Adjuvant concurrent chemoradiotherapy for early-stage cervical cancer patients with intermediate-risk factors in the era of intensity-modulated radiation therapy

Abstract

Objective

To investigate whether adjuvant concurrent chemotherapy with pelvic intensity-modulated radiotherapy (IMRT) improves the outcome of patients with early-stage cervical cancer with intermediate risk factors (IRFs) compared with IMRT alone.

Methods

This study included a cohort of 244 women from January 2012 to October 2023 who underwent radical hysterectomy along with pelvic lymph node dissection (PLND) for stage IB–IIA cervical cancer with IRFs. One hundred thirty-nine patients were postoperatively treated with pelvic IMRT (45.0–50.4 Gy) and adjuvant concurrent chemoradiotherapy (CCRT) (weekly cisplatin or weekly or triweekly combination of paclitaxel and cisplatin). One hundred five patients were treated with pelvic IMRT alone. The significance of clinical parameters, overall survival (OS), and recurrence-free survival (RFS) of each arm were analyzed.

Results

The median follow-up period was 61.2 months, with a range spanning from 6.3 to 144.4 months. The 5-year OS rates in the patients in the radiotherapy (RT) and CCRT arms were 96.7% and 94.4%, respectively (p=0.271). Five-year RFS was 88.1% in the RT arm and 86.9% in the CCRT arm (p=0.740). Multivariate analysis identified adenocarcinoma (AC) histology as the only significant prognostic factor affecting OS and RFS.

Conclusion

This study demonstrates that even in the era of IMRT, CCRT is not more effective than RT alone in patients with stage IB–IIA cervical cancer with IRFs following radical hysterectomy with PLND.

Synopsis

Postoperative concurrent chemoradiotherapy (CCRT) versus intensity-modulated radiotherapy (IMRT) for early cervical cancer with intermediate risk factors. No superior overall/recurrence-free survival with CCRT versus IMRT alone. Adenocarcinoma is the sole poor prognostic factor. IMRT’s low toxicity supports clinical application.

INTRODUCTION

Cervical cancer ranks as the fourth most prevalent cancer among women globally and the eighth most frequently diagnosed cancer across all genders. In China, radical hysterectomy with pelvic lymph node dissection (PLND) is the commonly utilized primary treatment for the majority of patients diagnosed with stage IB–IIA cervical cancer. This encompasses IB3 and IIA2 sub-stages, graded according to the International Federation of Gynaecology and Obstetrics (FIGO) staging system.

The management of intermediate risk factors (IRFs) after cervical cancer surgery has always been controversial. Currently, our center’s treatment for this group of patients follows the Sedlis standard in the United States, derived from the Gynecologic Oncology Group protocol 92 (GOG 92) study [1] and its long-term follow-up results [2].

Lai et al.’s study [3] stated that upon adjusting for confounding variables, histologic subtype—specifically adenocarcinomatous and adenosquamous carcinoma (ASC) versus squamous cell carcinoma—was identified as an independent prognostic indicator for decreased recurrence-free survival (RFS) (relative risk [RR]=1.28; p=0.009) and overall survival (OS) (RR=1.26; p=0.015) in patients with stage IB and II cervical cancer who underwent primary radical surgery. A previous study established a four-factor risk prediction model for adenocarcinoma (AC), a significantly useful reference in clinical practice [4]. Radiotherapy (RT) combined with cisplatin-based chemotherapy can improve patients’ OS and progression-free survival and has become the first-line treatment for locally advanced cervical cancer [5,6,7] and patients with high-risk post-surgery [8].

The GOG 92 study did not demonstrate the effect of chemotherapy. Previous retrospective studies have attempted to answer this question. Some studies yielded positive results [9,10,11]. Other studies showed that a combination of chemotherapy and RT did not offer any benefits compared to RT alone but rather increased toxicity [12,13,14,15]. Most of these previous studies used traditional 4-field box irradiation or 3-dimensional conformal radiotherapy (3D-CRT), which was largely obsolete, even in rural hospitals in China. The present study focused on the results of utilizing a combination of modern intensity-modulated radiotherapy (IMRT) techniques with chemotherapy. This retrospective cohort study aimed to assess the impact of incorporating chemotherapy into postoperative IMRT on survival and toxicity among patients with early-stage cervical cancer in a real-world setting, considering IRF for recurrence after precise pathological evaluation.

