Category: Journal

Posted on

Correlation of Lymphovascular Density with Histological Prognostic Parameters in Gastric Carcinoma

Correlation of Lymphovascular Density with Histological Prognostic Parameters in Gastric Carcinoma

*Hassan MI,1 Kabir AN,2 Poly FH,3 Ray S,4 Munmun UK,5 Mehjabin M6

 

  1. * Mohammed Imran Hassan, Assistant Professor (CC), Department of Pathology, National Institute of Diseases of Chest and Hospital, Mohakhali, Dhaka. imran.vua@gmail.com
  2. AKM Nurul Kabir, Associate Professor, Department of Pathology, BangabandhuSheikh Mujib Medical University, Shahbag,Dhaka.
  3. Ferdousi Hossain Poly, Medical Officer, OSD, DGHS, Attached- Department of Pediatric Neurology, National Institute of Neurosciences and Hospital, Dhaka.
  4. Subrata Ray, Assistant Professor, Department of Pathology, Jashore Medical College.
  5. Umme Kulsum Munmun, Lecturer, Department of Pathology, Dhaka Medical College, Dhaka.
  6. Monika Mehjabin, Lecturer, Department of Pathology, Shahid Suhrawardy Medical College, Dhaka.

*For correspondence

 Abstract

Background: Carcinoma of the stomach is a major cause of cancer mortality worldwide. Gastric cancer is actually a curable disease if it is detected at an appropriate stage and treated adequately. It rarely disseminates widely before it has involved the lymph nodes and, therefore, there is an opportunity to cure the disease prior to dissemination. Tumor-induced lymphangiogenesis plays a crucial role in metastasis and tumor progression. However, the significance of lymph vessel density has been controversial in gastric cancer. Specific lymphatic markers are now available making possible analysis of lymphatic in cancer. The aim of this study was to investigate the relation of lymph vessel density with histological prognostic factors in gastric cancer as potential indicator.
Objectives: To estimate of the lymph vessel density using Podoplanin immunostain in gastric cancer and then to correlate lymph vessel density with histological prognostic factors.
Methods: Sections from Paraffin blocks of gastric adenocarcinoma with routine stain were assessed to detect prognostic factors and then assessed with immunostain Podoplanin to count lymph vessel density. The mean number of lymph vessels in five hot spots was calculated in both intratumoral and peritumoral areas. Finally correlation was seen between histological prognostic factors and lymph vessel density in gastric carcinoma.
Results: A positive significant correlation was found between lymph node metastasis with lymph vessel density in both intratumoral and peritumoral area. Similarly, a positive significant correlation was found between tumor stage with lymph vessel density in both intratumoral and peritumoral area.

 [Journal of Histopathology and Cytopathology, 2020 Jan; 4 (1):12-22]

 Keywords: Podoplanin, Lymphovascular density, Gastric carcinoma

 Introduction

Carcinoma of the stomach is a major cause of cancer mortality worldwide. The prognosis of gastric carcinoma is poor with cure rates little better than 5–10%, although better results are obtained in Japan where the disease is common. Gastric cancer is actually a curable disease provided that it is detected at an appropriate stage and treated adequately. It rarely disseminates widely before it has involved the lymph nodes and, therefore, there is an opportunity to cure the disease prior to dissemination.1

Worldwide gastric cancer ranks fourth in frequency and second in cancer mortality rate; with a 5-year survival rate not exceeding 30% in Western countries.2 Gastric cancer incidence varies markedly with geography. In Japan, Chile, Costa Rica, and Eastern Europe the incidence is up to 20-fold higher than in North America, northern Europe, Africa, and Southeast Asia.1 The incidence of gastric carcinoma in Bangladesh is 5.2 per 100,000.3

 

Gastric cancer is a multi-factorial disease. Epidemiological studies point to a role for Helicobacter pylori, although its importance is disputed. Helicobacter pylori seem to be principally associated with carcinoma of the body of the stomach and the distal stomach rather than the proximal stomach. As Helicobacter is associated with gastritis, gastric atrophy and intestinal metaplasia, the association with malignancy is perhaps not surprising. Patients with pernicious anemia, gastric atrophy, duodenogastric reflux and reflux gastritis, intestinal metaplasia and cigarette smoking are at increased risk, as are those with gastric polyps. Diet appears to be important factor. The ingestion of substances such as spirits may induce gastritis and, in the long term, cancer. Excessive salt intake, deficiency of antioxidants and exposure to N-nitroso compounds are also implicated.1

The most useful classification of gastric cancer is the Lauren classification.4 In this system, there are principally two forms of gastric cancer: intestinal gastric cancer and diffuse gastric cancer. In intestinal gastric cancer, the tumor resembles carcinomas found elsewhere in the tubular gastrointestinal tract and forms polypoid tumors or ulcers. It probably arises in areas of intestinal metaplasia. In contrast, diffuse gastric cancer infiltrates deeply into the stomach without forming obvious mass lesions but spreading widely in the gastric wall.1

These tumors have several prognostic and predictive factors. The depth of invasion and the nodal and distant metastasis at the time of diagnosis remain the most important prognostic indicators for gastric carcinoma. Currently tumor size, nuclear grade and patient’s age are well known prognostic factors for patients with operable gastric cancer. These prognostic factors are widely used to determine whether to apply neo-adjuvant therapy in patient with gastric cancer.  In advanced cases, gastric carcinoma may first be detected as metastases to the supraclavicular sentinel lymph nodes, also called Virchow’s node. Gastric carcinoma can also metastasize to the periumbilical region to form a subcutaneous nodule, termed as Sister Mary Joseph nodule. Local invasion into the duodenum, pancreas, and retroperitoneum is also characteristic. In such cases efforts are usually focused on chemotherapy or radiation therapy and palliative care. However, surgical resection remains the preferred treatment for gastric carcinoma.5

Lymphangiogenesis is a critical process for tumor growth, invasion and metastasis.  That is why measurement of lymph vessel density may be clinically important in gastric cancer specimens. The induction of lymphangiogenesis by a tumor is controlled process, influenced by lymphangiogenic factors, which involve a complex interaction between tumor and endothelial cells.6 The pathologic approach to assess lymphangiogenesis involves microscopic estimation of lymph vessel density by using endothelial markers in immunohistochemistry.

