Predictive factors of pathologic good response after neoadjuvant chemoradiotherapy for rectal cancer

Abstract

Background

For patients with T3/T4N0, T(any)N+, or locally unresectable rectal cancer, the treatment approach involves preoperative chemoradiation (CRT) followed by radical surgical resection. Our objective was to identify the predictive factors associated with favorable pathologic response in these individuals.

Methods

We analyzed data from patients diagnosed with primary rectal cancer who underwent preoperative chemoradiation (CRT) followed by radical surgery at Changhua Christian Hospital. Information regarding patient demographics, clinical and pathological characteristics, laboratory data collected at baseline and during CRT, as well as perioperative details, was extracted from medical records. These patients were categorized into good-responder and poor-responder groups. We employed Kaplan–Meier curve analysis to evaluate overall survival (OS) and disease-free survival (DFS), and logistic regression to assess the variables related to treatment response.

Results

Overall, 98 patients were enrolled. A strong pathological response to preoperative CRT correlated with improved OS and DFS among patients with locally advanced rectal cancer. The introduction of simultaneous integrated boost (SIB) in long-course radiotherapy (RT) enhanced the pathological response while maintaining manageable acute toxicity.

Conclusions

Implementing SIB in long-course RT enhances treatment response while maintaining acceptable acute toxicity. Further studies are required to identify additional predictive factors.

Background

According to the 2018 Taiwan Cancer Registry Annual Report, colorectal cancer is the second most diagnosed cancer in Taiwan after breast cancer [1]. It commonly occurs during the fourth and fifth decades of life among Taiwanese people. Surgery remains the mainstay curative treatment for colorectal cancer. The German CAO/ARO/AIO-94 trial [2] and the American NSABP R-03 trial [3] recommended preoperative chemoradiation (CRT) followed by radical resection rather than surgical resection alone or postoperative concurrent chemoradiotherapy (CCRT) for T3/T4N0, T(any)N+, or locally unresectable rectal cancer. Preoperative CCRT effectively promotes tumor downstaging, preserves sphincter function, reduces the risk of local recurrence, and offers the potential for achieving a pathologic complete response (pCR) [2, 3].

The tumor regression grade reflects the response of a tumor to the treatment. Responses to preoperative CRT vary among individuals. The histopathological response to preoperative therapy has been shown to be a potential predictor of outcomes after surgery. Approximately 15% of the patients achieve pCR [3], and these patients have a good prognosis [4, 5]. Furthermore, the treatment strategy for these patients may differ from that for patients without pCR; therefore, predicting the response to neoadjuvant CRT is important. Maas et al. [4] identified the clinical factors that are predictors of tumor response to preoperative CRT, including pretreatment small tumor size, pretreatment CEA level < 5 ng/mL, and chemotherapy with capecitabine. In addition to pCR, response to preoperative CRT improves distant metastasis-free and disease-free survival rates [6].

To identify potential predictive factors for a good pathologic response, we retrospectively reviewed the medical literature at Changhua Christian Hospital and evaluated the outcomes of preoperative CRT for rectal cancer. We also evaluated the clinical and pathological responses of rectal cancer patients to neoadjuvant CRT. The predictive value of pathological response to preoperative CRT was also investigated, and the prognostic factors for overall survival (OS) and disease-free survival (DFS) were analyzed.

Methods

Patients

We retrospectively reviewed our institutional cancer registry. Patients with American Joint Committee on Cancer (AJCC) stage T3/T4N0 or T(any)N+ (AJCC Staging System, 7th Edition) or locally unresectable rectal cancer who were referred to the department of radiation oncology for preoperative CRT followed by radical surgery at Changhua Christian Hospital (CCH) between January 2010 and December 2018 were included. The observation period lasted from January 2010 to July 2023.

Eligible patients were ≥ 20 years old, diagnosed with histologically confirmed primary rectal cancer, and free of distant metastases. The staging workup included abdominal computed tomography (CT)/magnetic resonance imaging (MRI) and positron emission tomography/computed tomography (PET/CT). Patients who did not undergo radical surgery or were operated on at other hospitals after CRT, had a history of another malignancy before rectal cancer, or had a previous history of pelvic irradiation were excluded.

This study was conducted in accordance with the principles of the Declaration of Helsinki (2013 revision). Ethical approval was granted by the Ethics Committee of Changhua Christian Hospital (approval number: 200812). The requirement for individual consent was waived for this retrospective analysis, as the research involved no more than minimal risk.

