Identifying Risk Factors for Occult Lower Extremity Lymphedema using Computed

Transcription

1 Identifying Risk Factors for Occult Lower Extremity Lymphedema using Computed Tomography in Patients undergoing Lymphadenectomy for Gynecologic Cancers Miseon Kim a*, Dong Hoon Suh a*, Eun Joo Yang b, Myong Cheol Lim c, Jin Young Choi a, Kidong Kim a, Jae Hong No a, Yong-Beom Kim a a Department of Obstetrics and Gynecology, Seoul National University Bundang Hospital, Seongnam, Republic of Korea b Department of Rehabilitation Medicine, Seoul National University Bundang Hospital, Seongnam, Republic of Korea c Gynecologic Cancer Branch and Center for Uterine Cancer, Research Institute and Hospital, National Cancer Center, Goyang, Republic of Korea This was presented at the 47th annual meeting of the Society of Gynecologic Oncologists, San Diego, April *The first two authors equally contributed to this work. Corresponding author: Yong-Beom Kim Department of Obstetrics and Gynecology, Seoul National University Bundang Hospital 82, Gumi-ro 173 Beon-gil, Bundang-gu, Seongnam-si, Gyeonggi-do 13620, Republic of Korea Phone: Fax: ybkimlh@snubh.org

2 Abstract Objective: To identify risk factors for lower extremity lymphedema (LEL) using computed tomographic (CT) scan in patients undergoing lymphadenectomy for gynecologic cancers. Methods: We retrospectively reviewed 511 consecutive gynecologic cancer patients undergoing lymphadenectomy. Mean difference (3.77±3.14 mm) of subcutaneous layer thicknesses between preoperative and postoperative 1-year CT scans of 106 patients with clinical LEL was used as an objective criterion for regrouping all the patients into those with mean difference >3.77 mm and 3.77 mm. Risk factors for clinical LEL and significant increase of subcutaneous layer thickness on CT were evaluated using a logistic regression model. Results: A total of 106 (20.7%) patients were clinically diagnosed with LEL by a physician. Total number of lymph nodes (LNs) retrieved >30 (Odds ratio [OR] 3.2; 95% Confidence interval [CI] ; p<0.001) and adjuvant radiotherapy (OR 2.7; 95% CI ; p<0.001) were risk factors for clinical LEL. One hundred-nineteen (23.3%) had subcutaneous layer thickness increase of >3.77 mm. In addition to number of LNs retrieved >30 (OR 2.3; 95% CI ; p=0.001) and adjuvant radiotherapy (OR 1.6; 95% CI ; p=0.047), open surgery (OR 1.9; 95% CI ; p=0.014), long operation time (OR 1.7; 95% CI ; p=0.034), and no use of intermittent pneumatic compression (IPC) (OR 2.0; 95% CI ; p=0.017) were risk factors for thick subcutaneous layer on postoperative CT. Conclusions: In addition to high LN retrieval and adjuvant radiotherapy, open surgery, long operation time, and no IPC use could be risk factors for occult LEL after lymphadenectomy in gynecologic cancers. Key words: Lower Extremity Lymphedema; Risk Factor; Computed Tomography; Lymphadenectomy; Gynecologic Cancer

3 Highlights Total number of lymph nodes retrieved >30 and adjuvant radiotherapy are risk factors for clinical lower extremity lymphedema (LEL). Open surgery, long operation time and no use of intermittent pneumatic compression were risk factors for occult LEL.

