Abdominal Myomectomy Versus Uterine Fibroid Embolization in the Treatment of Symptomatic Uterine Leiomyomas

Abdominal Myomectomy Versus Uterine Fibroid Embolization in the Treatment of Symptomatic Uterine LeiomyomasAbdominal Myomectomy Versus Uterine Fibroid Embolization in the Treatment of Symptomatic Uterine Leiomyomas
OBJECTIVE.
The purpose of this study was to compare treatment efficacy and complica- tions of abdominal myomectomy with those of uterine fibroid embolization in women with symptomatic uterine fibroids.
MATERIALS AND METHODS.
We analyzed the outcomes of 111 consecutive patients who underwent abdominal myomectomy (
n
n
= 67) over a 30- month period. The mean ages of the two groups were 37.7 years (range, 28–48 years) and 44.2 years (range, 31–56 years), respectively. A questionnaire and review of medical records assessed all procedure-related complications and changes in symptoms. Length of hospital stay, time until resumption of daily activities, and pain medication requirements after the pro- cedure were also analyzed.
RESULTS.
Follow-up times for the myomectomy and embolization groups were 14.6 and 14.3 months, respectively. The respective observed success rates in abdominal myomectomy and uterine fibroid embolization patients were 64% versus 92% for menorrhagia (
p
< 0.05), 54% versus 74% for pain (not significant), and 91% versus 76% for mass effect (
p
< 0.05). The complication rates were 25% (abdominal myomectomy) and 11% (uterine fibroid embo- lization) (
p
< 0.05). The respective secondary end points for the two procedures were 2.9 ver- sus 0 days mean hospital stay, 8.7 versus 5.1 days of narcotics use, and 36 versus 8 days until resumption of normal activities. These differences were all statistically significant.
CONCLUSION.
Uterine fibroid embolization is a less invasive and safer treatment op- tion in women with symptomatic leiomyomas than myomectomy. Menorrhagia may be better controlled with embolization, and myomectomy may be a better option in patients with mass effect. Both procedures were equally effective in controlling pain.
terine fibroids are a common finding among women of reproductive age, with an estimated incidence of
20–25% [1]. When associated with symptoms, uterine fibroids are commonly treated with hys- terectomy. Traditionally, uterus-sparing alterna- tives have included hormonal therapy and myomectomies using hysteroscopy, laparos- copy, and laparotomy. Recently, uterine fibroid embolization has emerged as an alternative to these surgical procedures, with a reported success rate of 81–96% [2–5]. Although many fibroid embolizations are performed successfully each year, the role of this procedure in the treatment of women with symptomatic fibroids remains con- troversial. We reviewed our experience with fi- broid embolization and abdominal myomectomy to compare the therapeutic efficacy of these two uterus-sparing procedures in controlling fibroid- related symptoms.
Materials and Methods
The medical records of 111 consecutive patients who underwent abdominal myomectomy (
n
n
= 67) over the same 30-month period (July 1998–December 2000) in an academic institution were retrospectively reviewed. These patients were identified by searching the pro- cedure records of the interventional radiology and gynecology divisions for uterine artery embolization and abdominal myomectomy, respectively. Five pa- tients in the embolization cohort were excluded from analysis because they underwent planned lap- aroscopic myomectomy within 3 months of uterine fibroid embolization to reduce the risk of bleeding at surgery. Similarly, four patients from the myomec- tomy group were excluded because their primary reason for surgery was the treatment of infertility with no other symptoms.
The mean age of the myomectomy cohort was 37.7 years (range, 28–48 years) and that of the fibroid embolization patients was 44.2 years (range, 31–56 years). The mean follow-up times for the myomec-
Mahmood K. Razavi
Received September 5, 2002; accepted after revision October 30, 2002.
1
Department of Vascular and Interventional Radiology, H-3651, Stanford University Vascular Center, 300 Pasteur Dr., Stanford, CA 94305. Address correspondence to M. K. Razavi.
