Carbon dioxide embolism during salvage transanal total mesorectal excision in “frozen pelvis”: Two case reports

Abstract

BACKGROUND

Transanal total mesorectal excision (TaTME) is increasingly used for salvage surgery in local rectal cancer recurrence. However, the fibrotic “frozen pelvis” presents a unique physiological hazard. We report two cases of massive carbon dioxide (CO₂) embolism that occurred during salvage TaTME, highlighting the mechanism of “non-collapsible veins” and a specific capnographic diagnostic pattern.

CASE SUMMARY

Two male patients (64 years old and 68 years old, respectively) underwent salvage TaTME for recurrent rectal cancer. Both procedures utilized a simultaneous two-team approach with standard 15 mmHg insufflation. During deep pelvic dissection within the fibrotic field, both patients experienced sudden hemodynamic collapse (severe hypotension in case 1 and asystole in case 2). A distinct biphasic capnographic pattern was observed in both scenarios: An initial precipitous drop in end-tidal CO₂ marking the embolic event, followed by a rapid paradoxical surge (> 50 mmHg) during resuscitation. This “rebound hypercapnia” confirmed the massive systemic CO₂ absorption. Immediate cessation of insufflation, fluid resuscitation, and conversion to open surgery were performed. One patient recovered without sequelae, while the other patient achieved return of spontaneous circulation after cardiopulmonary resuscitation.

CONCLUSION

Enhanced vigilance for CO₂ embolism is strongly advised in the “frozen pelvis”. The biphasic capnographic signature can serve as a critical observational point for early recognition, and a rapid multidisciplinary response involving the nursing team is essential for patient survival.

Key Words: Carbon dioxide embolism; Transanal total mesorectal excision; Frozen pelvis; Salvage surgery; Perioperative nursing; Case report

Core Tip: This case series serves as a critical warning on the significant intraoperative hazard of salvage transanal total mesorectal excision: Dense fibrosis resulting from prior multimodality treatment creates “non-collapsible veins” that remain open when injured, allowing massive carbon dioxide (CO₂) influx. The investigators highlight the diagnostic biphasic capnographic signature-initial hypocapnia followed by rebound hypercapnia-as a key observational point for differentiating CO₂ embolism from other causes of intraoperative collapse, and emphasize the vital role of perioperative nursing collaboration in the successful management of such crises.

INTRODUCTION

Transanal total mesorectal excision (TaTME) offers superior visualization for distal mesorectal dissection[1-3]. Recently, this technique has been extended to “redo” or salvage surgeries for local recurrence[4,5]. However, this procedure confronts the hostile environment characterized by distorted anatomy and dense fibrosis that resulted from prior surgical intervention and radiotherapy-a condition often clinically referred to as the “frozen pelvis”[6,7].

Gas embolism is a known risk in TaTME[8], with the registry data reporting an incidence of approximately 0.4%[9]. Unlike virgin tissue, where veins collapse upon injury, fibrotic veins in a radiated pelvis may remain tethered open, creating fixed conduits for gas entry-the “non-collapsible vein” phenomenon[10,11]. This report aims to alert surgeons to the high risk of carbon dioxide (CO₂) embolism in salvage TaTME. The investigators analyzed two life-threatening events, describing the characteristic biphasic physiological response, and discussing its multidisciplinary crisis management.

CASE PRESENTATION

Chief complaints

Case 1: Elevated carcinoembryonic antigen (CEA) levels and anastomotic recurrence was detected by colonoscopy for one month.

Case 2: The patient was diagnosed with recurrent rectal cancer for three months, and admitted for planned salvage surgery.

History of present illness

Case 1: The patient was a 64-year-old male with a complex history of recurrent rectal cancer. This patient initially underwent radical resection for rectal cancer in February 2018. Following local recurrence in October 2019, the patient received radiotherapy and chemotherapy. In January 2023, due to sacral invasion (S3-S4), the patient underwent trans-sacrococcygeal excision of the rectal lesion and partial sacrectomy with ileostomy. In August 2023, the follow-up examination revealed elevated CEA levels. The colonoscopy confirmed a hard, irregular mass at the anastomosis, 6 cm from the anal verge, and this was diagnosed as recurrent adenocarcinoma. The patient was admitted for salvage surgery.

