The largest single-center report on intravenous leiomyomatosis and development of a classification to guide surgical management

Abstract

Background

Intravenous leiomyomatosis (IVL) is a rare neoplasm; the accumulated knowledge about the characteristics and prognosis of this tumor has been derived mainly from isolated case reports with no comprehensive research. In this study, we reviewed our institution's experience with IVL over a 20-year period and developed a classification system that can be used to guide surgical management.

Methods

The study had a retrospective cohort design and included patients who underwent resection of IVL at our institution between January 2002 and December 2022. Perioperative parameters were then collected among four stages of our proposed classification. The long-term outcomes, oncologic prognosis, and factors associated with recurrence were analyzed.

Results

A total of 216 patients were included (stage 1, n = 92; stage 2, n = 39; stage 3, n = 76; stage 4, n = 9). The mean follow-up duration was 26.34 months, during which 18 patients (9.7%) in the complete resection group had recurrence, and 12 (39.0%) in the incomplete resection group showed disease progression. Recurrence or progression of residual disease was associated with adjuvant aromatase inhibitor therapy and maximum tumor thrombus diameter but not with total hysterectomy and bilateral salpingo-oophorectomy, age, or postoperative treatment with a gonadotropin-releasing hormone agonist therapy.

Conclusions

This is the largest single-center report on IVL published to date and provides valuable information on its clinical features and long-term outcomes, as well as surgical technique. Our classification system can be used to evaluate the extent of lesion involvement and guide surgical management.

Article Highlights.

• •Type of Research: Single-center retrospective cohort study
• •Key Findings: Surgical resection is the main treatment for intravenous leiomyomatosis, and it can be performed safely without perioperative mortality and minimal postoperative complications. Among total 216 females in the cohort, 18 patients (9.7%) in the complete resection group had recurrence, and 12 (39.0%) in the incomplete resection group showed disease progression. Recurrence or progression was associated with adjuvant aromatase inhibitor therapy and maximum tumor thrombus diameter but not with total hysterectomy and bilateral salpingo-oophorectomy, age, or postoperative treatment with a gonadotropin-releasing hormone agonist therapy.
• •Take Home Message: The long-term overall outcome after surgical resection is acceptable. A classification provides evaluation of the extent of an intravenous leiomyomatosis and can well guide its surgical management.

Intravenous leiomyomatosis (IVL) is a rare neoplasm that is histologically benign but biologically aggressive and originates either from a uterine leiomyoma with vascular invasion¹ or in the smooth muscle wall of a vein.² IVL has unique biological behavior and a highly aggressive component. The lesion tends to spread along venous channels, but without invading the vessel wall itself and is confined to the uterus in the early stages. However, it can progress to involve the inferior vena cava (IVC) via the iliac vein or gonadal/renal vein, reaching the right cardiac chambers and the main pulmonary artery. IVL can even metastasize to the lung via the venous circulation.

The clinical manifestations of IVL are various and nonspecific. IVL may be asymptomatic in the earliest stages, but can cause sudden death by obstruction of the entire outflow tract or pulmonary embolism if it extends into the heart. Surgical resection is the primary treatment, but removal of tumor thrombus is a complex surgical procedure and difficult to perform in institutions lacking experience with IVL, especially when the lesion has progressed to involve the heart.

In view of the rarity of IVL, the accumulated knowledge about this tumor to present has been derived mainly from isolated case reports or small cohorts with limited long-term follow-up. Hence, data on the features of IVL and its oncologic prognosis are limited, and there has been no consensus regarding the surgical strategy.

In this study, we reviewed our institutional experience with IVL over 20 years to demonstrate its clinical characteristics, evaluate the long-term outcomes of surgery, and identify factors associated with the prognosis in the long term. We also introduce our classification system whereby patients can be staged to guide surgical management. We obtained the patients' agreement to publish their case details and images.

Methods

This study was approved by the Institutional Review Board of Peking Union Medical College Hospital (approval number I-23PJ848; date of approval May 24, 2023).

Patient population and data collection

We reviewed the medical records of all patients who underwent surgery for IVL at Peking Union Medical College Hospital between January 2002 and December 2022. A total of 216 women were enrolled in the study. Information was collected on patient demographics, presenting symptoms, previous surgical history, diameter of the tumor in the IVC and heart, surgical technique, intraoperative blood loss, blood transfusion requirement, operation time, adjuvant therapy, and follow-up.

Definitions: Classification of stage and type

The presence and extent of the tumor were evaluated based on characteristics seen at baseline on imaging, namely, computed tomography (CT) scan, ultrasound examination, and/or echocardiography. Representative examples are shown in Fig 1.

Fig 1.