On previous studies detection of lymphangiogenesis was limited due to lack of specific lymphatic endothelial markers. The recent discovery of anti-Podoplanin antibody has a major impact on lymphatic studies. Podoplanin has become a valuable marker for detecting lymph vessels and for identifying lymph vessel invasion in tumor.7

Podoplanin is a specific marker of the lymphatic endothelium and is not expressed in blood vessels. It is expressed by both developing and mature lymphatic endothelial cells and seems to be a more specific marker of lymphatic endothelial cells.  By electron microscopy and immunoelectron microscopy, it was demonstrated that Podoplanin is mainly expressed on the luminal surface of lymphatic endothelial cells.  Peritumoral (PT) lymphatic vessels are larger and more irregular than the intratumoral (IT) lymphatic’s, with a significantly lower density. Intratumoral lymphatic vessels are found in a large variety of tumors and are usually small, flattened and irregular. It was suggested that Podoplanin has potential role in invasion and metastasis of tumor. This hypothesis is mainly based on the observation that high expression of Podoplanin is consistently correlated with the presence of metastases. It was reported that Podoplanin expressing cells were found at the invasion.8 The use of an anti-Podoplanin-based therapeutic strategy could be suggested in the treatment of lymphatic metastases based on three considerations: (i) Podoplanin is a well-known marker of lymphatic endothelial cells; (ii) its expression seems to be associated with bad prognosis and high risk for lymph node metastases; (iii) it is involved in tumor invasion. In conclusion, Podoplanin is a sensitive marker of lymphatic endothelial cells and is very useful in evaluating lymphatic micro vessel density. Immunohistochemical detection of Podoplanin is helpful in the diagnosis of lymphovascular invasion.8 Some studies have suggested that lymph vessel density is associated with an increased risk of lymph node metastasis;9 however, this conclusion is not supported by all of the published studies.10

 

 

 

 

 

 

 

 

 

 

 

The study was conducted a) to investigate morphological prognostic factors of gastric adenocarcinoma, b) to estimate lymph vessel density by using immunostain Podoplanin and c) to correlate lymph vessel density with prognostic factors.

 Methods

It was a cross sectional observational study carried out from March, 2016 to July, 2018at the Department of Pathology, Bangabandhu Sheikh Mujib Medical University (BSMMU), Dhaka, Bangladesh. The study population was paraffin blocks of gastric adenocarcinoma from pathological laboratory in the same Institute with consecutive convenience sampling. Inclusion Criteria was paraffin blocks of gastric samples diagnosed as gastric adenocarcinoma in pathology laboratory and Exclusion criteria were with history of treated gastric carcinoma cases like neo adjuvant therapy, blocks containing autolyzed tissue, extensive necrosis, hemorrhage and desmoplasia, blocks having no tissue or damaged blocks and blocks containing other cancers like lymphoma, carcinoid tumor, malignant GIST etc. The sample size was 53. Monoclonal Anti-Human Podoplanin antibody, clone D2-40, Ready to use (code- IR072) was used for detecting lymph vessel. In this study, sections of normal vermiform appendix were taken as positive control (as recommended by Dako). Podoplanin antigen expression was detected by immunohistochemistry in the submitted blocks (paraffin blocks with maximum tumor bulk were chosen) and were performed using DAKO Cytomation at immunohistochemistry laboratory, department Pathology, BSMMU.

Histopathological categorization of tumor and grading of all the cases were done. Lymphovascular invasion was recorded. Each lymph node was histologically examined to determine metastasis. Numbers of lymph vessels were counted by Weidner’s method in peritumoral (PT) and intratumoral (IT) areas.11

 Micro vessel Quantification

Determination of lymph vessel density (LVD) were performed according to Weidner et al.11Brown staining of cytoplasmic membrane of endothelial cells of lymph vessels were considered as positive reaction. Lymph vessel density was calculated according to Weidner’s method by Olympus microscope model BH51. The immunostained sections were scanned by light microscopy at low magnification (10x) and the areas of tissue with the greatest numbers of distinctly highlighted lymph vessels (“hot spots”) were selected. So, “hot spot” is an area with the most dense lymph vessels growth. Five hot spots were selected in intratumoral (IT) and peritumoral (PT) areas. When the hot spots were defined, lymph vessel count were performed by counting the individual stained lymph vessel (at power 20x) representing a field size of 0.74mm2 (i.e. – 20x objective lens, 10x ocular lens; equivalent to 0.7386 mm2 per 200x field).11 The mean number of lymph vessels in each case were determined and considered as the lymph vessel density.7So, at first five hot spots were chosen in intratumoral (IT) and peritumoral (PT) area. In each hot spot, lymph vessel count was performed at power x20. Finally lymph vessel density was calculated as the mean of the total number of lymph vessels in these five hot spots. Lymph vessel density (LVD) assessed in this manner may be not equivalent to LVD seen in multiple areas of multiple samples of the stomach; however, it is well accepted that these hot spots, not arbitrary areas, are representative of the entire tumor. Although there is some variability in the distribution of lymph vessels within the layers of the gastric wall, the five ‘hot spot’ areas were chosen to obtain an objective assessment and to avoid observer variation.7

Results

A total number of fifty three diagnosed cases of gastric adenocarcinoma were taken for this study. Forty six samples were partial/subtotal gastrectomy specimen and seven samples were total gastrectomy specimen. All fifty three samples contain epigastric lymph nodes ranging one to twenty five. Thirty seven cases had lymph node metastasis and the rest sixteen were free of tumor metastasis. Thirty seven cases were LVI positive in H&E routine stain and forty two cases were LVI positive with Podoplaninimmunostain.

