Abstract
Study Design
Retrospective Cohort Study.
Objectives
To determine the incidence of symptomatic retroperitoneal fluid collections (RFCs) following anterior lumbar interbody fusion (ALIF), characterize clinical presentation, and identify independent predictors and associated complications.
Methods
Between 2014 and 2023, adult patients who underwent 1- to 4-level primary or revision ALIF at a single academic university-affiliated hospital were retrospectively reviewed for postoperative RFCs. Asymptomatic RFCs, <1 year follow-up, and surgical indications for trauma, malignancy, or infection were excluded. Symptomatic RFCs, including lymphoceles, hematomas/seromas, abscesses, and urinomas, were identified on postoperative MRI with final classification based on fluid aspiration when available. Multivariable logistic regression was performed to identify independent predictors of RFCs and associations with postoperative complications.
Results
Among 553 included patients, 44 (7.97%) developed symptomatic postoperative RFCs, most commonly seromas, presenting with abdominal pain and distension. Multi-level surgeries (aOR: 4.472), estimated blood loss ≥725 mL (aOR: 4.213), intraoperative transfusion (aOR: 3.347), L4-5 fusion (aOR: 3.186), higher ASA class (aOR: 2.291), and venous injury (aOR: 2.051) (all P < 0.05) were independent predictors of RFCs. L3-4 fusion provided a protective effect (aOR: 0.163). RFCs were associated with increased risk of subsequent deep vein thrombosis (DVT; aOR: 4.087, P = 0.007) and incision and drainage (aOR: 9.593, P = 0.005).
Conclusions
Symptomatic RFCs occurred postoperatively in 7.97% of ALIF cases. Multi-level fusion, blood loss ≥725 mL, intraoperative transfusion, L4-5 fusion, higher ASA class, and venous injury were independently associated with RFC development. Notably, RFCs increased risk of postoperative DVT by 4 times, underscoring the benefit of early recognition in high-risk patients.
Introduction
Anterior lumbar interbody fusion (ALIF) is widely used at the lower lumbar levels, where the anterior retroperitoneal approach enables thorough disc-space preparation and placement of large lordotic implants while minimizing posterior muscle dissection.1,2 The approach presents specific advantages and limitations distinct from other fusion strategies. Inherent to the anterior, retroperitoneal exposure are unique complications such as venous, arterial, lymphatic, peritoneal, or ureteral damage, or injury to the sympathetic hypogastric plexus leading to retrograde ejaculation in men.3-6 A 2024 meta-analysis of 8066 procedures documented a complication rate of 13.1%, and a 2025 U.S. database audit of 18,000 lumbar fusions demonstrated comparable figures. 7 Within this spectrum, retroperitoneal fluid collections (RFCs)–lymphoceles, hematomas, seromas, abscesses, and urinomas–have emerged as a clinically meaningful, yet understudied morbidity.
Current studies focus on the incidence of individual RFCs by etiology. A large single-center imaging series found a 2.1% incidence of retroperitoneal lymphocele following ALIFs. 4 Hematomas and seromas, although less frequent, carry significant consequences. Contemporary national and institutional datasets cite a 0.5-0.9% incidence of retroperitoneal hematomas following an anterior case, which have been associated with higher odds of readmission and prolonged hospital stays.8,9 Seromas may manifest months post-operatively and require percutaneous or surgical evacuation. 10 Hyper-coagulability, complex exposures, and prolonged operative time are risks, emphasizing the role of patient selection and meticulous hemostasis.8,11 Sterile accumulations may provoke ileus, radicular pain, or deep-vein thrombosis, while infected body cavities threaten the underlying fusion construct, underscoring the need for precise incidence reporting and understanding risk factors for RFC development.3,11,12
Current knowledge remains fragmentary because most studies pool heterogeneous anterior and lateral approaches, focus solely on one etiology of RFC, or include asymptomatic patients. 4 Studies looking at the incidence of all etiologies of RFCs following ALIF remain rare. Accordingly, a retrospective cohort analysis of ALIFs at a single high-volume academic spine center was performed. The objectives were to (1) determine the incidence of radiologically confirmed symptomatic retroperitoneal fluid collections; (2) analyze symptom presentation and complications; and (3) identify independent peri-operative predictors.
