See related editorial on pages 3100–2, this issue.
Uterine leiomyosarcoma (ULMS) is identified in 0.1% to 0.2% of hysterectomy specimens of presumed leiomyoma. To date, there is no preoperative technique that reliably differentiates ULMS from uterine leiomyoma. Increasing use of minimally invasive approaches for the management of leiomyomas may result in inadvertently morcellated ULMS with resultant intraperitoneal dissemination of tumor. The objective of this study was to assess the impact of intraperitoneal morcellation on the outcomes of patients with ULMS.
In this retrospective cohort study, all patients with ULMS who attended the authors' institutions from 2007 to 2012 were reviewed. Demographics and outcomes were compared between those who underwent morcellation or total abdominal hysterectomy (TAH) as their first surgery for uterus-limited ULMS.
In total, 58 patients were identified, including 39 who underwent TAH and 19 who underwent intraperitoneal morcellation. Intraperitoneal morcellation was associated with a significantly increased risk of abdominal/pelvic recurrences (P = .001) and with significantly shorter median recurrence-free survival (10.8 months vs 39.6 months; P = .002). A multivariate adjusted model demonstrated a >3 times increased risk of recurrence associated with morcellation (hazard ratio, 3.18; 95% confidence interval, 1.5-6.8; P = .003).
Uterine leiomyomas constitute one of the most frequent gynecologic problems among women in the United States. When symptomatic, surgical resection is a frequent management choice. The use of minimally invasive techniques has increased dramatically during the last 2 decades because of decreased morbidity and length of hospital stay compared with conventional laparotomy or total abdominal hysterectomy (TAH).[3, 4] Intraperitoneal uterine morcellation is one such procedure that involves resection of uterine masses with either a spinning blade or a scalpel and extraction of resected tissue from the uterine cavity through a port incision. Although limited morbidity is expected,[3, 4] the incidence of uncommon but serious complications remains largely unknown. One complication is the inadvertent disruption and dissemination of uterine leiomyosarcoma (ULMS) in women who were presumed preoperatively to have benign uterine leiomyoma.[5-7]
ULMS is the most common subtype of uterine mesenchymal malignancy, with a risk of recurrence of approximately 50% after complete removal by TAH and a median overall survival (OS) of approximately 1 or 2 years in patients with metastatic disease. It is estimated that between 1 in 500 and 1 in 1000 hysterectomy specimens of presumed benign leiomyoma will reveal leiomyosarcoma.[5, 7] In addition, before removal, it is often impossible to predict which uterine masses will represent benign leiomyomas and which will represent ULMS. The primary prognostic factor related to cure of adult soft tissue sarcoma of any site, including leiomyosarcoma, is complete en bloc surgical resection of primary, localized disease with negative margins. On the basis of this concern for inaccurate preoperative assessment of ULMS, the objective of this study was to assess the impact from intraperitoneal morcellation of ULMS on outcomes in patients with a preoperative diagnosis of uterine leiomyoma.
MATERIALS AND METHODS
We conducted an institutional review board-approved, retrospective cohort study at Dana-Farber Cancer Institute, Brigham and Women's Hospital, and Massachusetts General Hospital. All patients with ULMS who attended one of these institutions between 2007 and 2012 were reviewed. Patients who had disease confined to the uterus at presentation, at least 6 months of follow-up, and who underwent either intraperitoneal tumor morcellation (defined as either power morcellation or scalpel morcellation) or TAH without tumor disruption as their first surgery were included. Patients who underwent intraoperative tumor morcellation in an extraction bag or myomectomy were excluded. Patient, disease, recurrence, and survival characteristics were collected.
Clinical Outcome Assessments
The primary outcome was recurrence-free survival (RFS) among patients who underwent intraperitoneal morcellation compared with those who underwent TAH. RFS was defined as the time from initial surgical procedure to first reported recurrence. Patients without recurrence were censored at the date of last disease assessment. The secondary outcome was OS in the 2 groups. OS was defined as the time from primary surgery to death or censoring.
Descriptive statistics were used to describe patient characteristics by surgery type. Groups were compared using a Wilcoxon rank-sum test for continuous variables and chi-square and Fisher exact tests for categorical variables. Kaplan-Meier plots and step-down proportional hazard models were constructed to identify the factors associated with recurrence and survival. A one-sided mid-P test was used to explore difference in incidence rates per person-months of follow-up between groups. Two-sided P values <.05 were considered statistically significance. STATA version 13 (StataCorp, College Station Tex) was used for all analyses.
From 2007 to 2012, 58 patients were identified. In total, 19 patients underwent morcellation, and 39 patients underwent TAH. Patient characteristics are summarized in Table 1.
Table 1. Patient Characteristics and Recurrence and Survival Outcomes
Type of Surgery: No. of Patients (%)
Total, n = 58
TAH, n = 39
Morcellation, n = 19
Abbreviations; FIGO, International Federation of Gynecology and Obstetrics; hpf, high-power fields; NA, not applicable; SD, standard deviation; TAH, total abdominal hysterectomy.
Accurate pathologic staging is not possible in morcellated specimens (see Takamizawa et al, 19997).
P values that account for censoring and patient survival are detailed in Table 2 and in Figures 1 and 2.