In the group of 53 patients with gastric carcinoma, mean patients’ age was 49.87±11.38 years (range 20-70).33 patients ware male and 20 patients ware female with a male/female ratio of 1.6:1. According to Lauren’s classification, 30 cases ware diffuse type and 23 cases were intestinal type. Two third (60.4%) samples belonged to tumor size ≤ 5 cm. The mean tumor size was 5.61±1.92 cm, ranged from (3.0 – 11.0) cm. Lymph vessel density (LVD) was more frequent in cases having tumor size > 5 cm in diameter.

The histological grade was assessed according to WHO grading system. 33(62.3%) samples belonged to grade-III (poorly differentiation), followed by grade-II (moderate differentiation) (24.5%) and grade-I (well differentiation)(13.2%). Lymph vessel density (LVD) was most frequent in grade- II group.

Cases were grouped according to depth of invasion (tumor stage) as T1, T2, T3 and T4. 41.5% samples belonged to subserosal invasion (T3). T3 was followed by T2(32.1%), T4 (20.8%) and T1(5.7%). Lymph vessel density(LVD) was most frequent in T3 group.

Numbers of lymph node ranged from 1-25. Cases were grouped according to the numbers of lymph node metastasis as N0, N1, N2 and N3. 30.2% samples had no lymph node metastasis (nodal stage N0). 28.3% samples belonged to nodal stage N1, followed by nodal stage N2 (20.8%) and N3 (20.8%). Nodal stages were increasing with increase of Lymph vessel density (LVD).

Lymph vessels in intratumoral (IT) area ranged from 1-12 and peritumoral (PT) area ranged 2-11. Mean LVD in IT area was 5.93 ±2.79 cm and mean LVD in PT area was 5.50±2.38 cm.

A positive correlation was found between Lymph vessel density(LVD) and tumor stage in intratumoral (IT) area (Figure 7).

The value of Pearson’s correlation coefficient was 0.377 and it was significant (p=0.005). Therefore, there was linear association between tumor stage and lymph vessel density (LDV) in intratumoral (IT) area with Podoplanin stain.

A positive correlation was found between Lymph vessel density(LVD) and tumor stage in peritumoral (PT) area (Figure 8).  The value of Pearson’s correlation coefficient was 0.334 and it was significant (p=0.014). Therefore, there was linear association between tumor stage and lymph vessel density (LVD) in peritumoral (PT) area with Podoplanin stain.

A positive correlation was found between Lymph vessel density (LVD) and Lymph node metastases inintratumoral (IT) area (Figure 9).

The value of Pearson’s correlation coefficient was 0.855 and it was significant (p=0.000). Therefore, there was linear association between lymph node metastasis and lymph vessel density (LVD) in intratumoral (IT) area with Podoplanin stain.

A positive correlation was found between Lymph vessel density (LVD) and Lymph node metastases in peritumoral (PT) area (Figure 10).

The value of Pearson’s correlation coefficient was 0.791 and it was significant (p=0.000). Therefore, there was linear association between lymph node metastasis and lymph vessel density (LVD) in peritumoral (PT) area with Podoplanin stain.

A positive correlation was found between LVD in IT and PT.

The value of Pearson’s correlation coefficient was 0.937 and it was significant (p=0.000). Therefore, there was linear association between LVD in intratumoral (IT) and peritumoral (PT) area with Podoplanin stain.

Discussion

Gastric carcinoma is the most common leading tumor of the world. Although great efforts have been made in the field of early diagnosis and adjuvant therapy, the incidence and overall mortality of gastric carcinoma continues to increase. Since gastric tumor cells commonly infiltrate into the lymphatic system, lymph node status is routinely used to identify a patient’s prognosis, tumor stage, and treatment modality. Inhibition of tumor cells to lymph node metastasis (LNM) is a promising way to prevent distant metastasis. However, the relationship between lymphangiogenesis, lymphovascular invasion and lymph node metastasis (LNM) remains ambiguous.

This cross sectional study was carried out with an aim to estimate the density of expression of Podoplanin in lymphangiogenic vessel in gastric adenocarcinoma and to see the correlation between the percentages of area covered by lymphangiogenic vessels with prognostic factors. The present study findings were discussed and compared with previously published relevant studies.

In this study, 62.3% sample had grade-III (poorly differentiation) followed by 24.5% grade-II (moderately differentiation) and 13.2 % grade-I (well differentiation). Significant relation was found between lymph vessel density-intratumoral (LVD-IT) area (p=0.013) and lymph vessel density-peritumoral (LVD-PT) area (p= 0.006) with different histological differentiation. In their study, Pak et al., 201512 found that 53.3 % sample had grade-III and significant relation (p= 0.021) between LVD-IT area with histological differentiation, which is comparable with the current study.

 

It was observed that 41.5 % of samples belonged to T3, followed by 32.1% T2, 20.8% T4  and 5.7% T1 and significant relation found between LVD-IT area (p=0.019) and LVD-PT area (p=0.041) with different tumor stage. Gresta, Rodrigues Jr and Cabral, 201413 found 63.5% cases were T3, followed by 19.2% T2, 13.4% T1 and 3.8% T4. Pak et al., 2015 found 39.3% samples were T3 followed by 16.6 % T2 and T4. They also found significant association between LVD-IT area (p=0.024) with different tumor stage. Raica et al., 20088 found significant correlation between LVD with tumor stage (p<0.002) and Nakamura et al., 20067 found significant positive correlation between LVD with tumor stage (p=0.0008). So, present study result is consistent with previous study.

