Pathology of Power Morcellation: A 12-Year Retrospective Analysis

Introduction

Minimally invasive approaches to gynecologic surgery have clearly established advantages, including reduced morbidity, shorter hospital stays, and quicker recovery.^1,2 However, these advantages of minimally invasive laparoscopy are limited by the difficulty associated with removing large specimens. Since its introduction in 1993 as a cylindrical device with a coning knife, morcellation has been commonly utilized to facilitate the delivery of large uteri or uterine myomas without the use of a conventional laparotomy.^3,4 While morcellation is associated with risks inherent to many gynecologic operations, including vascular and bowel injury, a specific concern is morcellation's potential to disseminate tissue. ⁵ Several case reports have shown that fragments of tissue remaining inside the abdominal cavity could result in complications such as pain and persistent disease. ⁶ Instances of iatrogenic implants secondary to morcellation have included endometriosis, parasitic leiomyomas, and malignancies including leiomyosarcomas.^7–9

In April 2014, with rising public concern, the U.S. Food and Drug Administration (FDA) released a report discouraging the use of power morcellation in laparoscopic hysterectomy and myomectomy. ¹⁰ This report raised concerns about if dissemination is a common risk among patients undergoing all types of laparoscopic procedures with morcellation and if overall survival is therefore compromised. After this report was released, the number of laparoscopic supracervical hysterectomies declined and the opportunity to pursue alternative routes of hysterectomy, such as vaginal hysterectomy, ensued. ¹¹ The present study set out to evaluate further the risk associated with dissemination in laparoscopic procedures with morcellation.

Materials and Methods

Following institutional review board approval, a retrospective observation was performed for a case series of women who underwent laparoscopic myomectomy or supracervical hysterectomy at a single academic institution between January 2002 and May 2014. Current Procedural Terminology codes for laparoscopic myomectomy (58551) and laparoscopic supracervical hysterectomy (58541, 58542, 58543, 58544) were used to identify patients. Cases were not included if the specimen had been removed without the use of power morcellation; this procedure was confirmed via identification in the pathology report of a morcellated specimen.

Demographics of the included patients were compiled including each patient's age, body mass index (BMI), and menopausal status if documented. The indications for surgery were varied, and the preoperative diagnoses were noted. Preoperative endometrial and uterine evaluations were also assessed which included biopsy; dilation and curettage; and imaging modalities, including ultrasound (US), computed tomography (CT), and magnetic resonance imaging (MRI). Finally, the procedure lengths, specialty of the attending surgeons, and postoperative diagnoses were reviewed. Final uterine pathology reports were also obtained, and charts of patients with pathology results that were of concern were evaluated further for follow-up information.

Statistical analysis was undertaken using SPSS software (version XX.0; IBM Corp., Armonk, NY). Chi-squared tests were used to compare nominal variables and Pearson's correlations were used to compare continuous variables.

Results

There were 456 women identified who underwent laparoscopic myomectomy or supracervical hysterectomy with the use of power morcellation during the study period, with the majority of patients undergoing hysterectomy (82%). As shown in Table 1, the majority of patients (n = 355; 78%) were premenopausal with a median BMI of 26 (range: 19–50). The most common preoperative diagnoses were uterine fibroids (n = 260; 57%), abnormal uterine bleeding (AUB; n = 106; 22%), and pelvic organ prolapse (n = 97; 21%).

General gynecologists performed the majority of surgeries (58%), followed by urogynecologists (22%) and reproductive endocrinologists (19%). Three hundred and three patients underwent preoperative uterine evaluation with US (50%), endometrial sampling (30%), or MRI (8%), with 21% of overall patients having undergone more than one testing modality. Patients with the preoperative diagnosis of fibroids were more likely to undergo preoperative endometrial sampling than patients with other diagnoses (32% versus 16%; p < 0.05). Urogynecologists were less likely to perform preoperative endometrial evaluations, compared to generalists or reproductive endocrinologists; however this finding was not statistically significant (12% versus 80%; p = 0.23). In addition, the most common preoperative diagnosis for patients with urogynecologists was prolapse (n = 74; 86%), in comparison to fibroids (n = 193; 72%) as the most common diagnosis among patients with generalists. The median age among the patients operated on by urogynecologists was 55 (range: 31–70) years, and the median age among the patients operated on by generalists was 46 (range: 34–65) years.

One morcellated uterine specimen (0.2%) had an undiagnosed atypical leiomyoma, with moderate-to-severe cytologic atypia, and increased mitosis up to 5/10 high-power fields, with no caseating tumor necrosis. Further characterization was completed via immunostaining that revealed diffuse immunoreactivity for p16 and Ki-67 indexes up to 15%. The 37-year-old patient had not undergone preoperative endometrial sampling. Her case was discussed with the attending reproductive endocrinology and infertility (REI) surgeon, and was also reviewed at the gynecology oncology intradepartmental quality assurance conference. By reviewing the Bell Criteria based upon a Stanford series of 213 patients, her tissue stratified her into Group 2, which warranted the term atypical leiomyoma with a low risk of recurrence. Follow-up imaging was recommended. ¹² She was reevaluated with repeat MRI annually for 2 years and had no further sequelae.

