Regional blood flow occlusion during extensive pelvic procedures for ovarian cancer: a randomized trial

Transcription

1 Int J Gynecol Cancer 2004, 14, SURGEON S CORNER Regional blood flow occlusion during extensive pelvic procedures for ovarian cancer: a randomized trial S. M. EISENKOP*, N. M. SPIRTOS, W. M. LIN*, F. RAFIDI, G. M. GROSS *Women s Cancer Center: Encino-Tarzana, Tarzana; Women s Cancer Center: Palo Alto, Palo Alto; Vascular Group, Tarzana, CA; and Huntsville Vascular Specialists PC, Huntsville, AL Abstract. Eisenkop SM, Spirtos NM, Lin WM, Rafidi F, Gross GM. Regional blood flow occlusion during extensive pelvic procedures for ovarian cancer: a randomized trial. Int J Gynecol Cancer 2004;14: The objective of this study was to determine the effects of intraoperative aortic occlusion on blood loss and operative time when used during en bloc resection of internal reproductive organs, pelvic peritoneum, and rectosigmoid colon [modified posterior exenteration (MPE)] for primary cytoreduction of ovarian cancer. Patients undergoing MPE, without palpable distal aortic plaque or calcification, were randomized to: (a) complete distal aortic occlusion ( 60 min, with heparin and protamine reversal) with a vascular clamp immediately before MPE, (b) bilateral hypogastric artery occlusion, or (c) no regional blood flow occlusion. Outcomes were compared with respect to blood loss, operative time, and the transfusion rate (ANOVA analysis of variance). Fifty-six patients were accrued. Groups were equivalent with respect to age, disease severity, extent of upper abdominal surgery done, and cytoreductive outcomes. Aortic occlusion significantly reduced the total operative time (P =0.02), estimated blood loss (P = 0.01), transfusion rate (P =0.02), hospital stay (P = 0.05), and both operative time (P 0.001) and blood loss (P 0.001) specifically associated with MPE. There were no immediate or delayed complications due to aortic clamping. Aortic occlusion significantly reduces the blood loss and operative time for patients requiring MPE in the context of primary cytoreductive operations. KEYWORDS: morbidity, ovarian cancer cytoreduction. The pelvic phase of the operative management of advanced ovarian cancer often represents an extensive undertaking with potential morbidity (1 8). En bloc excision of internal reproductive organs, pelvic peritoneum, and rectosigmoid colon is sometimes necessary to achieve optimal cytoreduction (1 8). In a preliminary series, we reported that intraoperative Address correspondence and reprint requests to: Scott M. Eisenkop, MD, Women s Cancer Center, 5525 Etiwanda Ave. Suite 311, Tarzana, CA dobsncats@aol.com aortic clamping may reduce blood loss, operative time, and the transfusion rate for selected patients undergoing these and other pelvic procedures (9). There were no complications resulting from intraoperative aortic clamping. Hence, we performed a prospective, randomized investigation to confirm efficacy of intraoperative aortic clamping in reducing blood loss, operative time, and the transfusion rate for patients requiring a modified posterior pelvic exenteration (MPE) in the context of primary cytoreductive operations for ovarian cancer (7,9). Results of the randomized trial are presented in this report. # 2004 IGCS

2 700 S. M. Eisenkop et al. Fig. 1. Clamping of distal aorta. BIF, aortic bifurcation; IMA, inferior mesenteric artery; IVC, inferior vena cava. Patients and methods Patients undergoing an MPE during primary cytoreductive operations for ovarian cancer, without palpable plaque or calcification in the distal aorta, were randomized intraoperatively to: (a) complete occlusion of the aorta (distal to inferior mesenteric artery for 60 min) with a vascular clamp immediately before the MPE, (b) complete occlusion of the hypogastric arteries, or (c) performance of the procedure without regional blood flow interruption. Patients refusing blood products for religious reasons were excluded from the study and had aortic clamping. All patients had both their cytoreductive and vascular occlusive procedures performed by gynecologic oncologists within the Encino-Tarzana division of the Woman s Cancer Center. The trial was presented to Institutional Review Board and did not require specific approval, because all study arms previously were reported in other contexts, are within practice standards, and neither regional blood flow occlusion arm is associated with increased morbidity (5 13). Informed