Factors Affecting the Survival of Patients with Synchronous Metastatic Colorectal Cancer in a Tertiary Hospital in Indonesia: A Retrospective Study

Abstract

Introduction: Little is known about Indonesian metastatic CRC patients’ survival and prognostic factors, although this malignancy is one the most frequent cancer in the country. We aimed to investigate the survival outcome and the factors influencing local CRC patients presenting with a metastatic stage at diagnosis.
Methods: A retrospective cohort study was done on 441 data of synchronous metastatic CRC cases treated between January 2016 and December 2019 at Dr. Sardjito Hospital, Yogyakarta, Indonesia. Secondary data were collected from the CRC clinical registry database. Demographic, clinicopathology and the type of treatment were collected. Survival staus was obtained from the registry database and communication to patients or caring families. Kaplan-Meier curves were used to estimate the overall survival (O.S). The Cox proportional hazards regression mode was applied to analyzethe potential factors affecting survival.
Results: The median follow up of the recent study was 17 months, The median overall survival was 13 months. Two-year overall survival was 37% and the estimated 5-year overall survival was 16.1%. Multivariate cox analyses identified poor performance status (HR 2.639, 95% CI 1.438-4.842, p = 0.002), elevated carcinoembryonic antigen (CEA) (HR 2.795, 95% CI 1.509-5.176, p = 0.001) and higher histological grade (HR 2.019, 95% CI 1.112-3.667, p = 0.021) were associated with poorer overall.
Conclusion: Poor performance status, high CEA level, and higher histological grade as factors that were associated with unfavourable overall survival of patients with synchronous metastatic colorectal cancer in Yogyakarta, Indonesia.

Introduction

Colorectal cancer (CRC) is a malignancy with the third most common incidence globally based on the GLOBOCAN database 2020 [1]. In Indonesia, CRC is ranked 4^th for cancer incidence and 5^th for the cause of death due to cancer [2]. Approximately 22% of CRC patients are diagnosed at stage IV [3]. Stage 4 colorectal cancer patients have poor overall survival, around 10% for five years [3]. A retrospective cohort study using the Surveillance Epidemiology and End Results (SEER) database reported that age at diagnosis, marital status, race, primary tumor site, tumor grade, CEA level, T status, N status, surgery for the primary lesion, chemotherapy, and location of metastases (bone, brain, liver and lung) independently influenced prognosis [4].

There is little information on CRC survival in Indonesia and even less on mCRC. A study in eastern Indonesia reported an association between age, type of histopathology, stage and mode of surgery on overall survival of CRC [5]. Still, it did not focus on patients with metastatic disease. In Yogyakarta province, CRC is the third most common cancer according to the cancer registry data, with 39% of patients have been diagnosed at stage 4 [6]. The purpose of this study is to determine the survival of this group of patients and identify factors affecting it.

Materials and Methods

Data collection

This study used secondary data from the hospital based cancer registry (HBCR) of colorectal in Dr. Sardjito General Hospital Yogyakarta. HBCR abstracted 32 variables using canreg5 software. Based on initial data from canreg5, more detailed data were collected through medical records as the CRC clinical registry, including patients diagnosed in 2016-2019. The present study recruited all patients with synchronous metastatic disease data from the clinical registry database, including demographic, clinical, pathology and first line treatment information. Data collection was undertaken between August 2020 and February 2022. We included 441 mCRC patients’ data from the registry database. The joint ethics committee from the Faculty of Medicine, Public Health and Nursing, Universitas Gadjah Mada/Dr. Sardjito Hospital, Yogyakarta, has approved this study (reference number KE/FK/0549/E.C./2020). Based on ethical clearance policy, informed consent is not required for retrospective studies

Key Variables

We determined age as “young” if <50 years and “old” if ≥50 years. gender (male versus female), education (5 ng/ml) were also collected.

Tumor location was categorized into right sided colon (caecum, ascending colon, hepatic flexure, and transverse colon) and left sided colon (splenic flexure, descending colon, sigmoid colon and rectosigmoid colon). Other tumor parameters included histological grade (1, 2, 3, or 4), morphological subtypes (adenocarcinoma, mucinous carcinoma, or signet ring cell carcinoma), T status (1, 2, 3, 4, and X if it was not determined or unknown), N status (0, 1, 2, and X if it was not determined or unknown), M status (0, 1, and X if it was not determined or unknown), and metastatic stage (A or B-C). Disease stage was determined according to TNM classification of the 8^th edition American Joint Committee of Cancer (AJCC) [9]. Primary tumor resection was categorized into two, namely yes and no.