4 Introduction Lower extremity lymphedema (LEL) is a chronic complication after pelvic lymphadenectomy that lasts a lifetime in most cases. Incidence rates of LEL after pelvic and/or para-aortic lymphadenectomy for the treatment of gynecologic cancers have been reported to be as high as 67% [1-3]. Many patients with LEL suffer from physical and psychological pain, as well as economic burdens, because lymphedema is difficult to cure, especially when it progresses to the irreversible stage. Therefore, LEL is one of the main causes of poor quality of life in postoperative patients with gynecologic cancers. For this reason, prophylaxis is more important than treatment in the management of LEL. There are many studies evaluating risk factors of LEL, such as number of lymph nodes retrieved, removal of distal iliac lymph nodes, and adjuvant radiotherapy [4-8]. Patients with vulvar cancer after inguinal lymphadenectomy frequently suffer from severe lymphedema [3]. However, most of the previous reports were based on retrospective study populations, in which patients were not screened for the diagnosis of LEL, and therefore, the diagnosis may have been missed in some patients with symptoms and/or signs of LEL. Salani et al. reported that only 22% of patients who had swelling of the lower limb were diagnosed with LEL [9]. Patients who had symptoms and/or signs of LEL but were not diagnosed were defined as occult LEL in our study. Furthermore, there might be patients who neither had symptoms nor signs of LEL, and who were not diagnosed with LEL, but have some risk factor of LEL; these patients were defined as potential LEL in our study (Fig. 1). Ideal risk factors are the ones identified from the genuine LEL population, including the occult LEL group. Those risk factors could identify potential LEL patients, as well. Therefore, we thought of an objective method to postoperatively monitor every patient who underwent pelvic lymphadenectomy, in order to identify risk factors for postoperative LEL (including occult LEL) more accurately and help in the early detection of LEL before it progressed.

5 Computed tomography (CT) is a routine follow-up imaging study after surgery for gynecologic cancers. Almost all patients undergo serial CT scans, which have been shown to provide non-invasive measurements of edema accumulation [10, 11]. Therefore, we conducted this study to identify risk factors for occult LEL using CT scans in patients undergoing lymphadenectomy for gynecologic cancers. Methods Study population The medical records of 511 patients undergoing lymphadenectomy for gynecologic cancers in Seoul National University Bundang Hospital between June 2003 and March 2015 were retrospectively reviewed. Of the 511, 131 (25.6%) were diagnosed with lymphedema, whereas 405 (74.4%) were not. Among the 131 subjects with a diagnosis of lymphedema, 25 patients were excluded for having genital lymphedema; thus, a total of 106 patients had a diagnosis of LEL (Table 1 and Fig. 1). Every patient had results of both preoperative and postoperative 1-year (±6 months) abdominopelvic CT. Patients with any cause of LEL other than lymphadenectomy for gynecologic cancers were excluded. We obtained Institutional Review Board approval of this study (B ). Determination of cut-off value of subcutaneous layer thickness on CT scan of patients with LEL Criteria for screening or diagnosis of LEL using imaging modalities have not yet been established. We arbitrarily used the mean difference of subcutaneous layer thicknesses between preoperative and postoperative 1-year CT scans in the 106 patients who had clinical diagnosis of LEL (3.77±3.14 mm) as the cut-off value (Fig. 2). This cut-off value was used as an objective criterion for grouping all the patients according to the CT scan findings. We

6 measured the subcutaneous layer thicknesses over the sartorius muscle of the right anterior thigh at the level of lesser trochanter of femur, which was the lowest indicator of the lower extremity covered by abdominopelvic CT (Supplementary Fig. 1). If the patient had a diagnosis of LEL on one side (right or left) the relevant side was selected. The reason why we chose postoperative 1-year CT scan for evaluation of subcutaneous layer thickness was because half of the LEL was reported to occur within 1-year of the surgery, and the remaining cases developed sporadically beyond 4 years after the surgery, according to previous studies [5, 6]. Surgical procedures Patients who underwent lymph node dissection with total abdominal hysterectomy and/or bilateral salpingo-oophorectomy were included in this study. Lymph node dissection included pelvic lymph node dissection (PLND) and/or para-aortic lymph node dissection (PaLND). PLND included dissection of external iliac nodes, internal iliac nodes, obturator nodes, presacral nodes, and common iliac nodes. Although dissection of PaLND occurs at the surgeon s discretion, PaLND usually included dissection of para-aortic nodes between the level of the bifurcation of the aorta and bilateral renal veins. Assessment of risk factors for LEL Patient information including age, body mass index, origin of cancer, surgical approach, operation time, total number of lymph nodes retrieved, duration of surgical drain placement, amount of drain, adjuvant radiotherapy, early ambulation within 24 hours, and perioperative application of elastic stockings or intermittent pneumatic compression (IPC) were extracted from the medical records. Two sets of univariate and multivariate analyses were performed for assessing the associations of these variables with (1) a diagnosis of LEL by a physician