MATERIALS AND METHODS

A retrospective review was conducted using data from January 2012 to October 2023, which included 602 patients who underwent surgery after being diagnosed with early-stage cervical cancer (stage IB–IIA) preoperatively. Among them, 526 patients underwent postoperative RT or concurrent chemoradiotherapy (CCRT) at Fudan University Zhongshan Hospital. The eligibility criteria included patients (n=482) diagnosed histologically with squamous cell carcinoma (SCC), AC, or ASC of the uterine cervix, those accurately staged as per the FIGO 2018 guidelines with stage IB1–IIA2 disease confirmed by postoperative pathology, and those who underwent radical hysterectomy with PLND followed by pelvic RT or CCRT at a minimum dose ≥40 Gy. From this cohort, patients (n=199) exhibiting high-risk factors, such as pelvic and/or para-aortic lymph node metastasis, parametrial invasion, and/or positive resection margins (encompassing high-grade intraepithelial neoplasia of the vaginal margins and tumors located within 5 mm of the margins), were excluded. Women who received neoadjuvant and unknown adjuvant therapies and those for whom complete information was not available were also excluded. Finally, 244 patients with IRFs were eligible for the study.

1. Definition of IRF

The criteria for IRFs were defined using the below methodology. Patients with SCC were identified according to the standards set by GOG [1]. A three-factor model for categorizing patients with intermediate risk was established using the following criteria [1]. Patients were assigned to category 1 if they exhibited positive capillary lymphatic space involvement with deep one-third stromal invasion, regardless of tumor size. Category 2 included those patients with positive capillary lymphatic space involvement and middle one-third stromal invasion, provided the tumor size was at least 2 cm. Category 3 included patients who had positive capillary lymphatic space involvement with superficial one-third stromal invasion and tumors measuring 5 cm or larger. Category 4 encompassed patients with negative capillary lymphatic space involvement who had either deep or middle one-third stromal invasion and tumors measuring 4 cm or larger.

Patients diagnosed with AC or ASC were identified based on the four-factor model. The criteria for identification required the presence of at least 2 of the following factors: tumor size greater than 3 cm, deep stromal invasion involving the outer third of the cervix, lymphovascular space invasion (LVSI), and histological confirmation of AC or ASC [4].

2. Surgery

All patients underwent open radical hysterectomy and PLND. The types of radical hysterectomy were classified using the “Querleu–Morrow” classification system [16]. Type B2 corresponds to a modified radical hysterectomy combined with PLND, type C1 to transection of the paracervix at its junction with the internal iliac vascular system with nerve-sparing, and type C2 to transection of the paracervix at its junction with the internal iliac vascular system with no preservation of autonomic nerves. None of the enrolled patients underwent type A or D surgery.

3. RT

Adjuvant RT generally employed IMRT (6-MV photon beam). The delineation of the clinical target volume (CTV) adhered to the target delineation guidelines specified by the Radiation Therapy Oncology Group [17]. Target area delineation was conducted following the latest version of the Consensus Guidelines after February 2021 [18]. The planning target volume (PTV) was established by applying a margin expansion of 0.5–1.0 cm to the CTV. The prescribed radiation dose regimen comprised 45.0–50.4 Gy/1.8 Gy or 46 Gy/2.0 Gy, administered five times weekly. The acceptable range for the PTV dose was between 95% to 110% of the prescribed dose, with a median of 45 Gy. Each patient underwent pelvic cavity radiation field treatment without para-aortic extension fields. Additionally, none of the patients received vaginal brachytherapy or a boost at the vaginal stump.

Treatment was initiated 3 to 4 weeks post-surgery and was completed within 6 to 8 weeks. The organ-at-risk constraints were the same as our team’s prior study [19]. The maximum dose to the spinal cord was restricted to less than 45 Gy. For the small intestine, the volume receiving 40 Gy (V₄₀) was maintained below 30%, and the volume receiving 30 Gy (V₃₀) was below 40%. The average dose to bilateral femoral heads was kept below 30 Gy. The bladder volume receiving 45 Gy (V₄₅) was limited to <40%. The rectal V₄₀ was maintained below 40%, and the volume receiving 20 Gy (V₂₀) of both left and right kidneys was constrained to <33%.