Data collection

The relevant demographic characteristics and clinical and pathological variables of each patient were extracted. Pretreatment/during CRT/perioperative hematologic parameters, including white blood cell count (WBC), hemoglobin (Hb) level, platelet (PLT) count, neutrophil percentage, lymphocyte percentage, monocyte percentage, systemic immune-inflammation index, neutrophil-to-lymphocyte ratio, lymphocyte-to-neutrophil ratio, and platelet-to-lymphocyte ratio, were collected. Overall survival (OS) was defined as the time from the date of the diagnostic biopsy until death. Disease-free survival (DFS) was defined as the time from the date of radical surgery to the date of local recurrence, distant metastasis, or death.

Treatment

Preoperative CRT is recommended in patients with T3/T4N0, T(any)N+, or locally unresectable rectal cancer [2, 3]. Based on the clinical situation and physician’s experience, the patients received 5-fluorouracil (5FU)-based chemotherapy orally or via intravenous infusion, including oral tegafur/uracil(UFUR), oral capecitabine, 5-fluorouracil with leucovorin(FL), 5-fluorouracil/leucovorin/oxaliplatin (FOLFOX), 5-fluorouracil/leucovorin/irinotecan(FOLFIRI) or capecitabine/oxaliplatin (CapeOX).

Concurrent radiotherapy, including three-dimensional conformal radiation therapy (3D-CRT), intensity-modulated radiation therapy (IMRT), image-guided intensity modulated radiation therapy (IG-IMRT), volumetric-modulated arc therapy (VMAT), or image-guided volumetric-modulated radiation therapy (IG-VMAT), was delivered to the whole pelvis, with or without intensity-modulated radiotherapy with simultaneous integrated boost (IMRT-SIB)/volumetric-modulated arc therapy with simultaneous integrated boost (VMAT-SIB) to the primary gross tumor. Using a 10-MV linear accelerator with patients in the supine position, radiotherapy (RT) was administered in fractions of 1.8–2 Gy/day, 5 days a week for 5–6 weeks. A three-dimensional treatment planning system was used for treatment planning. The target volume included the primary tumor, perirectal fat tissue, mesorectum, and presacral and pelvic lymph nodes. The patients underwent surgical intervention 12 weeks after the completion of RT. The surgical method was based on the clinical judgment of the surgeon.

Evaluation of therapeutic response

Pathological assessment of preoperative CRT response was conducted using the AJCC tumor regression grading system as follows: TRG0, no residual tumor cells; TRG1, single cell or small group of cells; TRG2, residual cancer with desmoplastic response; and TRG3, minimal evidence of tumor response. We divided the patients into two groups based on their AJCC tumor regression grade and pathological findings: good responders (TRG0-1) and poor responders (TRG2-3). The primary endpoint of the study was the pathological tumor response according to the AJCC cancer tumor regression grading system.

Statistical analysis

Predictors of a good pathological tumor response were analyzed using logistic regression. Univariate factors suggesting an association with a good pathologic tumor response were selected as variables in the multivariate logistic regression analysis. Kaplan–Meier curve analysis and Cox regression models were used to estimate OS and DFS. Statistical significance was set at p < 0.05. IBM^® SPSS^® Statistics for Windows, Version 25.0 (Armonk, NY: IBM Corp.), was used for all data analyses.

Results

A total of 98 patients were included in the study, comprising 71 males (72.4%) and 27 females (27.6%). The enrollment process is illustrated in Fig. 1. Among the participants, 48 (49.0%) were diagnosed with rectal cancer before 2015, while 50 (51.0%) were diagnosed in 2015 or later. The median age of all patients was 58 years (range: 25–83 years). Of these patients, 44.9% had an Eastern Cooperative Oncology Group (ECOG) performance status of 0, 53.1% had a score of 1, and 2% had a score of 2. Most patients (89.8%) had moderately differentiated histology. The median initial carcinoembryonic antigen (CEA) level was 4.7 ng/dL (range: 0.7–949.6 ng/dL). Regarding tumor stage, 19 patients had T2 tumors, 72 had T3 tumors, and 7 had T4 tumors (Table 1).

Fig. 1.

Patient enrollment flowchart

Table 1.