7 and (2) a significant increase of subcutaneous layer thickness of the thigh on postoperative 1- year CT scan. Independent risk factors for the diagnosis of LEL and a significant increase of subcutaneous layer thickness of the thigh on postoperative 1-year CT scan were identified, respectively. Patients were clinically diagnosed using the grading of LEL severity proposed by the International Society of Lymphology (ISL) [12]. On the basis of medical records, we identified patients with postoperative LEL. Unilateral or bilateral lower limb edema detected by a physician was defined as LEL, whereas patients with subjective complaints of swelling in their legs without an assessment by a physician were not included in the clinically diagnosed LEL group. Statistical analysis Student's t-test and Chi-square test were used for continuous and categorical variables, respectively. For corresponding non-parametric statistics, Mann-Whitney U and Fisher s exact test were used, respectively. In this study, variables with p<0.2 in univariate analysis were selected to enter multivariate analysis to identify independent risk factors for LEL. We used SPSS version 22.0 (IBM Inc, Armonk, NY, USA) and p<0.05 was considered statistically significant. Results Table 1 shows the patient characteristics of the two groups according to the clinical diagnosis of LEL. A total of 106 (20.7%) patients were clinically diagnosed with LEL by a physician. The rest of the subject population (79.3%), which might include occult LEL patients, were not diagnosed with LEL. Of the evaluated variables, the total number of lymph nodes retrieved was significantly higher in patients diagnosed with LEL than in those who

8 were not (30.5±13.0 vs. 21.6±11.6, p<0.001). Patients with a diagnosis of LEL received adjuvant radiotherapy more frequently than those without LEL diagnosis (50.0% vs. 22.5%, p<0.001). Early ambulation 24 hours after surgery was more frequently observed in patients who were not diagnosed with LEL than in those who were (56.2 % vs. 44.8%, p=0.036). Risk factors for clinically diagnosed LEL were described in Table 2. Multivariate analysis showed that >30 lymph nodes retrieved (odds ratio [OR] 3.2; 95% confidence interval [CI] ; p<0.001) and adjuvant radiotherapy (OR 2.7; 95% CI ; p<0.001) are independent risk factors for the clinical diagnosis of LEL. In Table 3, the variables were evaluated for their associations with significant increases in subcutaneous layer thickness between preoperative and postoperative 1-year CT scans. One hundred nineteen (23.3%) patients had increases in subcutaneous layer thickness of >3.77 mm. Similar to patients with a clinical diagnosis of LEL, total number of lymph nodes retrieved >30 (28.7±14.3 vs. 21.9±11.4, p<0.001) and adjuvant radiotherapy (40.3% vs. 24.5%, p=0.001) were significantly associated with greater increases in subcutaneous layer thickness on postoperative 1-year CT scan. Early ambulation 24 hours after surgery was more frequently observed in patients with subcutaneous layer thickness difference 3.77 mm than those with subcutaneous layer thickness difference >3.77 mm (56.7% vs. 44.5%, p=0.020). In addition, open surgery (75.6% vs. 57.7%, p<0.001), long operation time (64.7% vs. 30.4%, p=0.004), and prolonged placement of surgical drain after surgery (53.0% vs 42.1%, p=0.030) were more commonly seen in the group with >3.77 mm increase in subcutaneous layer thickness on CT scan. Patients with subcutaneous layer thickness difference 3.77 mm used IPC more frequently than those with difference >3.77 mm (25.3% vs. 15.1%, p=0.021). Risk factors for greater increases in the thickness of the subcutaneous layer of the thigh between preoperative and postoperative 1-year CT scans are shown in Table 4. In the

