Sage Journals: Discover world-class research

Abstract

Few studies have assessed whether cases of asymptomatic pulmonary embolism (PE) in the early postoperative phase are subsegmental versus more proximal. In this study, we investigated whether asymptomatic PE occurring just after gynecologic surgery was subsegmental, and we examined the background characteristics of patients who experienced PE within 2 months postoperatively. All hospital records were reviewed, yielding a total of 2052 women who had undergone surgeries performed by the gynecologic oncology team between 2003 and 2013 in the National Kyushu Cancer Center. Asymptomatic and symptomatic postoperative PE cases diagnosed by multidetector computed tomography angiography or lung scan were identified; after excluding 2 cases of preoperative PE, there were 15 (0.73%) cases of postoperative PE among 2050 women. Of the 15 cases, 9 (60%) were diagnosed on postoperative day 1 or 2. Of the 9 women, 4 had no or minor symptoms/signs other than decreased oxygen saturation as measured by pulse oximetry (Spo ₂), and PE was segmental or more proximal in 3 cases. Only 1 of the 9 cases showed dyspnea. The remaining 4 cases showed dizziness or perspiration, suggesting PE. Univariate analysis showed age, operation time, hypertension, and preoperative d-dimer elevation to be associated with postoperative PE. Multivariate analysis demonstrated that a high (≥1 µg/mL) preoperative d-dimer level was associated with postoperative PE (odds ratio, 6.331; 95% confidence interval, 1.567-25.589). Most asymptomatic PE cases occurring within 2 days postoperatively were segmental or more proximal. Identification of early, asymptomatic postoperative PE may be clinically significant because most of these emboli are proximal to the subsegmental arteries.

Keywords

gynecologic surgery perioperative management pulmonary embolism subsegmental artery

Introduction

In a previous study of patients who had undergone orthopedic surgery or had end-stage arteriosclerotic disease, only 4 of the 48 experienced symptoms suggestive of pulmonary embolism (PE); in general, asymptomatic PE is common.¹ Regarding gynecologic diseases, in a prior study involving women with ovarian cancer, PE was identified in 11.1% of the women prior to any treatment, and all of them were asymptomatic.² Following gynecologic surgeries, another study found that 35.6% of women diagnosed with postoperative PE were asymptomatic and had no signs except for a low oxygen saturation as measured by pulse oximetry (Spo ₂) level.³ The PE was mainly diagnosed by lung scan in these studies,^1
–3 and subsegmental PE cases were hard to precisely diagnose.

Multidetector computed tomography angiography (MDCTA) has a higher sensitivity for PE within the subsegmental pulmonary arteries compared with single detector computed tomography (CT) angiography, and the use of MDCTA has the potential to increase the rate of subsegmental PE diagnosis.⁴ The risk profile and short-term clinical course of patients with symptomatic subsegmental PE appear to mimic those of patients with segmental or more proximal PE.⁵ However, asymptomatic subsegmental PE may not be clinically relevant; in a previous study, the 3-month thromboembolic risk for patients with suspected PE who were left untreated was similar in those who underwent single CT and those who underwent multidetector CT pulmonary angiography.⁴ In our previous study, 51.1% of PE cases occurring after gynecologic surgeries were detected on postoperative days 1 and 2; 70.4% of patients were asymptomatic and had no signs except for a drop in Spo ₂.³ However, the importance of identifying PE that occurs just after surgery is not clear, because few studies have assessed whether cases of asymptomatic PE in the early postoperative phase are subsegmental versus more proximal. In this study, we investigated the location of, and symptoms and signs associated with, PE diagnosed on postoperative days 1 and 2, and we also investigated the background characteristics of women who were diagnosed with postoperative PE within 2 months after surgery.

Methods

Mechanical prophylaxis with/without medical prophylaxis was established in our department around June 2003; thereafter, PE cases have been recorded when reported by attending doctors, and we have created a digital database of PE cases. For this study, we reviewed the surgical database to identify women who had undergone surgery performed by the gynecologic oncology team through the Gynecology Service of the National Kyushu Cancer Center from June 2003 through December 2013. A total of 2052 women who had undergone laparotomy, laparoscopic surgery, vaginal surgery, or vulvar surgery were identified from this database. Women who had undergone conization, dilatation and curettage, or left supraclavicular lymph node biopsy were not included, and there were no PE cases in this population. Patients’ background characteristics, such as height, weight, operation time, blood loss, and concomitant diseases, were obtained from the digital database of the Department of Anesthesiology. Digital data for laboratory tests were available starting in March 2004, and digital Spo ₂ levels were available starting in November 2005. Laboratory data from June 2003 to March 2004 and Spo ₂ data from June 2003 to November 2005 were obtained by reviewing article-based records from the National Kyushu Cancer Center.