At the date of analysis (November 15, 2013), in total, 34 recurrences were identified, including 14 in the morcellation cohort and 20 in the TAH cohort (Table 1). In patients who underwent morcellation, abdominal and/or pelvic recurrence was significantly more common compared with recurrences in other sites (P = .001). The median RFS among patients who underwent morcellation was significantly shorter compared with those who underwent TAH (10.8 vs 39.6 months, respectively; P = .002) (Fig. 1). The incidence rate of recurrence was significantly higher for patients who underwent morcellation (0.048 vs 0.015 recurrences per person-month of follow-up, respectively; one-sided mid-P = .001). The incidence rate ratio of recurrence was 3.14 (95% confidence interval [CI], 1.5-6.5).
Table 2 summarizes the univariate, age-adjusted, and multivariate models we used to assess the association between surgery type and RFS. In univariate analysis, patients in the morcellation cohort had a 3 times increased risk of recurrence compared with patients in the TAH cohort (hazard ratio [HR], 2.95; 95% CI, 1.5-6.0; P = .003). When the model was adjusted for age, the morcellation cohort had an almost 4 times increased risk of recurrence (HR, 3.92; 95% CI, 1.8-8.6; P = .001). A multivariate adjusted model demonstrated that patients in the morcellation cohort had a >3 times increased risk of recurrence when adjusted for mitotic rate of the primary tumor (HR, 3.18; 95% CI, 1.5-6.8; P = .003). Age, postoperative adjuvant treatment, oophorectomy, and other potential confounders were not statistically significant in the multivariate RFS model and were excluded.
In total, 21 deaths (8 patients who underwent morcellation and 13 patients who underwent TAH) occurred during the follow-up period (Table 1). All patients developed recurrences before death, and all but one death was related to recurrent ULMS. The median OS for patients who underwent morcellation procedures was 48 months, whereas the median OS for patients who underwent TAH was not reached. The OS rate at 36 months was 64% in the morcellation cohort and 73% in the TAH cohort (Fig. 2).
An age-adjusted model and a multivariate model adjusted for the risk factor of primary tumor mitotic rate demonstrated that the patients who underwent morcellation had an almost 2 times increased risk of death compared with the patients who underwent TAH, although the difference was not statistically significant (multivariate HR, 1.85; 95% CI, 0.7-4.7; P = .20).
Our current data demonstrate that intraperitoneal morcellation of undiagnosed ULMS worsens the natural history of this malignancy. The procedure is associated with a markedly increased risk of tumor recurrence (particularly in the peritoneum) and significantly decreased RFS compared with TAH. This is most likely because of procedure-related fragmentation and dissemination of ULMS in the peritoneal cavity.[6, 11] Worse OS also was noted in association with intraperitoneal morcellation, although, in this small series, the difference did not reach levels of statistical significance.
This study of a rare cancer is limited by the retrospective methodology, with patients selected from tertiary care referral institutions and a relatively small number of cases in each group. In this retrospective cohort analysis, there were some imbalances in clinical factors between the morcellation and TAH cohorts. Specifically, patients in the morcellation cohort tended to be younger than those who underwent TAH. However, younger age at diagnosis is associated with better outcomes in ULMS; therefore, the difference in age in the current study is unlikely to explain the inferior outcomes of the morcellation group. Also, patients in the TAH cohort more frequently underwent oophorectomy; however, oophorectomy has not consistently been associated with a difference in outcome in patients with ULMS and, thus, this is not likely to explain the differences observed between cohorts. Other factors, including parameters in the validated nomogram of ULMS survival, including tumor size, grade, and mitotic index, were similar between the morcellation and TAH cohorts. The step-wise multivariate models constructed for this analysis accounted for these and other potential prognostic variables, including postoperative treatment, and morcellation retained a significant association with inferior outcomes. In addition, our results are consistent with prior studies demonstrating worsened outcomes in women who undergo morcellation of ULMS.[11, 14-16]
Leung et al have suggested that these negative outcomes should not deter the practice of minimally invasive resection of presumed leiomyomas because of the rarity of ULMS and its inherent poor prognosis.[17, 18] Given the rarity of uterine sarcomas, there have been few studies comparing different surgical modalities, but all of them have consistently demonstrated that tumor disruption impacts negatively on local control and OS; therefore, this procedure is clearly leading to worsened outcomes in a subset of patients with ULMS. Clinical practice guidelines from the European Society of Medical Oncology and the National Comprehensive Cancer Network recommend en bloc tumor resection, without tumor disruption, as the standard of care for localized leiomyosarcoma, which is consistent with the accepted management principles of soft tissue sarcomas arising at any anatomic location.
Although minimally invasive laparoscopic surgical techniques may provide many benefits, in this case, it is critical to note that the risk is associated with the inability to diagnose this rare and potentially life-threatening malignancy with accuracy in advance of definitive surgical resection. It is accepted that morcellation should not be performed in cases of known uterine malignancy. In practical application, however, there are no validated tests or clinical factors that can be used by clinicians to identify preoperatively those patients whose uterine mass represents a sarcoma.[22-24]
Our study, together with previously reported studies, demonstrates that intraperitoneal morcellation of presumed uterine leiomyoma places women who harbor undiagnosed ULMS at risk for worse outcomes. Given the increasing popularity of minimally invasive surgery in the management of presumed uterine leiomyomas, the negative impact on outcomes for women with occult ULMS, and the inability to preoperatively identify ULMS with certainty, patients should be counseled on this rare but clinically significant risk of minimally invasive surgical resection of presumed uterine leiomyoma. For patients in whom the potential benefits of minimally invasive surgery are judged greater than the risks of standard open-exposure surgery, all efforts to eliminate the risks of intraperitoneal dissemination of malignant tissue must be considered.
Dr. George was supported by funding from the Driscoll Family Leiomyosarcoma Fund.