consent was obtained concerning both aortic cross-clamping and cytoreductive surgery. Virtually, all patients had a pulmonologist/intensivest involved in their postoperative care, and those with a cardiac history had a cardiologist involved perioperatively. The objective of all procedures was a visibly diseasefree cytoreductive outcome. All patients had upper abdominal cytoreduction completed before their pelvic procedures. After achieving hemostasis, operative sites were packed before proceeding to the pelvis. Upon committing to an MPE, the peritoneum lateral to the ovarian vessels was opened to expose retroperitoneal structures and also permit the ovarian vessels to be transected with an endoscopic vascular stapler. A segment of rectosigmoid colon proximal to the pelvic disease was subsequently transected with a stapling device. The mesentery was then transected with a vascular stapling device to permit additional retroperitoneal access and the palpation of the aorta for plaque and calcification. A palpable plaque or calcification within the wall of the distal aorta, at the specific site of potential cross-clamping, excluded the patient from participation in the study. Vascular disease remote from the site of possible clamping was acceptable. Immediately after ascertaining the absence of plaque and calcification at the site of possible aortic occlusion, randomization was accomplished by using the random number generation function on SPSS 11.0 software (SPSS Inc., Chicago, IL). If randomized to aortic occlusion, the aorta was first separated from the vena cava with blunt and sharp dissection and then dissected from the vertebrae with a right-angle or tonsil clamp pointed away from the vena cava. The aorta was completely occluded with a large adjustable bull-dog vascular clamp, placing the spring-like securing mechanism to the left lateral side of the aorta (Fig. 1). The minimal tension that was needed to eliminate a common iliac pulse was used. If randomized to hypogastric artery occlusion, the hypogastric arteries were dissected away from the adjacent hypogastric veins before complete occlusion with a vascular clamp. After achieving regional blood flow occlusion, if randomized to do so, the MPE was subsequently completed (7). Pedicles such as the rectal pillars and parametria were transected with a vascular endoscopic stapler (14,15). The clamp time began upon aortic occlusion, ended with release of the clamp, and was never permitted to exceed 1 h. If necessary, additional time was dedicated to completing the MPE and achieving hemostasis after releasing the vascular clamp. The pelvic procedure time for the group having aortic clamping was defined as the clamp time plus any additional time needed to complete the MPE, irrigate, and achieve hemostasis. The pelvic procedure time for the group undergoing hypogastric artery occlusion began at the time of arterial clamping and ended when the MPE was completed and hemostasis attained. Finally, the pelvic procedure time for the control group without any regional blood flow interruption began at the time when aortic

3 Prophylactic aortic clamping 701 Table 1. Patient characteristics* Characteristic Aortic occlusion (%) Hypogastric occlusion (%) No occlusion (%) Number of patients Age (median, range) (years) 57.2 ( ) 58.1 ( ) 56.6 ( ) Gynecological Oncology Group performance status (21.2) 4 (22.2) 4 (21.1) (78.8) 14 (77.8) 15 (78.8) Stage III 15 (78.8) 14 (77.8) 14 (73.7) IV 4 (21.2) 4 (22.2) 5 (26.3) Site of largest metastatic disease Omentum/other 7 (36.8) 6 (33.3) 8 (42.1) Pelvis 12 (63.2) 12 (66.7) 11 (57.9) Volume of ascites (l) 1 5 (26.3) 6 (33.3) 6 (31.6) >1 14 (73.7) 13 (68.4) 12 (66.7) Largest dimensions of metastatic disease (cm) 10 3 (15.8) 2 (11.1) 4 (21.1) >10 16 (84.2) 16 (88.9) 15 (78.8) Operative procedures (in addition to modified posterior exenteration and omentectomy) Bowel resection 9 (47.4) 9 (50.0) 10 (52.6) Diaphragm stripping 5 (26.3) 6 (33.3) 6 (31.6) Diaphragm resection 3 (15.8) 2 (11.1) 3 (15.8) Splenectomy distal pancreatectomy 3 (15.8) 2 (11.1) 4 (21.2) Peritoneal implant ablation 17 (89.5) 17 (94.4) 16 (84.2) Pelvic and aortic lymph node dissection 18 (94.8) 16 (88.9) 17 (89.5) Urologic resection (partial cystectomy and/ 2 (10.5) 0 1 (5.3) or ureter resection) Hepatic resection 2 (10.5) 2 (11.1) 1 (5.3) Cytoreductive outcome Visibly disease free 17 (89.5) 14 (77.8) 16 (84.2) 1 cm residual disease 2 (10.5) 4 (22.2) 3 (15.8) *Differences between groups