Predictors of Tubal Rupture in Ectopic Pregnancy: A Retrospective Analysis of Clinical and Laboratory Factors

Summary

This retrospective study analyzed 150 patients with tubal pregnancy to identify predictors of tubal rupture. A history of pelvic inflammatory disease (PID), serum human chorionic gonadotropin (β-hCG) > 3,000 IU/L, neutrophil-to-lymphocyte ratio (NLR) ≥ 4, and ultrasonographic detection of a yolk sac were independent risk factors.

Abstract

Tubal rupture is a major cause of intra-abdominal hemorrhage and maternal mortality in early pregnancy. Early identification of high-risk patients is clinically important; however, standardized predictive tools integrating multidimensional indicators remain limited, particularly in Chinese populations. The aim of this study was to investigate clinical, laboratory, and ultrasound predictors of tubal rupture in ectopic pregnancy and provide evidence for early risk assessment. This retrospective study included 150 patients with tubal pregnancy treated surgically or medically in our hospital between January 2023 and December 2025. Based on intraoperative findings or ultrasound diagnosis, patients were classified into a ruptured group (n = 51) and an unruptured group (n = 99). Clinical characteristics, laboratory indicators [serum human chorionic gonadotropin (β-hCG), progesterone, hemoglobin, and neutrophil-to-lymphocyte ratio (NLR)], and ultrasound features were analyzed. Univariate and multivariable logistic regression analyses were performed to identify independent risk factors for tubal rupture. Patients in the ruptured group had significantly higher rates of pelvic inflammatory disease (PID) history and moderate-to-severe abdominal pain (P < 0.05). Serum β-hCG and NLR levels were significantly elevated in the ruptured group (P < 0.05). Ultrasound findings showed a larger adnexal mass diameter, greater pelvic fluid depth, and a higher proportion of fetal cardiac activity in ruptured cases (P < 0.05). Multivariable logistic regression identified PID history (odds ratio [OR] = 2.983, 95% confidence interval [CI]: 1.306–6.811, P = 0.009), serum β-hCG > 3,000 IU/L (OR = 4.119, 95% CI: 1.661–10.214, P = 0.002), NLR ≥ 4 (OR = 3.597, 95% CI: 1.613–8.023, P = 0.002), and ultrasonographic detection of a yolk sac (OR = 2.506, 95% CI: 1.015–6.190, P = 0.046) as independent risk factors for tubal rupture. These factors may aid early clinical risk stratification and decision-making.

Introduction

Ectopic pregnancy refers to the implantation and development of a fertilized ovum outside the uterine cavity, with approximately 95% occurring in the fallopian tubes¹. Ectopic pregnancy is a common cause of acute abdominal emergencies in obstetrics and gynecology, with an incidence of approximately 1%–2% among women of reproductive age². Despite the widespread use of high-resolution transvaginal ultrasound (TVUS) and dynamic monitoring of serum human chorionic gonadotropin (β-hCG) in recent years, which have enabled early diagnosis and conservative treatment (such as methotrexate or expectant treatment) for most ectopic pregnancies, tubal rupture remains the most serious and fatal complication of ectopic pregnancy^3,4,5.

When the fallopian tube wall ruptures due to the invasion of trophoblasts and the growth of embryos, it can lead to severe intra-abdominal bleeding, hemorrhagic shock, and even maternal death. According to statistics, deaths caused by ectopic pregnancy rupture account for 6% to 13% of all maternal deaths in early pregnancy⁶. Therefore, in the clinical management of ectopic pregnancy, accurately and quickly identifying patients at high risk of rupture before rupture occurs is a major challenge faced by obstetrics and gynecology emergency departments. Accurate prediction can help physicians determine whether emergency surgical intervention is needed while avoiding unnecessary overtreatment of low-risk patients and preserving fertility⁷. Previous studies have shown that the occurrence of ectopic pregnancy rupture is a complex pathophysiological process involving multiple factors, including the anatomical characteristics of the fertilized egg implantation site, the invasive ability of trophoblasts, the local microenvironment of the patient's body, and the systemic immune-inflammatory response^8,9.

