Pilot Study

Comparison of Clinical Outcomes with Low-Voltage (Cut) VersusHigh-Voltage (Coag) Waveforms during Hysteroscopic EndometrialAblation with the Rollerball: A Pilot Study

Paul T. Chang, MD*, George A. Vilos, MD, Basim Abu-Rafea, MD, Jackie Hollett-Caines, MD,Zoreh Nikkhah Abyaneh, MD, and Fawaz Edris, MDFrom St. Joseph’s Health Care, Obstetrics and Gynaecology (all authors), London, Ontario, Canada.

ABSTRACT Study Objective: To compare efficacy of rollerball endometrial ablation with low-voltage (cut) versus high-voltage

The authors have

products or comp

This pilot study

Invasive Gynaeco

Corresponding au

and Gynaecology

E-mail: pticog@g

Submitted Novem

Available at www

1553-4650/$ - see

doi:10.1016/j.jmig

(coag) waveforms.

Design: Pilot comparative clinical study (Canadian Task Force Classification II-1).

Setting: University-affiliated teaching hospital.

Patients: Fifty premenopausal women with menorrhagia.

Intervention: Women with menorrhagia were allocated randomly to thermal destruction of the endometrium by a 5-mm

rollerball with unmodulated cutting current or modulated coagulating current. Complication rate, clinical outcomes, and

need for reintervention were evaluated.

Measurements and Main Results: At 2 years of follow-up, the reintervention rate was 26.3% in the cutting waveform group

versus 31.4% in the coagulating waveform group. This difference was not statistically significant. Hysterectomy was performed

in 3 (14%) women in the cutting waveform group and 5 (20%) women in the coagulating waveform group. There were no

complications in either group.

Conclusion: Both cutting and coagulating waveforms are equally effective for hysteroscopic endometrial ablation with the

rollerball. Journal of Minimally Invasive Gynecology (2009) 16, 350–3 � 2009 AAGL. All rights reserved.

Keywords: Hysteroscopy; Rollerball endometrial ablation; Radiofrequency waveform; Menorrhagia

Menorrhagia is a common problem in women of repro-

ductive age, especially around the time of menarche and

menopause [1,2]. Causes include intrauterine disease, coa-

gulopathy, and hormonal imbalance. Management options

are medical and surgical: medical options include antifibri-

nolytic, nonsteroidal antiinflammatory drugs, combined oral

contraceptives, and the levonorgestrel-releasing intrauterine

system; surgical options include endometrial ablation or

resection and hysterectomy [2].

no commercial, proprietary, or financial interest in the

anies described in this article.

was presented at the 27th Global Congress of Minimally

logy. October 28-November 1, 2008, Las Vegas, Nevada.

thor: Paul Chang, MD, St. Joseph’s Health Care, Obstetrics

, 268 Grosvenor St, London, Ontario N6A 4V2, Canada.

mail.com

ber 20, 2008. Accepted for publication March 2, 2009.

.sciencedirect.com and www.jmig.org

front matter � 2009 AAGL. All rights reserved.

.2009.03.001

Controversy exists regarding the most appropriate radio-

frequency waveform for hysteroscopic endometrial tissue

destruction to yield optimal clinical outcomes [3–5]. Cut

and coag waveforms differ in their modulation and peak-

to-peak voltage. Modulation is defined as the periodic

interruption of current flow. The cut waveform has no mod-

ulation; it is a pure sine wave with a relatively low peak-to-

peak voltage. Given the same wattage, coagulating

waveform is highly modulated with high peak-to-peak volt-

age. An electrosurgical generator provides the required

wattage by multiplying the voltage times the current. Theo-

retically, the low voltage cut waveform with slower heating

of tissue is more likely to result in deep and homogeneous

tissue penetration [6–9]. The high voltage modulated coag

waveform is more likely to result in superficial tissue pen-

etration as a result of rapid tissue desiccation and

subsequent buildup of tissue impedance. An in vitro study

supported these presumptions by showing maximal thermal

injury into the myometrium of 4.2 mm when using the cut

waveform versus 3.0 mm when using the coag waveform

Table 1

Patient characteristics

Cut (n 5 22) Coag (n 5 25)