Patient characteristics

Characteristic	n = 98 (%)
Diagnosis year
Before 2015	48(49.0)
After 2015	50 (51.0)
Age (years), median (range)	58 (25–83)
Median follow-up time (months, range)	64.5 (0–156)
Sex
Men	71 (72.4)
Women	27 (27.6)
BMI (kg/m2), mean (range)	23.2 (15.8–33.0)
Pretreatment ECOG
0	44 (44.9)
1	52 (53.1)
2	2 (2.0)
Differentiation
Well differentiated	3 (3.1)
Moderately differentiated	88 (89.8)
Poor differentiated	7 (7.1)
Distance from anal verge (cm)
≤ 5	49 (50.0)
6–15	49 (50.0)
Clinical T stage
T2	19 (19.4)
T3	72 (73.5)
T4	7 (7.1)
Lymph node involvement
N0	11 (11.2)
N1/N2	87 (88.8)
Clinical stage
I	4 (4.1)
II	7 (7.1)
III	87 (88.8)
Initial CEA level (ng/mL)
Median	4.7 (0.7–949.6)
≥ 5	46 (46.9)
< 5	52 (53.1)
Initial Hb level (g/dL)
Median	12.8 (6–16.5)
≥ 12	68 (69.4)
< 12	30 (30.6)
Interval between preoperative RT and surgery(days)
Median	46 (27–82)
< 42(days)	26 (26.5)
42–56(days)	57 (58.2)
56–84(days)	15 (15.3)
Pathologic response
Good	53 (54)
Poor	45 (46)
Pathologic complete response (pCR)	13 (13.2)
Adjuvant chemotherapy	37 (37.8)

BMI body mass index, ECOG the Eastern Cooperative Oncology Group performance status, CEA carcinoembryonic antigen

All patients received pelvic RT, either 50 Gy in 25 fractions or 50.4 Gy in 28 fractions. The median duration of RT was 38 days (range: 34–55 days). The median interval from completion of preoperative RT to surgery was 46 days (range: 27–82 days). Neoadjuvant chemotherapy regimens included UFUR/FL in 48 patients (49.0%), capecitabine in 26 (26.5%), and FOLFOX/FOLFIRI/CapeOX in 24 (24.5%) (Table 2). All patients underwent surgery within 12 weeks after completing RT. The median follow-up was 64.5 months (range: 0–156 months). Good pathological response was observed in 53 patients (54.0%), whereas 45 (46.0%) were classified as poor responders. The pathologic complete response (pCR) rate was 13.3% (n = 13).

Table 2.

Characteristics of neoadjuvant treatment

Radiation Technique	n = 98 (%)
3D-CRT	9 (9.2)
IMRT	21 (21.4)
IG-IMRT	1 (1.0)
VMAT	11 (11.2)
IG-VMAT	56 (57.1)
IMRT-SIB/VMAT-SIB	55 (56.1)
Median CTV-H (Gy)	54 (50–56)
≥Grade 3 toxicity	2 (2.0)
Concurrent Chemotherapy	n = 98 (%)
Oral UFUR/FL	48 (49.0)
Oral Capecitabine	26 (26.5)
FOLFOX/FOLFIRI/CapeOX	24 (24.5)

CRT chemoradiation, IMRT intensity-modulated radiation therapy, VMAT volumetric-modulated arc therapy, SIB simultaneous integrated boost, IG image-guided

The 3-year OS for the cohort was 84.5%. By subgroup, OS was 94.3% in good responders and 72.9% in poor responders (p < 0.001). The 3-year overall DFS rate was 74.1% overall, 84.9% in good responders, and 61.0% in poor responders (p = 0.002). Kaplan–Meier survival curves for OS and DFS are shown in Figs. 2 and 3. Results of univariate and multivariate analyses of predictors of good pathological response are presented in Tables 3 and 4.

Fig. 2.

Impact of pathologic response on overall survival. Patients with a good pathologic response showed better overall survival than poor responders (p < 0.001)

Fig. 3.

Impact of pathologic response on disease-free survival. Patients with a good pathologic response show better disease-free survival than poor responders (p = 0.002)

Table 3.