Representative examples. (A-C) Representative case of a patient with continuous tumor thrombus that reached the pulmonary artery. (A) Surgical specimen after resection of a pelvic cavity mass, the uterus, and the ovaries, and removal of a tumor from the right iliac vessel (RIV), inferior vena cava (IVC), right atrium (RA), tricuspid valve (TV), right ventricle (RV), right ventricular outflow tract (RVOT), and right pulmonary artery (PA). The tumor protruded through the TV and was adherent to it. (B) A computed tomography (CT) scan showing a continuous intravascular leiomyomatosis (IVL). The red arrow indicates the IVL inside the vessel and heart chambers. (C) Three-dimensional reconstructed CT image indicating the tumor inside the vessel and heart chambers. (D-F) Representative case of a patient with continuous tumor thrombus reached pulmonary artery. (D) Surgical specimen after resection of a mass from the right ovary (RO) and removal of a tumor from the left iliac vessel (LIV), IVC, RA, TV, RV, RVOT, and pulmonary artery. An isolated tumor thrombus is adherent to the TV. (E) A CT scan showing a continuous mass. The red arrow indicates an IVL inside the vessel and heart chambers. (F) Three-dimensional reconstructed CT image. (G and H) Surgical specimens from two cases with tumor thrombus arising from the RIV and extending into the RV. (I-O) A representative example of the use of CT scan, ultrasound examination, and echocardiography to evaluate a patient before surgery. (I) Surgical specimen after resection of a pelvic cavity mass and ovaries and removal from the left gonadal iliac vessel (LGV), left renal vein (LRV), IVC, RA, and RV. (J) CT scan. (K) Three-dimensional reconstructed CT image. (L) Echocardiogram showing an intracardiac tumor located in the right cardiac chamber and crossing the TV. (M) Ultrasound image showing a tumor inside IVC. (N) Ultrasonographic image showing a tumor inside the LRV in which the lesion is separate from the vessel wall and filled with colored blood flow. (O) Ultrasound image showing a tumor originating from the LGV and extending into the LRV. The blue arrow indicates where the LGV drains into the LRV. The yellow arrow indicates the tumor in the pelvic cavity. (P) Surgical specimen of tumor thrombus arising from the LIV and extending into the RV. (Q) Surgical specimen of tumor thrombus arising from the LGV and extending into the RV and RA. The specimen is particularly small in diameter. (R) Surgical specimen of a tumor thrombus arising from the RIV and extending into the LPA.

The patients were classified into the following four stages according to the extent of tumor thrombus involvement: stage 1 (tumor thrombus growing in the abdominopelvic cavity, including thrombus confined to the veins within the uterine myometrium and extending to the iliac vein or gonadal vein/renal vein from the uterine vein but not into the IVC); stage 2 (thrombus extending to the IVC but not reaching the right atrium); stage 3 (thrombus extending into the right atrium or protruding through the tricuspid valve into the right ventricle but below the pulmonary valve); or stage 4 (thrombus extending above the pulmonary valve and reaching the pulmonary artery).

Stage ≥2 IVC was divided into the following three types: type I (tumor thrombus continuous and intravenous lesions do not adhere to the vessel wall [ultrasound examination shows relative movement within the vessel wall or endocardium]); type II (tumor thrombus continuous but adherent to the vessel wall or endocardium [ultrasound examination does not show relative movement within the vessel wall or endocardium]); type III (tumor thrombus is isolated and attached to the vessel wall or endocardium via a pedicle, such as the entrance of the hepatic veins or IVC, coronary sinus, or tricuspid valve; the lesion is an independent metastasis from the uterine IVL to any of the above-mentioned locations via the bloodstream). A schematic diagram of this classification is shown in Fig 2. Perioperative parameters were then analyzed using this classification system.

Fig 2.

Classification of intravascular leiomyomatosis (IVL). Schematic diagram of our classification system for intravascular leiomyomatosis.

Surgical technique and adjuvant therapy

The surgical approach typically involves laparotomy or thoracolaparotomy, resection of the IVL in the abdominopelvic cavity, removal of tumor thrombus from the vessels and heart, and total hysterectomy and bilateral salpingo-oophorectomy (TH-BSO). Details of our surgical management based on our classification system are shown in the Appendix.

Adjuvant therapy is generally offered to patients who retain their ovaries and to those who have undergone TH-BSO but still have low circulating levels of estrogen. This treatment usually consists of a gonadotropin-releasing hormone agonist or an aromatase inhibitor. In the absence of standard guidelines, adjuvant therapy is based on the attending physician's preference.

Clinical end points and follow-up

Patients were followed every 6 months after surgery. If a new mass was detected on imaging during follow-up, the tumor was deemed to have recurred. Recurrence was divided into intravascular and extravascular. Residual progression was defined as residual tumor that had increased by >1 cm in diameter since the previous postoperative examination. Duration of follow-up was calculated from the date of surgery to the date of recurrence/residual disease progression or last follow-up.

Metastatic pulmonary nodules were characterized as concurrent multiple round-shaped nodules of varying size that were distributed randomly in the lung fields bilaterally. Metastasizing leiomyoma was confirmed on biopsy in four of these patients, other type(s) of pulmonary nodules could not be identified in these cases but were assumed to be caused by IVL. We also monitored the prognosis of pulmonary nodules by CT scan during follow-up and classified them as unchanged, progressed, or regressed.