2
AJR
AJR:180, June 2003
Razavi et al.
tomy and embolization groups were 14.6 and 14.3 months, respectively. The indications for treatment in- cluded menorrhagia; pain, pressure and pelvic dis- comfort; and mass effect causing abdominal distention or urinary tract symptoms. Patients in both groups expressed a strong desire to avoid hysterec- tomy. Patient characteristics are described in Table 1.
A telephone questionnaire was used to assess the change in symptoms in both groups. The changes in patients’ symptoms in the three catego- ries listed above (bleeding, pain, and mass effect) were rated on a 6-point scale (6, completely re- solved; 5, significantly improved; 4, moderately improved; 3, no change; 2, moderately worse; and 1, significantly worse).
Fibroid Embolization Patients and Procedure
Bleeding was present in 52 patients (84%), pelvic pain in 34 (55%), and mass effect in 37 (60%). All patients expressed a strong desire to avoid surgical intervention. Workup before the procedure included a gynecologic evaluation by a gynecologist, routine history and physical examination, renal function tests, and pelvic MR imaging or sonography. Fi- broids were determined to be the cause of the pa- tients’ symptoms in all cases. Selective uterine artery embolization was performed using 350–500 µm of polyvinyl alcohol or 500–700 µm of Embo-
sphere particles ([trisacryl gelatin microsphere] Bio- sphere Medical, Rockland, MA).
Myomectomy Patients and Procedure
Complaints included menorrhagia in 22 pa- tients (55%), pelvic pain in 26 (65%), and mass ef- fect in 23 (58%). Myomectomies were performed through a Pfannenstiel’s incision; as many fibroids as possible were removed. If blood loss was con- sidered excessive, pedunculated and small subse- rosal myomas were spared.
Measures of Outcome
Outcome of the procedures was tested with re- spect to change in patients’ symptoms. Primary end points were defined as successful control of symp- toms as expressed by a score of 5 or 6 in the patient questionnaire; major adverse events leading to death, additional procedures, prolongation of the hospital stay, any procedure-related undesirable out- come requiring treatment or clinic visits within 30 days of the index procedure as assessed by the ex- amination of medical records; and bleeding compli- cations requiring nonautologous blood transfusion.
Secondary end points were defined as the mean length of stay at the hospital, days on narcotic medications after the procedure, duration before resumption of normal daily activities, and number
of secondary interventions during the follow-up period. Success in control of each symptom was analyzed independently.
Statistical Analysis
t
test was used to compare mean differences between treatment groups. Fisher’s exact test or the chi-square test was used when comparing differences in proportions. Our calculation indicated that the sample size required per group to detect a 50% difference in proportion of patients who experi- enced significant improvement in symptoms is 31 [6]. Excel software (Microsoft, Redmond, WA) was used for statistical analysis.
Results
Uterine artery embolization was techni- cally successful, with stasis of flow in the main uterine artery (Fig. 1) in all patients but one who had unilateral uterine artery embo- lization because of difficult catheterization. Patient characteristics before the procedure are listed in Table 1. With the exception of mean age and an initial symptom of menor- rhagia, no differences existed in these char- acteristics between the two groups.
Table 2 summarizes the responses for each category of outcome for the primary end points. Successful outcome (defined as a cat- egory 5 or 6 response) was observed in 14 (64%) of the myomectomy patients and in 48 (92%) of the embolization patients with menorrhagia (
p
< 0.05). Pain was relieved in 14 (54%) and 25 (74%) of the patients (
p
> 0.05), and treatment of mass effect was suc- cessful in 21 (91%) and 28 (76%) women, respectively, in the myomectomy and embo- lization groups (
p
< 0.05). According to the definition of adverse
events in this analysis, 10 complications oc- curred in the myomectomy (25%) and seven
Note.—Numbers in parentheses are percentages unless otherwise noted. NS = not significant.