Chemotherapy treatments were categorized into non-oxaliplatin based and oxaliplatin based regimens. Included in the non-oxaliplatin based regimens are single capecitabine (common dose: 2000-2500 mg/m²/day day 1-14 every 21 days for 8 cycles), De gramont or 2FU/LV (common dose: fluorouracil 400 mg/m² IV bolus days 1-2, fluorouracil 600mg/m² continuous IV days 1-2, folinicacid 200 mg/m² every 14 days for 12 cycles), and FOLFIRI (common dose: irinotecan 180 mg/m² day 1, folinic acid 200 mg/m² day 1, fluorouracil 400-800 mg IV bolus day 1, fluorouracil 2400-3000 mg/m² IV continuously for 36 hours, every 14 days for 6-8 cycles). In comparison, the oxaliplatin based regimens are FOLFOX (common dose: oxaliplatin 85 mg/m² day 1, folinic acid 200 mg/m² day 1-2, fluorouracil 400 mg/m² IV bolus day 1-2, fluorouracil 600 mg/m² continuous IV for 22 hours days 1-2, repeated every 14 days for 12 cycles) and CAPOX (common dose: oxaliplatin 135 mg/m² day 1, capecitabine 2000-2500 mg/m² day 1-14, repeated every 21 days for 8 cycles). The use of targeted therapy was categorized into no targeted therapy and with targeted therapy. Targeted therapy referred to bevacizumab (5 mg/kg day 1, usually administered with FOLFOX 12 cycles maximum or cetuximab (400 mg/kg on first administration, then 200 mg/kg, repeated every week, 12 cycles maximum (based on Indonesia national insurance regulation) according to the KRAS mutation status. The addition of bevacizumab or cetuximab was based on the consideration of the oncologist.

Overall survival (O.S.) was determined as months difference between date of diagnosis and date of death from any causes or the date from current information from medical record. If the patients did not visit the outpatient clinic for more than six months, we contacted them or their families by telephone or mail. We used estimation date in 7 patients who died and the family could only remember the year of death. We determined Jun 30 of the death year as the date of death. While we could not contact the patients or their families, their survival status was determined by the date of their last visit to the hospital.

Statistical analysis

Variables were compared by Cox proportional hazards regression model. Variables with a p-value <0.05 were included in the multivariate analysis. The Kaplan-Meier method was used to calculate the O.S. Comparisons between groups of interest were analyzed using a log-rank test. A p-value of <0.05 was considered statistically significant. All statistical analyses were performed using SPSS statistical software (IBM SPSS Statistics for Windows, version 28 (IBM Corp., Armonk, N.Y., USA).

Results

We analyzed 441 eligible patients. The majority of subjects were ≥50 years old (318, 72.1%), males (244, 55.3%), ≥ junior high school educational attainment (245, 55.6%) and being married (391, 88.7%). Sixty-seven percent had private or premium payer or government insurance for a civil servant. Most subjects had ECOG 0-1 performance status (228, 51.7%). At total of 183 cases (41.5%) had 18.5-23 kg/m² BMI and 310 cases (70.3%) had a pretreatment hemoglobin level ≥10 g/dL. CEA level >5 ng/mL were in 227 subjects (51.5%). Most subjects had left-sided tumor (333, 75.5%) and histological grading 1 (152, 34.5%). Adenocarcinoma was the prominent morphological subtypes (n =390, 88.4%). T3 was the majority of tumor status (213, 48.3%) and N1 was the majority of N status (142, 32.2%). A total of 257 (58.3%) patients did not undergo primary tumor resection, 291 patients received targeted therapy (66.0%). and 335 patients (76.0%) had a IV-A metastatic disease. The baseline characteristics of study subjects are listed in Table 1.

Table 1. Baseline Characteristics of Study Subjects (n=441).

Variables	Frequency (N)	Percentage (%)
Age (years)
<50	123	27.9
≥50	318	72.1
Gender
Male	244	55.3
Female	197	44.7
Education
< junior high school	134	30.4
≥ junior high school	245	55.6
Unknown	62	14.1
Marital status
Single	12	2.7
Married	391	88.7
Widower/widow	31	7
Unknown	7	1.6
Insurance
Government insurance for the poor	114	25.9
Private insurance, premium payer, or government insurance for civil servant	297	67.3
Self-pocket	23	5.2
Unknown	7	1.6
ECOG
0-1	228	51.7
2	94	21.3
04-Mar	46	10.4
Unknown	73	16.6
BMI (kg/m2)
<18.5	146	33.1
18.5-23	183	41.5
≥23	81	18.4
Unknown	31	7
Hemoglobin level (g/dL)
<10	98	22.2
≥10	310	70.3
Unknown	33	7.5
CEA (ng/mL)
≤5	56	12.7
>5	227	51.5
Unknown	158	35.8
Tumor location
Right	82	18.6
Left	333	75.5
Unknown	26	5.9
Histological grading
1	152	34.5
2	131	29.7
04-Mar	46	10.4
Unknown	112	25.4
Morphological subtypes
Adenocarcinoma	390	88.4
Mucinous carcinoma	19	4.3
Signet ring cell carcinoma	8	1.8
T status
1	4	0.9
2	23	5.2
3	213	48.3
4	120	27.2
X	81	18.4
N status
0	138	31.3
1	142	32.2
2	50	11.3
X	111	25.2
Targeted therapy
No	291	66
Yes	150	34
Resection of primary tumor
No	257	58.3
Yes	75	17
Unknown	109	24.7
Metastatic stage
IVA	335	76
IVB	91	20.6
IVC	15	3.4