Statistical analysis

Continuous variables are shown as the mean ± standard deviation or as the median (range) depending on whether the data were distributed normally. Patients were stratified according to stage and subsequent comparative analyses were undertaken. Groups were compared using one-way analysis of variance or the Mann-Whitney U test. Categorical variables are reported as the frequency (percentage) and were compared between groups using the χ² test or Kruskal-Wallis rank-sum test.

Survival was analyzed using the Kaplan-Meier method, with the log-rank test used to assess associations between recurrence or progression and various prognostic factors. The end point was defined as recurrence or progression, and loss to follow-up was treated as a censoring event. For censored observations, the duration of follow-up was calculated from the date of surgery to the date of the last recorded follow-up. Univariable Cox proportional hazards models were used to investigate the influence of various patient-related and surgery-related covariates on recurrence or progression of IVL. Variables with a P value of <.05 in univariate analysis were included in the multivariable Cox proportional hazards regression analysis.

All statistical analyses were performed using R version 4.3.1. P values of <.05 were considered statistically significant.

Results

Patient characteristics

The study population included 216 women with a mean age of 45 ± 7 years (stage 1, n = 92; stage 2, n = 39; stage 3, n = 76; stage 4, n = 9) (Fig 3). At diagnosis, 78 patients (36.0%) were asymptomatic. The most frequently presenting symptoms were hypogastralgia (13.0%, n = 27), chest tightness (10.0%, n = 22), abdominal distension (8.8%, n = 19), and an abdominopelvic mass or edema in the lower extremities (7.9%, n = 17), which varied significantly in frequency according to disease stage. The majority of patients (59.0%) had a history of gynecologic surgery. At the time of diagnosis, the mean maximum tumor diameter was 1.88 ± 1.08 cm in the IVC and 2.56 ± 1.26 cm in the heart. Seventy-six patients (36.0%) had type I, 32 (15.0%) had type II, and 9 (4.3%) had type III disease. Pulmonary nodules were seen in more than one-half (58.0%) of the patients at diagnosis. There was no significant difference in the distribution of age or in growing pulmonary nodules between the four stages (P > .05). The baseline patient characteristics are summarized in Table I.

Fig 3.

Distribution of stages, Distribution of stages in our total patient population with intravascular leiomyomatosis. IVC, inferior vena cava.

Table I.

