Uterine artery pulsatility index: A predictor of methotrexate resistance in gestational trophoblastic neoplasia

Neo-angiogenesis is an early sign of cancers. This study assessed if among low-risk gestational trophoblastic neoplasia (LR-GTN) patients, the uterine artery pulsatility index (UAPI), which is a measure of tumor vascularity, can be an independent predictor to resistance to methotrexate chemotherapy (MTX-R).

UTERINE ARTERY PULSATILITY INDEX: A PREDICTOR

OF METHOTREXATE RESISTANCE IN GESTATIONAL

TROPHOBLASTIC NEOPLASIA

Nguyen Thai Giang¹, Vuong Tien Hoa¹,

Vu Ba Quyet 2 , Tran Trung Hieu¹

¹Hanoi Medical University

²National hospital of Obstetrics and gynecology (NHOG) Neo-angiogenesis is an early sign of cancers This study assessed if among low-risk gestational trophoblastic neoplasia (LR-GTN) patients, the uterine artery pulsatility index (UAPI), which is a measure of tumor vascularity, can be an independent predictor to resistance to methotrexate chemotherapy (MTX-R)

A prospective observational study of 204 LR-GTN patients (FIGO scores ranged from 0 to 6) were treated with methotrexate from January 2015 to September 2017 at National Hospital of Obstetrics and Gynecology (NHOG) Patients underwent a single Doppler ultrasound to assess the UAPI The median UAPI was lower (higher vascularity) in MTX-R compared with that of MTX-sensitive patients (1.60 vs 2.05, p= 0.03) UAPI, hCG and the largest tumor size were all predictors of MTX-R on univariate analysis Moreover, the UAPI remained

as a significant independent predictor of MTX-R on multivariate logistic regression analysis The odds ratio of MTX-R for patients with a UAPI ≤1.2 was 2.26 (95% CI, 1.00 – 5.09; p=0.049), compared to patients with a UAPI

>1.2 In conclusion, UAPI, as an indirect in vivo measure of functional tumor vascularity, could independently predict the response to MTX-R in LR-GTN patients Patients with FIGO score of 5-6 and a UAPI<1.2 might

be considered to be upstaged and offered combination chemotherapy as the initial regimen rather than MTX.

I INTRODUCTION

Keywords: low-risk GTN; UAPI; MTX resistance; angiogenesis.

Gestational trophoblastic neoplasia

(GTN) is a group of pathologies caused by

malignant neoplasia or high potential for

malignancy of trophoblasts The prevalence

of GTN is more common in Asia due to lower

socioeconomic status and poor nutrition in the

region compared to that of Europe and North

America The disease has a good prognosis

with the overall cure rate up to 98% if detected

early, but only 87% if detected and treated late

In their groundbreaking paper of 1956, Li et

al successfully treated GTN by Methotrexate (MTX) and opened a new era in the treatment

of GTN [1] Since then, GTN has been widely investigated until and in a major advance in

2002, International Federation of Gynecology and Obstetrics (FIGO) established a revised classification system as well as a risk factor scoring system for GTN Specifically, the scoring system classifies patients into high-risk and low-high-risk groups Low-high-risk patients are recommended to treat with single-agent chemotherapy, such as MTX However, several previous works have called into question that

Corresponding author: Nguyen Thai Giang,

Hanoi Medical University

Email: thaigianghmu@gmail.com

Received: 27/11/2018

Accepted: 12/03/2019

about 30% of these low-risk patients become

resistant to single-agent chemotherapy and

are required to change their regimen into

combination chemotherapy [2 - 4] Switching

chemotherapy regimen due to MTX resistance

will unnecessarily lengthen the overall

duration of chemotherapy, cause considerable

psychological distress to patients and slow

down the recovery of their normal procreation

ability Consequently, additional research is

needed to precisely identify the group of patient

that would develop MTX resistance in a timely

manner

Neo-angiogenesis, the formation of new

blood vessels, is a critical step in tumorogenesis

Neo-angiogenesis is associated with increased

tumor growth, acquisition of metastatic

potential and drug resistance The assessment

of angiogenesis is based on histological

pathology of increasing microvessel density

As opposed to a biopsy, which is frequently

contraindicated due to the risk of life-threatening

hemorrhage in GTN, Doppler ultrasonography

is considered an appropriate non-invasive

test to assess tumor vascularity and vascular

characteristics of the main artery supply in

GTN (uterine artery) Argawal, Long and Hsieh

proposed that the uterine artery pulsatility index

(UAPI) can predict resistance to methotrexate

chemotherapy (MTX-R) in GTN patients [5],[6]