4. Concurrent chemotherapy regimen

The choice between CCRT or RT was made through collaborative discussions involving the physician, patient, and their families, considering factors such as patient age, number and degree of IRFs, performance status, and comorbidities. The study included several options for CCRT: weekly cisplatin at 25–40 mg/m² or weekly paclitaxel at 75 mg/m² combined with cisplatin at 25 mg/m². The patients treated with weekly chemotherapy for 5 scheduled consecutive weeks were concurrently treated with IMRT. The patients treated with triweekly combination chemotherapy comprised 135 mg/m² of intravenous paclitaxel followed by 60 mg/m² of cisplatin every 3 weeks for 4 cycles, with the first and second cycles given concurrently to RT. Chemotherapy was temporarily discontinued if the patient’s leukocyte, neutrophil, or platelet count or creatinine clearance decreased to <2,000/mm³, <1,000/mm³, <75,000/mm³, or <50 mL/min, respectively, on the day of administration and was restarted as soon as the patient recovered. Administration of short-acting granulocyte-monocyte colony-stimulating factor and erythropoietin was supervised by a qualified physician.

5. Follow-up and outcomes

The follow-up time and content of the patients were consistent with those detailed in our team’s prior study [19]. Disease recurrence was characterized by the histological identification of tumor cells via biopsy or fine-needle aspiration cytology, the presence of any lymph node measuring over 1 cm along the short axis diameter on a computed tomography (CT) scan, or a tumor size increase exceeding 30% on the follow-up CT scan. Pelvic recurrence denoted loco-regional recurrence, including vaginal recurrence, as well as recurrences at any site within the radiation field. Distant metastases were characterized by lymph node involvement beyond the radiation field area, as well as metastases to the lungs, bones, or other organs. Survival times were calculated from the date of surgery until the date of the event or, in patients who did not experience any adverse events, until death or last follow-up.

6. Toxicity

Side effects were documented by the radiation oncologist and gynecologist in the electronic patient file. Toxicity grading followed the guidelines outlined in the Common Terminology Criteria for Adverse Events version 3.0, with the highest observed grade of toxicity reported for each patient. Gastrointestinal (GI), genitourinary (GU), and lymphedema toxicities were assessed during follow-up appointments. Hematologic (HT) toxicities were evaluated based on complete blood counts obtained during treatment and at each follow-up visit. Treatment toxicity was classified as acute if it occurred within the first 3 months post-RT and as late if it occurred or persisted beyond 3 months from the start of RT. In case of recurrence, toxicity data from 3 months before recurrence was excluded.

7. Statistical analysis

Statistical analysis was performed using the Statistical Package for the Social Sciences (SPSS) software (release 12.0.1; SPSS Inc., Chicago, IL, USA). Continuous variables such as age was tested using independent samples t-tests. χ² or Fisher’s exact tests compared discrete variables such as LVSI and pathological type among the arms. To account for differences in baseline characteristics and risk between RT and CCRT patients, propensity score (PS) matching was performed. Using a caliper value of 20% of the standard deviation of the logit of the PS [20], we matched study patients undergoing CCRT with replacement 1:1 with RT patients. The standardized group mean difference was used to compare continuous and dichotomous variables in matched and unmatched data. An absolute value of 10% is considered a meaningful imbalance between groups. The propensity score matching was implemented using R package MatchIt [21]. OS was defined as the time from surgery to death from any cause or the most recent documented follow-up. RFS was defined as the time from surgery to the first documented recurrence or the most recent follow-up. Survival curves were generated using the Kaplan–Meier method, with comparisons between curves performed using the log-rank test. Independent prognostic factors were identified using Cox proportional hazards regression analysis. A p-value ≤0.050 was considered statistically significant for 2-tailed tests.

8. Ethics statement

The study was ethically reviewed and approved by the Ethics Committee of Zhongshan Hospital Affiliated to Fudan University (B2021-814R). Informed consent was given by accepting the invitation to participate and follow-up for each patient.