all nonsignificant. occlusion would have occurred and ended when the MPE was completed and hemostasis attained. Blood losses were estimated separately for the pelvic procedure and the entire cytoreductive operation by the anesthesiologists and circulating nurses who used separate containers to collect suctioned blood and accounted for laparotomy pads used during the MPE. All patients had rectal anastomosis with an end-to-end stapling technique (6,7). Use of a loop colostomy was individualized based on adequacy of the bowel-prep and medical status. Pelvic and paraaortic lymph node dissections were performed to remove grossly involved nodes and possible occult nodal disease after completing the pelvic procedure and achieving hemostasis for all patients with minimal or no residual disease who were hemodynamically stable. For patients randomized to aortic clamping, heparin was used in a dose of approximately 100 U/kg of body weight intravenously, 5 min prior to occlusion of the aorta distal to the inferior mesenteric artery (10). If indicated, due to the failure to observe clotting prior to the use of heparin, a prothrombin time was obtained to exclude a coagulopathy. The anesthesiologist was informed 5 min prior to and at the time when the clamp was released to allow preemptive administration of additional fluid boluses and the adjustment of medications that could adversely reduce the blood pressure upon removal of the vascular clamp. Subsequently, hemostasis was confirmed while protamine was given in a dose of mg/kg of body weight intravenously over 5min (10). Ankle and/or dorsal pedal pulses were confirmed to be palpable both preoperatively and immediately prior to closure of the abdomen. Comparison of strategies in this investigation with other techniques was accomplished with a literature review using the Unabridged MEDLINE as accessed through Version 4.28 of KNOWLEDGE FINDER software (Aries Systems Corporation, North Andover,

4 702 S. M. Eisenkop et al. MD). Variation of disease characteristics among the groups was analyzed with the Chi-square test for discrete and binomial data and one-way ANOVA analysis of variance for continuous data. One-way ANOVA analysis of variance was also used to evaluate differences in operative outcomes for the study groups. P-values greater than 0.05 are reported as not significant. Statistical analyses were accomplished using SPSS version 11.0 (SPSS Inc.). Results Fifty-six patients were accrued, most requiring extensive cytoreductive operations for advanced disease. All three study groups were equivalent with respect to clinical characteristics such as age, disease severity, extent of upper abdominal surgery done, and cytoreductive outcomes (Table 1). For patients undergoing aortic occlusion, the median clamp time was 32 min (range 18 60). All patients required additional time (1 5 min) to irrigate and assure hemostasis, and one (5.6%) patient required additional time (14 min) to both complete the MPE and achieve hemostasis. The total operative times, pelvic procedure times, total estimated blood losses, blood losses specifically associated with the pelvic procedures, units of blood transfused, and hospital stays were significantly reduced for patients who had aortic clamping (Table 2). Outcomes of the hypogastric artery occlusion group were virtually identical to the control group (Table 2). There were no immediate or delayed vascular complications due to aortic clamping. There was no postoperative mortality. Discussion Survival correlates with the completeness of primary cytoreductive surgery for patients with advanced epithelial ovarian cancer (16,17). Procedures are described to facilitate complete excision of disease at virtually all intra-abdominal locations (2,6,7,18 23). However, cytoreductive operations often require multiple procedures in one setting and predictably have risk of morbidity. Among procedures available for cytoreduction, en bloc resection of internal reproductive organs, pelvic peritoneum, and rectosigmoid for Table 2. Operative outcomes Clinical parameter No occlusion Hypogastric occlusion Aortic occlusion P-value* Pelvic procedure operative time (min) Mean Median Range Total operative time (min) Mean Median Range Pelvic procedure estimated blood loss (ml) Mean Median Range Total estimated blood loss (ml) Mean Median Range Units of packed red blood cells transfused Mean Median Range Hospital stay (days) Mean Median Range *One-way ANOVA analysis over all groups.