Summary of patients' clinical characteristics

Variable	Overall (n = 216)	1 (n = 92)	2 (n = 39)	3 (n = 76)	4 (n = 9)	P value
Age	45 ± 7	43 ± 8	45 ± 7	46 ± 7	45 ± 5	.070b
Presenting symptoms and signs
Asymptomatic	78 (36)	41 (45)	20 (51)	16 (21)	1 (11)	<.001b
Dyspnea	3 (1.4)	0 (0)	0 (0)	3 (3.9)	0 (0)	.2b
Palpitation	12 (5.6)	1 (1.1)	0 (0)	10 (13)	1 (11)	.002b
Abdominopelvic mass	17 (7.9)	7 (7.6)	4 (10)	5 (6.6)	1 (11)	.7b
Dizziness	5 (2.3)	1 (1.1)	0 (0)	4 (5.3)	0 (0)	.4b
Vaginal bleeding	16 (7.4)	10 (11)	0 (0)	5 (6.6)	1 (11)	.10b
Menostaxis	9 (4.2)	8 (8.7)	1 (2.6)	0 (0)	0 (0)	.028b
Hypermenorrhea	16 (7.4)	11 (12)	2 (5.1)	3 (3.9)	0 (0)	.2b
Dysmenorrhea	1 (0.5)	0 (0)	0 (0)	1 (1.3)	0 (0)	.6b
Frequent micturition	10 (4.6)	6 (6.5)	1 (2.6)	3 (3.9)	0 (0)	.8b
Incontinence	2 (0.9)	1 (1.1)	1 (2.6)	0 (0)	0 (0)	.5b
Dysuria	1 (0.5)	0 (0)	0 (0)	0 (0)	1 (11)	.042b
Constipation	1 (0.5)	0 (0)	1 (2.6)	0 (0)	0 (0)	.2b
Oedema of the bilateral lower extremities	17 (7.9)	0 (0)	3 (7.7)	12 (16)	2 (22)	<.001b
Swelling of right lower limb	3 (1.4)	1 (1.1)	2 (5.1)	0 (0)	0 (0)	.2b
Abdominal distension	19 (8.8)	7 (7.6)	0 (0)	10 (13)	2 (22)	.023b
Hypogastralgia	27 (13)	14 (15)	8 (21)	5 (6.6)	0 (0)	.088b
Lumbodorsalgia	5 (2.3)	2 (2.2)	1 (2.6)	2 (2.6)	0 (0)	>.9b
Syncope	10 (4.6)	0 (0)	0 (0)	9 (12)	1 (11)	<.001b
Cough	2 (0.9)	0 (0)	0 (0)	1 (1.3)	1 (11)	.046b
Chest tightness	22 (10)	0 (0)	2 (5.1)	16 (21)	4 (44)	<.001b
Chest pain	2 (0.9)	0 (0)	0 (0)	2 (2.6)	0 (0)	.2b
Shortness of breath	13 (6.0)	0 (0)	1 (2.6)	11 (14)	1 (11)	<.001b
Suffocated	5 (2.3)	0 (0)	0 (0)	4 (5.3)	1 (11)	.025b
Fatigue	7 (3.2)	0 (0)	1 (2.6)	6 (7.9)	0 (0)	.028b
Decreased activity endurance	2 (0.9)	0 (0)	0 (0)	2 (2.6)	0 (0)	.2b
Shock	1 (0.5)	0 (0)	0 (0)	1 (1.3)	0 (0)	.6b
Previous gynecological surgery
None, coexisting uterine leiomyoma	88 (41)	48 (52)	8 (21)	29 (38)	3 (33)
Myomectomy	48 (22)	31 (34)	4 (10)	12 (16)	1 (11)
Partial hysterectomy	9 (4.2)	4 (4.3)	3 (7.7)	2 (2.6)	0 (0)
Total hysterectomy	33 (15)	5 (5.4)	6 (15)	20 (26)	2 (22)
TH-USO	26 (12)	2 (2.2)	12 (31)	9 (12)	3 (33)
TH-BSO	12 (5.6)	2 (2.2)	6 (15)	4 (5.3)	0 (0)
Previous nongynecological surgery						.001b
Resection of cardiac mass	7 (3.3)	0 (0)	0 (0)	7 (9.2)	0 (0)
Resection of IVC mass	2 (0.9)	0 (0)	2 (5.3)	0 (0)	0 (0)
Resection of abdominopelvic cavity mass	15 (7.0)	4 (4.3)	6 (16)	4 (5.3)	1 (11)
Resection of cardiac and IVC mass	3 (1.4)	0 (0)	0 (0)	3 (3.9)	0 (0)
Resection of cardiac and abdominopelvic cavity mass	1 (0.5)	0 (0)	0 (0)	1 (1.3)	0 (0)
Extension route vein
Iliac vein	74 (35)	-	27 (71)	41 (56)	6 (67)
Gonadal vein	32 (15)	-	10 (26)	21 (29)	1 (11)
Both	11 (5.2)	-	1 (2.6)	8 (11)	2 (22)
Extension route side
Right	76 (36)	-	26 (68)	46 (63)	4 (44)
Left	30 (14)	-	9 (24)	17 (23)	4 (44)
Both	11 (5.2)	-	3 (7.9)	7 (9.6)	1 (11)
Maximal diameter of tumor in IVC	1.88 ± 1.08	-	2.12 ± 1.40	1.85 ± 0.95	1.37 ± 0.83	.2a
Maximal diameter of tumor in heart	2.56 ± 1.26	-	-	2.60 ± 1.25	2.21 ± 1.42	.3a
Type
I	76 (36)	-	27 (77)	47 (64)	2 (22)
Ⅱ	32 (15)	-	8 (23)	20 (27)	4 (44)
Ⅲ	9 (4.3)	-	-	6 (8.2)	3 (33)
Pulmonary nodules	83 (58)	28 (58)	13 (48)	35 (59)	7 (88)	.3b

IVC, Inferior vena cava; TH-BSO, total hysterectomy and bilateral salpingo-oophorectomy; TH-USO, total hysterectomy and unilateral salpingo-oophorectomy.

Values are mean ± standard deviation or number (%).

^aKruskal-Wallis rank sum test.

^bFisher's exact test.

Surgical parameters and outcomes

Operative details and complications

The operative details and postoperative complications according to stages are shown in Table II. Sixty-one patients (28.0%) underwent thoracolaparotomy, 151 (70.0%) underwent laparotomy alone, and 4 (1.9%) underwent thoracotomy alone. One hundred eighty-six patients (86.0%) underwent TH-BSO (including myomectomy/TH/TH-unilateral salpingo-oophorectomy previously at another medical center with subsequent resection of the remaining ovary at our institution) and 30 (14.0%) refused TH-BSO. The mean operation time and estimated blood loss were 151 minutes and 590 mL, respectively, for stage 1 IVL, 392 minutes and 1603 mL for stage 2, 450 minutes and 1174 mL for stage 3, and 500 minutes and 1550 mL for stage 4; the differences between the four stages were statistically significant (P < .05). One hundred eighty-two of the 216 patients (85.0%) underwent complete resection and 33 (15.0%) underwent incomplete resection.

Table II.