Case 2: The patient was a 68-year-old male. This patient underwent radical resection for rectal cancer with ileostomy in March 2023 for an ulcerative, moderately differentiated adenocarcinoma with proficient mismatch repair. Following surgery, the patient received adjuvant chemoradiotherapy. In June 2023, stoma reversal was performed. In February 2024, the follow-up examinations revealed elevated CEA levels, and the colonoscopy combined with positron emission tomography-computed tomography (CT) suggested an anastomotic recurrence. The patient subsequently underwent six cycles of palliative chemotherapy (raltitrexed + irinotecan + bevacizumab). In May 2024, the patient was admitted for salvage surgery.

History of past illness

Case 1: The patient has a history of heavy pretreatment, including pelvic radiotherapy in 2019, which contributed to the significant pelvic fibrosis. The patient also received multiple lines of chemotherapy (FOLFOX + cetuximab, FOLFIRI + bevacizumab) between July 2021 and April 2023.

Case 2: The patient has a 5-year history of hypertension and a 1-year history of coronary heart disease. Notably, this patient had significant underlying cardiac pathology, which was later confirmed by CT to include severe stenosis (> 80%) of the left anterior descending artery. In addition, the patient has a diagnosis of alcoholic fatty liver. The patient’s regular medication regimen included a fixed-dose combination of valsartan and amlodipine (Exforge), one tablet daily.

Personal and family history

Case 1: The patient had no history of smoking or alcohol consumption. Furthermore, there was no family history of hereditary colorectal cancer syndromes or other malignancies.

Case 2: The patient has a history of alcohol consumption, but had quit. The patient had no smoking history. The patient’s father has a history of colon cancer.

Physical examination

Case 1: The patient had a height of 170 cm and a weight of 57 kg, corresponding to a body mass index (BMI) of 19.7 kg/m². Furthermore, the patient had an ileostomy in the right lower quadrant from the previous surgery. The digital rectal examination (DRE) revealed a hard, fixed mass at the anastomosis, approximately 6 cm from the anal verge. The abdomen was soft with no signs of peritonitis.

Case 2: The patient had a height of 169 cm and a weight of 61.5 kg, corresponding to a BMI of 21.5 kg/m². The vital signs were stable [blood pressure (BP): 102/70 mmHg, heart rate (HR): 100 bpm]. The DRE revealed a hard palpable mass at the anastomosis, approximately 3 cm from the anal verge. The finger could pass through the stenosis, but the glove was stained with blood upon withdrawal.

Laboratory examinations

Case 1: The preoperative serum CEA was elevated at 26.04 ng/mL. The baseline hemoglobin was 123 g/L. Other routine biochemical markers were within normal limits.

Case 2: Preoperative CEA was 5.27 ng/mL. Baseline hemoglobin was 94 g/L. The cardiac ultrasound revealed an ejection fraction of 78%, but noted tachycardia and valvular regurgitation.

Imaging examinations

Case 1: The pelvic CT and magnetic resonance imaging (MRI) revealed a soft tissue mass at the anastomotic site with postoperative changes in the sacrum. The colonoscopy confirmed a 4 cm × 3 cm ulcerated mass that infiltrated the adventitia.

Case 2: The abdominal CT and pelvic MRI with diffusion-weighted imaging confirmed the thickening and enhancement of the rectal anastomotic wall, which was highly suspicious for recurrence.

FINAL DIAGNOSIS

Case 1

Massive CO₂ embolism, recurrent rectal cancer (pT3N0M1), and respiratory failure (postoperative).

Case 2

Massive CO₂ embolism, cardiac arrest, recurrent rectal cancer, and coronary artery disease.

TREATMENT

Case 1

In August 2023, a salvage surgery that involved TaTME and abdominal adhesiolysis was performed. The perineal phase utilized a transanal access platform with a gel cap (GelPOINT Path, Applied Medical). Standard American Society of Anesthesiologists monitoring was employed without proactive transesophageal echocardiography or precordial Doppler. During the transanal dissection of the fibrotic presacral tissue (pneumopelvis 15 mmHg), the patient experienced sudden hemodynamic collapse, with the HR dropping below 40 bpm and undetectable BP. The direct visualization of the avascular surgical field immediately confirmed the absence of severe hemorrhage. The transanal platform was immediately removed to cease insufflation. Then, fluid resuscitation and metaraminol support were initiated, which led to rapid hemodynamic recovery. Given the immediate temporal relationship to insufflation, the procedure was converted to open laparotomy with Miles resection (abdominoperineal resection) and creation of a colostomy.