Note, ECOG, Eastern Cooperative Oncology Group; BMI, body mass index; CEA, carcinoembryonic antigen, T, tumor; N, nodal.

The median duration of follow-up for the whole cohort was 17 months (ranged 0.0-72 months). The median overall survival was 13 months. The 2-year survival was 37%, and the estimated 5-year survival was 16.1% (Figure 1).

Figure 1. Kaplan-Meier curve of OS in mCRC.

Univariate cox regression showed that ECOG, BMI, CEA, hemoglobin, tumor location, histological grading, N status, resection of the primary tumor, targeted therapy and metastatic stage were potential survival predictors.

Multivariate cox regression demonstrated that ECOG 3-4 was associated with poorer prognosis (HR =2.639, 95% confidence interval/CI 1.438-4.842, p 0.002). A higher CEA level was associated with increased mortality (HR = 2.795, 95% CI 1.509-5.176, p = 0.001).

A higher histological grade was also associated with poorer survival (histological grade 3-4, HR 0.021, 95%CI 1.112-3.667, p = 0.021). Due to missing data, only in 177 cases (40.1%) can multivariate analysis be performed. The univariate and multivariate analyses for overall survival are listed in Table 2.

Table 2. Univariate and Multivariate Analyses for Overall Survival.

Variables		Univariate analysis			Multivariate analysis
	HR	95% CI	p*	HR	95% CI	p*
Age (years)
<50
≥50	1.179	0.877-1.584	0.275
Gender
Male
Female	1.037	0.802-1.341	0.781
Education
≥junior high school	1.043	0.777 -1.402	0.778
Marital status
Single
Married	0.952	0.563-1.609	0.855
Widower/Widow	1.085	0.481-2.447	0.844
Insurance type
Government insurance for the poor
Private insurance, premium payer, or government insurance for civil servant	0.998	0.737-1.351	0.99
Self-pocket	0.727	0.359-1.471	0.375
ECOG
0-1
2	1.336	0.955-1.87	0.091	1.245	0.734-2.111	0.417
04-Mar	2.471	1.655-3.688	0	2.651	1.444-4.865	0.002
BMI (kg/m2)
<18.5
18.5-22.9	0.878	0.651-1.183	0.392	0.782	0.502-1.217	0.276
≥23	0.774	0.53-1.13	0.185	0.904	0.508-1.609	0.731
CEA (ng/mL)
<=5
>5	1.854	1.214-2.83	0.004	2.762	1.49-5.121	0.001
Hemoglobin (mg/dL)
<10
≥10	0.524	0.384-0.716	<0.001	0.767	0.446-1.319	0.338
Tumor location
Right
Left	0.819	0.595-1.126	0.219	0.728	0.417-1.271	0.264
Histological grading
1
2	0.81	0.587-1.116	0.198	0.755	0.482-1.184	0.221
04-Mar	1.428	0.945-2.157	0.091	2.01	1.107-3.649	0.022
Pathological morphology
Adenocarcinoma
Mucinous carcinoma	0.962	0.51-1.813	0.904
Signet ring cell carcinoma	1.12	0.278-4.515	0.874
T status
1
2	0.661	0.148-2.957	0.588
3	0.637	0.157-2.587	0.528
4	0.818	0.2-3.342	0.78
x	0.804	0.193-3.346	0.764
N status
0
1	1.057	0.767-1.456	0.734	0.807	0.5-1.304	0.382
2	1.195	0.766-1.863	0.432	0.93	0.482-1.795	0.829
x	1.528	1.078-2.166	0.017	1.053	0.57-1.944	0.869
Resection of primary tumor
No
Yes	1.374	0.976-1.934	0.068	1.197	0.676-2.12	0.536
Unknown	1.023	0.661-1.583	0.919	1.41	0.641-3.103	0.393
Targeted therapy
No
Yes	0.787	0.605-1.023	0.073	1.066	0.704-1.615	0.762
Metastatic stage
A
B-C	1.544	1.544-2.098	0.006	1.345	0.817-2.215	0.243

Statistical analysis performed by Cox proportional hazard model with p*< 0.05; ECOG, Eastern Cooperative Oncology Group; BMI, body mass index; CEA, carcinoembryonic antigen, T, tumor; N, nodal.