The traditional risk factors for rupture mainly include prolonged amenorrhea, a history of ectopic pregnancy, high serum β-hCG levels, and ultrasound findings of large adnexal masses or the presence of fetal poles and heartbeats. However, the sensitivity and specificity of a single indicator in predicting tubal rupture are often limited. At present, many studies have examined the predictive factors for tubal rupture, but the results are not entirely consistent. Some studies suggest that serum β-hCG levels are associated with the risk of rupture¹⁰, while others have not found such an association. In addition, clinical and ultrasound features such as abdominal pain, pelvic fluid accumulation, and mass size are considered potential predictive factors¹¹. In recent years, some inflammatory indicators, such as the neutrophil-to-lymphocyte ratio (NLR) and platelet-to-lymphocyte ratio, have also been found to be associated with tubal rupture¹².

In recent years, the use of systemic inflammatory response indicators to predict acute abdominal conditions and obstetric complications has attracted widespread attention. Among them, the NLR, as a nonspecific systemic immune-inflammatory marker, has been confirmed by multiple studies to be closely associated with the pathological progression of ectopic pregnancy due to its ease of acquisition, low cost, and ability to sensitively reflect the body's inflammatory status¹³. The invasion of trophoblasts into the muscular layer of the fallopian tube can cause local inflammatory reactions and tissue necrosis. Cytokine release from this local damage may increase circulating neutrophils and reduce lymphocytes. However, there are still few large-scale and systematic studies on the comprehensive value of laboratory indicators such as NLR combined with traditional ultrasound and clinical features in predicting ectopic pregnancy rupture, particularly in Chinese populations, and the existing research conclusions are still controversial^14,15,16.

Although several studies have investigated risk factors for tubal rupture, many were limited by small sample sizes, the evaluation of a single biomarker, or limited evidence from Chinese populations. The neutrophil-to-lymphocyte ratio (NLR) is a readily available inflammatory biomarker that has not been systematically evaluated alongside clinical symptoms and ultrasound parameters in Chinese cohorts. Accordingly, this study retrospectively analyzed clinical, laboratory, and ultrasound data from 150 patients with tubal pregnancy treated at our hospital between January 2023 and December 2025. The aim was to identify factors associated with tubal rupture using multivariable logistic regression and to provide evidence to support early risk assessment and clinical decision-making.

Protocol

The study protocol was approved by the Ethics Committee of Jinhua Municipal Central Hospital (2026, Ethics approval No.277). Because of the retrospective study design, the requirement for informed consent was waived. Data collection and retrospective analysis were conducted in accordance with the Declaration of Helsinki¹⁷. The study used deidentified medical records and involved no direct contact with participants.

Research subjects
Patient identification and screening
This retrospective study reviewed the hospital electronic medical record system for records coded as tubal pregnancy (International Classification of Diseases, 10th Revision [ICD-10] code O00.1) from January 2023 to December 2025. The initial search identified 187 potentially eligible records. Two trained reviewers independently screened all records against the inclusion and exclusion criteria, and disagreements were resolved by a senior attending physician. In total, 150 patients met all criteria and were enrolled (Figure 1).

Classification into ruptured and unruptured groups
Patients were classified using direct intraoperative visualization as the primary reference standard for surgically treated cases. For surgically treated patients, rupture was defined as a visible full-thickness defect in the tubal wall with active bleeding or hemoperitoneum estimated at ≥500 mL that required hemostasis. An unruptured tube was defined as an intact tube without a wall defect, regardless of dilation or congestion. For patients initially managed conservatively, rupture was assigned only when ultrasound showed an ill-defined mass with a large volume of free fluid (>3 cm depth), hemodynamic instability (shock index ≥ 1.0), and subsequent emergency surgery. Patients were classified as unruptured when the mass remained intact without progressive fluid accumulation, and they underwent methotrexate or expectant management, followed by a decline in β-hCG to nonpregnant levels without surgery. When surgical and ultrasound findings conflicted, intraoperative findings took precedence. Two senior attending physicians adjudicated ambiguous cases.

Inclusion criteria
Patients were included if tubal pregnancy was confirmed by postoperative pathological examination or by a combination of clinical findings, serum β-hCG results, and transvaginal ultrasonography (TVUS). Complete clinical history and physical examination records at admission were required. Complete blood count, coagulation function, serum β-hCG, and progesterone measurements obtained within 24 h of admission were required. A gynecological transvaginal or transabdominal ultrasound examination performed within 24 h of admission was required.