Methods
Study Population
This retrospective cohort study was approved by the University of Southern California Institutional Review Board (IRB HS-23-00758) on January 2, 2024. All consecutive ALIF procedures performed at a single high-volume academic university-affiliated hospital between January 1, 2014 and May 31, 2024 were reviewed. Adult patients who underwent one-to four-level primary or revision ALIF with or without posterior instrumentation were included. Patients required at least 1 year of follow-up for inclusion. Exclusion criteria included: patients less than 18 years old, less than 1 year follow-up, and those with indications related to trauma, malignancy, or infection were excluded.
Patients were stratified according to the development of a symptomatic RFC, including lymphoceles, hematomas, seromas, abscesses, and urinomas up to 1 year postoperatively. RFCs were determined by analyzing all pertinent postoperative imaging reports (Magnetic Resonance Imaging, Computed Tomography, and Ultrasound reports) through the electronic medical record for documentation of fluid collections in the retroperitoneum within 1 year of surgery. Imaging conducted at the previously mentioned academic spine center and outside hospitals were assessed. Postoperative imaging was not obtained routinely and was performed only when clinically indicated, at the primary team’s discretion with imaging modality based on clinical presentation and team preference. Formal radiology reports from imaging studies obtained during routine clinical care were reviewed for documentation of retroperitoneal fluid collections. Symptomatic RFCs were ascertained based on formal radiology report documentation of a retroperitoneal fluid collection; no study-specific manual measurement protocol or minimum size threshold was used for case definition. RFCs were considered symptomatic if they required intervention (drain placement or return to operating room) or were associated with abdominal pain or distension, anemia, hypotension, fever of unknown origin, lower extremity swelling, urinary issues, constipation, back pain, weight loss, or early satiety. Incidentally detected, asymptomatic collections and fluid collections in all other compartments were excluded. If the symptomatic RFCs required drainage, they were further classified by etiology based on fluid aspiration results.
Risk Factors and Complications RFC
Evaluated risk factors included age, American Society of Anesthesiologist (ASA) class, body mass index (BMI), tobacco use, hypertension, diabetes, coronary artery disease (CAD), chronic kidney disease (CKD), previous abdominal surgery, chronic obstructive pulmonary disease (COPD), hyperlipidemia, intraoperative transfusion, estimated blood loss (EBL), operative time, venous injury, revision ALIF, multi-level ALIF, staged procedures, BMP use, and DuraSeal use. Venous injury was defined as damage to the iliac veins, inferior vena cava, or iliolumbar vein requiring primary suture repair; minor injuries controlled solely with hemostatic agents were excluded. 13
Postoperative complications, including deep vein thrombosis (DVT), pulmonary embolism (PE), ileus, acute kidney injury (AKI), surgical site infection (SSI), wound complications, and incision and drainage (I&D), were monitored up to 1 year postoperatively.
Statistical Analysis
All analyses were performed using International Business Machine (IBM) Statistical Package for the Social Sciences (SPSS), version 30 (IBM corporation, Armonk, NY). Pearson Chi-Squared test of independence was performed to analyze differences in patient sex, comorbidities, occurrence of venous injury, and number of levels fused. Independent samples’ t-test was used to assess differences in patient age, BMI, EBL, and operative time. The discriminatory performance of EBL for predicting RFCs was assessed using a receiver operating characteristic (ROC) curve. The area under the curve was 0.736 and an EBL threshold of 725 mL was identified by maximizing Youden’s J statistic. To assess predictors of RFC, univariate and multivariable regression analyses were performed, adjusting for patient age, sex, ASA, BMI, and number of levels fused. To assess complications associated with RFC, a similar multivariable regression model was employed, which additionally adjusted for intraoperative transfusion and venous injury. These covariates were selected as previous research has demonstrated their impact on complications following ALIF.8,14-16 Two-tailed P-values <0.05 were considered statistically significant.