Therefore, we carried out this research with the

aim to assess the value of UAPI and related

factors to prognostic methotrexate resistance

in low risk- gestational trophoblastic neoplasia

II METHODS

1 Study population

All the patients were diagnosed low-risk

GTN based on FIGO 2002 scoring system,

treated with single-agent methotrexate and

preserved the uterus in the Department of

Gynecologic Oncology from January 2015 to September 2017 at NHOG

Inclusion criteria: Diagnosed GTN post

molar according to FIGO 2002 criteria: a rise of hCG on three consecutive weekly measurements, a plateau of hCG lasting for four measurements taken throughout three weeks or longer, the hCG level remains elevated for six months or more (>5 IU/l) Diagnosed GTN if there is a histologic diagnosis of choriocarcinoma FIGO prognostic scores ranging from 0 to 6 GTN after other gestational events (term pregnancy, abortion, ectopic pregnancy)

2 Methods

This was a prospective observational study

Sample size:

In which: α = 0.05, Z = 1.96, choosing ε = 0.25 p: resistance rate to MTX treatment in GTN patients is 24.5 % [7] Add 10% patients The final sample size is 204 patients

Research process

The research clinical record was designed and tested GTN patients with FIGO prognostic scores from 0 to 6 were chosen In all cases, patients’ consent were obtained At the beginning of the study, all patients underwent a single abdominal pelvic ultrasound to measure total uterine volume and local invasion of GTN tumor UAPI was measured in both uterine arteries The lowest UAPI from either uterine artery was used for analysis, since it is a reflection of the maximal deviation from the normal impedance Doppler assessments of the uterine arteries were performed by only one experienced sonographer through ultrasound machine Voluson 730 Pro (GE Healthcare Technologies, USA) device equipped with

n z1 ( p) p.q .

2

2

2

transabdominal 3,5 MHz curvilinear probe MTX/FA regime: Patients were initially treated with fortnightly cycles of 50 mg of methotrexate i.m on days 1, 3, 5, and 7, with 5 mg of folinic acid i.m with rescue on days 2, 4, 6, and 8 Response to chemotherapy was monitored by measurements of hCG concentrations fortnightly The development of MTX-R was defined by a plateau or a rise in two consecutive hCG concentrations The MTX-R patients will be treated by combination chemotherapy

3 Statistical Analysis

Stata 13 was used for the statistical analyses Univariate analysis was performed using the Mann-Whitney U test or χ2 test, and the correlation between prognostic variables was tested using Spearman’s correlation coefficient A ROC curve was used to maximize sensitivity and specificity and find a cut-off point of UAPI These UAPI subgroups were used in the subsequent multivariate analyses of predictors of MTX-R, using binary logistic regression with forward stepwise selection All significant prognostic factors on univariate analysis were initially included

4 Ethical consideration

The study protocol was approved by the Institutional Review Board at NHOG on 25 January 2015

Table 1 FIGO scoring system for gestational trophoblastic neoplasia, 2002 [8]

Age (years) < 40 ≥ 40

Antecedent pregnancy mole abortion term

Interval (months) < 4 4 - 6 7 - 12 >12 βhCG (IU/1) < 103 103 - 104 104 - 105 > 105 Largest tumor diameter (cm) < 3 3 - 4 ≥ 5

Site of metastases lung Spleen, kidney Gastrointestinal tract Brain, liver Number of metastases 1 - 4 5 - 8 > 8 Previous chemotherapy Monotherapy Combined

therapy

Patients with FIGO scores ranging from 0 to 6 were deemed low-risk GTN patients, and were initially treated by single agent chemotherapy as the first regimen Patients who had FIGO scores equal to 7 and above were deemed high-risk GTN patients, and were initially treated by combination chemotherapy as the first regimen

III RESULTS

1 Patient characteristics

204 GTN patients were eligible for the study Their median age was 26 years (16-39 years), 84.3% had less than one child, 81.4% followed a molar pregnancy, medium latent interval was 2.1 months (1-14 months), 1.5 % had previous MTX chemotherapy, and 4.4% patients had either lung

or vaginal metastases The medium β-hCG concentration was 11667 UI/l (9 – 213180 UI/l) One third of patients had uterine local invasion with medium diameter 33.5 mm (13 - 69 mm) and 9.8% had FIGO score equal 4 or above

2 Resistance rate to single-agent chemotherapy MTX

27.45% (56/204) of the patients had MTX resistance, therefore they needed to switch regimen

to combination chemotherapy (EMACO)

