Perspective on the true incidence of bowel perforations occurring with extreme lateral lumbar interbody fusions. How should they be treated?

BACKGROUND: What is the risk of bowel perforation (BP) with open or minimally invasive (MI) extreme lateral lumbar interbody fusion (XLIF)? What is the truth? Further, if peritoneal symptoms/signs arise following XLIF/MI XLIF, it is critical to obtain an emergent consultation with general surgery who can diagnose and treat a potential BP. LITERATURE REVIEW: In multiple series, the frequency of BP ranged markedly from 0.03% (i.e. 1 of 2998 patients), to 0.08% (11/13,004), to 0.5%, to 8.3% (1 in 12 patients), up to 12.5% (1 in 8 patients). BPs attributed to different causes carry high mortality rates varying from 11.1% to 23%. For the 11 (0.08%) BP occurring out of 13,004 patients undergoing XLIF in one series, there was one (9.09%) death due to uncontrolled sepsis. In another series, where 31 BP were identified for multiple lumbar surgical procedures identified through PubMed (1960-2016), including 10 (32.2%) for lateral lumbar surgery including XLIF, the overall mortality rate was 12.9% (4/31).
CONCLUSION: The incidence of BPs occurring following XLIF/MI XLIF procedures ranged from 0.03% to 12.5% in various reports. What is the true incidence of these errors? Certainly, it is more critical that when spine surgeons' patients develop acute peritoneal symptoms/signs following these procedures, they immediately consult general surgery to both diagnose, and treat potential BP in a timely fashion to avoid the high morbidity (87.1%) and mortality rates (12.9%) attributed to these perforations. Copyright:

INTRODUCTION

What is the true risk of bowel perforation (BP) following open or minimally invasive (MI) extreme lateral lumbar interbody fusion (XLIF) [Table 1]?[1,2,5-22] Several studies identified vastly different frequencies of BP following XLIF/MI XLIF; the range varied from 0.03% (1/2988 patients), to 0.08% (11/13,004), to 0.5% (3/590), to 8.3% (1/12 patients), to 12.5% (1 of 8 patients) of cases. [7,10,15,16,21,22] Whatever the true frequency of BP with XLIF/MI XLIF, it is most critical that spine surgeons recognize that new postoperative peritoneal symptoms/signs of BPS so that general surgery can be immediately consulted to both diagnose, and potentially treat this medical error in a timely fashion.[19,22] Delays in diagnosing any type of BP are associated with high mortality rates ranging from 11.1-23%.[3,4] Notably, specific mortality rates for BP occurring in patients undergoing XLIF/MI XLIF ranged from 9.09% (1/11 BP in 13,004 XLIF patients) to 12.9% (4/31 BP occurring in a series of 18 patients following diskectomies/microdiskectomies, and 10 lateral procedures including XLIF).[19,22]

Table 1:

Summary of literature.

FREQUENCY OF SPONDYLOSIS WITH ISTHMIC SPONDYLOLYSIS AND SPONDYLOLISTHESIS

In 2020, Aoki et al. reviewed 580 cases of lumbar spondylosis; 37 (6.4%) patients had spondylolysis, with 19 of 37 additionally demonstrating spondylolisthesis (51.4%) [Table 1].[1] They emphasized that spondylolysis/spondylolisthesis defects were most typically reported in adolescent male athletes. Notably, these are the patients who may be specifically targeted for canal distraction/decompression with XLIF/MI XLIF supplemented with posterolateral fusions (PLF) performed utilizing pedicle screw instrumentation as these combined procedures offer preservation of the posterior elements.

XLIF SUPPLEMENTED WITH POSTERIOR LATERAL FUSION (PLF) WITH PEDICLE SCREW INSTRUMENTATION FOR SPONDYLOLYSIS/ SPONDYLOLISTHESIS

In select patients with isthmic spondylolysis/ spondylolisthesis, XLIF/MI XLIF may be combined with posterolateral instrumented pedicle screw fusions to provide simultaneous anterior indirect canal decompression with posterior stabilization/fusion (PLF).[20] This combination of procedures avoids disruption of the posterior elements (i.e. facet joints, laminae, spinous processes) [Table 1].[20] In Tamburrelli et al. (2018), MI XLIF were performed to address isthmic spondylolysis with spondylolisthesis, and were effectively supplemented with percutaneous pedicle screw fusions (PLF). This provided; “a reliable and safe option to the most common open procedures,” that usually included TLIF, while allowing for “good correction of the listhesis.”[20]