9 multivariate analysis, total number of lymph nodes retrieved >30 (OR 2.5; 95% CI ; p<0.001) and adjuvant radiotherapy (OR 1.7; 95% CI ; p=0.041) were independent risk factors for greater increases in subcutaneous layer thickness. Open surgery (OR 1.7; 95% CI ; p=0.045) and long operation time (OR 1.8; 95% CI ; p=0.026) were additional independent risk factors. Use of IPC was an independent protective factor against increased in subcutaneous layer thickness (OR 0.5; 95% CI ; p=0.024). Discussion Because of its chronic and intractable nature, LEL causes physical and psychological problems for gynecologic cancer survivors, thereby reducing their quality of life. There were many efforts for preventing postoperative LEL, including identifying risk factors for postoperative LEL. For example, harvesting a large number of lymph nodes and extensive lymphadenectomy are commonly reported as risk factors of LEL [4, 8, 13, 14]. Postoperative adjuvant radiotherapy is also frequently associated with developing LEL in gynecologic cancer survivors [3, 5-8, 15]. However, most of the previous reports evaluating risk factors of LEL were based on study populations who were clinically diagnosed with the condition. There may be a substantial number of occult cases who already have a certain degree of lymphedema but have not been diagnosed. Our data was consistent with previous study results in that a total number of lymph nodes retrieved of >30 and adjuvant radiotherapy are independent risk factors for LEL after lymphadenectomy [4-8]. Using the aforementioned criterion, open surgery, long operation time, and no IPC use were found to be associated with greater increases in subcutaneous layer thickness on postoperative 1-year CT scan, which suggests that those three factors could also be considered as additional independent risk factors for occult LEL (Fig. 1). Association of long operation time with greater increases in subcutaneous layer thickness could be related to

10 extensive surgery including lymphadenectomy, which was reported in previous studies [1, 2, 13]. The association of the kind of surgical approach (laparoscopy or open surgery) as a potential risk factor for occult LEL, which was noted in this study, was not demonstrated in previous studies [16-18]. This inconsistency might be due to the difference in study populations (clinical LEL and occult LEL), although the mechanism underlying the association between the open approach and LEL is obscure. Although there is still a debate on the efficacy of and adverse reactions to IPC therapy for LEL [12, 19-21], to the best of our knowledge, there is no study with high level evidence that evaluated the prophylactic effects of IPC for LEL in postoperative patients with gynecologic cancer. In this study, IPC was used for a certain period of time after surgery, not for the purpose of treatment of LEL but for the purpose of prophylaxis, and IPC was shown to have a protective effect against LEL. An objective method for diagnosis of lymphedema has not yet been established. The most commonly used diagnostic method is dependent on physical examination performed by experienced specialists, and there are currently no effective screening methods for early detection of LEL. To our knowledge, this is the first study that used CT scan as a screening tool for identifying risk factors of LEL in patients undergoing lymphadenectomy for the treatment of gynecologic cancer. There were a few reports demonstrating the clinical usefulness of CT scan in the diagnosis and management of lymphedema. Monnin-Delhom et al. showed that the sensitivity and specificity of CT scan for the diagnosis of LEL was 93% and 100%, respectively. They found that CT readily demonstrated edema accumulation in the subcutaneous layers; specific CT features of the subcutaneous fat and muscle compartments were also noted in LEL [10]. In addition, the clinical value of CT in determining the target area for treatment and monitoring the effects of breast cancer treatment on lymphedema was demonstrated in patients with upper extremity lymphedema [11]. In a randomized controlled trial of breast cancer patients who underwent surgery including