For conservatively managed patients, tubal pregnancy was confirmed by all three of the following criteria: serum β-hCG above the institutional discriminatory zone (>1,500 IU/L) with a suboptimal increase (<66% at 48 h); TVUS showing an adnexal mass distinct from the ovary without an intrauterine gestational sac; and clinical symptoms consistent with ectopic pregnancy. Patients with β-hCG < 1,500 IU/L were diagnosed using serial β-hCG measurements showing abnormal kinetics together with ultrasound findings suggestive of ectopic pregnancy, including an adnexal mass, an empty uterus, and/or free fluid.

Exclusion criteria
Patients with cervical, ovarian, abdominal, rudimentary horn, heterotopic, or other nontubal ectopic pregnancies were excluded. Patients were excluded if they had concurrent conditions that could affect inflammatory markers or bleeding risk, including hepatic dysfunction (alanine aminotransferase or aspartate aminotransferase >3 times the upper limit of normal), renal dysfunction (creatinine >2.0 mg/dL), New York Heart Association class III or IV heart failure, active malignancy, or a hematological disorder (hemoglobin <7 g/dL or platelet count <50 × 10⁹/L). Patients with acute or chronic infectious diseases, such as pneumonia or acute gastroenteritis, or autoimmune diseases that could affect the peripheral white blood cell count were excluded. Patients who had received immunosuppressants, glucocorticoids, or a blood transfusion within 2 weeks before admission were excluded. Patients missing any key variable required for analysis, including the admission visual analog scale score, serum β-hCG, complete blood count measures (white blood cell count, absolute neutrophil count, and absolute lymphocyte count), or ultrasound measurements (adnexal mass diameter and pelvic fluid depth), were excluded before group allocation.

Data collection and indicator definitions
Demographic characteristics and medical history
The recorded variables were age, body mass index (BMI), duration of amenorrhea, parity, history of pelvic inflammatory disease (PID), history of ectopic pregnancy, history of pelvic or abdominal surgery, use of assisted reproductive technology for conception, and smoking. PID history was defined as a documented clinical diagnosis in the medical record rather than patient self-report alone. Smoking was defined as any tobacco use within 3 months before admission, regardless of daily quantity.

Clinical manifestations
Abdominal pain severity was assessed using the visual analog scale (VAS)^18,19, with a score >5 indicating moderate-to-severe pain²⁰. The admitting nurse recorded the VAS score within 2 h of arrival using a 10 cm horizontal line, where 0 indicated no pain, and 10 indicated the worst imaginable pain. Vaginal bleeding was recorded as present or absent from the admission examination; present bleeding was defined as active bleeding observed during speculum examination or patient-reported vaginal blood loss within the preceding 24 h, including minimal spotting. The shock index (SI) was calculated as heart rate divided by systolic blood pressure using the first documented vital signs at admission. Because an SI ≥ 0.81 has been associated with rupture²¹, the number of patients with SI > 0.8 was recorded in each group.

Laboratory tests
Venous blood test results from the emergency visit or first admission were extracted. Peripheral venous blood samples were collected within 30 min of admission. Complete blood count parameters were measured in dipotassium ethylenediaminetetraacetate (K₂-EDTA) anticoagulated tubes using an automated hematology analyzer. Serum samples for β-hCG and progesterone analysis were obtained in separator tubes, kept at room temperature for 30 min to permit clot formation, and then centrifuged at 1500 × g for 10 min under ambient conditions. Serum β-hCG and progesterone were measured using an electrochemiluminescence immunoassay analyzer. Recorded variables included serum β-hCG, serum progesterone, hemoglobin (Hb), total white blood cell count (WBC), absolute neutrophil count (NEU), absolute lymphocyte count (LYM), and NLR, calculated as NEU/LYM. Daily calibration and three-level internal quality controls were performed with each batch for both analyzers. Total coefficients of variation were ≤4.2% for β-hCG and ≤5.1% for progesterone during the study period. The β-hCG assay measurement range was 0.100–10,000 IU/L; samples above 10,000 IU/L were automatically diluted and reassayed. Laboratory reference ranges were Hb 110–150 g/L, WBC 3.5–9.5 × 10⁹/L, NEU 1.8–6.3 × 10⁹/L, LYM 1.1–3.2 × 10⁹/L, β-hCG < 5 IU/L in nonpregnant patients, and progesterone 0.2–1.5 ng/mL in the follicular phase. Previous studies reported an increased risk of tubal rupture at serum β-hCG > 3,000 IU/L²² and NLR ≥ 4²³; therefore, these thresholds were recorded for both groups.

