Abstract
Introduction. Various clinicopathological, radiological, and molecular parameters are predictive of prognosis in patients with colorectal carcinoma and distant organ metastases continue to have a significant place among them. Recent studies reveal that not only the presence of metastases but also the histopathological growth pattern of the metastatic tumor significantly affects prognosis. This study aimed to investigate the prognostic significance of the histopathological growth patterns of metastatic tumors, the morphological findings in the peritumoral non-neoplastic liver, and its relationship with survival in patients who have metastatic colorectal carcinoma. Materials and Method. Hematoxylin and eosin-stained slides of the tumors were re-examined in terms of histopathological diagnosis, growth pattern, presence and degree of peritumoral lymphocytic infiltration, steatosis, cholestasis, and peritumoral ductular reaction in the non-neoplastic liver. Results. In terms of histopathological growth patterns, 24 (47%) tumors showed replacement, 19 (37%) showed desmoplastic and 8 (16%) showed pushing growth pattern. In terms of total survival, there was a significant difference (P = .011) between desmoplastic and replacement growth patterns, and the survival period was shorter in patients with replacement growth patterns. Conclusion. Recent studies show that histopathological growth patterns in metastatic liver tumors may be a promising prognostic and predictive parameter. It is important to include this parameter in the pathology reports as it does not require additional equipment for evaluation in routine pathology practice, does not bring additional costs, or takes a long time to evaluate. This feature can be evaluated standardly by every pathologist.
Introduction
Colorectal carcinoma is the third most common cancer in men and the second most common in women. Despite improved diagnostic modalities and therapeutic methods, colorectal carcinoma is responsible for approximately 693 900 annual deaths worldwide. 1 Various morphological, immunohistochemical, and molecular parameters have been defined to predict prognosis in colorectal carcinomas, and studies are still ongoing to identify new biomarkers. Currently, tumor stage is still the most important prognostic parameter, while others such as histological subtype, grade, lymphovascular invasion, perineural invasion, tumor budding, host immune response (tumor-infiltrating lymphocytes and Crohn-like lymphoid reaction), and neoadjuvant treatment response, guide treatment regimens. The presence of distant metastasis has a significant effect on prognosis, especially when determining the tumor stage. 2 The liver with its anatomical, diagnostic, clinical, and therapeutic properties, takes an important place in colorectal cancers. These tumors most frequently metastasize to the liver and approximately 50% of colorectal carcinoma patients have liver metastases at the time of diagnosis. 3
In metastatic colorectal carcinomas, the total excision of the metastatic focus has a significant effect on survival. 4 Recent studies emphasize the histopathological growth pattern of the metastatic focus as another important prognostic parameter. 5 Various histopathological features can be present in the non-neoplastic liver in metastatic tumors. Cholestasis, steatosis, and peritumoral ductular reaction are among them. Obesity has a role in the etiology of colorectal carcinomas and is associated with a poorer prognosis. This suggests that the presence of steatosis might also be a risk factor for metastasis. However, some studies suggest that steatosis can also prevent the development of metastasis. 6
In this study, we aimed to investigate the prognostic significance of the histopathological growth pattern of the metastatic tumor, the morphological findings in non-neoplastic liver, and its relationship with survival in metastatic colorectal carcinoma patients.
Materials and Method
Patients diagnosed with metastatic colon carcinoma from liver resection materials (metastasectomy, segmentectomy, sectorectomy, and wedge resection) in our center between 2011 and 2020 were evaluated. Patients whose pathology slides were not suitable for re-evaluation and patients who underwent radiofrequency ablation before surgery and/or neoadjuvant treatment were excluded from the study. A retrospective study group of 51 patients, with available hematoxylin and eosin slides and blocks, was established.
Demographic information (age, gender), diagnosis of primary and metastatic tumor, serum carcinoembryonic antigen levels, survival status, total survival, and follow-up periods of the patients were obtained from the hospital database and pathology reports. Hematoxylin and eosin-stained slides of the tumors included in the study were re-examined in terms of histopathological diagnosis, growth pattern of the metastatic tumor, presence and degree of peritumoral lymphocytic infiltration, steatosis, cholestasis and peritumoral ductular reaction in non-neoplastic liver.
Histopathological growth patterns of the metastatic tumors were established based on the article “International consensus guidelines for scoring the histopathological growth patterns of liver metastasis” by van Dam et al, published in 2017. According to the article, the growth pattern is considered “desmoplastic” if the metastatic focus is separated from the non-neoplastic liver by a desmoplastic rim. If tumor cells are continuous with the liver parenchyma and are situated perpendicular to the tumor-liver interface “replacement pattern” and if tumor cells are pushing and compressing the normal parenchyma without invading “pushing pattern” is given. “Sinusoidal pattern” is when the tumor cells are only seen in sinusoidal blood vessels or in perisinusoidal spaces. “Portal pattern” is considered if tumor cells are only in the portal areas. 5 Evaluation of peritumoral lymphocytic infiltration was done by dividing into 3 groups: mild, moderate, and prominent. All slides including non-neoplastic liver were evaluated in terms of steatosis, cholestasis, and ductular reaction. The presence of steatosis, cholestasis, or ductular reaction in the non-neoplastic liver was reported as present or absent. Steatosis above 5% was considered “present” and below 5% was considered “absent.” The ductular reaction was assessed as present or absent in the non-neoplastic liver, regardless of the size or extent of the affected area.