A schematic illustration of the diagnostic protocol used for identifying postoperative PE is presented in Figure 1. The Spo ₂ levels of all the patients were routinely monitored by pulse oximetry during and after surgery until the patients achieved complete ambulation. During the preoperative period and the period after complete ambulation was achieved, pulse oximetry was performed once daily throughout hospitalization. Pulse oximetry was also performed whenever patients felt unwell. We considered Spo ₂ readings of 93% or less to represent hypoxia because a 94% Spo ₂ reading is equivalent to a Pao ₂ of 71 mm Hg. The PE was confirmed by MDCTA.^6
–8 Among patients with a past allergic reaction to contrast media or those who were not able to undergo MDCTA because of availability, PE was confirmed by lung scan based on the criteria of the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED).⁹ Only high-probability cases according to the PIOPED criteria were diagnosed as PE. All imaging studies were interpreted by the attending radiologists.

Figure 1.

Schematic illustration of the diagnostic protocol for postoperative pulmonary embolism. When hypoxia (Spo ₂ ≤ 93% on room air) or suspected symptoms and/or signs of PE were detected, additional diagnostics were performed. CXP indicates chest X-ray photography; MDCTA, multidetector computed tomography angiography; PE, pulmonary embolism; Spo ₂, oxygen saturation as measured by pulse oximetry.

During the study period, 2052 patients were given graduated compression stockings (GCSs) and/or intermittent pneumatic compression (IPC) devices during and after surgery. The GCS and IPC were removed after complete ambulation. Most of the patients with malignant diseases and some of the patients with benign diseases were given low-molecular-weight heparin (LMWH; ie, dalteparin and/or enoxaparin) plus/minus fondaparinux after surgery. Five patients with preoperative deep vein thrombosis (DVT) above the calf received preoperative LMWH or unfractionated low-dose heparin. Under the Japanese health insurance system, LMWH use is usually restricted to malignant diseases in gynecological surgery.

Patients’ background characteristics included age, d-dimer value, body mass index (BMI), operation time, blood loss, GCS use, IPC use, parenteral anticoagulant use, scheduled vaginal or vulvar surgery, scheduled laparoscopic surgery, malignant diseases, disease types, d-dimer measurement, and d-dimer elevation (Table 1). The evaluated concomitant diseases were hypertension, diabetes mellitus, and hyperlipidemia (Table 1).

Table 1.