Serum β-hCG monitoring
Serum β-hCG was measured at admission in all patients with suspected ectopic pregnancy. For patients managed expectantly or with methotrexate, serial measurements were obtained on days 4 and 7 and then weekly until the concentration declined to <5 IU/L, according to the institutional protocol. A decline < 15% between days 4 and 7, or a plateau or increase, prompted reassessment and consideration of surgery. For patients who underwent immediate surgery, only the baseline β-hCG value was used in the analysis. All measurements were performed using the same electrochemiluminescence immunoassay analyzer with a range of 0.100–10,000 IU/L; samples above 10,000 IU/L were automatically diluted and reassayed.

Ultrasound imaging features
The first gynecological ultrasound report obtained at admission was reviewed. Recorded parameters were the maximum diameter of the adnexal mass, the depth of pelvic free fluid, the detection of a yolk sac, and the detection of a fetal pole with cardiac activity. The maximum diameter of the adnexal mass was recorded as the largest linear dimension among three orthogonal planes. Pelvic fluid depth was measured as the maximum anteroposterior diameter of an anechoic or echogenic fluid collection in the pouch of Douglas on a sagittal view. Yolk sac detection and fetal cardiac activity were recorded as binary variables. Fetal cardiac activity was defined as visible pulsation > 100 beats/min confirmed by motion mode or color Doppler. All images were stored in the picture archiving and communication system. A random 20% sample was remeasured offline by a second blinded reviewer; the intraclass correlation coefficient was >0.90.

Data extraction and quality control
Two trained reviewers independently extracted data using a standardized case report form and were blinded to the outcome group during extraction. The reviewers completed a 2 h training session covering form completion, variable definitions, and source-document navigation. Interreviewer agreement was assessed in a random 10% sample (κ ≥ 0.90 for all categorical variables). Discordant entries were resolved by review of the source records, with adjudication by a senior attending physician when disagreement persisted; fewer than 5% of variables required adjudication.

Statistical analysis
Distributional assumptions for continuous variables were evaluated with the Kolmogorov-Smirnov test. Variables showing a normal distribution were summarized as mean ± standard deviation and compared between groups with an independent-samples t-test. Variables that were not normally distributed were expressed as median (first quartile [Q1], third quartile [Q3]) and analyzed with the Mann-Whitney U test. Categorical variables were presented as n (%) and assessed using either the chi-square test or Fisher’s exact test, depending on cell frequencies. Potential predictors of tubal rupture were examined by univariate and multivariable logistic regression. Statistical significance was defined as a two-sided P value < 0.05.

Because patients with missing key variables were excluded according to the exclusion criteria, no imputation was needed (complete-case analysis). Candidate predictors with P < 0.05 in univariate analysis were entered into the multivariable logistic regression model. The variance inflation factor (VIF) was calculated to evaluate multicollinearity, and values below 5 were interpreted as indicating no meaningful collinearity among the predictors.

Results

Baseline characteristics
This study included 150 patients with tubal pregnancy, including 51 (34.0%) in the ruptured group and 99 (66.0%) in the unruptured group. As shown in Table 1, there were no statistically significant differences between groups in age, BMI, number of pregnancies, parity, duration of amenorrhea, history of ectopic pregnancy, history of pelvic or abdominal surgery, use of assisted reproductive technology, or smoking (P > 0.05). However, the proportion of patients with a history of PID was significantly higher in the ruptured group than in the unruptured group (P = 0.036).

Clinical manifestations and vital signs
As shown in Table 2, moderate-to-severe abdominal pain (VAS score > 5) was more frequent in the ruptured group than in the unruptured group (58.8% vs. 40.4%, P = 0.032). The frequency of vaginal bleeding was comparable between the two groups (P = 0.919), whereas the shock index at admission was significantly elevated in patients with tubal rupture (P < 0.001).