Univariate analysis of predictors for a good pathologic response

Characteristic	Good response n = 53 (54%)	Poor response n = 45 (46%)	p value
Diagnosis year
Before 2015	20(41.7)	28(58.3)	0.017*
After 2015	33(66.0)	17(34.0)
Age, years
≥ 65	14 (46.7)	16 (53.3)	0.329
< 65	39 (57.4)	29 (42.6)
Sex
Men	41 (57.7)	30 (42.3)	0.240
Women	12 (44.4)	15 (55.6)
Pretreatment ECOG
0	25 (56.8)	19 (43.2)	0.801
1	26 (50.0)	26 (50.0)
2	2 (100.0)	0 (0.0)
Differentiation
Well differentiated	2 (66.7)	1 (33.3)	0.379
Moderately differentiated	49 (55.7)	39 (44.3)
Poor differentiated	2 (28.6)	5 (71.4)
Distance from anal verge (cm)
≤ 5	29 (59.2)	20 (40.8)	0.312
6–15	24 (49.0)	25 (51.0)
Clinical T stage
T2	13 (68.4)	6 (31.6)	0.353
T3	37 (51.4)	35 (48.6)
T4	3 (42.9)	4 (57.1)
Lymph node involvement
N0	7 (63.6)	4 (36.4)	0.753
N1	23 (54.8)	19 (45.2)
N2	23 (51.1)	22 (48.9)
Clinical stage
I	4 (100.0)	0 (0.0)	0.986
II	3 (42.9)	4 (57.1)
III	46 (52.9)	41 (47.1)
Initial CEA level (ng/mL)
≥ 5	17 (37.0)	29 (63.0)	0.002*
< 5	36 (69.2)	16 (30.8)
Initial Hb level (g/dL)
≥ 12	38 (55.9)	30 (44.1)	0.591
< 12	15 (50.0)	15 (50.0)
I nterval between preoperative RT and surgery (median, days)	47 (27–82)	46 (33–75)	0.654
Radiation Technique
3D-CRT	3 (33.3)	6 (66.7)	0.202
IMRT	9 (42.9)	12 (57.1)	0.202
VMAT	7 (63.6)	4 (36.4)	0.040*
IG-VMAT	34 (60.7)	22 (39.3)	0.130
IG-IMRT/ IG-VMAT	34 (59.6)	23 (40.4)	0.194
IMRT-SIB/VMAT-SIB
1	38 (69.1)	17 (30.9)	0.001*
0	15 (34.9)	28 (65.1)
CTV-H (Median, Gy)	54.0 (50–56)	50.4 (50–56)	0.001*

*, p-value < 0.05; CEA carcinoembryonic antigen, ECOG the Eastern Cooperative Oncology Group performance status, CRT chemoradiation, IMRT intensity-modulated radiation therapy, VMAT volumetric-modulated arc therapy, SIB simultaneous integrated boost, IG image guided

Table 4.

Multivariate analysis of predictors of good pathologic response

Characteristic	Univariate analysis			Multivariate analysis
	p value	OR	95% CI	p value	OR	95% CI
Sex	0.240	1.708	0.699–4.174	0.957	0.970	0.321–2.927
Age	0.128	0.976	0.945–1.007	0.022*	0.948	0.905–0.992
Clinical stage	0.986			0.982
Initial CEA level ≥ 5	0.002*	0.261	0.113–0.603	0.001*	0.172	0.058–0.509
VMAT	0.040*	2.497	1.042–5.983	0.376	0.549	0.146–2.071
IMRT-SIB/VMAT-SIB	0.001*	4.173	1.786–9.750	< 0.001*	9.692	2.791–33.660

*p-value < 0.05

Analysis of pathologic tumor response predictors

In univariate analysis, patients diagnosed after 2015 had a higher response rate compared to those diagnosed before 2015 (p = 0.017). Male sex was associated with a higher response rate, but the difference was not statistically significant (p = 0.240). Patients treated with VMAT had higher response rates compared to those treated with 3D-CRT or IMRT (p = 0.040). However, IG-IMRT/ IG-VMAT was not a significant predictor (p = 0.194). In contrast, RT utilizing IMRT-SIB/ VMAT-SIB to gross tumors with a high-risk clinical target volume (CTV-H) dose > 52 Gy and a high CTV-H dose were strongly associated with a good response (all p = 0.001). An initial CEA concentration < 5.0 ng/mL was also significantly associated with a good response (p = 0.002).

Among the 55 patients who received IMRT-SIB or VMAT-SIB, 49 were diagnosed after 2015. A strong correlation was observed between CTV-H dose and the use of IMRT-SIB/VMAT-SIB. To minimize multicollinearity, “diagnosis year” and “CTV-H dose” were excluded from the multivariate analysis. Logistic regression analysis identified younger age (p = 0.022), IMRT-SIB/VMAT-SIB with CTV-H dose > 52 Gy (p < 0.001), and pretreatment CEA < 5.0 ng/mL (p = 0.001) as independent predictors of good pathologic response. Male sex, clinical stage, and VMAT use were not significant predictors.