Patients’ Background Characteristics and Univariate Analysis.^a

		Total (n = 2050), n	Without PE (n = 2035), n (%)	With PE (n = 15), n (%)	P
Malignant tumor	−	706	703 (99.6)	3 (0.4)	.287
Malignant tumor	+	1344	1332 (99.1)	12 (0.9)
Disease type					.431
Benign		702	699 (99.6)	3 (0.4)
Cervical cancer		397	395 (99.5)	2 (0.5)
Endometrial cancer		516	509 (98.6)	7 (1.4)
Ovarian cancer^b		375	372 (99.2)	3 (0.8)
Other malignancy		60	60 (100.0)	0 (0.0)
Concomitant diseases
Hypertension	−	1681	1672 (99.5)	9 (0.5)	.038^c
Hypertension	+	369	363 (98.4)	6 (1.6)
Diabetes mellitus	−	1926	1912 (99.3)	14 (0.7)	.609
Diabetes mellitus	+	124	123 (99.2)	1 (0.8)
Hyperlipidemia	−	1236	1227 (99.3)	9 (0.7)	.608
Hyperlipidemia	+	121	120 (99.2)	1 (0.8)
Type of surgery
Vaginal surgery^d	−	1955	1940 (99.2)	15 (0.8)	1.000
Vaginal surgery^d	+	95	95 (100.0)	0 (0.0)
Laparoscopic surgery	−	1985	1970 (99.2)	15 (0.8)	1.000
Laparoscopic surgery	+	65	65 (100.0)	0 (0.0)
Lymphadenectomy	−	986	981 (99.5)	5 (0.5)	.305
Lymphadenectomy	+	1064	1054 (99.1)	10 (0.9)
Prophylaxis
GCS	−	34	33 (97.1)	1 (2.9)	.222
GCS	+	2016	2002 (99.3)	14 (0.7)
IPC	−	71	70 (98.6)	1 (1.4)	.412
IPC	+	1979	1965 (99.3)	14 (0.7)
Anticoagulants	−	858	855 (99.7)	3 (0.3)	.115
Anticoagulants	+	1192	1180 (99.0)	12 (1.0)
d-Dimer measurement	−	1337	1331 (99.6)	6 (0.4)	.054
d-Dimer measurement	+	713	704 (98.7)	9 (1.3)
		Total (n = 713), n	Without PE (n = 704), n (%)	With PE (n = 9), n (%)	P
d-Dimer elevation	−	538	535 (99.4)	3 (0.6)	.009^c
d-Dimer elevation	+	175	169 (96.6)	6 (3.4)
		Total (n = 2050)	Without PE (n = 2035)	With PE (n = 15)	P
Age, years			52.90 ± 12.97	61.93 ± 11.32	.007^c
BMI, kg/m²			22.58 ± 6.02	23.51 ± 3.83	.549
Operation time, minutes			187.61 ± 88.59	251.60 ± 109.39	.005^c
Blood loss, g			293.36 ± 411.17	396.67 ± 437.86	.333
		Total (n = 713)	Without PE (n = 704)	With PE (n = 9)	P
d-Dimer value, µg/mL			1.39 ± 3.99	2.21 ± 2.16	.535

Abbreviations: BMI, body mass index; GCS, graduated compression stocking; IPC, intermittent pneumatic compression; PE, pulmonary embolism.

^aTwo cases of preoperative PE were excluded.

^bOvarian cancer includes peritoneal cancer, tubal cancer, and borderline tumor of the ovary.

^cSignificant difference (P < .05) according to Fisher’s exact test (categorical variables) or the Student’s t-test (continuous variables).

^dVaginal surgery includes vulvar surgery.

Patients with cervical cancer underwent radical hysterectomy and pelvic lymphadenectomy with/without bilateral salpingo-oophorectomy and with/without para-aortic lymphadenectomy. Patients with endometrial cancer underwent total abdominal hysterectomy and bilateral salpingo-oophorectomy, with/without pelvic and para-aortic lymphadenectomy. Some patients with endometrial cancer, such as women with serous adenocarcinoma, underwent omentectomy with/without appendectomy. Patients with ovarian cancer, including those with peritoneal cancer, tubal cancer, and borderline tumor of the ovary, underwent total abdominal hysterectomy, bilateral salpingo-oophorectomy, omentectomy, and appendectomy, with/without pelvic lymphadenectomy and para-aortic lymphadenectomy and with/without resection of disseminated lesions. Patients with vulvar cancer underwent radical vulvectomy or radical local excision of the vulvar tumor with/without inguinal lymphadenectomy. Patients with vaginal cancer underwent the same procedures as patients with cervical cancer. Patients with uterine sarcoma underwent total abdominal hysterectomy and bilateral salpingo-oophorectomy with/without lymph node sampling and with/without resection of disseminated tumors. Some women with microinvasive cancer, small-sized cancer, serious concomitant diseases, and obesity—as well as women hoping to become pregnant in the future—underwent more minimal procedures. The lymphadenectomy rate among women with malignant diseases was 79.2% (1064 of 1344), which was calculated using the data in Table 1.

Compression ultrasonography (CUS) of the lower limbs has been available since November 2006 in the Department of Clinical Laboratory Medicine at National Kyushu Cancer Center. Before this date, attending doctors in our department were instructed to avoid using IPC among patients with a high preoperative d-dimer level. Thereafter, patients with high preoperative d-dimer levels were supposed to receive CUS for DVT diagnosis. The normal d-dimer level in this institution is less than 1 µg/mL. Attending doctors have the option of whether or not to perform CUS for patients with a d-dimer level between 1 and 1.4 µg/mL; 1.5 µg/mL is used as the cutoff value, providing a high sensitivity (100%) and high negative predictive value (100%).² Measurement of d-dimer level has typically been allowed by the Japanese health insurance system for patients with malignant diseases and some with benign diseases. In our department, d-dimer measurement has been actively performed since December 2009; it was sporadically performed before this date.