Results
Study Population, Presentation, and Management Strategies
A total of 553 ALIF patients were included, of whom 44 (7.97%) developed a symptomatic RFC within one year postoperatively. The majority of surgical exposures were performed by a vascular surgeon (542 patients, 98.0%), while 13 cases (2.4%) were performed by an orthopaedic spine surgeon. All procedures utilized a retroperitoneal approach; no cases involved a transperitoneal approach.
Differences in Demographics, Comorbidities, and Surgical Variables
RFC = Retroperitoneal Fluid Collection, ASA = American Society of Anesthesiologists, BMI = Body Mass Index, HIV = Human Immunodeficiency Virus, COPD = Chronic Obstructive Pulmonary Disorder, ALIF = Anterior Lumbar Interbody Fusion, BMP = Bone Morphogenetic Protein.
Bold values indicate statistically significant results (p < 0.05).
Patients who developed RFC had higher rates of intraoperative transfusions (59.09% vs 23.97%, P < 0.001) and venous injury (31.82% vs 15.32%, P = 0.006). The RFC cohort also had longer operative times (8.18 ± 3.26 hours vs 6.54 ± 2.86 hours, P < 0.001) and greater EBL (1657 ± 2066 mL vs 634.15 ± 1136 mL, P < 0.001). Additionally, one-level ALIFs were less common in the RFC cohort (22.73% vs 60.31%), while two-level ALIFs were more frequent (75.00% vs 35.17%) (P < 0.001). The distribution of three-level (2.27% vs 3.73%) and four-level (0% vs 0.79%) ALIFs was similar between groups. In total, multi-level ALIFs were more common in the RFC cohort (75.0% vs 39.69%, P < 0.001). Fusion at L5-S1 was the most common disc level in both cohorts.
Symptom Presentation of Patients With Symptomatic RFC
Symptomatic RFC Classification by Etiology
Risk Factors and Complications of RFC
Univariate Analysis of Risk Factors for RFC
CI = Confidence Interval, RFC = Retroperitoneal Fluid Collection, ASA = American Society of Anesthesiologists, BMI = Body Mass Index, HIV = Human Immunodeficiency Virus, COPD = Chronic Obstructive Pulmonary Disorder, ALIF = Anterior Lumbar Interbody Fusion, BMP = Bone Morphogenetic Protein.
Bold values indicate statistically significant results (p < 0.05).
Multivariate Analysis of Risk Factors for RFC
CI = Confidence Interval, RFC = Retroperitoneal Fluid Collection, ASA = American Society of Anesthesiologists, BMI = Body Mass Index, HIV = Human Immunodeficiency Virus, COPD = Chronic Obstructive Pulmonary Disorder, ALIF = Anterior Lumbar Interbody Fusion, BMP = Bone Morphogenetic Protein.
Bold values indicate statistically significant results (p < 0.05).
Forest plot for predictors of retroperitoneal fluid collection
Complications Associated With RFC
RFC = Retroperitoneal Fluid Collection, OR = Odds Ratio, DVT = Deep Vein Thrombosis, PE = Pulmonary Embolism, AKI = Acute Kidney Injury, SSI = Surgical Site Infection, I&D = Irrigation and Debridement.
Bold values indicate statistically significant results (p < 0.05).
Discussion
The study aimed to evaluate the incidence, risk factors, and complications associated with RFCs following ALIF. In this retrospective review of 553 ALIF cases, a 7.97% incidence of symptomatic RFCs was identified. Independent predictors included multi-level ALIF (aOR: 4.472), EBL ≥725 mL (aOR: 4.213), intraoperative transfusion (aOR: 3.347), L4-5 fusion (aOR: 3.186), higher ASA class (aOR: 2.291), and venous injury (aOR: 2.051). Fusion at L3-4 provided a protective effect (aOR: 0.163). RFC was also significantly associated with increased risk of DVT (aOR: 4.087) and the need for I&D (aOR: 9.593). These findings provide valuable clinical insight into identifying high-risk patients and, for the first time, establish a direct association between RFC and DVT, which may guide perioperative decision-making and surveillance strategies.