RESULTS

1. Patient and tumor characteristics

In total, 244 cases were eligible in the present study, with 139 assigned to the RT arm and 105 to the CCRT arm. The median follow-up duration was 61.2 months (interquartile range [IQR]: 6.3–144.4 months). The median age (to the date of surgery) of the population was 53.4 months (IQR: 24.5–76.4 months). After propensity score matching, both the RT group and the CCRT group included 77 patients. Patient characteristics are shown in Table 1. The number of people who completed different chemotherapy regiments in the CCRT group is shown in Table S1. Propensity score matching achieved balanced baseline characteristics between the groups.

Table 1. Clinicopathologic characteristics according to treatment type.

Characteristics		Unmatched			Propensity-score matched
		RT (n=139)	CCRT (n=105)	SMD	RT (n=77)	CCRT (n=77)	SMD
Age (yr)		54.41±10.28	50.89±10.76	0.334	53.28±10.98	53.16±9.91	0.011
Performance status				0.185			0.097
	ECOG 0	124 (89.2)	99 (94.3)		70 (90.9)	72 (93.5)
	ECOG 1	15 (10.8)	6 (5.7)		7 (9.1)	5 (6.5)
FIGO stage				0.268			0.175
	IB1	18 (12.9)	7 (6.7)		4 (5.2)	5 (6.5)
	IB2	45 (32.4)	31 (29.5)		26 (33.8)	20 (26.0)
	IB3	24 (17.3)	20 (19.0)		14 (18.2)	16 (20.8)
	IIA1	29 (20.9)	22 (21.0)		18 (23.4)	20 (26.0)
	IIA2	23 (16.5)	25 (23.8)		15 (19.5)	16 (20.8)
Histology				0.164			0.047
	SCC	118 (84.9)	83 (79.0)		62 (80.5)	61 (79.2)
	AC	12 (8.6)	11 (10.5)		9 (11.7)	9 (11.7)
	ASC	9 (6.5)	11 (10.5)		6 (7.8)	7 (9.1)
Tumor size (mm)				0.314			0.197
	≤20	25 (18.0)	10 (9.5)		8 (10.4)	8 (10.4)
	21–40	75 (54.0)	53 (50.5)		45 (58.4)	38 (49.4)
	>40	39 (28.1)	42 (40.0)		24 (31.2)	31 (40.3)
LVSI				0.618			<0.001
	Negative	60 (43.2)	17 (16.2)		15 (19.5)	15 (19.5)
	Positive	79 (56.8)	88 (83.8)		62 (80.5)	62 (80.5)
Depth of invasion				0.314			0.060
	Superficial 1/3	16 (11.5)	5 (4.8)		5 (6.5)	4 (5.2)
	Middle 1/3	39 (28.1)	23 (21.9)		19 (24.7)	20 (26.0)
	Deep 1/3	84 (60.4)	77 (73.3)		53 (68.8)	53 (68.8)
Type of operation (Querleu)				0.293			0.062
	B2	10 (7.2)	3 (2.9)		3 (3.9)	3 (3.9)
	C1	117 (84.2)	98 (93.3)		70 (90.9)	71 (92.2)
	C2	12 (8.6)	4 (3.8)		4 (5.2)	3 (3.9)
Technique				0.190			<0.001
	IMRT	106 (76.3)	88 (83.8)		66 (85.7)	66 (85.7)
	Tomo-IMRT	33 (23.7)	17 (16.2)		11 (14.3)	11 (14.3)
Schedules of pelvic EBRT				0.177			0.111
	45 Gy (25×1.8 Gy)	117 (84.2)	83 (79.0)		65 (84.4)	62 (80.5)
	46 Gy (23×2.0 Gy)	11 (7.9)	8 (7.6)		5 (6.5)	7 (9.1)
	50.4 Gy (28×1.8 Gy)	11 (7.9)	14 (13.3)		7 (9.1)	8 (10.4)

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

AC, adenocarcinoma; ASC, adenosquamous carcinoma; CCRT, concurrent chemoradiotherapy; EBRT, external beam radiotherapy; ECOG, Eastern Cooperative Oncology Group; FIGO, International Federation of Gynecology and Obstetrics; IMRT, intensity modulated radiotherapy; LVSI, lymphovascular space invasion; RT, radiotherapy; SCC, squamous cell carcinoma; SMD, standardized mean difference.