Laboratory parameters
As shown in Table 3, the mean serum β-hCG concentration was significantly higher in the ruptured group than in the unruptured group (P < 0.001). Serum progesterone was slightly lower in the ruptured group, but the difference was not statistically significant (P = 0.065). Hemoglobin was significantly lower in the ruptured group (P = 0.029). The ruptured group also had a higher neutrophil count, a lower lymphocyte count, and a higher NLR than the unruptured group (all P < 0.05).

Ultrasound imaging features
Table 4 compares transvaginal and transabdominal ultrasound findings. The maximum diameter of the adnexal mass was larger in the ruptured group than in the unruptured group (P = 0.006). Pelvic free-fluid depth was also greater in the ruptured group (P = 0.002), and yolk sac detection and fetal cardiac activity were more frequent in the ruptured group (both P < 0.05).

Predictive factors for tubal rupture
Univariate logistic regression analysis
Tubal rupture was entered as the dependent variable (rupture = 1). Univariate logistic regression identified 11 variables associated with rupture: PID history, moderate-to-severe abdominal pain, SI > 0.8, β-hCG > 3,000 IU/L, WBC, LYM, NLR ≥ 4, maximum adnexal mass diameter, pelvic fluid depth, yolk sac detection, and fetal cardiac activity (Table 5; all P < 0.05).

Multivariable logistic regression analysis
All variables with P < 0.05 obtained from the single factor logistic regression analysis were included in the multiple factor logistic regression model, and Backward: LR was used to obtain the following results (Table 6): PID history (OR = 2.983, 95% CI: 1.306–6.811, P = 0.009), β-hCG > 3,000 IU/L (OR = 4.119, 95% CI: 1.661–10.214, P = 0.002), NLR ≥ 4 (OR = 3.597, 95% CI: 1.613–8.023, P = 0.002), and ultrasonographic detection of a yolk sac (OR = 2.506, 95% CI: 1.015–6.190, P = 0.046) were independent risk factors for tubal rupture (Table 6).

Data Availability Statement
The deidentified data supporting the findings of this study are provided in Supplementary File 1.

Figure 1. Study flowchart. A total of 187 patients with tubal pregnancy were identified in the electronic medical record system from January 2023 to December 2025. After 37 patients were excluded, 150 patients were enrolled and classified into the ruptured (n = 51) and unruptured (n = 99) groups. Please click here to view a larger version of this figure.

Variables	Ruptured group (n=51)	Non-ruptured group (n=99)	χ²/t/Z	P
Age (years)	29.4±3.2	28.7±4.8	1.126	0.262
BMI (kg/m²)	22.3±2.6	21.9±2.4	0.859	0.392
Duration of amenorrhea (days)	51.5±6.6	49.9±7.1	1.299	0.196
Pregnancy times (times)	2 (1, 3)	2 (1, 2)	0.801	0.423
Delivery times (times)	1 (0, 1)	0 (0, 1)	0.833	0.405
PID history	22 (43.1)	26 (26.3)	4.405	0.036
Ectopic pregnancy history	8 (15.7)	12 (12.1)	0.37	0.543
Pelvic and abdominal surgery history	10 (19.6)	18 (18.2)	0.045	0.832
IVF-ET conception	5 (9.8)	9 (9.1)	0.024	0.878
Smoking history	4 (7.8)	6 (6.1)	0.005	0.945

Table 1: Baseline characteristics [mean ± SD, n (%), or median (Q1, Q3)]. Continuous data are summarized as mean ± standard deviation or as median (first quartile [Q1], third quartile [Q3]), while categorical data are reported as n (%). Between-group comparisons were performed using the independent-samples t-test, Mann-Whitney U test, chi-square test, or Fisher’s exact test, as applicable. Abbreviations: SD = standard deviation; Q1 = first quartile; Q3 = third quartile; BMI = body mass index; PID = pelvic inflammatory disease; IVF-ET = in vitro fertilization and embryo transfer.

Variables	Ruptured group (n=51)	Non-ruptured group (n=99)	χ²/t	P
Moderate to severe abdominal pain (VAS>5)	30 (58.8)	40 (40.4)	4.588	0.032
Vaginal bleeding	38 (74.5)	73 (73.7)	0.01	0.919
SBP (mmHg)	105.1±7.8	108.6±6.8	2.766	0.006
Beats per minute (times)	86.2±7.7	82.7±7.4	2.626	0.01
SI	0.824±0.095	0.765±0.083	3.916	<0.001
SI>0.8	29 (56.9)	34 (34.3)	7.007	0.008

Table 2: Clinical symptoms and vital signs [mean ± SD or n (%)]. Data are presented as mean ± SD or n (%). Abbreviations: SD = standard deviation; VAS = visual analog scale; SBP = systolic blood pressure; SI = shock index.