Surgical details and outcomes

Variable	Overall (n = 216)	1 (n = 92)	2 (n = 39)	3 (n = 76)	4 (n = 9)	P value
Surgical path						<.001b
Thoracolaparotomy	61 (28)	0 (0)	4 (10)	48 (63)	9 (100)
Laparotomy alone	151 (70)	92 (100)	35 (90)	24 (32)	0 (0)
Thoracotomy alone	4 (1.9)	0 (0)	0 (0)	4 (5.3)	0 (0)
TH-BSO						<.001b
No	30 (14)	24 (26)	2 (5.1)	4 (5.3)	0 (0)
Yes, including	186 (86)	68 (74)	37 (95)	72 (95)	9 (100)
Underwent at this visit	120 (56)	60 (65)	14 (36)	42 (55)	4 (44)
Previously resected	12 (5.6)	2 (2.2)	6 (15)	4 (5.3)	0 (0)
Myomectomy/TH/TH-USO previously, resected remaining ovary at this visit	54 (25)	6 (6.5)	17 (44)	26 (34)	5 (56)
Complete resection of tumor						.002b
Yes	182 (85)	87 (95)	27 (69)	62 (82)	6 (75)
No, leaving lesions in the small vessels	9 (4.2)	0 (0)	4 (10)	5 (6.6)	0 (0)
No, leaving lesions in abdominopelvic cavity	21 (9.8)	5 (5.4)	8 (21)	6 (7.9)	2 (25)
No, leaving lesions in the small vessels and abdominopelvic cavity	2 (0.9)	0 (0)	0 (0)	2 (2.6)	0 (0)
No, leaving lesions in heart	1 (0.5)	0 (0)	0 (0)	1 (1.3)	0 (0)
Hepatic mobilization	31 (14)	1 (1.1)	6 (15)	22 (29)	2 (22)	<.001b
Cardiopulmonary bypass						<.001b
No	152 (70)	92 (100)	35 (90)	25 (33)	0 (0)
Yes	64 (30)	0 (0)	4 (10)	51 (67)	9 (100)	<.001b
Beating heart	5 (2)	0 (0)	0 (0)	5 (7)	0 (0)
Arrested heart	59 (27)	0 (0)	4 (10)	46 (61)	9 (100)
Deep hypothermia circulatory arrest						<.001a
Surgery time	49 (23)	0 (0)	2 (5.3)	40 (53)	7 (78)
Total operative time, minutes	307 ± 187	151 ± 85	392 ± 144	450 ± 155	500 ± 107
CPB, minutes	111 ± 41	-	135 ± 33	111 ± 42	104 ± 36	.3a
AoC, minutes	50 ± 20	-	60 ± 13	49 ± 19	49 ± 27	.3a
DHCA, minutes	24 ± 13	-	27 ± 3	25 ± 14	17 ± 11	.2a
Bleeding	1018 ± 1364	590 ± 1001	1603 ± 1806	1174 ± 1333	1550 ± 1490	<.001a
Blood transfusions
RBC, U	2.45 ± 7.89	0.83 ± 1.96	2.19 ± 2.69	4.41 ± 12.68	3.22 ± 2.54	<.001a
FFP, mL	282 ± 440	104 ± 266	312 ± 385	442 ± 558	578 ± 186	<.001a
PLT (U)	0.22 ± 0.53	0.00 ± 0.00	0.24 ± 0.54	0.40 ± 0.67	0.78 ± 0.83	<.001a
GnRHa after surgery	16 (7.8)	13 (15)	1 (2.6)	2 (2.8)	0 (0)	.020b
AI after surgery
Early complications	32 (16)	11 (13)	10 (26)	9 (13)	2 (25)	.2b
Acute liver failure	2 (0.9)	1 (1.1)	0 (0)	1 (1.3)	0 (0)	>.9b
Hemorrhage	2 (0.9)	0 (0)	0 (0)	2 (2.6)	0 (0)	.2b
Urinary fistula	1 (0.5)	0 (0)	0 (0)	0 (0)	1 (11)	.042b
Ileus	2 (0.9)	0 (0)	2 (5.1)	0 (0)	0 (0)	.078b
Hepatic dysfunction	1 (0.5)	0 (0)	0 (0)	1 (1.3)	0 (0)	.6b
Acute renal failure	3 (1.4)	0 (0)	1 (2.6)	2 (2.6)	0 (0)	.3b
Urinary tract infection	5 (2.3)	0 (0)	2 (5.1)	2 (2.6)	1 (11)	.035b
Pneumothorax	1 (0.5)	0 (0)	0 (0)	1 (1.3)	0 (0)	.6b
Bloodstream infection	1 (0.5)	0 (0)	1 (2.6)	0 (0)	0 (0)	.2b
Delayed wound healing	2 (0.9)	1 (1.1)	0 (0)	0 (0)	1 (11)	.11b
Postoperative hoarseness	1 (0.5)	0 (0)	0 (0)	1 (1.3)	0 (0)	.6b
Abdominal distension	19 (8.8)	7 (7.6)	0 (0)	10 (13)	2 (22)	.023b
Pleural effusion	1 (0.5)	0 (0)	0 (0)	1 (1.3)	0 (0)	.6b
Delirium	1 (0.5)	0 (0)	1 (2.6)	0 (0)	0 (0)	.2b
Ventricular arrhythmia	1 (0.5)	0 (0)	1 (2.6)	0 (0)	0 (0)	.2b
Uroschesis	1 (0.5)	0 (0)	1 (2.6)	0 (0)	0 (0)	.2b
Pelvic infection	1 (0.5)	1 (1.1)	0 (0)	0 (0)	0 (0)	>.9b
Vesicovaginal fistula	1 (0.5)	0 (0)	0 (0)	1 (1.3)	0 (0)	.6b
Late complications						.4b
Uracratia	1 (0.5)	0 (0)	0 (0)	1 (1.3)	0 (0)
Scar itching	1 (0.5)	0 (0)	0 (0)	1 (1.3)	0 (0)
Urinary tract infection	1 (0.5)	0 (0)	1 (1.3)	0 (0)	0 (0)
Herniation	1 (0.5)	0 (0)	1 (2.6)	0 (0)	0 (0)
Atrial flutter	1 (0.5)	0 (0)	0 (0)	1 (1.3)	0 (0)
Keloidosis	1 (0.5)	0 (0)	0 (0)	1 (1.3)	0 (0)
Ileus	3 (1.4)	1 (1.1)	2 (5.1)	0 (0)	0 (0)
Vesicovaginal fistula	1 (0.5)	1 (1.1)	0 (0)	0 (0)	0 (0)
Pulmonary nodules outcome						.8b
Unchanged	104 (91)	35 (92)	18 (86)	44 (92)	7 (100)
Progressed	1 (0.9)	0 (0)	0 (0)	1 (2.1)	0 (0)
Regressed	9 (7.9)	3 (7.9)	3 (14)	3 (6.3)	0 (0)