According to number of lymph node metastasis (LNM), it was observed that 30.2% sample belonged to nodal stage 0 (N0), followed by 28.3% stage-I (N1), 20.8% stage II (N2) and stage III (N3). There was significant relation was found between LVD-IT area (p<0.001) and LVD-PT area (p<0.001) with nodal status. Yan et al., 200814 reported that LVD was obviously higher in the colorectal cancer sample with metastasis (12.08 ± 4.96) than in those without (8.26 ± 4.08) (p<0.001). Schoppmann et al., 200115 demonstrated that lymph vascular invasion (LVI), assessed by anti-Podoplanin immunostain, has been strongly associated with presence of lymph node metastasis. Pak et al. 201512 found LVD-IT area was a higher than LVD-PT area (12.29± 4.36 vs. 11.01±3.62), Nodal stage N0 cases were more frequent and significant relation between LVD-PT in area (p=0.040) with nodal status. Nakamura et al., 20067 found significant positive correlation between LVD with nodal metastasis (p= 0.0094). Raica et al., 20088 found significant correlation between LVD with nodal metastasis (p<0.031) which is comparable with current study.

In this study, there was a positive significant Pearson’s correlation (r= 0.855, p= 0.000) was found between lymph node (LN) metastasis with LVD in IT area. Similarly, there was a positive significant Pearson’s correlation (r= 0.791 p= 0.000) was found between LN metastasis with LVD in PT area. Pak et al., 201512 found in their study that LVD in PT area was significantly associated with LN metastasis (p=0.040). Nakamura et al., 20067 found significant correlation between LN metastasis with LVD (p= 0.0094). Raica et al., 20088 reported that there was significant positive correlation between both LVD in IT and PT areas with LN metastasis (P<0.031). So current study result is consistent with previous study.

In this current study, there was a positive significant Pearson’s correlation (r= 0.377, p= 0.005) was found between tumor stages with LVD in IT area. Similarly, there was a positive significant Pearson’s correlation (r= 0.334, p= 0.014) was found between tumor stage with LVD in PT area. Nakamura et al., 20067 found significant positive correlation between tumor stage with LVD (p= 0.0008). Raica et al., 20088 reported that there was significant positive correlation between both LVD-IT area and LVD-PT area with tumor stage (P<0.02).

In this current study, there was a positive significant Pearson’s correlation (r= 0.937, p= 0.000) was found between LVD-IT areas with LVD-PT areas. Pak et al., 201512 found significant correlation (p=0.028) between LVD-IT area with LVD-PT area.

Conclusion

Total fifty three cases of gastric adenocarcinoma were examined for lymphangiogenic vessels in peritumoral (PT) and intratumoral (IT) areas. Thirty seven cases had lymph node metastasis and the rest sixteen were free of tumor metastasis. Both peritumoral and intratumoral lymphangiogenic vessel count (density) stained by Podoplanin antibody correlated with lymph node metastasis and tumor stage. Lymphatic vessel count is more in the intratumoral area.  The specific lymphatic endothelial marker Podoplanin proved to be a valuable tool in highlighting lymph vessel density (LVD) and lymphovascular invasion, and therefore a predictor of lymph node metastasis.

Limitation 

The study population was selected from one institute in Dhaka city, so that the results may not be reflect the exact picture of the country.  The study period was short. Sample size was small. Podoplanin was only marker used to detect lymph vessels. Sometimes differentiation between blood vessels and lymph vessels were difficult in tissue sections. Differentiation between blood vessels and lymph vessels was also a limitation of this study.  Use of blood vessel immunostain like factor VIIIa, CD31, and CD 34 etc. could have overcome the limitation and overall no control was used.

 

References

  1. Williams NS, Bulstrode CJ, O’connell PR. Bailey & Love’s short practice of surgery. Crc Press; 2008. pp.450-456.
  2. Carl-McGrath S, Ebert M, Röcken C. Gastric adenocarcinoma: epidemiology, pathology and pathogenesis. Cancer therapy. 2007; 5(2):877-94.
  3. Rahman R, Asombang AW, Ibdah JA. Characteristics of gastric cancer in Asia. World journal of gastroenterology: WJG. 2014;20(16):4483.
  4. Lauren P. The two histological main types of gastric carcinoma: diffuse and so called intestinal type carcinoma: an attempt at a histoclinical classification. Acta Pathologica Microbiologica Scandinavica. 1965; 64(1):31-49.
  5. Kumar V, Abbas AK, Fausto N, Aster JC. Robbins and Cotran pathologic basis of disease, professional edition e-book. elsevier health sciences. 2014 Aug 27. pp.760-777.
  6. Joo YE, Sohn YH, Joo SY, Lee WS, Min SW, Park CH, Rew JS, Choi SK, Park CS, Kim YJ, Kim SJ. The role of vascular endothelial growth factor (VEGF) and p53 status for angiogenesis in gastric cancer. The Korean journal of internal medicine. 2002; 17(4):211.
  7. Nakamura Y, Yasuoka H, Tsujimoto M, Kurozumi K, Nakahara M, Nakao K, Kakudo K. Importance of lymph vessels in gastric cancer: a prognostic indicator in general and a predictor for lymph node metastasis in early stage cancer. Journal of clinical pathology. 2006; 59(1):77-82.
  8. Raica M, Ribatti D, Mogoanta L, Cimpean AM, Ioanovici S. Podoplanin expression in advanced-stage gastric carcinoma and prognostic value of lymphatic micro vessel density. Neoplasma. 2008;55(5):455-60.
  9. El-Gohary YM, Metwally G, Saad RS, Robinson MJ, Mesko T, Poppiti RJ. Prognostic significance of intratumoral and peritumoral lymphatic density and blood vessel density in invasive breast carcinomas. American journal of clinical pathology. 2008;129(4):578-86.
  10. Zhao YC, Ni XJ, Li Y, Dai M, Yuan ZX, Zhu YY, Luo CY. Peritumorallymphangiogenesis induced by vascular endothelial growth factor C and D promotes lymph node metastasis in breast cancer patients. World journal of surgical oncology. 2012;10(1):165.
  11. Weidner N, Folkman J, Pozza F, Bevilacqua P, Allred EN, Moore DH, Meli S, Gasparini G. Tumor angiogenesis: a new significant and independent prognostic indicator in early-stage breast carcinoma. JNCI: Journal of the National Cancer Institute. 1992;84(24):1875-87.
  12. Pak KH, Jo A, Choi HJ, Choi Y, Kim H, Cheong JH. The different role of intratumoral and peritumoral lymphangiogenesis in gastric cancer progression and prognosis. BMC cancer. 2015; 15(1):498.
  13. Gresta LT, Júnior IA, Cabral MM. Micro vessel density quantification in gastric cancer: comparing methods for standard measures. J Cancer SciTher. 2014;6(10):401-5.
  14. Yan G, Zhou XY, Cai SJ, Zhang GH, Peng JJ, Du X. Lymphangiogenic and angiogenicmicrovessel density in human primary sporadic colorectal carcinoma. World Journal of Gastroenterology: WJG. 2008; 14(1):101.
  15. Schoppmann SF, Birner P, Studer P, Breiteneder-Geleff S. Lymphatic microvessel density and lymphovascular invasion assessed by anti-podoplanin immunostain in human breast cancer. Anticancer research. 2001; 21(4A):2351-5.
Posted on