Statistical Analysis

Student’s t-test was used for univariate analysis of continuous variables, Fisher’s exact test was used for univariate analysis of categorical variables, and multiple logistic regression was used for multivariate analysis. P-values less than .05 were considered to be significant. Dr SPSS II software, version 11.0.1 (SPSS Japan, Tokyo, Japan) was used for analyses.

Results

Follow-Up Period and Exclusion of Preoperative PE

Among the 2052 women who underwent surgeries from June 2003 to December 2013, the median follow-up period was 902 days (range, 5-4217 days) and the median hospitalization duration was 16 days (range, 2-249 days). Of these patients, 2 women with International Federation of Gynecology and Obstetrics (FIGO) stage IIIC and IC1 ovarian cancer were diagnosed with preoperative PE by MDCTA. The subtype was clear cell adenocarcinoma in both cases. Their Spo ₂ levels were 96% and 97%. One of the 2 women also had preoperative DVT. Both women were excluded from statistical analysis (Figure 1).

Identification of Decreased Spo ₂ (≤93%) and Suspected Symptoms/Signs of PE

Using the diagnostic protocol outlined in Figure 1, 15 (0.73%) cases of postoperative PE were identified among the 2050 women remaining after exclusion of the 2 with preoperative PE (Figure 1). There were 6 women who underwent MDCTA without chest X-ray photography (CXP), and PE was not detected among them (Figure 1). Their low Spo ₂ levels recovered during hospitalization. Of the 533 cases with recorded decreases in Spo ₂, 236 (44.3%) did not undergo CXP (Figure 1); the main reason for not performing CXP was that the decrease in Spo ₂ was temporary, which occurred in 128 (54.2%) of these 236 women. All 533 women recovered Spo ₂ levels of ≥94% during hospitalization.

Symptoms and Signs Associated With Postoperative PE

The 15 cases of postoperative PE are listed in Table 2. Of the 15 cases, 4 were diagnosed by lung scan, and the rest were diagnosed by MDCTA (Table 2). Patients’ primary diseases were as follows: 7 cases of endometrial cancer, 3 cases of ovarian cancer, 2 cases of cervical cancer, and 3 cases of benign disease (Tables 1 and 2).

Table 2.

Postoperative Pulmonary Embolism Cases During the Study Period.

PE Case	Age, Years	POD	Disease^a	BMI, kg/m²	OT, Minutes	AC	GCS IPC	CD	d-Dimer,^b µg/mL	Diagnostic Modality	Location of Emboli	Symptoms and Signs Other Than Decreased Spo ₂ at Diagnosis
PE1	48	1	EMCA IVB	22.4	362	+	+		NT	MDCTA	R, lobar, and pl	Perspiration, face pallor, and coldness of the limbs
PE2	69	1	Benign	22.2	127	−	+		NT	MDCTA	R, lobar, and pl	Dizziness
PE3	69	1	Benign	19.0	118	−	+		0.51	MDCTA	R, lobar, and s	–
PE4	58	2	OVCA IIIC	21.9	455	+	+	HT	NT	Lung scan	R	Perspiration
PE5	89	2	EMCA II	28.0	90	−	+	HT	1.1	MDCTA	R, segmental, and s	Mouth pallor
PE6^c	58	2	OVCA IIIA1	22.5	187	+	+	HT	3.3	Lung scan	Bil	–
PE7	53	2	Benign	19.0	160	+	+		1	MDCTA	Bil, segmental, and s	–
PE8	56	2	EMCA IA	20.7	360	+	+		NT	MDCTA	R, segmental, and pl	Dyspnea and dizziness
PE9	62	2	CC IB1	25.2	213	+	+		0.4	MDCTA	R, pulmonary, lobar, and pl	Emesis and perspiration
PE10	61	3	EMCA IB	24.0	267	+	+	HT DM	NT	Lung scan	R	Perspiration and chill
PE11^c	51	5	CC IB1	22.2	427	+	+	HL	0.4	MDCTA	Bil, subsegmental, and pl	–
PE12	55	6	EMCA IVB	26.4	270	+	+		5.9	MDCTA	Bil, pulmonary, lobar, segmental, and pl	Tachypnea and dyspnea
PE13	55	28	EMCA IIIC1	32.7	263	+	+		1.9	MDCTA	Bil, pulmonary, lobar, segmental, and pl	Dyspnea, dizziness, and nausea
PE14	63	49	EMCA IIIA	27.3	260	+	+	HT	NT	MDCTA	Bil, lobar, and pl	Tachycardia and malaise
PE15	82	50	OVCA IIIC	19.2	110	+	−	HT	5.4	Lung scan	R	General fatigue