11 axillary lymph node dissection, the effectiveness of early physiotherapy for prophylaxis of secondary lymphedema was clearly demonstrated [22]. Significantly fewer women who received prophylactic physiotherapy developed clinically significant lymphedema at 1-year follow-up compared to controls (7.0% vs. 25.0%; OR 0.28; 95% CI ). On the other hand, there is no systematic LEL prophylaxis program for gynecologic cancer survivors. Identifying accurate risk factors for selection of proper candidates to be enrolled in such a program should precede the development of the program. Thus, we highlighted CT as an objective method to assess every patient undergoing lymphadenectomy, in order to identify risk factors for postoperative LEL in gynecologic cancer patients. The present study has some limitations, most of which were due to the retrospective study design. First, we used abdominopelvic CT scans instead of lower extremity CT, and site selection for measuring subcutaneous layer thickness on CT scan was arbitrary. It may be possible that the cut-off value would be greater than 3.77 mm if lower extremity CT scans were used and more LEL-vulnerable sites were selected for measurement. Second, we could not control any factors that may affect the thickness of the subcutaneous layer, such as diet and general activities. Third, increasing subcutaneous layer thickness on CT scan does not always correspond to the development of LEL. Thus, patients who have significant differences in subcutaneous layer thicknesses between the two CT scans may not have any symptoms and signs of LEL, and would therefore not be diagnosed as LEL. The correlation between CT findings and clinical symptoms and signs of LEL should be evaluated, in order to validate CT findings as a surrogate marker for LEL. We conclude that in addition to a high number of lymph nodes retrieved and adjuvant radiotherapy, open surgery, long operation time, and no use of IPC were independent risk factors for greater increases in subcutaneous layer thickness of the thigh between preoperative and postoperative 1-year CT scans for patients undergoing pelvic

12 lymphadenectomy in gynecologic cancers. Our study results could provide a guide in the selection of proper candidates for prophylactic management of LEL, including occult LEL, which could eventually lead to improvement in quality of life. Further prospective validation of CT as a useful method for identifying proper candidates for LEL prophylaxis in postoperative gynecologic cancer patients is needed. Conflict of Interest The authors declare no conflicts of interest. Acknowledgment This work was supported by a grant (No ) from the Seoul National University Bundang Hospital Research Fund.

13 Reference [1] Beesley V, Janda M, Eakin E, Obermair A, Battistutta D. Lymphedema after gynecological cancer treatment: prevalence, correlates, and supportive care needs. Cancer. 2007;109: [2] Ryan M, Stainton MC, Slaytor EK, Jaconelli C, Watts S, Mackenzie P. Aetiology and prevalence of lower limb lymphoedema following treatment for gynaecological cancer. Aust N Z J Obstet Gynaecol. 2003;43: [3] Berger J, Scott E, Sukumvanich P, Smith A, Olawaiye A, Comerci J, et al. The effect of groin treatment modality and sequence on clinically significant chronic lymphedema in patients with vulvar carcinoma. Int J Gynecol Cancer. 2015;25: [4] Lutman CV, Havrilesky LJ, Cragun JM, Secord AA, Calingaert B, Berchuck A, et al. Pelvic lymph node count is an important prognostic variable for FIGO stage I and II endometrial carcinoma with high-risk histology. Gynecol Oncol. 2006;102:92-7. [5] Kim JH, Choi JH, Ki EY, Lee SJ, Yoon JH, Lee KH, et al. Incidence and risk factors of lower-extremity lymphedema after radical surgery with or without adjuvant radiotherapy in patients with FIGO stage I to stage IIA cervical cancer. Int J Gynecol Cancer. 2012;22: [6] Ohba Y, Todo Y, Kobayashi N, Kaneuchi M, Watari H, Takeda M, et al. Risk factors for lower-limb lymphedema after surgery for cervical cancer. Int J Clin Oncol. 2011;16: [7] Tada H, Teramukai S, Fukushima M, Sasaki H. Risk factors for lower limb lymphedema after lymph node dissection in patients with ovarian and uterine carcinoma. BMC Cancer. 2009;9:47. [8] Todo Y, Yamamoto R, Minobe S, Suzuki Y, Takeshi U, Nakatani M, et al. Risk factors for postoperative lower-extremity lymphedema in endometrial cancer survivors who had treatment including lymphadenectomy. Gynecol Oncol. 2010;119:60-4. [9] Salani R, Preston MM, Hade EM, Johns J, Fowler JM, Paskett EP, et al. Swelling Among