Mean 6 SD 95% CI Mean 6 SD 95% CI p Value

Age (y) 41.6 6 4.5 39.6–43.5 39.8 6 5.6 37.5–42.1 .23

BMI (kg/m3) 27.6 6 6.0 24.9–30.2 27.8 6 6.3 25.2–27.8 .91

Parity 1.9 6 1.3 1.8 6 1.0 .63

Dysmenorrhea 59.1% 60.0% .95

Table 2

Endometrial histologic sample obtained at D&C

Cut Coag

Inactive 4.6% 12.5%

Decidualized 0.0% 4.2%

Proliferative 45.5% 41.7%

Secretory 40.9% 33.3%

Menstrual 4.6% 8.3%

Chang et al. Clinical Outcomes of Two Radiofrequency Waveforms for Endometrial Ablation 351

[5]. To our knowledge, this is the first prospective clinical

study comparing clinical outcomes after rollerball endome-

trial ablation (REA) with cut versus coag waveforms. This

study was undertaken to serve as pilot for a larger

randomized trial.

Patients and Methods

From November 2004 through March 2005, 50 cases of

rollerball endometrial ablation were performed for manage-

ment of menorrhagia in premenopausal women. Exclusion

criteria included atypical endometrial hyperplasia determined

by office biopsy, intramural or submucosal myomas of

R3 cm diameter, and active pelvic inflammatory disease.

One experienced physician (G.A.V.) in hysteroscopic

surgery and 2 surgical fellows performed all procedures at

St. Joseph’s Health Care, London, Ontario, Canada. If not

doing the procedure himself, he was physically present in

all cases. The endometrium was not medically pretreated.

Instead, dilation and curettage (D&C) was performed just

before endometrial ablation. The interventions were carried

out with patients under general anesthesia with 5-mm roll-

erball electrodes by use of the ERBE ICC 350 electrico-

surgical unit. Glycine solution (1.5%) was used for uterine

distension via an electronic fluid-management system

(Endomat; Storz, Tuttlingen, Germany). In half of these

patients, the cut waveform was used and the other half

the coag waveform. Patients’ treatment allocation was

according to computer-generated randomization numbers.

Power was set at 100 watts for both groups. Clinical out-

comes measured were rates of menstrual reduction, need

for reintervention, and patient satisfaction. The information

was collected at 2 years after REA from office charts and

telephone interview. Three patients in the cut waveform

group were lost to follow-up and therefore excluded from

the study.

Data Analysis

After Health Sciences Research Ethics Board approval,

the office and hospital charts were reviewed, and data

were recorded anonymously. Data were analyzed on SAS

Software (SAS Institute Inc., Cary, NC), with Student’s ttest or Wilcoxon signed-rank test for interval data and c

2

analyses for comparison of nominal data. A p value ,.05

was considered statistically significant.

Results

Table 1 illustrates the patient characteristics of the 2

cohorts. In both groups, 60% of women had concurrent

dysmenorrhea. Endometrial histologic samples obtained

from D&C at the time of surgery revealed mostly prolifer-

ative or secretory endometrium. In the cut cohort, 45.5% of

the D&C samples were proliferative, and 40.9% were secre-

tory versus 41.7% and 33.3%, respectively, in the coag

cohort (Table 2).

No surgical complications occurred in either group. The

clinical outcomes, menstrual blood loss, patient satisfaction,

and reintervention rates are shown in Table 3. Three pa-

tients in the cut cohort and 5 patients in the coag cohort re-

quired hysterectomy whereas the rest had repeat resection

of the endometrium. In 1 instance, a patient in the coag

group had uterine perforation during repeat ablation. She

subsequently underwent hysterectomy for persistent pain

and bleeding.

The indications for hysterectomy in the cut group were

bleeding (n 5 1), pain (n 5 1), and both bleeding and pain

(n 5 1). In the coag group, the indications were bleeding

(n 5 2) and pain (n 5 3). In 5 patients, the histologic study

of the hysterectomy specimen revealed fibroid, adenomyo-

sis, or both. No statistically significant differences were

found in the primary outcomes between the 2 groups. On

the basis of this pilot study, the sample size needed to

show a difference in hypomenorrhea and amenorrhea rates

between the 2 waveforms is 216 in each group for alpha

of 0.05 and power of 80%.