Ethics approval, dated 03 November 2021 and numbered 2021-16/7, was obtained from the local Clinical Research Ethics Committee.
Statistical Analysis
A Shapiro-Wilks test was performed to determine the normal distribution of the variables. For non-normally distributed continuous variables, median (minimum and maximum) values were reported. For categorical variables, frequency or percentage values were given. A Kruskal-Wallis test was performed to compare the variables between growth patterns. Mann-Whitney U test was used for pairwise comparison. Fisher-Freeman-Halton or chi-square tests were performed for group comparisons of categorical variables and chi-square or Fisher’s exact test were used for 2 subgroup comparisons. Survival data were analyzed with the Kaplan-Meier method. All statistical analyses used in the study were performed using SPSS version 21.0. The level of significance was set at α = .05.
Results
The general characteristics of the 51 patients with metastatic colorectal carcinoma are summarized in Table 1. Sixteen (31%) patients were female and 35 (69%) were male. Female-male ratio was 0.4. Patients’ age ranged from 23 to 80 (mean: 58 years). All patients had metastasis at the time of diagnosis of colorectal carcinoma.
General Characteristics of the Patients (n = 51).
Histopathological evaluation of liver excision materials morphologically showed adenocarcinoma in 43 (84%) and mucinous adenocarcinoma in 8 (16%) patients. In terms of histopathological growth pattern, 24 (47%) tumors showed replacement, 19 (37%) showed desmoplastic and 8 (16%) showed a pushing growth pattern (Figures 1–3). None of the tumors had sinusoidal or portal growth patterns.
Images of the replacement histopathological growth patterns (A, ×40 and B, ×200). T, tumor; IM, invasive margin; P, liver parenchyma.
Serum carcinoembryonic antigen levels were measured in 42 patients at the time of admission. Carcinoembryonic antigen levels ranged from 0.69 to 5718.19 (mean: 9.76). When serum carcinoembryonic antigen levels were compared, the mean value was 25 µg/L in patients showing a pushing growth pattern, 12 µg/L in patients with replacement, and 6.1 µg/L in patients with a desmoplastic growth pattern.
Peritumoral lymphocytic infiltration was mild in 16 (31%), moderate in 9 (18%), and prominent in 19 (37%) tumors, while 7 tumors had no lymphocytic infiltration. Prominent lymphocytic infiltration was detected in 53% of the tumors with desmoplastic pattern, 33% of the tumors with replacement pattern, and 12% of the tumors with pushing growth pattern. A statistically significant correlation was found between peritumoral lymphocytic infiltration and the growth patterns (P = .030). In subgroup analyses, a significant difference was observed between desmoplastic and replacement (P = .034) and desmoplastic and pushing (P = .022) growth patterns. In cases with a desmoplastic growth pattern, peritumoral lymphocytic infiltration was observed more intensely. However, there was no significant difference between replacement and pushing growth patterns (P = .575).
Peritumoral ductular reaction around metastatic focus was present in 26 (51%) tumors. It was observed in 54% of the tumors with a desmoplastic pattern, 38% of the tumors with replacement pattern, and 8% of the tumors with a pushing growth pattern. A significant relationship between the presence of peritumoral ductular reaction and growth patterns (P = .036) was found. Subgroup analyses showed a significant difference between growth patterns of desmoplastic and replacement (P = .036) and desmoplastic and pushing (P = .033). The peritumoral ductular reaction was observed more frequently in cases with a desmoplastic growth pattern. However, there was no significant difference between replacement and pushing growth patterns (P = .676).
Evaluation of the non-neoplastic liver revealed steatosis in 39% and cholestasis in 35% of the patients. There was no significant relationship between steatosis or cholestasis and the growth patterns (P = .090 and P = .926).
The mean follow-up period was 18 months (range: 0.5-90 months). The total survival period ranged from 0.5 to 95 months, and the mean total survival period was 27 months. Thirty-five patients died during the follow-up period. The survival rates of patients with desmoplastic, pushing, and replacement growth patterns were 47%, 37%, and 17%, respectively. A significant difference between growth patterns in terms of survival was present (P = .042) (Figure 4A). There was a significant difference (P = .011) between desmoplastic and replacement growth patterns and the survival period was shorter in patients with a replacement growth pattern. Other subgroup evaluations did not differ significantly (Figure 4B).
Images of the desmoplastic histopathological growth patterns (A, ×40 and B, ×200). T, tumor; IM, invasive margin; P, liver parenchyma.