Variables	Ruptured group (n=51)	Non-ruptured group (n=99)	χ²/t	P
β-hCG (IU/L)	3818.1±766.6	3167.5±999.5	4.07	<0.001
β-hCG>3000 IU/L	43 (84.3)	55 (55.6)	12.291	<0.001
Progesterone (ng/mL)	12.5±4.8	14.2±5.6	1.862	0.065
Hb (g/L)	105.4±15.6	110.5±12.2	2.22	0.029
WBC (×109/L)	9.8±2.5	9.0±1.8	2.232	0.027
NEU (×109/L)	7.4±1.9	6.8±1.4	2.428	0.016
LYM (×109/L)	1.8±0.6	2.2±0.5	3.152	0.002
NLR	4.7±2.6	3.4±1.2	4.333	<0.001
NLR≥4	26 (51.0)	22 (22.2)	12.793	<0.001

Table 3: Laboratory indicators [mean ± SD or n (%)]. Data are presented as mean ± SD or n (%). Abbreviations: SD = standard deviation; β-hCG = beta-human chorionic gonadotropin; Hb = hemoglobin; WBC = white blood cell count; NEU = absolute neutrophil count; LYM = absolute lymphocyte count; NLR = neutrophil-to-lymphocyte ratio.

Variables	Ruptured group (n=51)	Non-ruptured group (n=99)	χ²/t	P
Maximum diameter of adnexal mass (cm)	3.6±0.9	3.3±0.8	2.803	0.006
Depth of pelvic effusion (cm)	3.9±1.5	3.3±0.6	3.218	0.002
Yolk sac detected	16 (31.4)	16 (16.2)	4.641	0.031
Fetal heart pulsation detected	9 (17.7)	5 (5.1)	4.911	0.027

Table 4: Ultrasound imaging features [mean ± SD or n (%)]. Continuous data are reported as mean ± standard deviation, whereas categorical data are expressed as n (%). Abbreviation: SD = standard deviation.

						95% CI for OR
	B	SE	Wald	P	OR	Lower	Upper
PID history	0.756	0.363	4.328	0.037	2.130	1.045	4.343
Moderate to severe abdominal pain	0.745	0.351	4.520	0.033	2.107	1.060	4.189
SI>0.8	0.924	0.353	6.849	0.009	2.520	1.261	5.035
β-hCG>3000 IU/L	1.459	0.435	11.247	0.001	4.300	1.833	10.085
WBC	0.186	0.086	4.694	0.030	1.204	1.018	1.425
LYM	-0.956	0.321	8.855	0.003	0.384	0.205	0.722
NLR≥4	1.292	0.370	12.193	<0.001	3.640	1.763	7.517
Maximum diameter of adnexal mass	0.578	0.215	7.240	0.007	1.783	1.170	2.716
Depth of pelvic effusion	0.542	0.181	8.988	0.003	1.719	1.206	2.450
Yolk sac detected	0.863	0.407	4.502	0.034	2.371	1.068	5.265
Fetal heart pulsation detected	1.393	0.588	5.619	0.018	4.029	1.273	12.75

Table 5: Univariate logistic regression analysis. Each candidate predictor was entered separately into the model. P < 0.05 was considered statistically significant. Abbreviations: OR = odds ratio; CI = confidence interval.

						95% CI for OR
	B	SE	Wald	P	OR	Lower	Upper
PID history	1.093	0.421	6.727	0.009	2.983	1.306	6.811
β-hCG>3000 IU/L	1.416	0.463	9.338	0.002	4.119	1.661	10.214
NLR≥4	1.28	0.409	9.783	0.002	3.597	1.613	8.023
Yolk sac detected	0.919	0.461	3.967	0.046	2.506	1.015	6.19
Constant	-2.705	0.495	29.839	<0.001	0.067	-	-

Table 6: Multivariable logistic regression analysis. Nagelkerke R² = 0.273; χ² = 5.218; df = 6; P = 0.516. Abbreviations: OR = odds ratio; CI = confidence interval; df = degrees of freedom.