2. Patterns of recurrence and survival outcomes according to adjuvant treatment

Each arm recorded 5 cases of pelvic recurrence involving vaginal stumps, pelvic lymph nodes, and pelvic masses located near the bladder and rectum. There were 5 cases of distant metastasis in the RT arm and seven in the CCRT arm. The most common site of metastasis was the lung, followed by lymph nodes adjacent to the abdominal aorta. Metastases were also observed in the pleura, psoas major, inguinal lymph nodes, and supraclavicular lymph nodes. Additionally, 2 cases with both pelvic recurrence and distant metastasis were noted in the CCRT arm. There was no statistically significant difference in recurrence patterns between the 2 arms (p=0.248; Table 2). The 5-year OS rates were 96.7% and 94.4% in the RT and CCRT arms, respectively (p=0.271; Fig. 1A). The 5-year RFS rates were 88.1% and 86.9% in the RT and CCRT arms, respectively (p=0.740; Fig. 1B). In univariate analysis, OS was significantly associated with histology (p=0.041), as was RFS (p=0.012). In multivariate analysis, AC histology emerged as the only significant prognostic factor affecting both OS and RFS (Table 3).

Table 2. Recurrence patterns according to adjuvant treatment.

Recurrence patterns		Total (n=244)	RT (n=139)	CCRT (n=105)	p-value
No relapse		220 (90.2)	129 (92.8)	91 (86.7)	0.248
Relapse		24 (9.8)	10 (7.2)	14 (13.3)
	Reginal	10 (4.1)	5 (3.6)	5 (4.8)
	Distant	12 (4.9)	5 (3.6)	7 (6.7)
	Reginal and distant	2 (0.8)	0 (0)	2 (1.9)

Values are presented as number (%).

CCRT, concurrent chemoradiotherapy; RT, radiotherapy.

Fig. 1. Kaplan-Meier OS (A) and RFS (B) curves of patients according to adjuvant treatment.

CCRT, concurrent chemoradiotherapy; HR, hazard ratio; OS, overall survival; RFS, recurrence-free survival; RT, radiotherapy.

Table 3. Univariable and multivariable Cox proportional hazard regression analyses of OS and RFS.

Variables		OS				RFS
		Univariable		Multivariable		Univariable		Multivariable
		HR (95% CI)	p-value	HR (95% CI)	p-value	HR (95% CI)	p-value	HR (95% CI)	p-value
Age (yr)			0.362				0.694
	≤50	Reference				Reference
	>50	0.611 (0.212–1.763)				1.179 (0.515–2.695)
Performance status			0.831				0.629
	ECOG 0–1	Reference				Reference
	ECOG 2	0.801 (0.027–5.100)	0.801			0.699 (0.164–2.989)	0.629
FIGO stage			0.724				0.374
	IB1	Reference				Reference
	IB2	0.871 (0.159–4.765)	0.874			0.673 (0.179–2.537)	0.559
	IB3	0.348 (0.064–1.900)	0.223			0.257 (0.086–0.969)	0.045
	IIA1	1.120 (0.280–4.477)	0.873			0.550 (0.165–1.825	0.328
	IIA2	0.977 (0.243–3.919)	0.974			0.735 (0.255–2.122)	0.570
Histology			0.041				0.012
	SCC	Reference				Reference
	AC	0.753 (0.096–5.851)	0.787	8.455 (1.911–37.403)	0.005	1.379 (0.183–10.396)	0.755	4.338 (1.389–13.550)	0.012
	ASC	2.997 (0.350–25.682)	0.317	0.972 (0.100–9.488)	0.981	5.132 (0.631–41.738)	0.126	0.513 (0.060–4.411)	0.543
Tumor size (mm)			0.255				0.127
	≤20	Reference				Reference
	21–40	0.985 (0.296–3.281)	0.981			0.643 (0.210–1.974)	0.440
	>40	0.383 (0.115–1.272)	0.117			0.388 (0.155–0.972)	0.043
LVSI			0.419				0.355
	Negative	Reference				Reference
	Positive	0.644 (0.221–1.874)	0.419			0.658 (0.271–1.958)
Depth of invasion			0.324				0.163
	Superficial 1/3	Reference				Reference
	Middle 1/3	0.459 (0.060–3.517)	0.454			0.328 (0.044–2.445)	0.277
	Deep 1/3	0.353 (0.080–1.567)	0.171			0.361 (0.107–1.215)	0.100
Type of operation (Querleu)			0.435				0.644
	B2	Reference				Reference
	C1	2.537 (0.565–11.387)	0.224			1.544 (0.361–6.597)	0.558
	C2	1.588 (0.354–7.117)	0.546			0.492 (0.066–3.659)	0.488
Technique			0.503				0.428
	IMRT	Reference				Reference
	Tomo-IMRT	1.662 (0.376–7.346)	0.503			1.632 (0.486–5.483)
Adjuvant treatment			0.777				0.107
	RT alone	Reference				Reference
	CCRT	0.865 (0.425–3.145)	0.777			0.509 (0.225–1.156)