3 Doppler ultrasound characteristics of MTX-R and response group

Mean peak systolic velocity (PSV) of MTX-R group was significantly higher than that of response group (72.86 ± 37.03 cm/s vs 60.73 ± 27.15cm/s, p = 0.04) Mean end diastolic velocity (EDV) of MTX-R group was also higher than PSV of response group (25.58 ± 26.40 vs 15.39 ± 19.86 cm/s,

p = 0.01)

4 UAPI cut-off point and meaning resistance to MTX prediction

Figure 1 Median of UAPI in MTX-R group lower than response group

(1.60 vs 2.05, p = 0.03, Mann –Whitney U test)

Figure 2 ROC curve of UAPI to predict MTX-R

The ROC curve and sequential analysis of the UAPI demonstrated that UAPI at the cut-off point 1.2 showed moderate predicted probabilities of MTX resistance (AUC=0.59, sensitivity 71.62%, and specificity 48.21%)

Table 2 UAPI and resistance to MTX-R at cut-off point 1,2

UAPI

Response MTX (n = 148)

MTX-R

0.007*

The group of patients with UAPI ≤ 1.2 had a higher MTX-R rate than that of the group with UAPI

>1.2 (p=0.007)

Table 3 Univariate logistic regression analysis factors related to MTX-R

βhCG pre-treatment

< 1000 IU/ml 1

1000 - < 10000 IU/ml 1.38 0.64 – 2.96 0.39

10000 - < 100000 IU/ml 2.42 1,11 – 5.26 0.025*

> 100000 IU/ml 3.83 0.50 – 29.11 0.19

Largest tumor size (cm)

< 3 cm 1

3 - 4 cm 1.02 0.44 – 2.36 0.96

≥ 5 cm 5.16 1.43 – 18.56 0.012*

UAPI

> 1.2 1

≤ 1.2 2.42 1.25 – 4.69 0.008*

Three significant factors related to MTX-R are βhCG pre-treatment, the largest tumor size and UAPI

5 Results of combination UAPI to FIGO score

Table 5 Resistance rate to MTX using FIGO score in combination with the cut-off point

UAPI = 1.2

FIGO

score

Resistance rate to MTX FIGO FIGO + PI > 1.2 FIGO + PI ≤ 1.2

Table 4 Multivariate logistic regression analysis of MTX resistance and dependent factors

(UAPI, βhCG and tumor size)

UAPI

> 1.2 1 ≤ 1.2 2.26 1.00 – 5.09 0.049*

βhCG pre-treatment

< 1000 IU/ml 1

1000 - < 10000 IU/ml 1.21 0.55 – 2.64 0.62

10000 - < 100000 IU/ml 1.67 0.67 – 4.12 0.26

≥ 100000 IU/ml 2.18 0.23 – 20.64 0.49

Largest tumor size (cm)

< 3 cm 1

3 – 4 cm 0.55 0.20 – 1.48 0.24

≥ 5 cm 2.31 0.53 – 10.07 0.26

As shown in table 4, UAPI was the only independent predictor of MTX-R (p < 0.05)

IV DISCUSSION

Since tumor angiogenesis is a significant

step in the growth and metastasis of solid

tumors [9], evaluating tumor vascularity could

predict metastatic risk and overall survival rates in several types of cancer, such as breast cancer, melanoma, response to radiation in cervical cancer [10] While previous literature pointed out that the tumor vascularity was

primarily assessed by histological examination,

Doppler ultrasound plays a vital role as a

useful tool to assess tumor vascularity and

clinical response due to its ability to identify

neo-angiogenesis characteristics of tumor

and vascular changes GTN, a solid tumor, is

especially fitted to study by Doppler ultrasound

as it allows researchers and clinicians the

ability to evaluate tumor vascularity and

hemodynamic changes in the main artery

supplying the uterus (uterine arteries) In GTN

patients, studies have highlighted that due to

low resistance downstream, hemodynamic

change in uterine arteries showed high velocity

and low impedance [11]

