/prf

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://jss.amegroups.com/article/view/10.21037/jss-23-143/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki (as revised in 2013). Given the minimal risk nature of this retrospective study, the ethical approval was deemed exempt from full review by the University of Pennsylvania Institutional Review Board. Informed consent requirements were waived due to the retrospective nature of the study.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Hyun SJ, Kim YJ, Rhim SC. Spinal pedicle subtraction osteotomy for fixed sagittal imbalance patients. World J Clin Cases 2013;1:242-8. [Crossref] [PubMed]
2. Ferrero E, Liabaud B, Henry JK, et al. Sagittal alignment and complications following lumbar 3-column osteotomy: does the level of resection matter? J Neurosurg Spine 2017;27:560-9. [Crossref] [PubMed]
3. Bianco K, Norton R, Schwab F, et al. Complications and intercenter variability of three-column osteotomies for spinal deformity surgery: a retrospective review of 423 patients. Neurosurg Focus 2014;36:E18. [Crossref] [PubMed]
4. Hu SS. Blood loss in adult spinal surgery. Eur Spine J 2004;13:S3-5. [Crossref] [PubMed]
5. Mikhail C, Pennington Z, Arnold PM, et al. Minimizing Blood Loss in Spine Surgery. Global Spine J 2020;10:71S-83S. [Crossref] [PubMed]
6. Haddad AF, Ames CP, Safaee M, et al. The Effect of Systemic Tranexamic Acid on Hypercoagulable Complications and Perioperative Outcomes Following Three-Column Osteotomy for Adult Spinal Deformity. Global Spine J 2022;12:423-31. [Crossref] [PubMed]
7. Smith JS, Shaffrey CI, Klineberg E, et al. Complication rates associated with 3-column osteotomy in 82 adult spinal deformity patients: retrospective review of a prospectively collected multicenter consecutive series with 2-year follow-up. J Neurosurg Spine 2017;27:444-57. [Crossref] [PubMed]
8. Myers SP, Kutcher ME, Rosengart MR, et al. Tranexamic acid administration is associated with an increased risk of posttraumatic venous thromboembolism. J Trauma Acute Care Surg 2019;86:20-7. [Crossref] [PubMed]
9. Nishihara S, Hamada M. Does tranexamic acid alter the risk of thromboembolism after total hip arthroplasty in the absence of routine chemical thromboprophylaxis? Bone Joint J 2015;97-B:458-62. [Crossref] [PubMed]
10. Purvis TE, Goodwin CR, Molina CA, et al. Percentage change in hemoglobin level and morbidity in spine surgery patients. J Neurosurg Spine 2018;28:345-51. [Crossref] [PubMed]
11. Raad M, Puvanesarajah V, Harris A, et al. The learning curve for performing three-column osteotomies in adult spinal deformity patients: one surgeon's experience with 197 cases. Spine J 2019;19:1926-33. [Crossref] [PubMed]
12. Lau D, Haddad AF, Joshi RS, et al. Cardiac Complications After 3-Column Osteotomy for Adult Spinal Deformity Patients With Formal Cardiac Clearance: Identifying Key Risk Factors and Threshold Cutoffs Via Recursive Partitioning Analysis. Neurosurgery 2022;91:562-9. [Crossref] [PubMed]
13. Zhou JJ, Hemphill C, Walker CT, et al. Adverse Effects of Perioperative Blood Transfusion in Spine Surgery. World Neurosurg 2021;149:73-9. [Crossref] [PubMed]
14. Alfonso AR, Hutzler L, Lajam C, et al. Institution-Wide Blood Management Protocol Reduces Transfusion Rates Following Spine Surgery. Int J Spine Surg 2019;13:270-4. [Crossref] [PubMed]
15. Qiao J, Liu Z, Xu L, et al. The pattern of blood loss during correction surgery combined with three column osteotomy for adult spine deformity. Preprint Research Square 2019 [cited 2023 Feb 23]. Available online: 10.21203/rs.2.12777/v110.21203/rs.2.12777/v1
16. Zhang F, Wang K, Li FN, et al. Effectiveness of tranexamic acid in reducing blood loss in spinal surgery: a meta-analysis. BMC Musculoskelet Disord 2014;15:448. [Crossref] [PubMed]