TABLE 1 Patient Characteristics Before Procedure
Characteristic Patients Undergoing
p Abdominal Myomectomy Uterine Fibroid Embolization
No. of patients 40 62 Age (yr) 37.7 (range, 28–48) 44.2 (range, 31–56) < 0.05 Menorrhagia 22 (55) 52 (84) < 0.05 Pain 26 (65) 34 (55) NS Mass effect 23 (58) 37 (60) NS Hematocrit (mg/dL) 36 (range, 27–44) 35.5 (range, 25–52) NS Mean follow-up (mo) 14.6 14.3 NS
TABLE 2 Outcomes of Primary End Points Expressed as Number and Percentage of Responders Who Underwent Abdominal Myomectomy or Uterine Fibroid Embolization
Response Category Menorrhagia Pain Mass Effect
Myomectomy Embolization Myomectomy Embolization Myomectomy Embolization
6, Completely resolved 6 (27) 31 (60) 10 (38) 10 (29) 15 (65) 7 (19) 5, Significantly improved 8 (36) 17 (33) 4 (15) 15 (44) 6 (26) 21 (57) 4, Moderately improved 5 (23) 3 (6) 10 (38) 8 (24) 2 (9) 7 (19) 3, No change 2 (9) 1 (2) 1 (4) 1 (3) 0 2 (5) 2, Moderately worse 1 (4.5) 0 1 (4) 0 0 0 1, Significantly worse 0 0 0 0 0 0
Total 22 52 26 34 23 37 Successful outcome (scores 5 and 6) 14 (64) 48 (92) 14 (54) 25 (74) 21 (91) 28 (76)
D ow
nl oa
de d
fr om
w w
w .a
jr on
lin e.
or g
by 2
7. 79
.7 5.
39 o
n 02
/1 1/
23 f
ro m
I P
ad dr
es s
27 .7
9. 75
.3 9.
C op
yr ig
ht A
R R
S. F
or p
er so
AJR:180, June 2003
p
n
n
= 1) among the myomectomy patients. We did not exam- ine the intraoperative conversion rate to hyster- ectomy in this study. Complications in the fibroid embolization group included en- dometritis requiring readmission for IV antibi- otics (
n
n
n
=
4). All who experienced menopause after fibroid embo- lization were older than 46 years. The mean hematocrit decrease in the surgical patients
was 6.4 mg/dL (range, 0.9–13.5 mg/dL). Hematocrit level after the procedure was not measured in any of the fibroid embo- lization patients.
The results of the secondary end points are listed in Table 3. The differences be- tween the two groups were statistically sig- nificant in all areas examined except in the number or patients undergoing secondary interventions. During the follow-up period, repeated procedures for the treatment of fi- broids were performed in four (10%) of the 40 myomectomy patients and five (8%) of the 62 fibroid embolization patients. The re- peated procedures were three myomecto- mies and one hysterectomy in the abdominal myomectomy group and three myomecto-
mies and two hysterectomies in the embo- lization group.
Discussion
In an attempt to better define the role of uterine fibroid embolization in women with symptomatic fibroids, we retrospectively compared this procedure with the standard uterus-sparing abdominal myomectomy. On the basis of the observations of this study, women in whom menorrhagia is the domi- nant problem do better with fibroid embo- lization. Bleeding resolved in 92% of the embolization group versus 64% of the myo- mectomy patients (
p
< 0.05). A feasible ex- planation for the failure of myomectomy to
A B
Fig. 1.—38-year-old woman with uterine fibroids. A, Before embolization, pelvic angiogram reveals enlarged uterine arteries (arrows) bilaterally. B, After embolization, pelvic arteriogram shows absence of uterine arteries. C and D, Selective right (C) and left (D) uterine angiograms show hypervascular tumor. E, Selective left uterine arteriogram after embolization shows stasis for flow in main uterine artery after embolization.
C D E
AJR:180, June 2003
Razavi et al.
control bleeding may be that the culprit fi- broid may not always be removed.