AC, adenocarcinoma; ASC, adenosquamous carcinoma; CCRT, concurrent chemoradiotherapy; CI, confidence interval; ECOG, Eastern Cooperative Oncology Group; FIGO, International Federation of Gynecology and Obstetrics; HR, hazard ratio; IMRT, intensity modulated radiotherapy; LVSI, lymphovascular space invasion; OS, overall survival; RFS, recurrence-free survival; RT; radiotherapy; SCC, squamous cell carcinoma.

3. Toxicities according to adjuvant treatment

During follow-up, 17 patients (6.9%) exhibited acute grade 3 or higher treatment-related complications (Table 4). Specifically, in the RT arm, 5 patients (3.6%) exhibited grade 3 complications with no grade 4 or higher side effects. In the CCRT arm, 7 patients (6.7%) exhibited grade 3 complications, and 5 patients (4.8%) exhibited grade 4 complications with no grade 5 side effects. There was no statistically significant difference in complication rates between the 2 arms (p=0.124).

Table 4. Toxicities according to adjuvant treatment.

Type of adverse event			Total (n=244)	RT (n=139)	CCRT (n=105)	p-value
Acute adverse effects*
	Hematologic toxicity					0.065
		G3	8 (3.3)	3 (2.2)	5 (4.5)
		G4	3 (1.2)	0 (0.0)	3 (2.7)
	Gastrointestinal toxicity					0.840
		G3	1 (0.4)	1 (0.7)	0 (0.0)
		G4	1 (0.4)	0 (0.0)	1 (0.9)
	Genitourinary toxicity					0.397
		G3	3 (1.2)	1 (0.7)	2 (1.8)
		G4	1 (0.4)	0 (0.0)	1 (0.9)
Chronic adverse effects†						0.478
	Lymphedema
		G2	7 (2.8)	5 (3.7)	2 (1.8)
		≥G3	0 (0.0)	0 (0.0)	0 (0.0)

Values are presented as number (%).

CCRT, concurrent chemoradiotherapy; RT, radiotherapy.

^*No acute grade 5 toxicity; ^†No chronic ≥3 grade hematologic, gastrointestinal and genitourinary toxicity.

Grade 3 or higher HT toxicity was the most common complication, occurring in 4.5% of patients overall, with 2.2% in the RT arm and 7.7% in the CCRT arm. The incidence of hematotoxicity in the CCRT arm showed an increasing trend compared to the RT arm, but the difference was not statistically significant (p=0.065).

The incidence rates of grade 3 or higher GI and GU toxicities were relatively low, with no significant differences between the 2 arms (p>0.390). No chronic grade 3 or higher HT, GI, or GU toxicities were observed. Additionally, 5 patients (3.6%) in the RT arm and 2 patients (1.9%) in the CCRT arm exhibited grade 2 lymphedema (p=0.478).

DISCUSSION

Due to the widespread availability of radiation therapy equipment and the goal of preserving ovarian function, surgery is frequently employed in treating patients with early-stage cervical cancer. The treatment regimen for patients in the IRF group remains one of the most contentious issues in the current management of cervical cancer.

Presently, there is no standardized definition for IRF among clinical practitioners and researchers. In Sun et al. [10] and Kim et al.’s study [15], the intermediate-risk group comprised patients exhibiting any of the following IRFs: LVSI, stromal invasion exceeding one-half, or tumor size ≥4 cm. Ryu et al.’s study [9] classified IRFs using 2 criteria sets: the classic criteria and GOG criteria. Cao et al.’s study [22] defined IRF according to the Sedlis criteria in GOG 92. In Mahmoud et al.’s study [13], patients aged 18 years and older who underwent radical hysterectomy for pathologically confirmed stage IB–IIA non-metastatic invasive cervical cancer with negative pelvic lymph nodes and negative surgical margins formed the intermediate-risk group. However, the varied eligibility criteria across these studies certainly influenced the results.