Most patients tolerated preoperative CRT well. Gastrointestinal toxicity was the most common treatment-related complication. Two patients in the poor responder group experienced grade 3 proctalgia (one treated with IMRT and one with IG-VMAT). Adjuvant chemotherapy was administered to 37 patients (37.8%).

Discussion

Owing to revolutions in treatment modalities, including total mesorectal excision (TME), neoadjuvant radiotherapy, and chemotherapy, the management of locally advanced rectal cancer has improved significantly. With the emergence of neoadjuvant therapy as the standard of care for locally advanced rectal cancer, the degree of tumor regression has become an important prognostic factor [7, 8]. We investigated whether the use of IMRT-SIB/VMAT-SIB for gross tumors at a CTV-H dose > 52 Gy and a pretreatment CEA level < 5.0 ng/mL were significant predictors of a good pathologic response using multivariable analysis. Patients with locally advanced rectal cancer (LARC) who achieved a favorable pathological response to preoperative CRT exhibited improved OS and DFS. Moreover, the use of the SIB technique in long-course RT may enhance pathological response while maintaining tolerable acute toxicity.

IMRT-SIB/VMAT-SIB

Compared with 3D-CRT, IMRT/VMAT treatment improved tumor-dose conformity and adjacent organ-at-risk sparing and reduced the irradiation of the small bowel and treatment-related toxicity without hindering dose escalation to tumor volume. Compared to other elective procedures, IMRT-SIB/VMAT-SIB has the clinical and dosimetric advantages of increasing the fraction size of the boost volume with a low dose. Several phase II clinical trials using SIB-IMRT/SIB-VMAT reported good oncologic outcomes with acceptable toxicities [9, 10], corresponding to our results.

Initial CEA concentration

Related studies have reported that initial CEA concentration is a significant independent predictive factor for downstaging [11, 12]. A high pretreatment CEA level predicts a poor tumor response to CRT. Park et al. [13] reported that CEA level could be of clinical value as a predictor of the response to preoperative CRT and as an independent prognostic factor after preoperative CRT and curative surgery. Perez et al. [14] showed that a CEA level < 5 ng/mL was a favorable prognostic factor for rectal cancer and was associated with increased rates of early disease staging and complete tumor regression.

Importance of pCR

In 1995, Chari et al. [15] reported that patients with pCR or a good response to chemoradiation for rectal cancer had better long-term outcomes than those with a poor response. Since then, improvements in the local control of the primary disease have led to improved OS and progression-free survival [3]. If cancer cells are not identified in post-neoadjuvant CRT histopathological specimens, surgery would be an unnecessary overtreatment. A watch-and-wait approach may be appropriate and does not compromise oncological outcomes. The pCR rate in the present study was approximately 13.2%. However, to date, no specific predictive factors for pCR have been identified. In our study, low-lying rectal cancer was the only significant predictor of pCR.

Total neoadjuvant therapy

Recent advancements, particularly findings from the RAPIDO trial [16] and the UNICANCER-PRODIGE 23 trial [17], have redefined the management of LARC, positioning total neoadjuvant therapy (TNT) before radical surgery as the preferred standard. TNT provides notable advantages, including prolonged survival and improved local control, and the potential for non-operative management (NOM). Additionally, patients generally tolerate neoadjuvant therapy better, and the total duration of treatment may be shortened.

Neoadjuvant RT can be administered through short-course radiotherapy (SCRT) or long-course CRT. Evidence from the Polish I trial [18] and TROG 01.04 [19] indicated no marked differences in survival, local control, or late toxicities between SCRT and CRT. However, CRT has shown higher rates of pCR, reduced tumor staging, and lower incidences of positive radial margins [18].

TNT protocols can follow two distinct sequences: induction chemotherapy (INCT) followed by CRT/SCRT, or CRT/SCRT followed by consolidation chemotherapy. Both approaches have demonstrated superior therapeutic outcomes compared to conventional preoperative CRT with or without subsequent adjuvant chemotherapy [16, 17]. Of note, the local response to RT tends to improve over time, with higher rates of response observed in RT-first treatment arms [20].

The CAO/ARO/AIO-12 trial [21] and the OPRA trial [22] investigated the timing of chemotherapy relative to preoperative RT in TNT regimens, both utilizing long-course CRT. Both studies identified that initiating TNT with CRT, followed by systemic chemotherapy, resulted in superior pCR rates [21] and better organ preservation outcomes [22]. Moreover, the ongoing trial ACO/ARO/AIO-18.1 (NCT04246684) [23] aims to directly compare the effectiveness of CRT and SCRT as initial interventions within TNT, aiming to refine sequencing strategies for optimal therapeutic efficacy.