Conversely, patients with complaints about mass effect may have a better outcome with surgery than with embolization. Treat- ment of symptoms caused by the mass effect of fibroids was successful in 91% and 76% of myomectomy and fibroid embolization patients, respectively (
p
< 0.05). This finding is not surprising because fibroids are not physically removed with embolization. Fur- thermore, it may take many weeks before any significant reduction in fibroid size is re- alized with fibroid embolization, whereas re- duction in uterine size is immediate after surgery. For the same reason, those who seek treatment of fibroids only for cosmetic rea- sons may not be appropriate candidates for embolization. Although on the average the volume of the fibroids decreases by 30–60% after embolization [2–4, 7], the degree of re- duction varies from patient to patient and cannot be predicted. Relief of pain was simi- lar in both groups, being successful in 54% of the surgical and 74% of the embolization patients (
p
> 0.05). The safer nature of fibroid embolization as
compared with myomectomy is reflected in the complication rates (11% vs 25%) and the data on secondary end points (Table 3). Fewer inpatient hospital days (0 vs 2.9), shorter duration of the use of narcotics for pain (5.1 vs 8.7 days), and less time until re- sumption of normal daily activities (8 vs 36 days) were seen with fibroid embolization as compared with myomectomy.
Another symptom that would necessitate the removal of myomas is fibroid-related in- fertility. This issue has not been well studied in the setting of fibroid embolization. Most centers (including ours) do not offer embo- lization as a treatment for fibroid-related in- fertility. However, that does not mean
embolization is contraindicated in women desiring to preserve their childbearing poten- tial. Similar fertility rates to myomectomy have been reported after fibroid embolization [8]. A small risk exists of ovarian failure af- ter embolization; however, failure occurs mainly in women older than 45 years [9, 10].
The cumulative probability of recurrence of fibroids after myomectomy increases with time and has been reported to be 27–51% [11, 12]. In a study of outcome and resource use as- sociated with myomectomy, Subramanian et al. [13] reported repeated surgeries in 16.5% of women within 2 years of myomectomy; others have reported rates as high as 50% [14]. Among our patients who had abdominal myo- mectomy, the incidence of recurrence of fi- broids is not known because few had imaging follow-up. However, the number of repeated surgeries after myomectomy for recurrence of symptoms (10%) was lower than that in other studies. The efficacy rates of 54% and 64% for pain and bleeding after surgery were also lower than in some previous reports [15]. Sim- ilar discrepancies were also observed in the embolization group. Our embolization results in controlling menorrhagia paralleled those in published series [2–5, 7], but successful out- come was achieved in fewer patients in the pain and mass effect categories than in those reports. One explanation of these discrepan- cies might be that our study was a symptom- specific analysis evaluating changes in each one of the fibroid manifestations separately, whereas other studies grouped the symptoms together in their analyses. Furthermore, only scores of 5 or 6, corresponding to significant improvement or complete resolution of symp- toms, were considered successful. Those who had moderate improvement were not included in the calculations of treatment success. There- fore, our strict definition of treatment success, our relatively short follow-up time, and our po-
tential population differences from other stud- ies may explain the differences in our outcomes and those of others for both groups.
The limitations of our study stem mainly from its retrospective nature. The search for the abdominal myomectomy group yielded only those patients who actually underwent (rather than those who were planned to un- dergo) myomectomy. The reason for this is that we searched for the term “abdominal myomectomy” in the surgical logs of the gy- necology department, which document only the performed surgeries and not the intended ones. The intraoperative conversion rate to hysterectomy is therefore not known. This fact would underestimate the complications in the myomectomy patients by about 2–3.7%, which is the estimated rate of intraoperative conversion of myomectomy to hysterectomy [13, 14]. Another problem with our study de- sign is the potential inability of patients to accurately recall details such as the length of time taking pain medication or until resump- tion of normal daily activities. Because we had a similar follow-up time for both groups, however, the likelihood of this type of error should be the same for both groups. We have found that the data collected prospectively in patients who have had fibroid emboliza- tion since the termination of this study are similar to the data obtained here, which sug- gests that the error introduced by inaccurate recall is not likely to be significant. Further- more, duration of recuperation reported by our abdominal myomectomy patients was consistent with that for most laparotomies reported previously [16]. Finally, the differ- ences, such as the mean age, between the two groups studied could introduce a bias into the results.