PSV of MTX-R group (72.86 ± 37.03 cm/s)

was significantly higher than that of response

group (60.73 ± 27.15 cm/s, p < 0.01) Since

PSV of uterine artery exhibited high increase

blood flow into the uterine circulation to feed

neoplastic trophoblast, MTX-R patients

exhibited a hyperdynamic uterine circulation

by revealing significantly higher PSV in uterine

artery, compared with response patients We

also found a much higher value of EDV among

MTX-R group (25.58 ± 26.40 cm/s) , compared

to that among MTX response group (15.39

± 19.86 cm/s, p = 0.01) Our results were

similar to Hsieh et al , who also implied that

the increase of EDV implies low impedance

downstream resulting of neo-angiogenesis and

arterio-venous shunts within the trophoblastic

lesions [12] Therefore, MTX-R group exhibited

a sharp increment of EDV To the best of our

knowledge, there are two possible mechanisms

which would explain the relationship between a

low impedance to flow in the uterine arteries and

the development of drug resistance Firstly, the

low impedance reflects the presence of

arterio-venous shunts within the uterus As a result,

it could cause a decrease in tumor perfusion,

which has been shown to be avascular within itself Thus, the penetration and the effectiveness of cytotoxic cancer drugs would

be reduced and can lead to the development

of clinical drug resistance Second, drug resistance could develop alternatively due

to intrinsic characteristics of the tumor itself Specifically, internal vascular changes, which might be varied among different trophoblastic tumors, could also be related to the presumed biochemical basis of drug resistance For example, choriocarcinoma, which is derived from villus cytotrophoblast, is very responsive

to cytotoxic treatment but a corresponding tumor derived from interstitial cytotrophoblast has a higher possibility of being resistant to chemotherapy In a normal pregnancy, the interstitial cytotrophoblast is presumed to be responsible for the spiral artery dilatation and the reduction in the impedance to flow in the uterine arteries Thus, the low impedance to blood flow in the MTX resistance patient group could be one of the consequences of those tumors containing interstitial cytotrophoblastic originated cells, which would intrinsically reduce the sensitivity of the tumor to therapy [6]

Of the study population, UAPI in MTX-R group is significantly lower than that of the MTX response group It might be explained

by one of the consequences of hemodynamic changes in PSV and EDV of uterine arteries (Figure 1) The ROC curve of UAPI (Figure 2) and sequential analysis of the UAPI illustrated that UAPI at the cut-off point 1.2 show moderate predicted probabilities of MTX resistance (AUC = 0.59, sensitivity 71.62% and specificity 48.21%) At this cut-off point, UAPI was found to be independently and inversely related to the risk of MTX-R on both univariate and multivariate analysis Table 2 showed that

when assessing patients with UAPI alone, the

group of patient with UAPI≤ 1.2 has significant

higher MTX-R rate thanthe group with UAPI

>1.2 (p = 0.007)

On univariate analysis, three significant

factors related to MTX-R are β-hCG

pre-treatment, the largest tumor size and UAPI

The odds ratio for MTX-R was 2.42 for patients

with a β-hCG pre-treatment > 10000 UI/l,

which was relative to patients with β-hCG

pre-treatment < 1000 UI/l Patients with tumor size

≥ 5 cm raised the odds ratio for MTX-R 5.16

times, compared to the group of patient with

the tumor size < 3cm Interestingly, changes

in hemodynamic characters of the uterine

artery was a new prognostic factor of MTX-R

Those with UAPI ≤ 1.2 had an increment of the

odds ratio for MTX-R to 2.42 times, compared

to patients with UAPI > 1.2 (table 3) On

multivariate analysis, the data reported that

the UAPI was the only significant independent

predictor of MTX-R (p < 0.05) The odds ratio

for MTX-R was 2.26 for patients with UAPI ≤

1.2 (95% CI: 1.00 – 5.09), compared to those

with UAPI > 1.2 (table 4)

Our findings have further strengthened our

confidence in a combination of FIGO score

and UAPI To be more specific, UAPI might be

integrated into the FIGO score as a biological

marker of tumor vascularity and angiogenesis

Although our sample was not optimal, we

nevertheless believe that the combination

would allow better early detection of MTX-R

risk of GTN low-risk patients as shown in table

5 The criteria of UAPI ≤ 1.2 corresponded 1-2

point of FIGO score when assessing MTX-R

risk As anticipated, when combined with UAPI

≤ 1.2, we discovered 66.67% of patients with

FIGO score ≥ 4 and 100% of patients with FIGO

score ≥5 had MTX resistance The same FIGO

scores combined with UAPI > 1.2 resulted in a

lower MTX-R rate than that of the group using FIGO score alone Consequently, using the cut-off point UAPI achieved a better result in prognostic in upper and below the threshold The Doppler ultrasound may serve

as a useful method for in vivo functional assessment of tumor vasculature by its assessment to hemodynamic changes in the macrovasculature, due to the fact that those changes are an indirect reflection of the microvasculature (vessel diameter < 15 micrometers) used to define neoangiogenesis Our results seemed to confirm that UAPI is

a non-invasive method to assess in vivo the tumor vasculature and may predict single agent MTX-R in low-risk GTN patients