Case 2

In June 2024, the patient underwent laparoscopic radical resection. During the perineal phase, and while establishing the transanal platform and dissecting the rectum (pneumopelvis 15 mmHg), the patient suffered sudden cardiac arrest. Immediate cardiopulmonary resuscitation (CPR) was initiated, achieving return of spontaneous circulation (ROSC) after 2-3 minutes. Due to the immediate requirement for external chest compressions, central venous gas aspiration was not performed. The intraoperative blood gas analysis shortly after ROSC revealed a pH of 7.244 and a pCO₂ of 53.4 mmHg, which are consistent with the hypercapnic phase of CO₂ embolism. Immediate CPR was initiated alongside desufflation. Following ROSC, the surgery was completed with resection of the tumor and creation of a sigmoid colostomy. The patient was transferred to the intensive care unit (ICU) for post-resuscitation care, which included targeted temperature management and mechanical ventilation. The steep Trendelenburg position was maintained to minimize cerebral embolization risk and optimize hemodynamic recovery, with the nursing team ensuring secure patient fixation to prevent slippage and pressure injuries during the crisis.

OUTCOME AND FOLLOW-UP

Case 1

Following the hemodynamic collapse, the anesthesia monitor revealed a biphasic pattern: The end-tidal CO₂ (ETCO₂)[12] initially dropped from baseline, and subsequently surged to > 50 mmHg during resuscitation (Figure 1A). The postoperative electrocardiography and cardiac biomarker panels remained negative, definitively excluding acute myocardial infarction and pulmonary thromboembolism. The patient did not require ICU admission. The pathological examination confirmed the R0 resection margin. Postoperatively, the patient was alert with no focal neurological deficits, but developed type I respiratory failure that required high-flow nasal cannula oxygen therapy and transfusion for anemia, which was classified as a Clavien-Dindo grade II complication. The patient was discharged with a drainage tube after flap revision.

Figure 1 Perioperative hemodynamic and capnographic trends that demonstrate the biphasic signature of carbon dioxide embolism. A: Case 1: The graph depicts the sequence of events surrounding the transient hemodynamic collapse. Phase 1 (“gas lock”) was characterized by the precipitous drop in end-tidal carbon dioxide (ETCO₂, orange solid line) coincident with the significant decline in mean arterial pressure (blue dashed line) and heart rate (pink dash-dot line). Phase 2 (“rebound hypercapnia”) was manifested during resuscitation as a paradoxical surge in ETCO₂ (> 50 mmHg) due to the washout of dissolved carbon dioxide; B: Case 2: The graph illustrates the event that progressed to cardiac arrest, and the subsequent cardiopulmonary resuscitation (CPR). The values shown represent the last recorded measurements immediately prior to CPR initiation. A similar biphasic pattern was observed: An initial drop in ETCO₂ during the gas lock phase, followed by the rebound hypercapnia upon return of spontaneous circulation. CPR: Cardiopulmonary resuscitation; CO₂: Carbon dioxide.

Case 2

A similar biphasic pattern was retrospectively identified (Figure 1B). Following admission to the ICU, the patient required 17 hours of mechanical ventilation before successful extubation. The patient was safely transferred to the general surgical ward after a 65-hour ICU stay, representing a Clavien-Dindo grade IVa complication. The neurological examination revealed an alert status without focal deficits. The postoperative Holter monitoring revealed frequent ventricular premature beats, and the coronary CT angiography confirmed the severe coronary artery stenosis, which likely lowered the threshold for ischemia-induced arrest during the embolic event. The pathological examination confirmed an R0 resection margin. The patient was discharged on June 12, 2024, with normal stoma function.

DISCUSSION

The incidence of clinically significant CO₂ embolism during TaTME has been reported to be approximately 0.4%, according to the International TaTME Registry[9]. However, this figure likely underestimates its true prevalence[13]. Single-center series with rigorous monitoring have reported significantly higher rates, which ranged from 1.6%[14] to as high as 3.8%[15]. The focused review of these cases revealed a spectrum of risk factors, rather than a single etiology. For instance, Zhu et al[14] reported a massive embolic event in a patient with primary rectal cancer and no prior radiotherapy, which was directly triggered by increasing the pneumopelvis pressure to 15 mmHg to counteract smoke. Conversely, Shiraishi et al[15] described two severe events that occurred at pressures below 15 mmHg, but both patients had a history of neoadjuvant chemoradiotherapy and required complex multi-visceral or sacral resections, underscoring the peril of anatomical distortion. The discrepancy between the registry data and specific institutional reports suggests that in the subgroup of patients who underwent salvage surgery within a “frozen pelvis”, the fibrotic environment[16] amplified this intraoperative hazard. Therefore, CO₂ embolism should be considered as a notable risk that requires enhanced vigilance during salvage TaTME.