The preliminary data presented here suggest that fibroid embolization is safer than abdomi- nal myomectomy and has a shorter recovery time. Efficacy for the treatment of menor- rhagia appears to be greater with embolization; conversely, surgery may be a better choice for symptoms stemming from the mass effect of fibroids. Factors that should affect the choice of treatment in women with symptomatic uter- ine fibroids include the chief complaint, ana- tomic variables such as number and location of fibroids, patient expectations, and treatment goals. Depending on the chief complaint and the goals of treatment, fibroid embolization should be offered as a treatment option in this patient population.
Note.—Numbers in parentheses are ranges unless otherwise noted. NS = not significant.
TABLE 3 Outcome of Secondary End Points
End Point Patients Undergoing
p Abdominal Myomectomy Uterine Fibroid Embolization
Inpatient hospital days 2.9 (2–7) 0 < 0.05 Days taking pain medication 8.7 (2–47) 5.1 (1–21) < 0.05 Days until normal activity 36 (7–120) 8 (1–49) < 0.05 Secondary interventions (no.) 4 (10%) 5 (8%) NS Estimated blood loss (mL) 376 (50–2000) Minimal
D ow
nl oa
de d
fr om
w w
w .a
jr on
lin e.
or g
by 2
7. 79
.7 5.
39 o
n 02
/1 1/
23 f
ro m
I P
ad dr
es s
27 .7
9. 75
.3 9.
C op
yr ig
ht A
R R
S. F
or p
er so
AJR:180, June 2003
1575
References
1. Cramer SF, Patel A. The frequency of uterine lei- omyomas.
Am J Clin Pathol
1990
;94:435–438 2. Spies JB, Ascher SA, Roth AR, Kim J, Levy EB, Go-
mez-Jorge J. Uterine artery embolization for uterine leiomyomata.
Obstet Gynecol
2001
;98:29–34 3. Goodwin SC, McLucas B, Lee M, et al. Uterine
artery embolization for the treatment of uterine leiomyomata: midterm results.
J Vasc Interv Ra- diol
1999
Jr. Uterine arterial embolization for the manage- ment of leiomyomas: quality-of-life assessment and clinical response.
Radiology
1998
therapy for myoma-induced menorrhagia.
2000
Clinical trials.
:127–129
7. Spies JB, Roth AR, Jha RC, et al. Leiomyomata treated with uterine artery embolization: factors associated with successful symptom and imaging outcome.
Radiology
2002
;222:45–52 8. McLucas B, Goodwin S, Adler L, Rappaport A,
Reed R, Perrella R. Pregnancy following uterine fibroid embolization.
Int J Gynecol Obstet
2001
;74:1–7 9. Chrisman HB, Sakter MB, Ryu RK, et al. The im-
pact of uterine fibroid embolization on resump- tion of menses and ovarian function.
J Vasc Interv Radiol
2000
;11:699–703 10. Razavi MK, Wolanske KA, Hwang GL, Sze DY,
Kee ST, Dake MD. Angiographic classification of ovarian artery-to-uterine artery anastomoses: ini- tial observations in uterine fibroid embolization.
Radiology
2002
;224:707–712 11. Fedele L, Parazzini F, Luchini L, et al. Recur-
rence of fibroids after myomectomy: a transvagi-
nal ultrasonographic study.
; 10:1795–1796
12. Candiani GB, Fedele L, Parazzini F, et al. Risk of recurrence after myomectomy.
Br J Obstet Gy- naecol
1991
;98:385–389 13. Subramanian S, Clark MA, Isaacson K. Outcome
and resource use associated with myomectomy.
Obstet Gynecol
2001
;98:583–587 14. Iverson RE Jr, Chelmow D, Strohbehn K, Wald-
man L, Evantash EG. Relative morbidity of ab- dominal hysterectomy and myomectomy for management of uterine leiomyomas.
Obstet Gy- necol
;88:415–419 15. Buttram VC Jr, Reiter RC.…