Discussion

It has been shown that the satisfaction rate after endo-

metrial ablation, in general, is 80% to 90%, with menstrual

reduction rates between 85% and 95% [1,10,11]. Approxi-

mately 15% to 25% of patients require a second surgical

procedure. Up to 20% of patients have a subsequent hyster-

ectomy for pain, abnormal bleeding, or both [1,7,12]. In our

study, the patient satisfaction rate was lower (63.6% and

68.0% in the cut and coag cohort, respectively), and the re-

intervention rate was higher (36.4% and 32.0%,

respectively). This may be attributed to the small sample

sizes.

A recent study comparing the probability of hysterec-

tomy after endometrial ablation found that age was the

Table 3

Primary outcomes at 2 years’ follow-up

Cut Coag p Value

Amenorrhea 8 (36.4%) 7 (28%) .54

Hypomenorrhea 7 (38.1%) 13 (52%) .16

Satisfied or very satisfied 14 (63.6%) 17 (68%) .46

Reintervention 8 (36.4%) 8 (32%) .75

Repeat hysteroscopic ablation 5 (22.7%) 3 (12%)

Hysterectomy 3 (13.6) 5 (20%)

352 Journal of Minimally Invasive Gynecology, Vol 16, No 3, May/June 2009

most significant predictor, whereas the presence of myomas

and the generation of endometrial ablation techniques used

were nonpredictive [9]. Although not statistically signifi-

cant, our data showed the same with the coag group being

younger, and more of them went on to undergo hysterec-

tomy than the cut group.

We hypothesized that with unmodulated low-voltage cut

waveform, the slower heating of endometrial tissue would

cause deeper tissue penetration and greater reduction in

menstrual flow. The modulated high-voltage coag wave-

form would desiccate the superficial tissue and thus prevent

thermal destruction of deeper tissue due to the insulating

effect of attendant tissue drying. Supporting this theory in-

cludes work done by Duffy et al [3] that showed doubling

power output from 50 W to 100 W with a modulated wave-

form was not associated with increased thermal effect on in

vitro endometrial tissue. Their subsequent work [4] on in

vivo models confirmed this finding and concluded that ther-

mal damage with the modulated coag waveform appeared

confined, self-limiting, and not affected greatly by pro-

longed application (greater than 5 seconds) of the rollerball

to the endometrial lining.

One significant observation we made in this study was that

the rollerball electrode was rapidly covered with coagulated

tissue and debris when low-voltage (cut) waveform is used.

It is well-known that the easiest way to clean a ‘‘gummed

up’’ electrode is to step on the coag mode to deliver high volt-

age and blow away the coagulated tissue, producing a shiny

clean rollerball [13,14]. Because we were restricted to do so

using the cut waveform, we had to withdraw the resectoscope

and clean the rollerball manually several times during a single

procedure. This indeed was found to be a very frustrating

exercise by all surgeons.

Onbargi et al [5] performed endometrial ablation using

the rollerball on extirpated uteri with different power set-

tings and waveforms to determine how often a depth of

3 mm of myometrial thermal destruction was achieved.

Treatment to this depth is believed to destroy completely

both the endometrial glands and the blood supply. They

reported that the optimal power and waveform setting that

achieved at least 3 mm of myometrial thermal destruction

most consistently was the unmodulated waveform at greater

than 90 W.

Given the evidence that exists supporting a more consis-

tent thermal destruction to at least 3 mm of the myometrium

when REA is performed with the unmodulated cut wave-

form, we wondered why no clinical studies were done to

address this issue. Furthermore, a review of literature for

studies on endometrial ablation shows a lack of disclosure

with regard to the waveform settings that were used. Our

study is the first to address the effect of current waveform

on clinical outcomes in vivo. Limited by the small sample

sizes, no significant differences were found between the

cut and coag waveform cohorts with regard to hypomenor-

rhea and amenorrhea rates. However, a trend toward higher

rate of amenorrhea was observed in REA with the cut wave-

form. Conversely, a trend toward greater patient satisfaction

was seen in the coag waveform cohort that is reflected in the

lower reintervention rate and the higher combined amenor-

rhea and hypomenorrhea rate, even if more women in this

group went on to have a hysterectomy. A larger randomized

controlled clinical trial is underway to address the most ap-

propriate waveform for REA.