Abbreviations: AC, anticoagulants; Bil, bilateral; BMI, body mass index; CC, cervical cancer; CD, concomitant diseases; DM, diabetes mellitus; DVT, deep vein thrombosis; EMCA, endometrial cancer; GCS, graduated compression stockings; HL, hyperlipidemia; HT, hypertension; IPC, intermittent pneumatic compression devices; MDCTA, multidetector CT angiography; NT, not tested; OT, operation time; OVCA, ovarian cancer; PE, pulmonary embolism; pl, plural; POD, postoperative day; R, right; s, singular; Spo ₂, oxygen saturation as measured by pulse oximetry.

^aFor patients with cancer diagnoses, the FIGO stage is indicated after the type of cancer.

^bThe normal d-dimer value at the institute is <1 µg/mL.

^cPostoperative DVT was also detected in these cases.

Of the 15, 9 (60%) postoperative PE cases were diagnosed on postoperative day 1 or 2 (Table 2). Of the 9 women, 3 had no symptoms or signs except for decreased Spo ₂ (≤93%; PE3, PE6 and PE7; Table 2). Patient PE5 showed pallor around the mouth. Consequently, 4 (44.4%) of the 9 women diagnosed on postoperative day 1 or 2 had no or only minor symptoms and signs other than a drop in Spo ₂ (Table 2). Among the 9 women, only 1 woman (PE8) showed dyspnea. One woman (PE1) experienced face pallor, cold limbs, and perspiration. The rest of the 9 women (PE2, PE4, and PE9) showed dizziness or perspiration in addition to decreased Spo ₂ (Table 2). The symptoms and signs of PE1, PE2, PE4, and PE9 were suggestive of PE.

The cases with a late onset (PE13, PE14, and PE15) were diagnosed during the postoperative chemotherapy period, and they all showed symptoms and signs that were relatively easy to detect (Table 2). There were no recurrent, massive,^10,11 or fatal postoperative PE cases from June 2003 through December 2013.

Characteristics of the 15 Women With PE

The BMI, operation time, concomitant diseases, and prophylaxes are shown in Table 2. For patients with malignant disease, the FIGO stage is also indicated. Of the 3 patients with ovarian cancer, the subtype was serous adenocarcinoma in 2 cases and endometrioid adenocarcinoma in 1 case. Six women received IPC without preoperative d-dimer evaluation (Table 2).

Locations of Postoperative Pulmonary Emboli

The locations of the emboli detected in the 11 postoperative PE cases diagnosed by MDCTA are indicated in Table 2. In all, 7 cases were diagnosed on postoperative days 1 and 2, and the remaining 4 cases were diagnosed thereafter. Among the cases diagnosed on postoperative days 1 and 2, the 3 women who were asymptomatic or had minor symptoms (PE3, PE5, and PE7) had emboli located in the segmental or lobar arteries (Table 2 and Figure 2A, B, and C). Four symptomatic women had emboli in the segmental arteries or more proximal (PE1, PE2, PE8, and PE9; Table 2). Two of the symptomatic cases (PE2 and PE9) are shown in Figure 2D and E.

Figure 2.

The CT angiography imaging of pulmonary emboli detected on postoperative days 1 and 2. The CT angiography shows several emboli (arrows). A, Illustrates an embolus associated with minor symptoms (PE5). B and C, Represent asymptomatic cases (PE3 and PE7). D and E, Represent symptomatic cases (PE2 and PE9). A, An embolus is shown in the right segmental artery. B, An embolus is shown in the right lobar artery. C, An embolus is shown in the right segmental artery. D, Emboli are shown in the right lobar arteries. E, An embolus is shown in the right pulmonary artery. CT indicates computed tomography; PE, pulmonary embolism.