5 Prophylactic aortic clamping 703 extensive pelvic disease is reported to be used by 95.6% of gynecologic oncologists (11). However, all descriptions of such procedures report extensive blood loss for some of the patients, and some reports indicate significant blood loss to be common (1 7). Our purpose of investigating aortic occlusion for patients having an MPE is to both reduce morbidity and potentially improve cytoreductive outcomes. In our preliminary series, patients undergoing aortic occlusion during MPE in the context of primary and secondary cytoreductive operations had reduced operative times, blood losses, and transfusion rates compared to other reports (1 4,6 9). However, transfusion criteria vary amongst physicians. Case selection, operator experience, operative objectives, and specific techniques all influence operative time and blood loss. Patients in our preliminary series were selected and operated on by individuals with experience in performing an MPE (7 9,16). The endoscopic stapler was used to transect pedicles such as the parametria and rectal pillars in our preliminary series. Stapling reduces the blood loss and operative time for a radical hysterectomy (14,15). Hence, the possibility that stapling and other issues may have accounted for some of our preliminary observations compelled us to investigate aortic occlusion in a randomized setting. Hypogastric artery occlusion was included as a study arm, despite being reported as to not reduce blood loss during radical hysterectomies, due to both uncertainty about whether the results of a study intended to determine effectiveness in reducing blood loss during radical hysterectomy can be extrapolated to MPE and anticipated criticism for not including a procedure that is essentially free of complications in the study design (12). Additionally, hypogastric artery occlusion was not limited to 60 min, because unlike aortic occlusion, there are no reports to suggest that hypogastric occlusion >60 min increases the risk of complications (10). Furthermore, the study design was intended to determine the operative strategy associated with the least operative time and blood loss. The current trial confirms that hypogastric artery occlusion does not reduce operative time or blood loss when used to control regional blood flow for extensive procedures (Table 2). The current randomized trial clearly demonstrates that intraoperative aortic clamping significantly reduces the blood loss and operative time for an MPE (Table 2). Despite most patients requiring extensive upper abdominal and pelvic procedures, the highly significant reduction in operative time and blood loss associated with the pelvic phases of surgery resulted in a total operative time, total estimated blood loss, and transfusion rate that were significantly lower for the patients having aortic occlusion (Table 2). Regional blood flow occlusion diminishes blood loss and operative time by reducing the flow of blood to the immediate operative field. It is utilized in vascular surgery, trauma surgery, hepatobiliary surgery by occluding the portal triad, and urology by occluding the renal vessels during partial nephrectomy (10,24,25). Aortic occlusion diminishes the blood loss and operative time for an MPE by reducing the flow of blood to the pelvic operative field. Oozing from vascular tumor surfaces as well as backbleeding is minimized. Hence, ongoing blood loss is minimized and exposure is optimized, because the operative field remains essentially blood free. In the event of inadvertent transection or avulsion of small vessels, the reduction of blood flow to the operative field minimizes the immediate blood loss and optimizes exposure, thus facilitating rapid hemostasis. Optimal exposure also potentially reduces the risk of operative complications. Patients randomized to aortic occlusion had a marginally briefer hospital stay (Table 2). Reduction of blood loss and anesthesia time potentially minimizes the risk of infectious complications, fluid balance problems, ileus, electrolyte abnormalities, adult respiratory distress syndrome, pneumonia, coagulopathy, and other complications (13). In addition to defining the clinical benefits of aortic occlusion when used with an MPE, the current randomized trial also confirms that brief aortic occlusion can be performed without ischemic complications. The maximum clamp time of 1 h was chosen, because ischemic damage to extremities is rare if the aorta is clamped for less than 1 h (10). Prolonged clamping is associated with other complications, such as thromboembolic events, hypotension when the clamp is released, and effects from systemic release of accumulated lactic acidosis in the extremities. Risk of these events is minimized by informing the anesthesiologist prior to both occlusion and clamp release so that the medications that induce vasospasm or decrease blood pressure can be avoided and additional fluid may be administered if indicated. Heparin minimizes the risk of a thrombotic complication. Protamine effectively reverses the effect of heparin. However, it may cause hypotension if administered too rapidly (10).Therisk of an embolic event or damage to the aorta increases with the presence of any plaques and calcifications. Plaques and calcifications are reliably identified by palpation (10). Hence, we limited our process of randomization to patients without palpable plaque or