Malignant Solid Childhood Tumors: Morphological Study in a Tertiary Hospital

Original Contribution

Malignant Solid Childhood Tumors: Morphological Study in a Tertiary Hospital

 *Rahman SS,1 Hossain MI,2 Nasreen S,3 Ahamad MU,4 Bhattecharjee P,5 Rahman Z,6 Khan MA7

 *Dr. Sharmin Sultana Rahman, Assistant Professor, Department of Pathology, Tairunnessa Memorial Medical College, Gazipur, Bangladesh. path@gmail.com

  1. Mohammad Ismail Hossain, Lecturer, Department of Pathology, Chittagong Medical College, Chattogram, Bangladesh. ismail.tushar@gmail.com.
  2. Sayeeda Nasreen, Assistant Professor, Department of Pathology, Chittagong Medical College, Chattogram, Bangladesh.
  3. M. Shahabuddin Ahamad, Associate Professor, Department of Pathology, Chittagong Medical College, Chattogram, Bangladesh.
  4. Pradip Bhattecharjee, Associate Professor, Department of Pathology, Chittagong Medical College, Chattogram, Bangladesh, 4203.
  5. Zillur Rahman, Professor, Department of Pathology, Chittagong Medical College, Chattogram, Bangladesh, 4203.
  6. Md. Morshed Alam Khan, Public Health Graduate. West Chester University of Pennsylvania, USA.

*For correspondence

Abstract

Background: Malignancy is the second most common childhood mortality in the developed world. The types of cancer that develop in children are often different  types that develop in adult with respect to incidence, type of tumor, underlying familial or genetic aberration and tendency to regress spontaneously or cytodifferentiation. In recent years, identification of specific genes, oncogenes, tumor markers and other biological and pathological factors have played an important role in staging and classifying risk categorization of specific tumors as low, intermediate and high-risk lesions.

Objectives: This study was done to evaluate the incidence and morphological patterns of solid malignant tumors in children.

Method: This hospital based cross-sectional descriptive study was conducted the Department of Pathology, Chattogram Medical College, Chattogram referred from Department of Pediatric Surgery, Chattogram Medical College Hospital, Chattogram. All the cases were subjected to examine histopathological slides with haematoxylin and eosin stain, and IHC done in malignant small round cell tumors and few tumors cases.
Results: An analysis of 43 cases of childhood solid malignant tumors, over a period of a year, was done. The study found that 28 out of 43cases of malignant tumors were seen in boys, whereas 15 casesin girls. The commonest tumor was lymphoma (11 cases) with non-hodgkin’s predominance (10 cases), followed by nephroblastoma (10 cases). In addition, the common age group was found 0-4 years (22 cases), with male predominance.
Conclusion: Histopathological diagnosis could be confirmed with IHC in cases B cell non-hodgkin lymphoma, infantile fibrosarcoma, malignant fibrous histocytoma, Ewing’s sarcoma / PNET. The frequency of tumors and their distribution was comparable to that report from other studies.

[Journal of Histopathology and Cytopathology, 2020 Jan; 4 (1):3-11]

 Keywords: Morphology, Malignant, Childhood, Solid tumour.

 

 Introduction

Malignancy is the 2nd most common cause of childhood mortality in the developing world 1st being malnutrition and infection. About 1/650 children develops malignancy before their 15thbirthday.1

In Bangladesh, there are no perfect cancer registries. Childhood cancers are expected to be high in Bangladesh because of the young population structure – about 30% (47.4 million) of the population is under 15 years old.2

Common childhood malignant tumor (almost 80%) arises from haemopoietic elements, lymphnodes, bones and soft tissues. The childhood cancer often the result of DNA changes in the cells that take places very early in life, sometimes even before birth. A few environmental factors, such as radiation exposure, have been linked with some types of childhood cancer. Some children inherit DNA mutation from parents that increases their risk of certain types of cancer. The reason for DNA changes that cause most childhood cancers are not known.3

Childhood malignancies differ biologically and histopathologically from those of adults with respect to incidence, type of tumor, underlying familial or genetic aberration and tendency to regress spontaneously or cytodifferentiate4. In recent years, identifications of specific gene, oncogenes, tumor markers and other biological and pathological factors have played an important role in staging and risk categorization of specific tumors. Hence there is a need for accurate histopathological reporting in conjugation with ancillary methods.5 Histological type is important for understanding etiology and progression of disease.