DIFFERENT FREQUENCIES OF BOWEL PERFORATIONS REPORTS FOR XLIF/MI XLIF

For multiple series, the incidence of BPs attributed to XLIF/MI XLIF ranged between 0.03% to 12.5%; this left us questioning the true incidence of this surgical error [Table 1].[2,7,9,15,17,19,21,22] Specifically, BPs occurred in one case report, and with higher numbers of cases reported in other clinical series [Table 1].[2,5,7,9,15,17,19,21,22] In 2015, Balsano et al. described a 70-year-old male who, following a L3-L4/ L4-L5 XLIF, developed a BP.[2] Tormenti et al. (2010) found 1 (12.5%) BP occurring after performing just 8 MI XLIF (2007–9).[21] Uribe et al. (2015) noted that out of 13, 004 MI XLIF, there were 11 BPs. [22] Surgeons in this study were experienced with XLIF/MI XLIF as they had performed between 11 to 50 cases or over 50 cases.[22] When Fujibayashi et al. (2017) looked at 1995 patients undergoing XLIF with another 1003 having oblique lateral lumbar interbody fusions (OLIF), there was just one case (0.03% of 2988 patients) of a BP.[7] Paterakis et al. (2018) later found that 1 of 12 patients (8.3%) undergoing XLIF for degenerative scoliosis sustained a BP (i.e. treated with a primary resection/anastomosis).[15] In 2018, Siasos et al. documented 31 BPs; 10 occurred after MI lumbar operations; (i.e. XLIF, anterior, lumbar interbody fusion [ALIF], and transforaminal lumbar interbody fusion [TLIF}), while 18 followed lumbar diskectomies/ microdiskectomies).[18] In 2019, Rustagi et al. found 3 (0.5%) of 590 XLIF resulted in BP.[17] Finally, Hwang et al. (2021) determined that 2 of 140 LLIF resulted in BP.[9]

STUDIES NOT SPECIFICALLY REPORTING THE INCIDENCE OF BOWEL PERFORATIONS

Four clinical series, involving between 23 and 96 patients per study, did not discuss BP as occurring following XLIF/ MI XLIF [Table 1].[8,10,16,18] In a prospective, multicenter combined randomized/observational series, Isaacs et al. (2016) looked at the 2 year outcomes for 1–2 adjacent-level L1-L5 MI XLIF (29 patients) versus MI TLIF (26 patients).[10] These were performed to address degenerative spondylolisthesis and stenosis; no bowel perforations were observed.[10] When Segawa et al. (2017) assessed the safety/ efficacy of microendoscopic 1–3 level XLIF performed in 96 patients, they reported no BPs or vascular injuries.[28] They attributed this to the introduction of the microendoscopy technique.[18] When Pereira et al. (2017) evaluated 23 XLIF, they too encountered no BPs.[16] Additionally, no BPs were cited in the Hiyama et al. series that compared the complications of 62 MI XLIF versus 44 MI TLIF.[8] Of interest, they also found that XLIF operations were nearly 1/3 shorter than TLIF, the average blood loss of MI XLIF was reduced by 2/3, while the XLIF/MI XLIF procedures offered the benefit of preserving the posterior elements.[8]

SIGNIFICANT INCIDENCE OF NEUROLOGICAL, VASCULAR, AND OTHER INJURIES WITH XLIF/MI XLIF

Many XLIF/MI XLIF studies cited high frequencies of new sensory (i.e. 5.15–75% including thigh paresthesias), new motor (5.15–40% - typically iliopsoas), and vascular (0.03– 0.04%) injuries [Table 1].[6,7,12,13,16,18,21] Other postoperative complications included; 0.7% plus incidence of surgical site infections, end plate injuries/fractures, retroperitoneal hematomas, sterile seromas, postoperative ileus, pseudarthrosis, and the need for additional surgery.[7,13,16,18]

LEARNING CURVE FOR XLIF VERSUS OLIF

In Li et al. (2019), the learning curve for the first 30 cases of XLIF versus OLIF were studied in patients undergoing lumbar surgery for disc disease, stenosis, degenerative spondylolisthesis, infection, trauma, or tumor.[13] There was a 10% incidence of adverse events (AE) occurring within the early learning curve for 1-3 level XLIF versus 33.3% for OLIF. Patients were on average in their late fifties, and although none developed BP in either group in this series, 3 undergoing OLIF had major vascular injuries, while 3 patients having XLIF sustained minor complications (i.e. 1 infection 1 end plate injury, and 1 new sensory deficit).