AI, aromatase inhibitor; AoC, aortic cross-clamp; CPB, cardiopulmonary bypass; DHCA, deep hypothermia circulatory arrest; FFP, fresh frozen plasma; GnRHa, gonadotropin-releasing hormone agonist; PLT, platelets; RBC, red blood cells; TH, total hysterectomy; TH-BSO, total hysterectomy and bilateral salpingo-oophorectomy.

Values are number (%) or mean ± standard deviation.

^aFisher's exact test.

^bKruskal-Wallis rank sum test.

There were no postoperative deaths during the study period. Early complications were seen in 21.0% of patients, the most common being abdominal distension (19.0%) and urinary tract infection (5.0%). One patient developed hemorrhage that required reoperation and another developed a urinary fistula while in hospital and underwent ureteral reimplantation before discharge. The other early complications resolved completely with conservative therapy.

The total incidence of late complications was 5%. Ileus was detected in three patients (1.4%), with one case each of urticaria, itching at the scar site, urinary tract infection, herniation, atrial flutter, keloidosis, and vesicovaginal fistula. One patient developed vesicovaginal leak and returned for repair of a vesicovaginal fistula 9 months after surgery. Another patient developed an abdominal hernia and underwent herniorrhaphy at a local medical center.

After surgery, a gonadotropin-releasing hormone agonist was prescribed for 16 patients (7.8%), most of whom wished to retain their ovaries for fertility preservation. Thirty-two patients (16.0%) received an aromatase inhibitor postoperatively, mainly because of the detection of estrogen during follow-up after TH-BSO.

Prognosis of metastatic pulmonary nodules

During follow-up, 98.9% of metastatic pulmonary nodules remained stable. The nodules progressed in one patient who did not undergo TH-BSO and regressed in nine patients (7.9%).

Oncologic outcomes

The oncologic prognosis was subjected to statistical analysis in the complete resection group and in the incomplete resection (residual disease) group.

Recurrence and residual progression rate

The distribution of patients according to whether resection was complete or incomplete is shown in Supplementary Table I (online only). Complete resection was achieved in 185 patients (86.0%) and incomplete resection in 31 (14.0%). The patients were followed for a mean of approximately 26.34 months (range, 1.0-210.0 months).

Eighteen of the 185 patients (9.7%) in whom resection was complete developed recurrence, with a statistically significant difference in the recurrence rate according to IVL stage. Twelve of the 31 patients (39.0%) in whom resection was incomplete were found to have progression of residual lesions (Supplementary Table I, online only). The median time to recurrence was 22 months (range, 4-72 months) and that to detection of residual disease progression was 11.5 months (range, 5-20 months). Kaplan-Meier curves showing the overall probability of no recurrence and no progression are depicted in Fig 4, A.

Fig 4.

Kaplan-Meier survival analysis. (A) Kaplan-Meier curves showing the overall probability of patients having no recurrence or progression was not significantly greater in the more advanced stages of intravascular leiomyomatosis but there was a trend for higher recurrence rates in both the complete resection group (B) and the incomplete resection group (C). TH-BSO (D and E) and type of intravascular leiomyomatosis (F and G) were not associated with recurrence/progression. TH-BSO, total hysterectomy and bilateral salpingo-oophorectomy.

Influence on recurrence and progression of residual disease

Kaplan-Meier survival analysis was performed to examine the associations between recurrence/progression and various factors. Kaplan-Meier curves were also analyzed for the complete resection group and incomplete resection (residual) group. The curve analysis showed that the recurrence rate was not significantly higher in patients with more advanced stage (P = . 36), but did indicate a trend for a higher recurrence rates in the complete resection group (Fig 4, B). There was a high rate of progression of residual lesions in the incomplete resection group (P = .46), particularly in patients with stage 4 IVL (Fig 4, C). Differences in the probability of progression of residual disease according to type of IVL did not reach statistical significance in either the complete resection group (Fig 4, D) (P = .10) or the incomplete resection group (Fig 4, E) (P = .46). There was no association between TH-BSO and recurrence (Fig 4, F) (P = .95) or progression (Fig 4, G) (P = .16).