The reported incidence of RFC classified by specific etiology following ALIF ranges from 0.6% to 2.1%.4,7,11,17,18 Prior literature has addressed the difficulty in establishing the differential diagnosis of a retroperitoneal fluid collection and tends to delineate between the different etiologies–hematoma, seroma, lymphocele, urinoma, abscess, or transudative effusion.19-21 To our knowledge, no studies aggregate all-etiology RFCs, with no reports on the incidence of urinomas or abscesses following ALIF. Although rare, this study reports a single case of a urinoma following ALIF in the setting of ureteral injury likely due to excessive scarring from a hysterectomy for cervical cancer. Management with percutaneous nephrostomy tube placement led to resolution of the RFC. This case underscores the critical importance of incorporating individualized medical and surgical history into preoperative planning and risk assessment. Due to similar symptomatology of abdominal pain (68.2%) and abdominal distension (25.0%), it remains difficult to diagnose etiology without percutaneous drainage for definitive diagnosis, as highlighted in recent case reports. 19 As such, this study aggregates symptomatic RFC etiologies and reports a higher overall incidence rate (7.97%) than prior literature. Furthermore, since not all symptomatic RFCs required drainage (45.5%), this study likely overreports incidence when compared to prior literature, which only reports incidence following confirmation.
This study identified multi-level ALIF, EBL ≥725 mL, intraoperative transfusion, L4-5 fusion, higher ASA class, and venous injury as independent predictors of RFC, with fusion at L3-4 having a protective effect. This aligns with prior literature, which has demonstrated that multi-level ALIF is associated with increased risk of venous injury, intraoperative transfusion, and greater EBL.11,13,14,22 Multi-level ALIF increases surgical complexity, operating time, EBL, and time under venous retraction. These factors increase the risk of venous injury, a known risk factor for developing retroperitoneal hematomas. Additionally, case reports have documented the development of retroperitoneal lymphoceles following multi-level ALIFs.23,24 With regards to surgery at the L4-5 vertebral level, studies have demonstrated that the majority of vascular injuries result from fusion at the L4-5 disc space, with the risk of venous injuries increasing by up to 4 times.11,13,25,26 This is attributed to the most common anatomic location of the aortic bifurcation at the L4 vertebral body, increasing risk of prolonged arterial and venous retraction at this disc space. Conversely, this study demonstrated that fusion at the L3-4 level had a protective effect. It is important to note that the limited sample size of patients who received L3-4 fusion (3 in the RFC cohort vs 38 in the cohort without RFC) may limit robust multivariable analysis at this specific level. However, important structures lie either cranially to the L3-4 vertebral level, such as the cisterna chyli or the inferior mesenteric artery, and caudally such as the aortic bifurcation or common iliac vein confluence, which when damaged may lead to a RFC.27–30 It is still important to recall that the aorta travels anterior to the L3-4 disc, limiting access to the disc in 90% of cases and necessitating significant retraction of the aorta. 30 Further, this study aligns with prior literature demonstrating that use of BMP to promote fusion did not increase risk of postoperative seromas or BMP-related fluid collections.31-34 Lastly, prior abdominal surgery was not an independent risk factor for symptomatic RFC following ALIF in this study; however, prior literature shows that each additional prior abdominal surgery increases intraoperative complication risk by 52%.9,35 Armed with this risk factor analysis, spine surgeons may have increased suspicion for an RFC when a patient presents with abdominal pain if they underwent multi-level or L4-5 ALIF complicated by venous injury or intraoperative blood loss ≥725 mL necessitating intraoperative transfusion.