14 Women Who Need Education About Leg Lymphedema: A Descriptive Study of Lymphedema in Women Undergoing Surgery for Endometrial Cancer. Int J Gynecol Cancer. 2014;24: [10] Monnin-Delhom ED, Gallix BP, Achard C, Bruel JM, Janbon C. High resolution unenhanced computed tomography in patients with swollen legs. Lymphology. 2002;35: [11] Kim SY, Bae H, Ji HM. Computed tomography as an objective measurement tool for secondary lymphedema treated with extracorporeal shock wave therapy. Ann Rehabil Med. 2015;39(3): [12] International Society of Lymphology. The diagnosis and treatment of peripheral lymphedema: 2013 Consensus Document of the International Society of Lymphology. Lymphology. 2013;46:1-11. [13] Hareyama H, Ito K, Hada K, Uchida A, Hayakashi Y, Hirayama E, et al. Reduction/prevention of lower extremity lymphedema after pelvic and para-aortic lymphadenectomy for patients with gynecologic malignancies. Ann Surg Oncol. 2012;19: [14] Bae HS, Lim MC, Lee JS, Lee Y, Nam BH, Seo SS, et al. Postoperative Lower Extremity Edema in Patients with Primary Endometrial Cancer. Ann Surg Oncol. 2016;23: [15] Mitra D, Catalano PJ, Cimbak N, Damato AL, Muto MG, Viswanathan AN. The risk of lymphedema after postoperative radiation therapy in endometrial cancer. J Gynecol Oncol. 2015;27(1):e4. [16] Abu-Rustum NR, Alektiar K, Iasonos A, Lev G, Sonoda Y, Aghajanian C, et al. The incidence of symptomatic lower-extremity lymphedema following treatment of uterine corpus malignancies: A 12-year experience at Memorial Sloan-Kettering Cancer Center.

15 Gynecol Oncol. 2006;103: [17] Hopp EE, Osborne JL, Schneider DK, Bojar CJ, Uyar DS. A prospective pilot study on the incidence of post-operative lymphedema in women with endometrial cancer. Gynecol Oncol Rep. 2016;15:25-8. [18] Ghezzi F, Uccella S, Cromi A, Bogani G, Robba C, Serati M, et al. Lymphoceles, lymphorrhea, and lymphedema after laparoscopic and open endometrial cancer staging. Ann Surg Oncol. 2012;19: [19] Zaleska M, Olszewski W, Durlik M. The Effectiveness of Intermittent Pneumatic Compression in Long-Term Therapy of Lymphedema of Lower Limbs. Lymphat Res Biol. 2014;12: [20] Blumberg S, Berland T, Rockman C, Mussa F, Brooks A, Cayne N, et al. Pneumatic Compression Improves Quality of Life in Patients with Lower-Extremity Lymphedema. Ann Vasc Surg. 2016;30:40-4. [21] Finnane A, Janda M, Hayes SC. Review of the evidence of lymphedema treatment effect. Am J Phys Med Rehabil. 2015;94: [22] Lacomba MT, Sánchez MJY, Goñi AZ, Merino DP, Moral OMd, Téllez EC, et al. Effectiveness of early physiotherapy to prevent lymphoedema after surgery for breast cancer: randomised, single blinded, clinical trial. BMJ. 2010;340:b5396.

16 Figure legends Fig. 1 Determination of computed tomography (CT)-based cut-off value of the difference in subcutaneous layer thicknesses between preoperative and postoperative 1-year CT scans

17 Fig. 2 Conceptual populations of lower extremity lymphedema with corresponding risk factors LEL, lower-extremity lymphedema Potential LEL has neither symptom/sign nor diagnosis, but could have risk factors; Occult LEL has symptom and/or sign, but does not have diagnosis; Clinical LEL has diagnosis of LEL based on obvious symptom and/or sign