Risk factors for recurrence and progression of residual disease

Univariate and multivariate analyses were used to identify risk factors associated with recurrence and disease progression. Factors were selected for inclusion in these analyses based on their clinical significance.

In the complete resection group (Supplementary Table II, online only), the only significant risk factor for recurrence identified in univariate analysis was adjuvant aromatase inhibitor therapy (hazard ratio [HR], 5.05; 95% confidence interval [CI], 1.83-14.0, P = .002). This association remained significant in multivariate analysis (HR, 5.20; 95% CI, 1.13-23.9, P = .034).

In the incomplete resection group (Supplementary Table III, online only), univariate analysis revealed a significant association of the maximal diameter of the IVC (HR, 0.43; 95% CI, 0.19-0.97; P = .042) and the chambers of the heart (HR, 0.49; 95% CI, 0.22-1.06; P = .068) with progression of residual disease. However, this association was not significant on multivariate analysis. There was no significant association of TH-BSO, age, or adjuvant gonadotropin-releasing hormone agonist therapy with recurrence or progression in either group.

Perioperative outcomes according to surgical entry approach in patients with type I IVL

Initially, thoracolaparotomy and extracorporeal circulation were planned in all patients with stage ≥2 IVL. However, with increasing experience, patients with stage ≥2 type I IVL were treated successfully without sternotomy or extracorporeal circulation. A less invasive non-open-heart surgical approach of laparotomy alone were performed to remove IVL in patients with stage ≥2 type I IVL. We also investigated the changes in perioperative outcomes in patients with stage ≥2 type I IVL, according to whether the surgical entry path was thoracolaparotomy or laparotomy alone (non-open heart). The operation time was significantly shorter (P < .01) with less transfusion of platelets (P < .017) in patients who underwent laparotomy alone. However, there was no significant difference in the bleeding rate or the complication rate between the two surgical approaches (Supplementary Table IV, online only).

Discussion

IVL was first described by Birch-Hirschfeld³ in 1896, and Dürck⁴ reported the first case of extension of IVL to the heart in 1907. To date, few cases have been reported, and there are even fewer reports of cases with cardiac involvement. A recent meta-analysis published in December 2022 showed that only 748 cases have been reported in the English literature worldwide.⁵ To the best of our knowledge, our present study is the largest ever published on IVL from a single center.

Given the rarity of IVL, its diagnosis, particularly in the early stages, poses notable challenges. Clinical symptoms lack specificity, and the initial discovery of an intravascular mass often raises the possibility of misdiagnosis as thrombosis or malignant tumor thrombus. Based on our clinical experience, the primary diagnostic modalities encompass ultrasound examination of the pelvis and IVC and transthoracic echocardiography in conjunction with contrast-enhanced CT scan of the abdomen and CT venography of the IVC. In women of childbearing age, a diagnosis can be established by assessment of imaging features, considering the distinct growth pattern of the tumor within the vascular system.

Previous studies of IVL had confused the progression of residual disease with recurrence. Our present study is the first to evaluate the outcome according to whether the resection of IVL is complete or incomplete. Our results demonstrate that surgical resection of IVL can be performed safely without perioperative mortality and minimal postoperative complications. We found that the recurrence rate during follow-up was 9.7% in the complete resection group and the residual disease progression rate was 39.0% in the incomplete resection group; these rates are substantially higher than in our previously reported study⁶ but lower than the recurrence rate of 27.8% during a mean follow-up of 20.5 months in 30 patients with IVC or cardiac involvement⁷ and a rate of 31.0% during a median follow-up of 11.5 months in 58 patients from our center.⁸ However, the results of these studies should not be compared because the present study includes a significantly larger number of patients who were followed for a longer period of time and also includes an incomplete resection group. Among all the patients who developed recurrence or progressive disease, only 10 underwent repeat surgery; the remaining patients did not experience severe discomfort and were kept under observation. Regular follow-up should be performed in all patients with IVL. Metastatic pulmonary nodules remained stable and rarely required further treatment.