The present study used the “Sedlis” criteria for SCC and the “four-factor model” criteria for patients with AC or ASC for adjuvant therapy. The latter approach, which accounts for pathological factors, may predict recurrence and poor survival more effectively than classic and GOG criteria in patients with cervical cancer who exhibit any IRF following radical hysterectomy. Our institution consistently employs a combination of the “Sedlis” and the “four-factor model” criteria to sensitively and specifically identify patients with IRFs who require postoperative adjuvant therapy.

This study identified AC histology as the only significant prognostic factor affecting all survival outcomes in multivariate analysis. However, ASC exhibits similarities to SCC and does not demonstrate a significant correlation with prognosis.

The 5-year OS rates among patients with SCC, AC, and ASC groups were 94.0%, 81.6%, and 87.5%, respectively, and the rates of RFS were 90.0%, 67.3%, and 83.3%, respectively (OS: p=0.041, RFS: p=0.012). Multivariate analysis identified AC histology as the sole significant prognostic factor influencing both OS and RFS. These results are consistent with previous findings [23,24]. Noh et al.’s study [23] showed that the 5-year RFS rates in patients with SCC, AC, and ASC histology were 83.7%, 66.5%, and 79.6%, respectively (p<0.001). Multivariate analysis identified AC histology as the sole significant prognostic factor influencing all survival outcomes. They concluded that AC histology was associated with poor survival outcomes in patients with FIGO stage IB–IIA cervical cancer who received adjuvant RT or CCRT. The prognosis for ASC histology was found to be more similar to that of SCC histology than to AC histology [23].

The present study’s data indicates that CCRT does not alter the prognosis of AC. Due to the relatively low incidence of AC/ASC, this pathological subtype accounted for only 17.6% (43/244) of patients in the present study. Although AC/ASC constituted a relatively higher proportion in the GOG 92 long-term follow-up study (21.3%, 59/277), unfortunately, this study failed to investigate the impact of concurrent chemotherapy administered during RT on the prognosis of patients with AC/ASC [2].

Previous studies have demonstrated that CCRT is more effective than RT alone for patients in the intermediate-risk group, indicating that adding concurrent chemotherapy improves survival outcomes [9,10,11]. Ryu et al.’s study [9] reported that a recurrence rate of 17.6% in the RT arm decreased to 2.2% using cisplatin-based chemotherapy concurrently with RT. This result was corroborated by a retrospective study demonstrating that CCRT improved survival in patients with early-stage cervical cancer with IRFs. Patients with 2 or more IRFs, including LVSI, may benefit from CCRT [10]. Another retrospective study indicated that adding concurrent chemotherapy to postoperative RT in patients with cervical cancer with IRFs improved RFS without increasing acute GI or chronic toxicities, although HT toxicities increased significantly [11]. However, the period for patient selection in the present study is too long, and there is significant heterogeneity in both RT and CCRT.

Contrarily, some studies have shown that CCRT did not yield improved survival outcomes compared to RT alone in the intermediate-risk group [12,13,14,15]. In Kim et al.’s study [15], the outcomes of 2 types of adjuvant treatments, RT versus CCRT, were compared in patients with intermediate-risk early-stage cervical cancers. The authors found no difference in the 5-year RFS or OS, concluding that chemoradiation is not superior to adjuvant RT alone. In addition to the absence of survival benefit, the combined treatment in the present study demonstrated a notably elevated risk of grade 3 toxicity. However, being a retrospective study, the baselines between the 2 arms are not identical and comparable. In the chemoradiation arm, risk factors were more frequently observed, including a 10% increase in tumors ≥4 cm, a 30% increase in LVSI, and a 20% increase in the combination of three risk factors.