Limitations

This retrospective study has several limitations. First, the chemotherapy regimens and surgical approaches were not standardized and varied according to physician preferences. In addition, some hematologic parameters were unavailable due to the constraints of retrospective data collection.

Second, at Changhua Christian Hospital, only a proportion of patients with LARC were referred for preoperative CRT (42.86% in 2019, with even fewer in earlier years), and some of these patients did not undergo surgery following CRT. Recently, Melton et al. [24] explored the feasibility and early toxicities of dose-escalated SCRT in rectal cancer patients receiving TNT. Their findings suggest that while SIB can enhance treatment efficacy, it may also increase the risk of acute toxicity.

Third, although previous studies (e.g., Melton et al. [24]) have explored the feasibility of dose-escalated SCRT with SIB, the potential benefit must be weighed against an increased risk of acute toxicity.

Finally, while TNT with SCRT followed by delayed surgery has shown promise in improving pCR rates, the role of SIB in this context remains uncertain. Further prospective, multi-center studies with larger patient cohorts are required to confirm the present findings and clarify the balance between efficacy and toxicity.

Conclusions

In this retrospective cohort, a favorable pathological response to preoperative CRT was associated with improved OS and DFS in patients with LARC. Delivery of long-course RT with SIB was associated with higher pathological response rates, with manageable acute toxicity. These results suggest a potential role for SIB in treatment intensification; however, the findings should be interpreted with caution given the retrospective design and limited cohort size. Future research should focus on validating predictive biomarkers and optimizing CRT protocols through well-designed prospective trials.

Abbreviations

3D-CRT

Three-dimensional conformal radiotherapy

BMI

body Mass index

CapeOX

Capecitabine/oxaliplatin

CCRT

Concurrent chemoradiotherapy

CEA

Carcinoembryonic antigen

CI

Confidence interval

CCH

Changhua Christian Hospital

CRT

Chemoradiation

CT

Computed tomography

CTV-H

High-risk clinical target volume

DFS

Disease-free survival

ECOG

The Eastern Cooperative Oncology Group performance status

FL

5-fluorouracil with leucovorin

FOLFIRI

5-fluorouracil/leucovorin/irinotecan

FOLFOX

5-fluorouracil/leucovorin/oxaliplatin

Hb

Hemoglobin

IG

Image-guided

IG-IMRT

Image-guided IMRT

IG-VMAT

Image-guided VMAT

IMRT

Intensity-modulated radiation therapy

IMRT-SIB

Intensity-modulated radiation therapy with simultaneous integrated boost

INCT

Induction chemotherapy

LARC

Locally advanced rectal cancer

MRI

Magnetic resonance imaging

NOM

Non-operative management

OR

Odds ratio

OS

Overall survival

pCR

Pathologic complete response

PET/CT

Positron emission tomography/computed tomography

PLT

Platelet

RT

Radiotherapy

SCRT

Short course radiotherapy

SIB

Simultaneous integrated boost

TME

Total mesorectal excision

TNT

Total neoadjuvant therapy

UFUR

Tegafur 100 mg/uracil 224 mg

VMAT

Volumetric-modulated arc therapy

VMAT-SIB

Volumetric-modulated arc therapy with simultaneous integrated boost

WBC

White blood cell count

Author contributions

Conceptualization, Y.-E. L. and J.-B. L.; Methodology, Y.-E. L. and J.-B. L.; Software, Y.-E. L.; Validation, T.-H. C. and T.-W. C.; Formal analysis, Y.-E. L.; Investigation, Y.-E. L.; Resources, T.-H. C. and L.-C. H.; Data curation, C.-C. H.; Writing—original draft preparation, Y.-E. L.; Writing—review and editing, J.-B. L. and J. L. ; Visualization, Y.-E. L.; Supervision, J.-C. L. and J. L; Project administration, L.-C. H. and J.-B. L. All authors have read and agreed to the published version of the manuscript.

Funding

This research received no external funding.

Data availability

Data is provided within the manuscript.

Declarations

Ethics approval and consent to participate

This study complied with the principles of the Declaration of Helsinki (revised in 2013). The Institutional Review Board of Changhua Christian Hospital approved the study (IRB No. 200812). Given the retrospective nature of the study, the need for individual informed consent to participate was waived by the IRB.

Consent for publication

For potentially identifying information and images of human participants included in this article, written informed consent for publication was obtained from the individual(s).

Competing interests

The authors declare no competing interests.