Discussion
Histopathological growth patterns have been identified in metastatic liver tumors and their prognostic importance has been investigated in recent studies. Based on the hypothesis that differences in angiogenesis of metastatic adenocarcinomas determine how they grow, Vermeulen et al evaluated the tumor-liver parenchyma interface and classified growth patterns of metastatic tumors into 4 groups: desmoplastic, pushing, replacement, and mixed. 7 With later studies these groups were rearranged and today, the histopathological growth patterns of metastatic liver tumors are divided into 5 patterns: desmoplastic, replacement, pushing, sinusoidal, and portal growth. The main criterion for determining the growth pattern remains to be the relationship between the tumor and the liver parenchyma. If there is a desmoplastic rim between tumor cells and the surrounding parenchyma “desmoplastic,” if tumor cells replace the parenchyma cells without destroying the main architecture “replacement,” if the tumor cells grow by pushing the parenchymal cells “pushing,” if the tumor is seen only in the sinusoidal blood vessels or perisinusoidal spaces “sinusoidal” and only in portal areas “portal” growth pattern is determined. Desmoplastic and replacement are the most common growth patterns while the others are rarely seen.5,7,8 Some researchers classify growth patterns only as desmoplastic and non-desmoplastic. 9 In this study, the most frequently seen growth pattern was replacement.
Images of the pushing histopathological growth patterns (A, ×40 and B, ×200). T, tumor; IM, invasive margin; P, liver parenchyma.
(A) Kaplan-Meier analysis of the relationship between total survival rate and histopathological growth pattern (P = .042). (B) Comparison of total survival in desmoplastic and replacement growth patterns by Kaplan-Meier analysis (P = .011).
The effects of growth patterns on prognosis vary. In one study, a better prognosis was found in tumors with desmoplastic pattern, and poor prognosis was found in tumors with a replacement growth pattern. 10 In their series of 732 patients, Galjart et al found better prognosis in tumors showing only desmoplastic pattern, and poorer prognosis when a replacement or pushing growth pattern was accompanied. They reported that this was independent of the percentage of the accompanying pattern. 11 Falcao et al also found that the pushing growth pattern negatively affected the survival rates of the patients. 12 In our study, tumors with desmoplastic growth patterns had better prognosis, similar to the results in the literature. This relationship between the histopathological growth pattern of the metastatic tumor and prognosis has been described, not only in colorectal cancers but also in malignant melanomas (uveal and skin) and breast carcinomas. 8
Five-year survival rates in metastatic colorectal cancer range from 46% to 58%. The combined use of targeted therapy methods and chemotherapeutic agents has recently led to prolongation of disease-free survival periods. 1 In the study of Buisman et al, in which histopathological growth pattern was determined as an independent predictor of survival, the 5-year total survival rate was 63.4% in patients with desmoplastic growth pattern, while it decreased to 45.9% in patients with non-desmoplastic growth pattern. 9 In our series, 53% of the patients with a desmoplastic growth pattern, 62% of the patients with pushing growth pattern, and 83% of the patients with a replacement pattern died during the follow-up period.
In addition, histopathological growth pattern has been defined not only as prognostic but also as a predictive parameter in colorectal carcinoma that metastasizes to the liver. Researchers report replacement growth patterns to be associated with resistance to systemic therapy. Moreover, the frequency of replacement growth patterns is higher in new metastatic tumors detected in patients actively receiving systemic treatment. 13
The rate of recurrence after metastasectomy is greater in tumors with replacement growth patterns compared to other growth patterns. 14 In their study investigating the relationship between histopathological growth patterns and surgical margins, Nierop et al found that the incidence of positive surgical margins in tumors with non-desmoplastic growth patterns was higher than with other growth patterns. In the same study, writers emphasized that these tumors had higher rate of positive lymph nodes in their primary resections and the preoperative carcinoembryonic antigen levels were higher (mean 14.3 µg/L). 15 In our study, serum carcinoembryonic antigen levels at the time of diagnosis did not show any correlation with histopathological growth patterns.
In patients with liver metastases, various histopathological changes can be seen both in the periphery of the tumor and in the non-neoplastic liver. Findings of intratumoral or peritumoral lymphocyte infiltration have an increasing importance in various tumors. Studies show that a high density of memory T-helper cells has a positive effect on total survival and disease-free survival rates. 16 Our study showed peritumoral lymphocytic infiltration to be more prominent in tumors with desmoplastic growth compared to non-desmoplastic growth. A similar result was obtained for the periductular reaction.
Recent studies have reported that histopathological growth patterns in metastatic liver tumors may be an important prognostic and predictive parameter. The results obtained in our study showed that the survival period was significantly shorter in patients with replacement growth patterns compared to patients with desmoplastic growth patterns. In addition, a significant relationship between growth patterns and peritumoral lymphocytic infiltration and peritumoral ductular reaction is present. In conclusion, the growth pattern of the metastatic tumor should be included in routine metastatic liver tumor pathology reports as it is shown to be effective in survival. Moreover, reporting tumor growth patterns does not require additional equipment in routine pathology practice, does not bring additional costs or take a long time to evaluate, and can be evaluated standardly by every pathologist.
Footnotes
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
Ethical Approval
Ethics approval, dated November 3, 2021 and numbered 2021-16/7, was obtained from the local Clinical Research Ethics Committee.
Informed Consent
Written informed consent was obtained for this study.
Trial Registration
Not applicable, because this article does not contain any clinical trials.