Among the cases diagnosed on or after postoperative day 3, 3 symptomatic women had emboli located in the pulmonary, lobar, and/or segmental arteries (PE12, PE13, and PE14; Table 2). One woman without symptoms had subsegmental PE (PE11; Table 2).

Univariate Analysis of the Patients’ Background Characteristics

Univariate analysis of suspected factors was performed for postoperative PE cases (Table 1). Statistically significant factors were hypertension, preoperative d-dimer elevation, age, and operation time (Table 1).

Age, Operation Time, Anticoagulant Use, and Preoperative d-Dimer Evaluation in Women With Malignant and Benign Diseases

The presence of a malignant tumor (as the operative indication during the study period) was not a significant factor in univariate analysis (P = .287; Table 1). The mean age of women with malignant diseases was 53.93 ± 12.61 years and that of women with benign diseases was 51.11 ± 13.46 years (P < .001). The operation time was 216.01 ± 90.25 minutes for women with malignant diseases and 134.91 ± 55.95 minutes for women with benign diseases (P < .001). Women with malignant diseases had more harmful factors in addition to malignancy itself than women with benign diseases. However, regarding thromboembolic prophylaxis, anticoagulants were administered to 1109 (82.5%) of the 1344 women with malignant diseases and 83 (11.8%) of the 706 women with benign diseases; the difference was statistically significant (P < .0001). Preoperative d-dimer level was evaluated in 490 (36.5%) of the 1344 women with malignant diseases and in 223 (31.6%) of the 706 women with benign diseases. This difference was also statistically significant (P = .028).

Multivariate Analysis of the Factors Identified as Significant in Univariate Analyses

The 4 variables significantly associated with postoperative PE based on univariate analysis were evaluated using multiple logistic regression (Table 3). Multivariate analysis demonstrated that preoperative d-dimer elevation was associated with postoperative PE (odds ratio, 6.331; 95% confidence interval [CI], 1.567-25.589; P = .010; Table 3). The other 3 variables were not found to be statistically significant risk factors in multivariate analysis (Table 3).

Table 3.

Multivariate Analysis of the Variables Influencing Pulmonary Embolism Incidence.

Variable	Comparison Category	Reference	Odds Ratio	95% CI	P
Age			1.046	0.985-1.110	.145
Operation time			1.003	0.996-1.010	.408
Hypertension	+	−	1.179	0.247–5.635	.836
DD elevation	+	−	6.331	1.567-25.589	.010^a

Abbreviations: CI, confidence interval; DD, d-dimer.

^aSignificant according to the multiple logistic regression (P < .05).

d-Dimer Measurement and CUS Performance

The d-dimer level was measured in 713 (34.8%) of the 2050 women remaining after excluding the 2 cases with preoperative PE (Table 1). d-Dimer elevation (≥1 µg/mL) was present in 175 cases (Table 1); in 124 cases, the d-dimer level was 1.5 µg/mL or more. Among the 175 women with high d-dimer levels, CUS of the lower limbs in their entirety was performed in 102 cases, and venography of the lower limbs was performed in 1 (58.9%) case.

Of the 2050 women included in the analysis, 15 were diagnosed with preoperative DVT; they all underwent CUS because of high d-dimer levels. No cases of postoperative PE arose from these 15 DVT cases.

Use of IPC Devices Among Women With High d-Dimer Levels

Of the 124 women, 21 (16.9%) with high d-dimer levels of ≥1.5 µg/mL did not undergo CUS or venography but did receive IPC. Of the 21 women, 7 had a decreased Spo ₂ level. Of the 7 women, 1 was diagnosed with postoperative PE (PE13; Table 2).

Discussion

The identification of asymptomatic PE cases just after surgery might not be clinically important if most of these emboli tended to be subsegmental, because patients with subsegmental PE can have a favorable clinical outcome even without receiving anticoagulant therapy.⁴ However, in the present study, 3 of the 4 patients who were asymptomatic or had minor symptoms when diagnosed with PE on postoperative days 1 and 2 had emboli in the segmental or more proximal arteries. Thus, diagnosing very early postoperative PE appears to be clinically relevant.