This study aims to provide a comprehensive recent overview on childhood cancers in the Department of Pathology, Chittagong Medical College, Chattogram which would contribute to the understanding of characteristics and provide a basis for the future strategies to deal with childhood cancer.

Method

It is a hospital based cross-sectional descriptive study carried out in the Department of Pathology, Chittagong Medical College, Chattogram over a one year period from 1st January 2017 to 31st December, 2017.

Children of 0-14 years old age diagnosed as malignant tumor, who had undergone operative treatment in the Department of Pediatric Surgery, were included in this study. Patients diagnosed as a case of malignant solid tumor was subjected to detail clinical history and thorough physical examination followed by relevant investigations.

All the specimens sent from Department of Pediatric Surgery were evaluated and stained by haematoxylin and eosin in all cases and immunohistochemistry (IHC) done in malignant small round cell tumors and (few tumors cases). All the necessary and relevant data regarding patients were recorded methodically in pre-designed data sheet.

 Statistical analysis

The data were collected from the filled data sheet and statistical analysis was done using the SPSS (Statistic Package for Social Science) Version-20 software package for windows.

 Result

Age incidence ranged from 0-14 years. Patients were divided into 3 age groups. The mean age was 6 years 9 months. 28 (65.1%) cases were seen in boys and 15 (34.9%) cases were seen in girls showing male preponderance. Male and female ratio: 1.87:1.

Lymphoma is the commonest malignant tumor and it constituted 25.3% of all tumors. 10 cases were Non-Hodgkin’s lymphoma and 01 case is Hodgkin lymphoma. Mean age is 8 years. Male is more prevalent than female with male female ratio 9:1, more frequently found in 0-4years age group. Immunohistochemistry CD20 and CD3 are done in selective cases. 6 cases of Non-Hodgkin’s lymphoma were B cell origin, CD20 positive and one case was T-cell origin which is CD3 positive. Hodgkin’s lymphoma was CD30 negative. 8 cases arose from lymph nodes and 2 cases were soft tissue origin. One of them was diagnosed as a case of malignant small round cell tumor.

Among the renal tumors, only Wilm’s tumor (10 cases) was found, with classical triphasic tumors were more prominent feature. The mean age was 4 years 7 months. Male to female ratio was 1:1.

Among the germ cell tumors, 4 cases were noted as immature teratoma: 3 of them arose from sacro-coccygeal region and 1 from ovary; 2 cases were dysgerminoma and 2 cases were yolk sac tumor; 1 case was extra gonadal origin. Female predominance with male female ratio 1:2.5. Mean age was 5 years 7 months.

 

In this study, four cases were diagnosed as malignant small round cell tumor. For further evaluation, CD20, Myogenin, Desmin, Neuron-specific enulose (NSE), Synaptophysin markers were done. 2 cases were diagnosed as B cell Non-Hodgkin lymphoma, 1 case is desmoplastic malignant small round cell tumor and other was neuroblastoma.

 

Soft tissue sarcoma comprised 9.3% (4 cases). Mean age was 8 years 4 months, with male female ratio 3:1. 2 cases were histologically diagnosed as infantile fibrosarcoma and malignant fibrous histocytoma. Infantile fibrosarcoma was found, six months old, male baby, arising from abdomen. Malignant fibrous histocytoma was noted in male, 14 years old. Immunohistochemically both cases showed strong positive for vimentin. A case of desmoplastic small round cell tumor also found. Histologically diagnosed as a case of malignant small round cells arising from intra abdominal cavity and confirmed by desmin. 1 case was PNET/ Ewings sarcoma family. For further evaluation CD99 was recommended.

Colorectal adenocarcinoma accounts for 7% (3 cases) of pediatric malignant tumors in the present study and commonly seen in 10-14years age group. Mean age is 11 years 7 months.

Bone tumor constituted 7% (3 cases): 1 case was metastatic osteosarcoma on the left chest wall; 1 case was Ewing’s sarcoma/PNET family; another was chondrosarcoma. Male were predominant with male female ratio 2:1. Immunophenotyping CD99 and S100 were applied for confirming the diagnosis in respectively Ewing’s sarcoma and chondrosarcoma.

 

 

 

 

 

Discussion

Most of the childhood malignant tumors occur below the age of eight years, although wide age variability exists in children.6 In present study, 14 years was considered as the pediatric age with infancy as a separate age group. The peak occurrence of tumors was found in 0-4 year’s age.

 

 

The overall incidence of malignant tumors of childhood was more in male (M: F=1.53:1) patients. This observation has been made uniformly in literature by author. Male predominance is noted in all age groups and female predominance in germ cell tumor by Lee et al10 and Miler  et al,11 In present study, male predominant in all age group of tumor except germ cell tumors. Male and female ratio: 1.86:1.

 

The commonest tumor comprised of lymphoma (11cases, 25.3%). Similar results were reported by Sharma et al,13 and Baneerjee et al.14 NHL was more frequent than Hodgkin. Non-Hodgkin lymphoma represented 23.3% whereas Hodgkin 2.3%. Non-Hodgkin lymphoma commonly expressed in 0-4 years age group in this study. Patients presented with swelling, pain, fever, and loss of weight with the duration ranging from one month to year. IHC CD20 was applied in 6 cases and revealed B cell origin Non-Hodgkin lymphoma.

 

Wilm’s tumor is common renal tumour that is 100% as compared to the 78.4% by Louisa et al. This difference may be due to small sample size in present study. Male and female ratio is 1:1. Whereas in the study of Paul et al,6 11 infantile Wilm’s tumor found the male to female ratio was 2.3:1. Husain et al.2found that Wilm’s tumor was slightly more common in girls in whom it tends to present at an older age.