Supplementary File 1. Deidentified patient-level data used in the analyses. The spreadsheet contains demographic, clinical, laboratory, and ultrasound variables for 150 patients with tubal pregnancy. Group was coded as 1 = ruptured and 0 = unruptured; binary variables were coded as 1 = yes/present and 0 = no/absent. Abbreviations: BMI = body mass index; PID = pelvic inflammatory disease; IVF-ET = in vitro fertilization and embryo transfer; SBP = systolic blood pressure; SI = shock index; β-hCG = human chorionic gonadotropin; Hb = hemoglobin; WBC = white blood cell count; NEU = absolute neutrophil count; LYM = absolute lymphocyte count; NLR = neutrophil-to-lymphocyte ratio. Please click here to download this file.

Discussion

Multivariable analysis in this study showed that serum β-hCG > 3,000 IU/L is an independent risk factor for rupture of tubal pregnancy. β-hCG is secreted by embryonic syncytiotrophoblasts, and its serum concentration may reflect the proliferation activity and metabolic level of trophoblasts²⁴. In a normal intrauterine pregnancy, the endometrium undergoes sufficient decidualization reaction, forming physical and immune barriers to limit the excessive invasion of trophoblasts. However, the fallopian tube mucosa lacks a complete submucosal layer and decidual tissue, which cannot effectively resist the erosion of trophoblasts⁹. Research has shown that high levels of β-hCG are often accompanied by high expression of matrix metalloproteinases (MMPs, especially MMP-2 and MMP-9). MMPs can effectively degrade the extracellular matrix (ECM) of the fallopian tube wall, leading to severe damage to the tubular muscularis and smooth muscle fibers²⁵. When β-hCG > 3,000 IU/L, highly active trophoblast cells not only rapidly penetrate the weak muscle layer of the fallopian tubes but also invade the vascular network within the mesosalpinx. Local microbleeding caused by erosion of the vascular wall may further increase intraluminal pressure and contribute to tubal rupture^26,27. This suggests that clinicians should consider the risk of rupture in patients with ectopic pregnancy and β-hCG > 3,000 IU/L, even when the ultrasonographic mass is small, and should carefully assess suitability for expectant or medical management.

In this study, an NLR ≥ 4 was identified as an important hematological indicator of tubal rupture. NLR, as a readily available and inexpensive systemic inflammatory marker, has attracted considerable attention in the prognostic evaluation of malignant tumors, cardiovascular diseases, and acute abdomen in recent years²⁸. NLR remained in the final multivariable model, whereas white blood cell count (WBC) and shock index did not. NLR may reflect systemic immune and inflammatory imbalance more comprehensively than WBC alone. During the development of a tubal pregnancy, ectopic implantation and growth of the embryo can cause local ischemia, necrosis, and microvascular rupture in the fallopian tubes. These necrotic tissues are released into the bloodstream as damage-associated molecular patterns (DAMPs), triggering strong systemic acute phase reactions²⁹. Neutrophils release large amounts of elastase and reactive oxygen species (ROS), which not only exacerbate tissue damage to the fallopian tube wall, but also weaken the integrity of blood vessels³⁰. On the other hand, as pain, bleeding, and psychological stress intensify, the body's sympathetic nervous system is activated, and endogenous cortisol secretion increases, leading to accelerated lymphocyte apoptosis and a decrease in the absolute count of peripheral blood lymphocytes^31,32. Therefore, a significant increase in NLR ≥ 4 not only represents severe tissue necrosis and inflammatory cascade reactions in the local fallopian tubes but also reflects the physiological stress response of the body before or in the early stages of rupture. These findings support consideration of NLR as an adjunctive indicator in emergency risk assessment for patients with ectopic pregnancy.

Multivariable analysis showed that a history of PID was an independent risk factor for tubal rupture (OR = 2.983, 95% CI: 1.306–6.811). PID-induced tubal injury is an important pathological basis for ectopic pregnancy. Previous studies have shown that damage to the tubal mucosa, impaired ciliary function, and luminal adhesions not only increase the risk of ectopic implantation but may also weaken the tubal wall's resistance to trophoblast invasion. Sivalingam et al.³³ reported that a history of PID increased the risk of tubal pregnancy and was associated with tubal rupture. Therefore, patients with a history of PID may require closer clinical assessment.