Plasma d-dimer level is highly sensitive (>95%) in excluding acute DVT or PE when it is below a certain threshold, usually 500 μg/L.⁶ This is true in gynecologic oncology. In previous studies, 1.5 μg/mL was chosen as a suitable cutoff value for detecting VTE in patients with ovarian cancer, endometrial cancer, and cervical cancer prior to treatment.^2,12,13 In our study, women with high preoperative d-dimer levels of 1 to 1.4 µg/mL developed postoperative PE, and preoperative d-dimer elevation was identified as a statistically significant risk factor for postoperative PE development in multivariate analysis. Accordingly, we should closely observe cases with preoperative d-dimer elevations even when the value is less than 1.5 µg/mL, and we should react promptly when we encounter low postoperative Spo ₂ levels among such patients. This is particularly important because patients who developed PE just after surgery rarely presented with respiratory symptoms in our study.

Arterial blood gas oxygen saturation levels have low sensitivity and specificity for PE diagnosis and are incorporated in neither the Wells score nor the revised Geneva score.⁶ Consequently, arterial blood gas saturation should not be routinely used to exclude PE.⁶ One of the reasons for this is that PE does not necessarily cause hypoxia. In a previous study, hypoxia was present in only 35% of PE cases.¹⁴ A high plasma d-dimer value on postoperative day 3 was found to be an independent risk factor for postoperative VTE among patients with gynecologic cancer in a prior study.¹⁵ Still, there are few readily available tools other than Spo ₂ monitoring to aid in identifying asymptomatic PE on postoperative days 1 and 2. New approaches may be needed to readily identify asymptomatic PE in the early postoperative phase.

Cancer is one of the major risk factors for PE.^6,16 In our previous study, which included the period before methodical prophylaxis was established, the overall incidence of postoperative PE was 2.14%; according to univariate analysis, the incidence was higher in women with malignant diseases than in women with benign diseases.³ However, in the present study, the incidence of postoperative PE did not statistically differ between women with malignant diseases and those with benign diseases, although the mean age was significantly older and the mean operation time was significantly longer among women with malignant diseases. The rate of anticoagulant use and the rate of d-dimer measurement were both significantly higher among women with malignant diseases compared with those with benign diseases. The significantly higher rate of prophylactic anticoagulant use among women with malignant diseases may be related to the lack of a statistically significant difference in postoperative PE incidence between women with malignant and benign diseases. In addition, the significantly higher rate of d-dimer measurement observed in women with malignant diseases may have aided in preoperative DVT diagnosis and may have prevented IPC misuse in women with preoperative DVT.

Among women who underwent vaginal or vulvar surgery, there were no cases of postoperative PE, although the type of surgery was not identified as a statistically significant factor. Opening the abdominal cavity and packing off the intestines may impede blood flow in the vena cava, potentially promoting more venous stasis than that which occurs in association with vaginal procedures.¹⁷

A strength of the present study was its focus on very early, asymptomatic postoperative PE. Our study does have several limitations. First, 108 of the 553 patients who had several episodes of low Spo ₂ (≤93%) did not undergo CXP. Second, we only examined 3 types of concomitant diseases. Third, although a high preoperative d-dimer level was identified as a significant variable associated with postoperative PE according to multivariate analysis, the 95% CI was rather wide. Further study in another population may be needed in the future. Finally, d-dimer evaluation was only performed in 34.8% (713 of 2050) of the patients, and some postoperative PE patients might have already had emboli preoperatively. Even so, the early detection of postoperative PE may have been important in reducing cases of massive or fatal postoperative PE, which were not seen in the present study. Preoperative d-dimer levels were less than 1.5 µg/mL among 5 of the 15 PE cases, indicating that PE likely occurred intra- or postoperatively in these cases.

In conclusion, patients who were asymptomatic or had minor symptoms at PE diagnosis on postoperative days 1 and 2 had emboli in the segmental or more proximal arteries. Identifying asymptomatic PE in the early postoperative period may be clinically significant because most of these emboli are proximal to the subsegmental arteries.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was supported by grants-in-aid for cancer research from the Ministry of Health, Labour and Welfare (nos. 11-1 and l5-6) of Japan.

References

Williams

Eikman

Greenberg

. Asymptomatic pulmonary embolism. A common event in high risk patients. Ann Surg. 1982;195(3):323–327.