 

Malignant germ cell tumors in the ovaries of very young children are rare in condition1. Weinblatt and Ortega, 1982 noted dysgerminoma as the commonest tumor.16 In this study, 7 cases were malignant germ cell tumors. Among them 3 cases were gonadal origins. Rest of them arose from extragonadal site, and two cases of dysgerminoma in ovary were documented. It is more common in gonadal tumor in patient with gonadal dysgenesis.17 Tumors with syncytiotrophoblastic giant cells have the same prognosis as tumors in which they are absent.18 

 

In this study, two cases of yolk sac tumor was locates at testis and intra abdomen. Both of them had shown elevated level of AFP. One case of yolk sac tumor, in testicular origin, at the age of 8 years and other one of female patient with 45 days arising from intra abdomen.

 

Soft tissue sarcoma comprises 9.3% of pediatric malignancy. Two cases are histologically diagnosed as infantile fibrosarcoma and malignant fibrous histocytoma. Infantile fibrosarcoma is found, six months old, male baby, arising from abdomen. Malignant fibrous histocytoma is seen in male, 14 years old. Immunohistochemically both cases show strong positive for vimentin. A case of desmoplastic small round cell tumor is also found. Histologically it is diagnosed as a case of malignant small round cells arising from intra abdominal cavity and confirmed by desmin.

 

Bone tumor constitutes 7% of pediatric malignant tumors in this study: one case is metastatic osteosarcoma arising from the left chest wall; one case is Ewing’s sarcoma/PNET family; another is chondrosarcoma. Male are predomident, and male female ratio is 2:1. Immunophenotyping CD99 and S100 are applied for confirming the diagnosis in respectively Ewing’s sarcoma and chondrosarcoma.

In this study, colorectal adenocarcinoma account for 7% (3cases) of pediatric malignant tumor and commonly seen in 10- 14 years age group which is similar to the study of Tonbary et al. 2012.19 Among the three cases, tumor arises from appendix, descending colon and rectum, respectively.

In this study, four cases of malignant small round cell tumor are documented .Male to female ratio is 1:1. Out of four cases of small round cell tumor, three cases arose from intra abdominal mass and one case from axillary swelling. For further evaluation of immunohistochemistry of CD20, CD3, Myogenin, Desmin, Vimentin, Neuron-specific enulose (NSE), Chromogranin, Synaptophysin markers are done. Two cases were diagnosed as B cell Non- Hodgkin lymphoma as well as one case was diagnosed as desmoplastic malignant small round cell tumor and other was as neuroblastoma.

In this study, one case is noted as anaplastic astrocytoma; Pilocytic astrocytoma, the most common type of astrocytoma in children, is a low grade tumor that typically arises in the cerebellum. One case of invasive squamous cell carcinoma is documented in the study, arose from retro-molar mass, male, 10 years old patient. No predisposing factor was elicited.

 Because of unwanted technical fault like delayed cold ischemic time and fixative time, negative result of IHC was found.

 Conclusion

In this study, histopathological diagnosis could be confirmed with IHC in following cases: B cell non-hodgkin lymphoma, infantile fibrosarcoma, malignant fibrous histocytoma, Ewing’s sarcoma / PNET. The frequency of tumors and their distribution is comparable to that report from other studies. The early onset and the embryonal nature of the major pediatric tumors suggest a prenatal origin and role of genetic factors. Infection, exposure to drugs and chemicals during pregnancy are other contributory factors. Accurate incidence of data is important in the planning and evaluation of clinical trial. Documentation of cases, advanced diagnostic methods like IHC, cytogenetic studies and treatment modalities with close follow up is needed to achieve better statistical evaluation of the problem.

 Acknowledgement

We express our heartiest gratitude to the staffs of the Department of Pathology of Chittagong Medical College, Chattogram.

 References

  1. Vandana G, Maruti D. The spectrum of malignant solid childhood tumors in the age group of 0-12 Years. Pediatrics and Neonatal Nursing. 2015 Oct 29; 2(3):85-90. Available from: http://dx.doi.org/10.17140/PNNOJ-2-114
  2. Hossain MS, Begum M, Mian MM, Ferdous SK, Sarker HK, Karim Sabina, et al. Epidemiology of childhood and adolescent cancer in Bangladesh, 2001–2014. BMC Cancer. 2016 Feb 15; 16 (104). Available from: DOI 10.1186/s12885-016-2161-0
  3. Annual cancer facts and figures, 2014. Special section: Cancer in children and adolescents. American Cancer Society. 2014. Available from: https://www.cancer.org › research › annual-cancer-facts-and-figures › 2014
  4. Nagaraja CT, Patil RB, Ugrappa G, Krishnamurthy R. Solid Malignant Tumors of Infancy and Childhood: A Histopathological Study. Scholars Journal of Applied Medical Sciences (SJAMS). 2015; 3(1C):167-174. Available from: https://pdfs.semanticscholar.org/8281/a1abfe44e013e5994218d9e032dccba26c50.pdf
  5. Variend S. Small cell tumors in childhood a review. Journal of Pathology. 1985; 45: 1-25.
  6. Paul V, Bagga A. Ghai Essential paediatrics. 8th ed. New Delhi: CBS publishers; 2013.
  7. Dawani GP, Tandon PL, Ghooi AM, Jain PK. Malignant tumors of infancy and childhood. Indian Journal of Surgery. 1972; 34: 460-468.
  8. Jussawala DJ, Yeole BB. Childhood cancer in greater Bombay.1973-84. Indian J Cancer. 1988; 25:197-206.
  9. Jain KK, Mathur GP, Srivastava JR. Malignancy in childhood. A clinico-pathological study of 45 cases. Ind J Paediatrics. 1975; 42(326):61-66.
  10. Lee CK, Lee SK. Malignant solid tumors of infancy and childhood in Korea. GANN Monograph on cancer research. Tokyo: University of Tokyo Press, 1976.
  11. Miller RW, Young JL, Novakovic B. Childhood cancer. Cancer. 1994; 75: 395-405.
  12. Sebastian O, Hyginus E, Boniface I, et al. The burden of pediatric malignant solid tumors in a developing country. Journal Tropical Pediatrics. 2010; 56(2): 111-114. Available from: doi:10.1093/tropej/fmp075
  13. Sharma S, Mishra K, Agarwal S, Khanna G. Solid tumors of childhood: The Indian Journal of Pediatrics. 2004; 71(6): 501-504.
  14. Baneerjee CK, Walia BNS. Pattern of neoplasms in childhood. The Indian Journal of Pediatrics. 1986; 53: 93-97.
  15. Louisa P, Durrane T, Suhail M, Irshad N, Zafar N, Sheema H. Clinicopathological profile of Wilms tumor. The Indian Journal of Pediatrics. 2000; 67(10): 765-767. Available from: doi: 10.1007/BF02723937
  16. Weinblatt ME and Ortega JA. Treatment of children with dysgerminoma of the ovary.Cancer. 1982 Jun 15; 49(12):2608-2611. Available from: https://doi.org/10.1002/1097-0142(19820615)49:12<2608::AID-CNCR2820491233>3.0.CO;2-0
  17. Kota SK, Gayatri K, Pani JP, Kota SK, Meher LK, Modi KD. Dysgerminoma in a female with turner syndrome and Y chromosome material: A casebased review of literature. Indian Journal of Endocrinology and Metabolism., 2012 May-June; 16(3): 436–440. Available from: http://www.ijem.in/temp/Indian J EndocrMetab163436-1232556_032525.pdf
  18. Fletcher CD. Germ cell tumors. In Diagnostic Histopathology of Tumors. 3rd Philadelphia: Elsevier Health Sciences; 2007. 605 p.
  19. Tonbary YA, Darwish A, Hussein AE, Fouda A. Adenocarcinoma of the colon in children: Case series and mini-review of the literature. Research Gate. 2012 November 24: 6(1): 29–33. Available from: DOI: http://dx.doi.org/10.1016/j.hemonc. 2013.02.003