Conclusion

Rollerball endometrial ablation with either unmodulated

cut or modulated coag waveform is safe for treatment of

menorrhagia in premenopausal women. No significant dif-

ferences were found in terms of reduction in menstrual

flow, need for reintervention, and patient satisfaction. There

was, however, a trend toward a higher amenorrhea rate when

the cut waveform was used. On the other hand, there was

a nonsignificant trend toward higher patient satisfaction in

the coag waveform cohort, which was also reflected by

a lower reintervention rate. These findings suggest that nei-

ther the cut nor coag waveform is better, but the debate

over the most appropriate waveform for use in rollerball en-

dometrial ablation will continue until a larger randomized

trial is done.

References

1. Vilos G. Second-Generation Technologies for Endometrial Ablation:

an Atlas of Operative Laparoscopy and Hysteroscopy. 2nd ed. New

York: Parthenon Publishing Group; 2001. p. 471-482.

2. Lethaby A, Hickey M, Garry R. Endometrial destruction techniques for

heavy menstrual bleeding. Cochrane Database Syst Rev. 2005;(4).

CD001501.

3. Duffy S, Reid P, Smith J, Sharp E. In vitro studies of uterine electrosur-

gery. Obstet Gynecol. 1991;78:213–220.

4. Duffy S. The tissue and thermal effects of electrosurgery in the uterine

cavity. Clin Obstet Gynecol. 1995;9:261–277.

5. Onbargi L, Hayden R, Valle R, Del Priore G. Effects of power and elec-

trical current density variations in an in vitro endometrial ablation

model. Obstet Gynecol. 1993;82:912–918.

6. Brill A. Principles for the Safe Practice of Electrosurgery and Ultrasonic

Surgery at Laparoscopy. Syllabus for the First Annual Comprehensive

Workshop on Gynecologic Endoscopy for Residents and Fellows.2003;74–86.

7. Soderstrom R. Principles of electrosurgery as applied to gynecology. In:

Rock J, Jones H, editors. Te Linde’s operative gynecology. 9th ed.

Philadelphia: Lippincott-Raven; 2003. p. 291–308.

Chang et al. Clinical Outcomes of Two Radiofrequency Waveforms for Endometrial Ablation 353

8. Vercellini P, Oldani S, Yaylayan L, et al. Randomized comparison of va-

porizing electrode and cutting loop for endometrial ablation. Obstet

Gynecol. 1999;94:521–527.

9. Longinoti M, Jacobson G, Hung Y, Learman L. Probability of hysterec-

tomy after endometrial ablation. Obstet Gynecol. 2008;112:1214–1220.

10. Furst S, Philipsen T, Joergensen J. Ten-year follow-up of endometrial

ablation. Acta Obstetr Gynaecol Scand. 2007;86:334–338.

11. Cooley S, Yhddand V, Walsh T, Geary M, McKenna P. The medium-

and long-term outcome of endometrial ablative techniques. Eur J Obstet

Gynecol Reprod Biol. 2005;121:233–335.

12. Munro M. Endometrial ablation for heavy menstrual bleeding. Curr

Opin Obstet Gynecol. 2005;17:381–394.

13. Papadopoulos N, Magos A. First-generation endometrial ablation:

roller-ball vs loop vs laser. Best Pract Res Clin Obstet Gynaecol.

2007;21:915–929.

14. Chang P, Vilos A, Abu-Rafea B, et al. Clinical comparison of low volt-

age (cut) and high voltage (coag) waveforms for endometrial ablation-

with the rollerball. Presented at the 37th Global Congress of

Minimally Invasive Gynecology, October 28-November 1, 2008, Las

Vegas, Nevada.