This study identified that RFCs are associated with an increased risk of further complications. Symptomatic RFCs have clinical significance, increasing the risk of subsequent incision and drainage by nearly 10 times in this study. Furthermore, RFCs increased the risk of subsequent DVT by 4 times, even after adjusting for intraoperative transfusion and vein injury. Prior literature surrounding ALIFs has speculated the relationship between venous injury, which was more common in the RFC cohort, and increased DVT risk (aOR: 3.33). 13 This study demonstrated an even greater risk of DVT (aOR: 4.087), which may be attributable to the RFC compressing the surrounding vasculature, leading to increased venous stasis and subsequent DVT. Importantly, this study demonstrates that symptomatic RFCs did not increase risk of other well-documented postoperative ALIF complications, such as ileus, SSIs, or wound complications.3,11,12 These findings highlight the need for high clinical suspicion of symptomatic RFCs followed by close monitoring and postoperative VTE prophylaxis in these high-risk patients.
Limitations
This study has several limitations. The retrospective nature of this study introduces selection bias and may not address all underlying confounders, which may limit the generalizability of this study. Furthermore, ALIFs in this study were performed collaboratively by multiple different vascular and spine surgeons, which introduces variability intraoperatively. As this study was performed in a single-center high-volume academic institution, generalizability to other settings, such as ambulatory surgery centers, may be limited. Additionally, due to the small sample of RFCs, robust multivariable risk factor analysis, especially pertaining to specific vertebral levels such as L3-4, may be limited, as previously mentioned. Furthermore, since not all RFCs were aspirated, complete classification by etiology was not feasible purely from imaging documentation. Additionally, because postoperative imaging was obtained only when clinically indicated and interpreted by different radiologists across imaging modalities, the study is subject to ascertainment bias and interobserver variability. The heterogeneous use of multiple imaging techniques may further increase risk of bias in RFC detection. Finally, it is possible that not all symptomatic patients received imaging due to early self-resolution of the RFC or symptom attribution to other etiologies. Future studies are needed to mitigate these limitations and develop standardized diagnostic and management pathways and protocols for the identification of suspected postoperative RFCs. Prospective studies will assist in validating the identified risk factors and the observed association between RFC and postoperative complications such as DVT, and provide insight on the temporal effect of early fluid collection detection on surgical outcomes.
Conclusions
In this large single-center institutional study, symptomatic retroperitoneal fluid collections occurred postoperatively in 7.97% of ALIF cases. Multi-level fusion, intraoperative transfusion, fusion at L4-5, higher ASA class, venous injury, and EBL ≥725 mL were independent predictors of symptomatic RFCs. Notably, symptomatic RFCs increased the risk of postoperative DVTs by 4 times, underscoring the clinical importance of this complication. These findings may assist in the identification of patients at elevated risk for developing symptomatic RFCs and emphasize the need for heightened postoperative vigilance.
Footnotes
ORCID iDs
Ethical Considerations
There are no human participants in this article and informed consent or IRB approval were not required.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
Declaration of Conflicting Interests
The authors declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: DR, JD, HA, AL, CF, RB, VS, MN, JK, WJK, NM, MCG, and GAM have nothing to disclose. JCW has received intellectual property royalties from Zimmer Biomet, NovApproach, SeaSpine, DePuy Synthes, and GS Medical; consulting fees from DePuy Synthes and Bioretec; and has stock options in Bone Biologics, Electrocore, PearlDiver, Surgitech, and Illuminant. RJH has received research support from ATEC and NuVasive; consulting fees from Medtronic, Globus/NuVasive, Orthofix, SI-Bone, Cerapedics, and Evolution Spine; and support for attending meetings from North American Spine Society. RKA has received consulting fees from Medtronic, Orthofix, and Globus; payment for lectures and presentations from Ecential Robotics; support for attending meetings from USC; has stock options in HIA Technologies, NeoOnc, and Globus; and has received research support from ATEC and Max BioPharma.
Data Availability Statement
Data is not publicly available but can be available upon request.