Another uterine specimen (0.2%) had benign leiomyoma pathology, but, upon reviewing her chart, the patient was noted to have extensive follow-up with her REI surgeon within 1 year of her supracervical hysterectomy, given that she had abdominal pain and CT findings that were suspicious for parasitic leiomyomas. This patient had been also evaluated by the gynecology oncology department and was thought to have benign imaging with low utility of biopsy. The patient's planned treatment was Lupron® with add-back therapy for her abdominal pain, but she was lost to follow-up.

Discussion

In the present study, 456 women who underwent laparoscopic procedures with morcellation were evaluated, revealing a low rate of occult malignancy (0.2%). In this study, only 1 woman was identified via pathology testing with atypical leiomyoma who had not undergone preoperative endometrial sampling. Since the 2014 FDA recommendation against laparoscopic power morcellation, several institutions have completed reviews of occult malignancies noted at the time of hysterectomy. The largest review, utilizing data from more than 10,000 hysterectomies at the University of Texas Southwestern Medical Center, found the incidence of unexpected uterine sarcoma to be 1 in 1124 (0.08%). ¹³ Of the 9 cases of unexpected uterine sarcoma, none followed the use of power morcellation. Furthermore, the most common primary indication for hysterectomy among the cases of occult sarcoma was AUB (78%).

A similar study published in 2015, provided an analysis of the Michigan Surgical Quality Collaborative database. Mahnert et al. noted a 2.7% incidence of occult malignancy among 7499 hysterectomies performed for benign indications but only a 0.22% rate of uterine sarcoma. ¹⁴ Similar to the study from Texas, the most common indications for hysterectomy was AUB and leiomyoma (65%). However, the Mahnert et al. study was limited by the lack of information regarding preoperative endometrial evaluation. It is likely that the higher rates of occult malignancies compared to the current study's findings are due to varied sampling of higher-risk patients and differences in preoperative evaluation. Thus, Mahnert et al.'s findings are not necessarily applicable to a population of women for whom preoperative evaluation and/or endometrial sampling led to the decision to utilize morcellation as these woman might have represented a lower-risk population.

The current study's pathology review was restricted to cases with morcellated specimens, only to find 1 case of atypical mitosis that was solely evaluated preoperatively via MRI. This raises the question of the utility of endometrial sampling and if it might have benefited the patient prior to her procedure.

The current study's chart review also showed dispersal of varied, nonmalignant pathologies via morcellation. Case reports have shown iatrogenic parasitic leiomyomas and other uterine pathologies, including iatrogenic endometriosis, parasitic adenomyomas, and peritoneal implants with complex atypical hyperplasias.^7,15,16 Parasitic myomas and other pathologies can present after small tissue fragments are released from morcellation and implanted onto peritoneal or omental tissue to obtain an alternative blood supply. ¹⁷ A 2007 review at Kaiser Permanente San Diego found 10 patients among 942 (1.1%) with the diagnosis of uterine sarcoma or parasitic myoma; 3 patients were diagnosed with uterine sarcoma on pathologic examination at the time of hysterectomy, another 3 were discovered after sampling 2–7 years later, and 4 patients had spindle-cell neoplasms suspicious for parasitic myoma. ¹⁸ This higher incidence rate could be accounted for by the inclusion of patients with delayed presentation of uterine sarcoma after initial disease at the time of procedure was benign, given the long-term nature of care at the researchers' institution. Conversely, the current study found 1 case of a parasitic myoma, but this is an incidence rate likely limited by loss of follow-up and the less-inclusive nature of the current authors' organization.

The current study's strengths include a large cohort of patients over a span of 12 years. The study was limited by the lack of a comparative group of women who underwent hysterectomy or myomectomy without the use of power morcellation. The current study also did not evaluate all potential uses of morcellation, including vaginal morcellation, morcellation through a mini-laparotomy incision, and morcellation within a bag. As a response to the FDA safety communication, multiple innovative methods have been established for enclosed morcellation, all of which do not have long-term data to confirm their theoretical benefits.^19–21

Another study limitation was the small occurrence of abnormal pathology and no occurrences of uterine sarcoma. This small occurrence of abnormal pathology was, however, consistent with the previous studies described above and with a recent national database study, which revealed a low prevalence of uterine cancer in patients who underwent morcellated myomectomy (0.09%), as well as without power morcellation (0.19%). ²² That study was able to demonstrate a low overall prevalence, with increase in age as a risk factor; however, that study was limited by an inability to verify pathology reports. ²²

Although an abundance of cohort studies have now been completed to estimate the risk of uterine sarcoma in patients undergoing laparoscopic surgery, further studies need to be done to estimate the implications. A 2015 literature review showed a lack of evidence to support an increased risk for upstaging of disease or decreased survival associated with peritoneal dissemination or occult malignancy after morcellation. ²³ This could weaken the discouragement power morcellation, but it does not justify the thoughtless use of the device and does emphasize the lack of complete evidence further.

Conclusions

The current study demonstrated that occult malignancy is rare in morcellated specimens following benign gynecologic surgery. Residual tissue is another outcome that should also be considered as a possible consequence of morcellation. Benign disseminated tissue may radiologically appear malignant or produce ensuing symptoms in a patient, prompting tissue diagnosis and surgical intervention. Maintaining the availability of power morcellation to aid in removal of specimens during minimally invasive hysterectomy and myomectomy should be at the discretion of the physician as well as the understanding of the patient.