6 704 S. M. Eisenkop et al. calcification within the distal aorta, at the specific site of anticipated cross-clamping, due to concern over vascular injury and plaque embolization. Most patients with significant vascular disease will be precluded from the benefits of intraoperative aortic occlusion. In addition to using aortic occlusion during an MPE for advanced ovarian cancer, we have selectively applied this strategy in other settings. It has been valuable while performing MPE for stage IV endometrial cancer with bulky pelvic disease, as well as exenterations for recurrent cervical cancer. Although unnecessary for most radical hysterectomies, aortic occlusion is useful for severely anemic patients refusing blood products for religious reasons. Finally, the procedure was helpful for a hypotensive patient with a placenta percreta requiring an intraoperative consultation to complete a cesarean hysterectomy and partial cystectomy. Although many patients who have aortic occlusion during an extensive pelvic procedure have a dramatic reduction in blood loss, benefit from aortic occlusion is reduced if the pelvic sidewall or presacral venous plexus is involved with tumor. Under such circumstances, clipping middle sacral and hypogastric veins provides additional hemostasis and exposure, thus permitting the resection of disease and suture placement for hemostasis. Hypogastric vein ligation has also been suggested to decrease the risk of pulmonary embolus (26). Exposing and clamping the aorta is rapidly accomplished with minimal risk of vascular injury. The time required to transect mesentery and to provide retroperitoneal exposure for palpation of the aorta was not included in the pelvic procedure time, because the process was necessary for all patients in the context of their pelvic procedures before the randomization and completion of the MPE. The time required to then isolate and clamp the aorta was not included in the clamp time because of the rapidity of the process after exposure and necessity of an equivalent dissection in the context retroperitoneal node dissections after the MPE was completed (Table 1). A large adjustable tension bull-dog clamp with atraumatic vascular teeth was used, because it can be securely positioned under packing and retractors that are placed to retract upper abdominal contents away from the pelvis (Fig. 1). A large clamp evenly distributes tension. Placement of the spring-like mechanism that secures the clamp lateral to the aorta prevents ongoing contact with and compression of the vena cava (Fig. 1). Although longer clamps such as DeBakey-type vascular clamps allow rapid access if dislodged, their protrusion through the abdominal incision, while not being the center of attention, subjects them to being more readily twisted or moved, both of which could traumatize the aorta. A potential criticism of the current series is that conclusions concerning efficacy and safety are made after accrual of a relatively small number of patients. However, reductions of operative time and blood loss associated with the pelvic phase of surgery are highly significant (P 0.001) (Table 2). Ischemic extremity damage, thromboembolic events, or vascular injury could theoretically occur from using aortic clamping for a very large number of patients. However, such events are rare with adherence to a maximum clamp time of 1 h and the exclusion of patients with distal aortic plaque or calcification from having this procedure (10). The importance of minimizing the time that the aorta is clamped may limit applicability of regional blood flow occlusion in some educational settings. In conclusion, aortic clamping dramatically decreases the blood loss and operative time for patients requiring MPE in the context of primary cytoreductive operations for advanced ovarian cancer. Application of this strategy might permit excision of extensive pelvic disease that would otherwise be considered to be too difficult to excise. Additionally, a significant reduction of the operating time and blood loss associated with the pelvic phase of surgery may indirectly facilitate more complete upper abdominal cytoreduction in some settings. After initiating any cytoreductive operation, determining whether to discontinue the effort is individualized based on the probable benefit, and the potential morbidity associated with continuing to operate. It is our hope that the reduction of the blood loss and operative time associated with the pelvic phases of extensive cytoreductive operations will facilitate complete cytoreduction and hence result in improved survival (16,17). References 1 Guidozzi F, Ball JH. Extensive primary cytoreductive surgery for advanced epithelial ovarian cancer. Gynecol Oncol 1994;53: Sopher JT, Couchman G, Berchuck A, Clarke-Pearson D. The role of partial sigmoid colectomy for debulking epithelial ovarian cancer. Gynecol Oncol 1991;41: Barnes W, Johnson J, Waggoner S, Barter J, Potkul R, Delgado G. Reverse hysterocolposigmoidectomy (RHCS) for resection of panpelvic tumors. Gynecol Oncol 1991;42: Hertel H, Diebolder H, Herrmann J et al. Is the decision for colorectal resection justified by histopathologic findings: a prospective study of 100 patients with advanced ovarian cancer. Gynecol Oncol 2001;83:481 4.