 

 

 

Posted on

Editorial

Editorial

Artificial Intelligence (AI) in Histopathology: An Emerging Technology in Diagnostic Fields

*Talukder SI

Microscopic examination of tissue sections on glass slides prepared from formalin fixed paraffin embedded tissue section stained with hematoxylin and eosin is histopathological  diagnostic technique. It is an old but is the gold standard diagnostic technique till now. Newer techniques, such as immunohistochemistry is used for refining the histopathological diagnosis. Both techniques are time consuming and laborious job. Artificial intelligence (AI) is a branch of computer science that deals with building smart machine capable of performing tasks that typically require human intelligence.1 AI technology of computer science and engineering has been incorporated in tissue level diagnosis.2 This technology is applied on histological glass slides. Microscopic images of glass slides are digitized by computer. These images are presented as digital matrix and is called virtual slide. A virtual slide allows calculation and related presentation of image information that otherwise can only be seen by individual pathologist. The digital world permits attachments of several (if not all) fields of view and the contemporary visualization on a screen. The presentation of all microscopic magnifications is possible if the basic pixel resolution is less than 0.25 microns. To start digital histopathological diagnosis into the daily routine practice of a histopathologist requires a new setup of workflow arrangement and procedures. The quality of digitized images is sufficient for diagnostic purposes; however, the time needed for viewing virtual slides exceeds that of viewing original glass slides by far. It is due to a slower and more difficult sampling procedure, which is the selection of information containing fields of view. By application of artificial intelligence, Kayser et al, mentioned, histopathological diagnosis in routine work can be managed automatically in steps as follows: 1. The individual image quality has to be measured, and corrected, if necessary. 2. A diagnostic algorithm has to be applied. An algorithm has be developed, that includes both object based (object features, structures) and pixel based (texture) measures. 3. These measures serve for diagnosis classification and feedback to order additional information, for example in virtual immunohistochemical slides. 4. The measures can serve for automated image classification and detection of relevant image information by themselves without any labeling. 5. The pathologists’ duty will not be released by such a system; to the contrary, it will manage and supervise the system, i.e., just working at a “higher level”.1 Virtual slides are already in use for teaching and continuous education in anatomy and pathology. Attempts to introduce them into routine histopathology laboratory work have been reported. Application of AI has been established by automated immunohistochemical measurement systems (EAMUS, www. diagnomX.eu). The performance of automated diagnosis has been reported for a broad variety of organs at sensitivity and specificity levels > 85%.

Histopathology has undergone major changes firstly with the introduction of Immunohistochemistry, and latterly with Genomic Medicine. Salto et al (2018) argued that a third revolution,  Artificial Intelligence (AI) is coming on Digital Pathology (DP).3 The introduction of AI has the potential to both challenge traditional practice and provide a totally new realm for diagnostic pathology.

Hereby, all pathologists should accept new revolutionary technique and be ready to innovate and actively engage in the creation, application and validation of technologies and oversee the safe introduction of AI into diagnostic histopathology practice.

References

  1. What is Artificial Intelligence? https://builtin.com/artificial-intelligence
  2. Klaus Kayser, JĂźrgen GĂśrtler, Milica Bogovac, Aleksandar Bogovac, Torsten Goldmann, Ekkehard Vollmer, Gian Kayser. AI (artificial intelligence) in histopathology-from image analysis to automated diagnosis. Folia histochemica et cytobiologica 2009; 47 (3), 355-361.
  3. Salto-Tellez, M., Maxwell, P., & Hamilton, P. W. (2018). Artificial Intelligence – The Third Revolution in Pathology. Histopathology. https://doi.org/10.1111/his.13760.

*Dr. Sadequel Islam Talukder, Associate Professor, Department of Pathology Mymensingh Medical College, Mymensingh. sadequel@yahoo.com