Multivariable analysis in this study showed that ultrasound detection of the yolk sac was an independent risk factor for tubal rupture (OR = 2.506). The appearance of a yolk sac marks a specific stage of embryonic development, indicating ongoing embryonic development and trophoblastic proliferation³⁴. In the context of tubal pregnancy, active embryonic development is often accompanied by greater invasive activity of trophoblasts, thereby increasing the risk of tubal wall rupture. In recent years, the role of ultrasound imaging in the risk assessment of ectopic pregnancy has been increasingly recognized³⁵. Multiple studies published between 2020 and 2025 have explored the association between different ultrasound features and tubal rupture. A prospective study of emergency point-of-care ultrasound showed that the detection of free fluid in Morison's pouch (the hepatorenal recess) has a high positive likelihood ratio for ectopic pregnancy requiring surgical intervention (112, 95% CI: 15–831)³⁶. Although the study sample size was limited (n = 242), its results emphasized the crucial role of ultrasound in rapidly identifying high-risk patients at risk of rupture. In this study, the depth of pelvic free fluid was greater in the ruptured group (P = 0.002), which is consistent with the above findings. Yolk sac detection is closely related to fetal cardiac activity. In this study, the detection rate of fetal heartbeat in the ruptured group was 17.7%, significantly higher than the 5.1% in the unruptured group (P = 0.027). Fetal cardiac activity indicates ongoing embryonic development and increased trophoblastic activity, consistent with the pathophysiological mechanism underlying elevated β-hCG levels³⁷. In clinical decision-making, detection of a yolk sac or fetal cardiac activity may warrant closer assessment when selecting expectant, medical, or surgical management³⁸.

Several limitations should be considered. First, the analysis was retrospective and conducted at a single center, and the cohort comprised only 150 patients. Second, the study period was January 2023 to December 2025, and follow-up may have been limited for some patients. Third, ultrasound examinations were performed by different physicians; although uniform criteria were used, operator dependence may have affected consistency. Fourth, NLR may be influenced by infectious or immune conditions. Patients with acute infections or autoimmune diseases were excluded, but subclinical inflammation could not be fully controlled.

Future studies should validate these findings in prospective, multicenter cohorts to confirm the generalizability of the identified predictors. Alternative analytical approaches, such as machine learning algorithms, could be employed to integrate multidimensional predictors and develop a more accurate risk stratification tool without relying on predetermined cut-offs for continuous variables. Additionally, serial measurements of serum β-hCG, NLR, and ultrasound parameters over time may better capture the dynamic process leading to rupture and could improve predictive performance. Biomarker discovery studies incorporating inflammatory cytokines or extracellular matrix remodeling markers (e.g., MMP-2, MMP-9) might further elucidate the mechanisms underlying loss of tubal wall integrity and identify novel therapeutic targets to prevent rupture. In conclusion, this study identified PID history, serum β-hCG > 3,000 IU/L, NLR ≥ 4, and ultrasonographic detection of a yolk sac as independent risk factors for tubal rupture. These findings may support early identification of high-risk patients and treatment decisions.

Materials

List of materials used in this article

Name	Company	Catalog Number	Comments
Automated hematology analyzer	Sysmex Corporation, Kobe, Japan	XN-9000	Used for complete blood count measurements
Electrochemiluminescence immunoassay analyzer	Roche Diagnostics, Mannheim, Germany	Cobas e601	Used for serum β-hCG and progesterone measurements
K2-EDTA vacuum blood collection tube	Becton Dickinson, Franklin Lakes, NJ, USA	Vacutainer K2-EDTA	Used for complete blood count collection
Sample size calculation software	Heinrich Heine University, Düsseldorf, Germany	G*Power, version 3.1.9.7; RRID: SCR_013726	Used for sample size estimation
Serum separator tube	Becton Dickinson, Franklin Lakes, NJ, USA	Vacutainer SST	Used for serum β-hCG and progesterone collection
Statistical software	IBM Corp., Armonk, NY, USA	SPSS Statistics, version 25.0; RRID: SCR_002865	Used for statistical analyses
Ultrasound system	GE Healthcare, Milwaukee, WI, USA	Voluson E8	Equipped with RIC5-9-D transvaginal probe (5-9 MHz) and C1-5-D transabdominal convex probe (3.5-5 MHz)