FREQUENCY OF BOWEL PERFORATIONS WITH XLIF/MI XLIF

Vastly different frequencies of BPs occurred in patients undergoing XLIF/MI XLIF, ranging from 0.03% to 12.5% [Table 1].[7,9,15,19,21,22] Most critically, when patients develop new postoperative peritoneal symptoms/signs (i.e. abdominal pain, hyperemesis, hypotension, peritonitis, and sepsis), spinal surgeons should immediately consult general surgeons to both diagnose and surgically manage BPs if diagnosed and warranted. In Siasos et al., 3 cases of BP were diagnosed intraoperatively, while the remainder were diagnosed/operated on between 2 days and 1 week/1 year postoperatively[19] For 3 of the 590 patients in Rustagi et al. series, the diagnosis of BPs led to surgery performed an average of 4.7 days (range 3–7) postoperatively.[17]

DIAGNOSTIC FINDINGS FOR BOWEL PERFORATIONS ON ABDOMINAL/PELVIC CT SCANS

Multiple studies also confirmed common findings on Abdominal/Pelvic CT scans (APCT) consistent with/diagnostic of BPs following XLIF/MI XLIF. [17,19] These findings included; fluid in the abdomen, pneumoperitoneum, instrumentation close to the bowel, extraluminal trapped air, loss of bowel continuity, and increased bowel wall thickness.[17,19] Of interest, Rustagi et al. (2019) noted: “The isolated presence of air/fluid in the retroperitoneal space is easily confounded by the presence of expected postoperative findings following TPIF (i.e. XLIF).”[17] Alternatively, after observing 2 instances of BP out of 140 patients undergoing LLIF, Hwang et al. (2021) prospectively performed APCT within 48 postoperative hours.[9] Out of 90 LLIF, they found that APCT studies performed in 5 patients (5.5%) with peritoneal symptoms had pneumoperitoneum; despite this, there were no documented BPs.[9] They further correlated an increased risk of BP with LLIF fusion of more than 3 levels, and surgery performed at the L2-L3, and/or L3-L4 levels.

TREATMENT OF BPs OF ALL ETIOLOGIES

High mortality rates are reported for BPs occurring due to many different factors (i.e. especially diveriticular disease, and non-diverticular disease exclusive of spinal surgery) [Table 1].[3,4] One series cited an 11.1% mortality rate where primary resections/anastomoses were possible, but a 22.2% mortality if these were not feasible.[3] The second study found an overall 23% mortality rate for BP, that included consideration of diverticular (133 patients) and nondiverticular-related (79%) perforations.[4]

HIGH MORTALITY RATE FOR BOWEL PERFORATIONS FOLLOWING XLIF/MI XLIF

High mortality rates (i.e. up to 12.9%) are reported for BPs occurring due to XLIF/MI XLIF.[14,17,19,22] This, therefore, warrants that patients presenting with new peritoneal signs following XLIF/MI XLIF should immediately undergo evaluation by general surgery to diagnose and potentially treat BPs. In Marzuez-Lara et al. (2014) National Inpatient Database involving 543,146 patients undergoing lumbar surgery, 30 patients had BPs/peritoneal injuries; they had a 20 fold greater mortality rate when compared to those without such intraoperative erros (i.e. 14.6 vs. 0.7/1000 cases).[14] In Uribe et al. (2015), 11 (0.08%) out of 13,004 patients following MI XLIF had BPs.[22] Repairs were performed intraoperatively (4 cases), or between postoperative days 2–5; the result was 1 death from sepsis, 5 full recoveries, one permanent colostomy, while 3 were lost to follow-up.[22] For the 31 BPs identified in Siasos et al. review (2018), including 10 due to XLIF, ALIF, and TLIF, surgery to address BPs were performed intraoperatively (3 cases), and up to 1 week/1 year postoperatively.[19] Further, when Rustagi et al. operated on 3 (0.5%) BPs out of a series of 590 patients undergoing XLIF, they were diagnosed and operated on an average of 4.7 days postoperatively, resulting in no deaths, but one permanent ileostomy.[17]

CONCLUSION

The incidence of BPs reported following XLIF/MI XLIF procedures ranged from 0.03% up to 12.5% in multiple studies.[7,8,15,17,19,21,22] Hence, it is apparent that the true incidence of BPs following these procedures is still not well defined [Table 1].[7,9,15,17,19,21,22] What is clear, however, is that patients who acutely develop peritoneal symptoms/signs following XLIF/MI XLIF should immediately undergo evaluation by general surgery to diagnose and potentially surgically treat BPs if warranted [Table 1].[3,4,14,17,19,22]