Kaplan-Meier curves also showed statistically significant differences based on ECOG, CEA, and histological grade (Figure 2-4).

Figure 2. Kaplan-Meier Curve of OS in ECOG 0-1, 2, and 3-4. OS, Overall Survival, ECOG, Eastern Cooperative Oncology Group.

Figure 3. Kaplan-Meier Curve of OS Based on CEA Level. OS, Overall Survival; CEA, Carcinoembryonic Antigen.

Figure 4. Kaplan-Meier Curve of OS Based on BMI. OS, overall survival; BMI, body mass index.

Discussion

Metastatic CRC is still a challenge in the field of oncology. In Yogyakarta, Indonesia, nearly 40% of CRC patients present at stage 4 (6). To the best our knowledge, this is the first study conducted to determine the factors that affect the overall survival of mCRC patients in Yogyakarta. We analyzed 411 mCRC patients treated at Dr. Sardjito Hospital, a tertiary hospital in Yogyakarta.

Five years O.S. in this study was 16.1% with a median survival of 13 months. There 41% of participants did not receive targeted therapy. The addition of targeted therapy was at the discretion of the oncologist. The reasons why these patients did not receive targeted therapy were not investigated further in this study. Although the resources for cancer services in Indonesia were limited, as was often found in developing countries, the 5-year overall survival in this study was comparable to the Surveillance, Epidemiology, and End Results (SEER) report, which showed 5-year O.S. for metastatic colorectal cancer in the USA was 14.7% [10,11]. Meanwhile, in Europe, the 5-year overall survival for mCRC was 22% [12].

Based on multivariate cox regression, the variables that independently affect overall survival were ECOG score, CEA and histological grade. The ECOG score, an assessment of performance status, has been known to be one of the most influencing factors for survival in cases of metastatic colorectal cancer [13]. In this study, patients with ECOG 3-4 were independently associated with a worse prognosis with HR 8,472 (p < 0.001, 95% CI: 3.748-19151). Since cases with metastases usually present with poor performance status associated with visceral crisis, clinicians tend to give less aggressive therapy or reduced chemotherapy doses [13]. However, we did not record whether the patients received a reduced dose or not in this study.

CEA can be measured in serum quantitatively and can be used as a diagnostic and prognostic marker [14]. Increased preoperative CEA has been known to be associated with poorer overall survival [15], as confirmed by our study. The relationship between histological grade based on tumor differentiation and disease prognosis has also been well-recognized, supporting our findings [16]. Those with colorectal adenocarcinomas that are grade 3 often have a worse prognosis than patients whose tumors are grade 1 or 2. Patients with grade 3 colorectal adenocarcinomas had a 45.5% 5-year survival rate, while the 5-year survival rate of grades 1 and 2 were 71.4% and 59.5%, respectively [17].

The strength of this study is that it is the first study on factors affecting survival in mCRC in Indonesia. Missing data is a weakness of retrospective studies; this study was no exception. There were 16.6% missing data on ECOG, 35.8% on CEA and 25.4% on histological grading. Thus only 40.1% of cases could be done multivariate analysis (n=177). Another limitation was that it was unknown why patients did not receive targeted therapy and whether patients with poor performance status received reduced chemotherapy doses. Despite these limitations, we found that ECOG score, CEA and histological grading showed their prognostic ability. In conclusion, this study showed ECOG score, CEA levels and histological grade were independent factors affecting survival in mCRC.

Funding

The authors received funding from the Institutional Links grant, ID 527558574, under the Newton Institutional Link-Indonesia KLN Fund partnership, funded by the UK Department for Business, Energy and Industrial Strategy and Indonesia Ministry of Research Technology & Higher Education and delivered by the British Council.

Institutional Review Board Statement

This study was conducted under Ethical clerence (KE/ FK/0549/E.C./2020; Medical and Health Research Ethics Committee (MHREC), FKKMK, UGM-Dr. Sardjito General Hospital).

Acknowledgments

The authors thank Yana Suryani, Rosita Yunanda Purwanto, Riani Witaningrum and Nugira Dinantia for technical assistance and coordination

Conflicts of Interest

All authors declare that they have no conflicts of interest.