Satoh

Oki

Uno

. High incidence of silent venous thromboembolism before treatment in ovarian cancer. Br J Cancer. 2007;97(8):1053–1057.

Okadome

Saito

Miyahara

Yamanaka

Kuroiwa

Kurihara

. Postoperative pulmonary embolism including asymptomatic cases in gynecologic oncology. Int J Gynecol Cancer. 2010;20(4):655–663.

Carrier

Righini

Wells

. Subsegmental pulmonary embolism diagnosed by computed tomography: incidence and clinical implications. A systematic review and meta-analysis of the management outcome studies. J Thromb Haemost. 2010;8(8):1716–1722.

den Exter

van Es

Klok

. Risk profile and clinical outcome of symptomatic subsegmental acute pulmonary embolism. Blood. 2013;122(7):1144–1149.

Goldhaber

Bounameaux

. Pulmonary embolism and deep vein thrombosis. Lancet. 2012;379(9828):1835–1846.

Perrier

Roy

Sanchez

. Multidetector-row computed tomography in suspected pulmonary embolism. N Engl J Med. 2005;352(17):1760–1768.

van Belle

Buller

Huisman

. Effectiveness of managing suspected pulmonary embolism using an algorithm combining clinical probability, d-dimer testing, and computed tomography. JAMA. 2006;295(2):172–179.

PIOPED Investigators. Value of the ventilation/perfusion scan in acute pulmonary embolism. Results of the prospective investigation of pulmonary embolism diagnosis (PIOPED). JAMA. 1990;263(20):2753–2759.

10.

Jaff

McMurtry

Archer

. Management of massive and submassive pulmonary embolism, iliofemoral deep vein thrombosis, and chronic thromboembolic pulmonary hypertension: a scientific statement from the American Heart Association. Circulation. 2011;123(16):1788–1830.

11.

Lin

Schreiber

Liu

. Therapy and outcomes in massive pulmonary embolism from the Emergency Medicine Pulmonary Embolism in the Real World Registry. Am J Emerg Med. 2012;30(9):1774–1781.

12.

Satoh

Matsumoto

Uno

. Silent venous thromboembolism before treatment in endometrial cancer and the risk factors. Br J Cancer. 2008;99(7):1034–1039.

13.

Satoh

Matsumoto

Tanaka

. Incidence of venous thromboembolism before treatment in cervical cancer and the impact of management on venous thromboembolism after commencement of treatment. Thromb Res. 2013;131(4):e127–e132.

14.

Bajaj

Bozarth

Guillot

. Clinical features in patients with pulmonary embolism at a community hospital: analysis of 4 years of data. J Thromb Thrombolysis. 2014;37(3):287–292.

15.

Kodama

Seki

Masahiro

. d-Dimer level as a risk factor for postoperative venous thromboembolism in Japanese women with gynecologic cancer. Ann Oncol. 2010;21(8):1651–1656.

16.

Lee

. Thrombosis in cancer: an update on prevention, treatment, and survival benefits of anticoagulants. Hematology Am Soc Hematol Educ Program. 2010;2010(1):144–149.

17.

Heilmann

von Tempelhoff

G-F

Schneider

. Prevention of thrombosis in gynecologic malignancy. Clin Appl Thromb Hemost. 1998;4(3):153–159.

Early Identification of Asymptomatic Pulmonary Embolism Proximal to the Subsegmental Arteries After Gynecologic Surgery

Abstract

Keywords

Introduction

Methods

Statistical Analysis

Results

Follow-Up Period and Exclusion of Preoperative PE

Identification of Decreased Spo 2 (≤93%) and Suspected Symptoms/Signs of PE

Symptoms and Signs Associated With Postoperative PE

Characteristics of the 15 Women With PE

Locations of Postoperative Pulmonary Emboli

Univariate Analysis of the Patients’ Background Characteristics

Age, Operation Time, Anticoagulant Use, and Preoperative d-Dimer Evaluation in Women With Malignant and Benign Diseases

Multivariate Analysis of the Factors Identified as Significant in Univariate Analyses

d-Dimer Measurement and CUS Performance

Use of IPC Devices Among Women With High d-Dimer Levels

Discussion

Footnotes

Declaration of Conflicting Interests

Funding

References

Identification of Decreased Spo ₂ (≤93%) and Suspected Symptoms/Signs of PE