Among our univariate and multivariate analyses, the factors not undergoing TH-BSO were not independent risk factors for recurrence or residual progression, which is inconsistent with our previous findings.⁶ However, this result should be interpreted with caution in view of the very small number of patients in this cohort who did not undergo TH-BSO. Nevertheless, in our experience, TH-BSO is necessary to avoid recurrence and progression of residual disease. A total of 116 patients had a history of gynecologic surgery with ovarian preservation (non-THBSO procedures) and 23 had a history of heart, IVC, and pelvic mass resections (without ovarian removal). Some of these patients experienced repeated recurrences and eventually sought treatment at our hospital. For example, three patients, having been misdiagnosed as having myxoma, presented to a local hospital for surgery and subsequently developed a recurrence. Similarly, Yu et al⁸ found that large vein involvement was associated with an increased risk of postoperative recurrence, whereas age, hysterectomy, resection of both ovaries, and hormone therapy were not. Based on our present findings, we suspect that the aggressiveness of the tumor is an independent risk factor for tumor progression regardless of the surgical procedure and adjuvant therapy used. However, this possibility should be confirmed by pathological and biological research in the future. Furthermore, the lesions can be reduced in size in some patients using a gonadotropin-releasing hormone agonist, which indicates that IVL is a hormone-dependent tumor. Therefore, whether supplementary hysterectomy and bilateral oophorectomy are necessary for IVL remains controversial. We believe that TH-BSO should be performed, despite our research suggesting it is not related to recurrence.

Our previous classification system for IVL⁶ included stages, but not types, for distinguishing whether a tumor is adherent to the vascular wall or the endocardium, which is crucial when planning the surgical strategy. Two previous studies, one by Liu et al⁹ and the other by Li et al,¹⁰ only included patients with IVL involving the heart. Therefore, based on our accumulated experience, we propose a comprehensive, simple, and convenient classification that is appropriate for all patients. This classification system is based on operative complexity and extent of tumor involvement. The diameter of the tumor and vessel and their interrelationship are not considered. As shown in Table II, most of the operative parameters are significantly different among the four stages. Therefore, this classification can be a better guide for surgery and convenient for multidisciplinary communication.

The multidisciplinary team, which involves a cardiac surgeon, gynecologist, vascular surgeon, general surgeon, urologist, imaging specialists, and an anesthesiologist, has become critical in the treatment of IVL. Initially, accurate classification and disease evaluation are highly reliant on high-quality CT scan, ultrasound examination, and echocardiography. Patients with a hypoechoic lesion in the uterine or systemic veins are often misdiagnosed as having myoma or venous thrombosis. Therefore, an experienced sonographer is invaluable for diagnosis of IVL and can provide detailed information on factors such as venous wall adhesion, the origin and extension route of the tumor, and surgical difficulty. A cardiac surgeon, gynecologist, and vascular surgeon are essential components of the team, managing the multidisciplinary approach and performing the operation. Furthermore, a general surgeon can help to perform hepatic mobilization if necessary, and a urologist can place temporary bilateral ureteral catheters before surgery and can sometimes deal with intraoperative complications, such as a vesicoureteral fistula. An anesthesiologist who offers real-time transesophageal echocardiographic monitoring can make it possible for surgeons to perform laparotomy only and evaluate the completeness of surgical excision.

Our previous studies^6,9 and studies by other teams^10,11 had successfully developed an adequate surgical strategy for removing the tumor, focusing mainly on a less invasive approach (Supplementary Video, online only). In a previous report,¹² we described the advantages of laparotomy alone without sternotomy and extracorporeal circulation in selected patients. We have also explored the outcomes of different surgical approaches using our classification system and concluded that the operation time is shorter with less blood loss in patients with stage 2 type 1 IVL when laparotomy is performed. Other advantages of non-open-heart surgery include avoidance of invasive mechanical ventilation after surgery and a stay in the intensive care unit and lower costs, although these factors were not assessed in this study. Furthermore, our study defines patients who are candidates for laparotomy alone. Given that tumor tissue is firm enough to endure high tensile strength without breaking and has plasticity, it is generally safe to perform this surgery for the majority of patients with type I, stage 3 IVL. However, a very small subset of patients may have tumor thrombus that is enlarged at the proximal end within the cardiac chamber while the part within the IVC is exceptionally small (diameter of <3 mm). In such cases, hasty extraction of tumor thrombus from below the vein can pose a risk of rupture of the tumor thrombus and sudden-onset fatal tumor embolism without the support of extracorporeal circulation. A thorough evaluation is essential, and when necessary, the tumor should be removed with extracorporeal support to ensure the patient's safety.

The limited number of cases involving adjuvant therapy results in insufficient statistical power to analyze the correlation between adjuvant therapy and recurrence or progression. Another limitation is that surgery was performed by various surgeons with varying experience in surgical decision-making.

In conclusion, the main finding of this study is that the long-term overall outcome after surgical resection is acceptable. Moreover, it provides valuable experience in terms of clinical features and surgical technique. Finally, our classification provides a system for evaluation of the extent of an IVL and can guide its surgical management.

Author Contributions

Conception and design: YW, GM, QM

Analysis and interpretation: YW, WL

Data collection: YW, JS, XL, CZ, JL, DC, NC, JW

Writing the article: YW

Critical revision of the article: YW, GM, QM, JS, WL, XL, CZ, JL, DC, NC, JW

Final approval of the article: YW, GM, QM, JS, WL, XL, CZ, JL, DC, NC, JW

Statistical analysis: YW, WL

Obtained funding: QM

Overall responsibility: GM

Funding

This study was funded by the National High Level Hospital Clinical Research Funding (2022-PUMCH-B-104) of Chinese government.

Disclosures

None.