7 Prophylactic aortic clamping Obermair A, Hagenauer S, Tamandl D et al. Safety and efficacy of low anterior en bloc resection as part of cytoreductive surgery for patients with ovarian cancer. Gynecol Oncol 2001;83: Berek JS, Hacker NF, Lagasse LD. Rectosigmoid colectomy and reanastomosis to facilitate resection of primary and recurrent gynecologic cancer. Obstet Gynecol 1986;64: Eisenkop SM, Nalick RH, Teng NNH. Modified posterior exenteration for ovarian cancer. Obstet Gynecol 1991;78: Eisenkop SM, Spirtos NM. Procedures required to accomplish complete cytoreduction for ovarian cancer. Is there a correlation with biological aggressiveness and survival? Gynecol Oncol 2001;82: Eisenkop SM, Spirtos NM, Lin WM, Gross GM. Reduction of blood loss during extensive pelvic procedures by aortic clamping a preliminary study. Gynecol Oncol 2003;88: Rutherford R. In: Vascular Surgery. Philadelphia: W.B. Saunders; Eisenkop SM, Spirtos NM. What are the current surgical objectives, strategies, and technical capabilities of gynecologic oncologists treating advanced epithelial ovarian cancer? Gynecol Oncol 2001;82: Arango HA, Hoffman MS, Roberts WS et al. Does ligation of the hypogastric artery at the time of radical hysterectomy and lymphadenectomy decrease blood loss? Results of a prospective randomized trial. Int J Gynecol Cancer 2000;10: Morrow CP, Curtin JP. Surgery for ovarian neoplasia. In: Gynecologic Cancer Surgery. New York: Churchill Livingstone; Patsner B. Radical abdominal hysterectomy using the ENDO-GIA stapler: report of 150 cases and literature review. Eur J Gynaecol Oncol 1998;19: Brewer CA, Chan J, Kurosaki T, Berman ML. Radical hysterectomy with the endoscopic stapler. Gynecol Oncol 1998;71: Eisenkop SM, Friedman RL, Wang H. Complete cytoreductive surgery is feasible and maximizes survival in patients with advanced epithelial ovarian cancer: a prospective study. Gynecol Oncol 1998;69: Bristow RE, Tomacruz RS, Armstrong DK, Trimble ELS, Montz FJ. Survival effect of maximal cytoreductive surgery for advanced ovarian carcinoma during the platinum era: a meta-analysis. J Clin Oncol 2002;20: Eisenkop SM, Nalick RH, Wang H, Teng NNH. Peritoneal implant elimination during cytoreductive surgery for ovarian cancer: impact on survival. Gynecol Oncol 1993;51: Montz FJ, Schlaerth JB, Berek JS. Resection of diaphragmatic peritoneum and muscle: role in cytoreductive surgery for ovarian cancer. Gynecol Oncol 1989;35: Rose PG. The cavitronal ultrasonic surgical aspirator for cytoreduction in advanced ovarian cancer. Am J Obstet Gynecol 1992;166: Spirtos NM, Gross GM, Freddo JL, Ballon SC. Cytoreductive surgery in advanced epithelial cancer of the ovary: the impact of aortic and pelvic lymphadenectomy. Gynecol Oncol 1995;56: Benedetti-Panici P, Scambia G, Baiocchi G. Technique and feasibility of radical para-aortic and pelvic lymphadenectomy for gynecologic malignancies: a prospective study. Int J Gynecol Cancer 1991;133: Nicklin JL, Copeland LJ, O Toole RV, Lewandowski GS, Vaccarello L, Havenar LP. Splenectomy as part of cytoreductive surgery for ovarian carcinoma. Gynecol Oncol 1995;58: Schwartz SI. Trauma and liver. In: Principles of Surgery. New York: McGraw-Hill; Marshall FF. Partial nephrectomy. In: Textbook of Operative Urology. Philadelphia: W.B. Saunders; Scarabelli C, Imparto E, Biffignandi F, Aspesi G. Hypogastric vein ligation during oncologic surgery as thromboembolic disease prevention. Eur J Gynaecol Oncol 1988;9: Accepted for publication October 3, 2003