Previous studies posed several shortcomings, including small sample sizes, excessively long periods for clinical studies, and considerable heterogeneity in both RT and CCRT protocols [9,10,11]. Additionally, differences in the number of IRFs [9,10,13,15,22], including patients with FIGO stage IIB cervical cancer potentially misclassified as high-risk, contributed to inconsistencies [10]. Further, an unbalanced baseline status between treatments [15], encompassing patients who had undergone preoperative neoadjuvant chemotherapy, was evident. Lastly, the use of obsolete treatment techniques may have influenced the assessment of CCRT efficacy in patients in the intermediate-risk group [5,6,7,12,13,14,15].

The conventional “four-field box” or 3D-CRT technique used in postoperative adjuvant therapy for gynecological tumors is considered sufficient to deliver a conformal radiation dose to the target at risk. However, it results in significant exposure to other organs in the pelvis with a certain toxicity [1,8,9,11,15]. However, IMRT permits delivering highly conformal therapeutic radiation doses to oncologic targets while constraining the dose to healthy tissue. Numerous studies have shown that IMRT can achieve good oncologic outcomes, such as improved RFS and OS and significantly decreased acute and chronic GI, GU, and HT toxicities compared to treatment with conventional radiation therapy as an adjuvant RT or CCRT for gynecological tumors after surgery [25,26,27,28]. Unfortunately, the long-awaited GOG 0263 trial had heterogeneous RT techniques owing to prolonged patient accrual, including both modern IMRT and conventional 3D-CRT. Our findings align with those of the recently published prospective NRG Oncology/GOG-263 trial (NCT01101451), which also demonstrated that CCRT is not superior to RT alone in patients with intermediate-risk early-stage cervical cancer after radical hysterectomy [29]. Notably, our study has several advantages over GOG-263. First, all patients received uniform pelvic IMRT with standardized target delineation and dosing, eliminating the radiation technique heterogeneity (e.g., 3D-CRT vs. IMRT) in GOG-263 from its prolonged enrollment. Second, we stratified IRF definitions by histology (Sedlis criteria for SCC; four-factor model for AC/ASC), enabling more precise high-risk subgroup identification and confirming AC as an independent prognostic factor—an underemphasized finding in GOG-263. Third, with a contemporary cohort (2012–2023) and chemotherapy doses adjusted for East Asian physical characteristics, our study better reflects real-world practice, improving generalizability to IMRT-standard clinical settings. These design strengths complement GOG-263 and confirm CCRT is not superior to RT alone even in the era of homogeneous IMRT.

In the present study, although no survival advantage was observed for CCRT over RT alone, the 5-year OS rate in all patients in the study was 93.8% and the 5-year RFS rate was 87.5%, respectively. The incidence of grade 3 and grade 4 blood toxicities in the present study is much lower (3.3% and 1.2%, respectively) than the data reported in the previous study with conventional RT [27,28]. The CCRT arm showed no increase in high-grade side effects compared to the RT arm except for the incidence of HT toxicity in the CCRT arm, which showed an increasing trend compared to the IMRT arm (p=0.065). The only grade 4 GU toxicity occurred in a 23-year-old female who had vaginal mutilation 1 year after surgery. However, the toxicity might be due to intense sexual activity based on the medical history provided by the patient herself. This may not only be due to the use of IMRT but also the dose regimen of 25–40 mg/m² cisplatin for chemotherapy based on the physical characteristics of East Asian individuals rather than an inflexible dose of 40 mg/m². The dose in the two-drug regimen was also appropriately reduced. Additionally, as there were no risk factors, the patient neither received brachytherapy nor a boost at the vaginal stump. Data on survival outcomes and side effects from the present study and existing literature collectively suggest that pelvic IMRT should be used more in postoperative RT or CCRT for cervical cancer.

The present study faced several limitations. First, the retrospective nature of the study and the relatively small sample size may have introduced selection bias. Second, the adjuvant treatment schedules were determined through discussions among doctors, patients, and family members, resulting in considerable uncertainty.

In conclusion, this study’s findings suggest that concurrent chemotherapy with pelvic IMRT does not demonstrate superior effectiveness compared to IMRT alone in real-world settings for patients with stage IB–IIA cervical cancer with IRFs following radical hysterectomy with PLND. A cautious decision is warranted regarding the administration of concurrent chemotherapy for patients with intermediate-risk early-stage cervical cancer following radical surgery in clinical practice.

SUPPLEMENTARY MATERIAL

Table S1

Number of patients with different regimens in the CCRT group