18 S1 Measurement of subcutaneous layer thickness on CT scan A patient who underwent radical hysterectomy with bilateral pelvic lymph node dissection for endometrial cancer, stage IB

19 TABLE 1. Patient characteristics according to the clinical diagnosis of lower extremity lymphedema (LEL) LEL (-) (n=405) LEL (+) (n=106) P-value Age, yrs 52.0± ± BMI, kg/m (67.4) 80 (75.5) > (32.6) 26 (24.5) Origin of cancer Cervix 155 (38.3) 48 (45.3) Uterine corpus 114 (28.1) 38 (35.8) Ovary and tube 133 (32.8) 20 (18.9) Vulva and others 3 (0.7) 0 Open surgery 250 (61.7) 66 (62.3) Operation time >3 hr 217 (53.6) 59 (55.7) Total number of lymph nodes retrieved 21.6± ±13.0 <0.001 Duration of drain, day (range) 5 (0-42) 6 (2-18) Amount of drain, ml ± ± Adjuvant radiotherapy 91 (22.5) 53 (50) <0.001 Early ambulation 24 hr 227 (56.2) 47 (44.8) Elastic stocking 191 (47.2) 50 (47.2) Intermittent pneumatic compression 97 (24.0) 20 (18.9) Values are presented as number (%) or mean±standard deviation, unless otherwise indicated. BMI, body mass index; LEL, lower extremity lymphedema

20 TABLE 2. Univariate and multivariate analyses of risk factors for clinical lower-extremity lymphedema Univariate Multivariate OR (95% CI) P-value OR (95% CI) P-value Cervix or corpus cancer 2.2 ( ) ( ) Total number of lymph nodes retrieved > ( ) < ( ) <0.001 Duration of drain > 5 days 1.4 ( ) ( ) Amount of drain > 1272 ml 1.7 ( ) ( ) Adjuvant radiotherapy 3.5 ( ) < ( ) <0.001 Early ambulation 24 hr 0.6 ( ) ( ) OR, odds ratio; CI, confidence interval Median

21 TABLE 3. Patient characteristics according to the difference in subcutaneous layer thickness of the thigh on computed tomography scan Difference 3.77 mm Difference > 3.77 mm (n=392, 76.7%) (n=119, 23.3%) P-value Age, yrs 52.2± ± BMI, kg/m (68.6) 84 (70.6) > (31.4) 35 (29.4) Origin of cancer Cervix 158 (40.3) 45 (37.8) Uterine corpus 112 (28.6) 40 (33.6) Ovary and tube 120 (30.6) 33 (27.7) Vulva and others 2 (0.5) 1 (0.8) Open surgery 226 (57.7) 90 (75.6) <0.001 Operation time >3 hr 199 (30.4) 77 (64.7) Total number of lymph nodes retrieved 21.9± ±14.3 <0.001 Duration of drain >5 days 165 (42.1) 63 (53.0) Amount of drain, ml ± ± Adjuvant radiotherapy 96 (24.5) 48 (40.3) Early ambulation 24 hr 221 (56.7) 53 (44.5) Elastic stocking 191 (48.7) 50 (42.0) Intermittent pneumatic compression 99 (25.3) 18 (15.1) Values are presented as number (%) or mean±standard deviation, unless otherwise indicated. BMI, body mass index

22 TABLE 4. Univariate and multivariate analyses for risk factors of greater increases in subcutaneous layer thickness of the thigh between preoperative and postoperative 1-year computed tomography scans Univariate Multivariate OR (95% CI) P-value OR (95% CI) P-value Open surgery 2.3 ( ) ( ) Total number of lymph nodes retrieved > ( ) < ( ) <0.001 Operation time >3 hr 1.9 ( ) ( ) Duration of drain >5 days 1.6 ( ) ( ) Adjuvant radiotherapy 2.1 ( ) ( ) Early ambulation 24 hr 0.6 ( ) ( ) Elastic stocking 0.8 ( ) ( ) Intermittent pneumatic compression 0.5 ( ) ( ) OR, odds ratio; CI, confidence interval Median 0 22