FIGO scoring system helps to stratify the initial treatment of GTN patient by combined chemotherapy instead of single one Currently,

up to one-third of low-risk GTN patient incorrectly received single-agent methotrexate and could develop drug resistance; it is necessary to improve the FIGO scoring system From some previous studies, Kohorn, Osborne and Taylor found that the elimination

of the moderate group was a mistake Many studies also corroborate the finding that the group of patients with FIGO scores ranging from 5 to 6 usually do not respond well to single-agent chemotherapy (resistance rate

up to 81%) and this group should have the first regimen as the combination chemotherapy, rather than delay until MTX–R has developed [13; 14]

At present, while the stratification of treatment for GTN patients is based on FIGO scoring (2002), UAPI can be considered as

a valuable factor that helps to predict the patient who may develop MTX-R and need combination chemotherapy right from the beginning

V CONCLUSION

This paper has stressed the importance of

hemodynamic characteristics of the uterine

artery as a novel factor contributing to predict

methotrexate resistance To be specific, UAPI

could be an independent prognostic factor of

single-agent methotrexate resistance

Taken together, FIGO scoring system

combined with a UAPI threshold of equal to 1.2

or lower would allow better early detection of

methotrexate resistance risk among low-risk

GTN patients

ACKNOWLEDGEMENTS

The authors would like to acknowledge

patient participation We appreciate the help

of participants who collaborated with us in this

study

REFERENCES

1 M C Li, R Hertz và D B Spencer (1956)

Effect of methotrexate therapy upon choriocarcinoma

and chorioadenoma Experimental Biology and

Medicine 93 (2), 361-366.

2 C Aghajanian (2011) Treatment of low-risk

gestational trophoblastic neoplasia Journal of Clinical

Oncology 29 (7), 786-788.

3 I McNeish, S Strickland, L Holden et al

(2002) Low-risk persistent gestational trophoblastic

disease: outcome after initial treatment with low-dose

methotrexate and folinic acid from 1992 to 2000

Journal of Clinical Oncology, 20 (7), 1838-1844.

4 M J Seckl, N J Sebire và R S Berkowitz

(2010) Gestational trophoblastic disease The Lancet,

376 (9742), 717-729.

5 R Agarwal, S Strickland, I A McNeish et

al (2002) Doppler ultrasonography of the uterine

artery and the response to chemotherapy in patients

with gestational trophoblastic tumors Clinical cancer

research, 8 (5), 1142-1147.

6 M G Long, J Boultbee, R Langley et al (1992) Doppler assessment of the uterine circulation

and the clinical behaviour of gestational trophoblastic

tumors requiring chemotherapy British journal of cancer, 66 (5), 883.

7 Phan Chí Thành (2012) Nghiên cứu tỷ lệ

kháng thuốc và yếu tố liên quan trong điều trị u nguyên

bào nuôi bằng MTX và acid folic tại BVPSTU Tạp chí Phụ sản, 4 (6), 49-54.

8 F O Committee (2002) FIGO staging for

gestational trophoblastic neoplasia 2000 International Journal of Gynecology and Obstetrics, 77 (3),

285-287

9 N Weidner, J P Semple, W R Welch et

al (1991) Tumor angiogenesis and metastasis—

correlation in invasive breast carcinoma New England Journal of Medicine, 324 (1), 1-8.

10 A Srivastava, P Laidler, R Davies et

al (1988) The prognostic significance of tumor

vascularity in intermediate-thickness (0.76-4.0 mm thick) skin melanoma A quantitative histologic study

The American journal of pathology, 133 (2), 419.

11 M Long, J Boultbee, M Hanson et al (1989) Doppler time velocity waveform studies of

the uterine artery and uterus BJOG: An International Journal of Obstetrics & Gynaecology, 96 (5), 588-593.

12 F.-J Hsieh, C.-C Wu, C.-A Chen et al (1994) Correlation of uterine hemodynamics with

chemotherapy response in gestational trophoblastic

tumors Obstetrics & Gynecology, 83 (6), 1021.

13 R J Osborne, V Filiaci, J C Schink et al (2011) Phase III trial of weekly methotrexate or pulsed

dactinomycin for low-risk gestational trophoblastic neoplasia: a gynecologic oncology group study

Journal of Clinical Oncology, 29 (7), 825-831.

14 F Taylor, T Grew, J Everard et al (2013)

The outcome of patients with low risk gestational trophoblastic neoplasia treated with single agent intramuscular methotrexate and oral folinic acid

European Journal of Cancer, 49 (15), 3184-3190.