The core pathophysiological mechanism in these cases is the “non-collapsible vein” phenomenon[17]. In virgin tissue, a venous injury typically results in vessel collapse, limiting gas entry. However, dense peri-vascular fibrosis from previous dissection and radiotherapy compromises this elasticity, transforming compliant vessels into rigid conduits[18]. This creates a “tethering effect” that holds the venous lumen patent even after transection. Consequently, a fixed, non-compliant conduit is established, connecting the high-pressure pneumopelvis (15 mmHg) directly to the low-pressure venous system (central venous pressure: 5-10 mmHg), and driving a continuous influx of CO₂ (Figure 2).

Figure 2 Schematic illustration of the mechanism of carbon dioxide embolism in the “frozen pelvis”. The sagittal cross-section of the male pelvis during salvage transanal total mesorectal excision demonstrates the transanal access platform (gel cap) tightly sealing the anal verge. The laparoscopic instruments were depicted to enter through this single port into the mesorectal plane. The presacral space was encased in dense, white scar tissue (labeled “Frozen Pelvis”) due to prior therapy. The transected presacral vein was held wide open (“Transected, Open Presacral Vein”) by the stiff fibrosis (tethering effect). Blue carbon dioxide bubbles flowed directly from the insufflated rectal cavity into the open venous lumen, driven by the pneumatic pressure gradient.

The critical clinical lesson from this series is the recognition of the biphasic capnographic signature, which distinguishes CO₂ embolism from air (nitrogen) embolism or other causes of shock. The first phase involves a precipitous drop in ETCO₂, which is caused by the “gas lock” effect in the right ventricle and pulmonary outflow tract, drastically reducing pulmonary blood flow and increasing the physiological dead space[19,20]. The second phase [observing a “rebound hypercapnia” (ETCO₂ > 50 mmHg)] occurs immediately after circulation is restored. Unlike nitrogen, CO₂ is highly soluble in blood (approximately 20 times more than oxygen)[21,22]. Once the gas lock is broken and forward flow resumes, the trapped CO₂ bubbles rapidly dissolve, and are washed out into the alveoli, causing a paradoxical surge in ETCO₂. This biphasic pattern is highly characteristic, and serves as a vital retrospective confirmation of the diagnosis.

The management of this crisis requires a departure from standard protocols. Although a pneumopelvis pressure of 12-15 mmHg is typical for primary cases, as an expert practice consideration, the investigators propose the utilization of a strict low-pressure protocol (8-10 mmHg) during the perineal phase of salvage TaTME[23,24]. This transanal pressure must be dynamically matched by the reduced abdominal pneumoperitoneum to prevent the “bellows effect” from collapsing the transanal working space[25]. Furthermore, in case of any unexplained hemodynamic instability, the “Stop, Pack, Flood” maneuver must be immediately executed[23]: Cessation of insufflation, packing of the pelvis with wet gauze to tamponade open veins, and flooding the field with saline to identify gas entry points. Moreover, the investigators advise the maintenance of the existing steep Trendelenburg position, rather than attempting the Durant maneuver, utilizing buoyancy for cerebral protection, while relying on the high solubility of CO₂ for rapid dissolution, thereby preventing sterile field disruption and critical delays in CPR[11,26]. The nursing team plays a pivotal role in this sequence, from rapidly shutting off the insufflation source to maintaining secure patient positioning during resuscitation efforts[27]. Effective execution relies on rapid multidisciplinary coordination and standardized nursing competencies[28], although the integration of intelligent anesthesia recovery prediction models may further enhance perioperative safety in these high-risk scenarios[29].

Causality in these cases was established through the perioperative event assessment approach. The diagnosis of massive CO₂ embolism was supported by the strict temporal association with the initiation of high-pressure pneumopelvis in the fibrotic field and the physiological plausibility of the non-collapsible vein mechanism. Crucially, the definitive clinical exclusion of primary cardiogenic shock, anesthetic anaphylaxis, or intractable vagal reflexes during the acute phase, combined with the distinct biphasic capnographic signature, provides robust diagnostic certainty. The postoperative evaluations further excluded acute myocardial infarction and pulmonary thromboembolism, confirming the embolic nature of the intraoperative collapse.

CONCLUSION

In salvage TaTME, the “non-collapsible vein” is a lethal trap. The biphasic capnographic pattern (initial hypocapnia followed by rebound hypercapnia) was the diagnostic hallmark of this crisis.

ACKNOWLEDGEMENTS

The authors would like to thank the operating room nursing team and the Department of Anesthesiology at Sir Run Run Shaw Hospital for their rapid response and collaboration during these critical events.