Estimation of tumor volume and its prognostic significance to study the biological behavior of carcinoma of cervix

Introduction

Carcinoma of cervix is the most common form of cancer in women in most developing countries. It affects women of middle age group. The lesion commences at the squamocolumnar junction in the transformation zone and spreads characteristically by direct extension to the vagina, corpus, parametrium, lower urinary tract, and uterosacral ligaments. Lymph node metastasis is also common, in a sequential fashion, and is related to the stage of the disease. Prognosis of cervical carcinoma is primarily determined by clinical stage (International Federation of Gynecology and Obstetrics [FIGO]) and presence or absence of nodal metastases. Few other prognostic indicators like tumor size, depth of invasion, vascular invasion, etc., are also taken into account. Keeping in view the biological behavior of the tumor, an attempt is made to study the tumor volume and the metastatic potential of the tumor in cases of carcinoma cervix. A study on the extracellular membrane (fibronectin) and macrophages (LN-5) as a part of tumor immunity has also been done.

Materials and Methods

A prospective study was done on 40 cases of carcinoma cervix to assess a correlation between tumor volume, the spread, and metastatic potential of the tumor. The study was done on material received in the tumor pathology laboratory from patients attending the gynecology department of Safdarjang Hospital. Clinical data were also received from the gynecology department itself. X-ray findings, cystoscopic findings (in selected cases), and per operative findings regarding the extent of the tumor was also recorded in all the patients. An informed consent for a radical hysterectomy was taken from all the patients. All the patients underwent radical hysterectomy with pelvic lymph node sampling with or without para-aortic lymph node sampling. The hysterectomy specimen and the lymph nodes were examined pathologically. For the preparation of histologic slides, specimens were fixed in formalin. The cut sections were dehydrated and paraffin embedded and the blocks so obtained were cut into very thin slices, placed in glass slides, and stained with hematoxylin and eosin.

Measurement of cervical tumor was done systematically. The specimen was cut longitudinally on the anterior surface from the fundus of the uterus to the external os of the cervix. The maximum limit of the grossly visible tumor was measured with a pair of calipers. Maximum horizontal and vertical distance was measured at three levels. Each specimen was cut into 4–12 sections. According to the size of the tumor mass, tissue was processed from each section. Cross-section slice was seen for depth of malignant infiltration. The surface area of each cross-section was calculated. The total volume (TV) of each section was calculated by multiplying the surface area by the thickness of each section. Finally, the total TV was calculated by adding the sectional volumes.

TV = π/6 × Length × Width × Depth.

Immunohistochemistry was done to detect fibronectin production by the tumor, and LN-5 to identify the macrophages in the inflammatory infiltrate in and around the tumor cells.

Results

A total of 40 patients attending the gynecology Department of Safdarjang Hospital within the diagnostic group of carcinoma cervix clinical stage less than or equal to IIb were included in the study group. The median age of the patients in this group was 47.5 years, with a range of 30–80 years Table 1 and Graph 1.

Close

Graph 1: Age distribution

Export to PPT

Majority of the patients attained menarche after 12 years (85%). A total of 55% had attained menopause at the time of presentation. 65% of the patients had 1–5 children. There was a positive history of tobacco smoking in 24 patients of which 6 smoked themselves.

The most common complaint in these patients was bleeding per vaginum. Other complaints were discharge per vaginum, postcoital bleeding. Clinical reports of the 40 patients revealed that 22 of them had tumor size more than 3 cm in diameter. The various parameters studied were the diameter of the tumor, vaginal involvement, and parametrial involvement. Finally, clinical staging was evaluated depending on these parameters Table 2.

Grossly, the tumor was seen restricted to the transition zone Figure 1, or a fungating tumor extending to the isthmus Figure 2, or an ulceroproliferative lesion extending to the vaginal cuff Figure 3. The major histological type of carcinoma among 40 cases is squamous cell carcinoma (SCC) (in 90% of cases) Figure 4, and 10% had adenocarcinoma Figure 5. Histopathological examination of the specimens revealed vaginal cuff involvement in 50% of cases. Parametrial and bladder involvement was also seen among few cases. Pathological staging of the carcinoma cervix showed stage Ib, IIa, IIb, and IVa (35%, 20%, 40%, and 5%) Table 3.

Close

Figure 1

Gross specimen showing cervical carcinoma restricted to the transition zone

Export to PPT

Close

Figure 2

Gross hysterectomy specimen showing fungating tumor extending to the isthmus

Export to PPT

Close

Figure 3

Gross hysterectomy specimen showing ulceroproliferative lesion extending to the vaginal cuff

Export to PPT

Close

Figure 4

H and E section showing well differentiated squamous cell carcinoma cervix

Export to PPT

Close

Figure 5

H and E section showing adenocarcinoma cervix

Export to PPT

Different tumors with different staging have been seen to show different tumor volume. Both cases with IVa staging showed tumor volume more than 15,000 cumm Table 4.

Tumor extension into the uterus was observed in 20 specimens. Tumor deposits in the parametrial tissue Figure 6 and serosal aspect of the fallopian tube Figure 7 was seen in few cases. Tumor spread to pelvic lymph nodes was observed in 12 cases, obturator group being the most commonly involved Figure 8. Tumor volume estimated on pathological specimens of 40 cases ranged from 230 cumm to 49,760 cumm with a mean of 14,844 cumm.

Close

Figure 6

H and E section showing tumor deposits in the parametrial tissue

Export to PPT

Close

Figure 7

H and E section showing tumor deposits in the serosal aspect of the fallopian tube

Export to PPT

Close

Figure 8

H and E section showing tumor deposits in the lymph node

Export to PPT

Correlation between tumor volume, lymph node metastasis, vaginal involvement, parametrial involvement, uterine extension of tumor, and tumor differentiation

12 (30%) cases had tumor volume more than 15,000 cumm, 12 (30%) cases had tumor volume < 5000 cumm, 16 (40%) cases had tumor volume between 5000 and 15,000 cumm Graph 2.

Close

Graph 2: Tumor volume in cases of carcinoma cervix

Export to PPT

Among 12 of the tumors with tumor volume <5000 cumm, only 2 (17%) showed lymph node metastases, whereas 67% (out of 12 cases) of cases with tumor volume more than 15,000 cumm showed lymph node metastases. In between tumor volume ranges 5000–10,000 and 10,000–15,000 cumm, cases did not show any significant difference in lymph node metastasis Graph 3.

Close

Graph 3: Correlation of tumor volume with lymph node metastasis

Export to PPT

Twenty (50%) cases showed vaginal involvement by the tumor. Among 12 of the tumors with tumor volume <5000 cumm, only 2 (17%) showed vaginal involvement, whereas all cases with tumor volume more than 15,000 cumm showed vaginal involvement. In between tumor volume ranges 5000–10,000 and 10,000–15,000 cumm, cases did not show any significant difference in vaginal involvement. Twenty (50%) cases showed the extension of tumor into the uterine corpus. Among 12 of the tumors with tumor volume <5000 cumm, only 4 (33%) showed uterine involvement, whereas 10 (83%) cases with tumor volume more than 15,000 cumm showed uterine involvement. In between tumor volume ranges 5000–10,000 and 10,000–15,000 cumm, cases did not show any significant difference in uterine involvement. Sixteen (40%) cases showed parametrial involvement. Among 12 of the tumors with tumor volume <5000 cumm, only 2 (17%) showed parametrial involvement, whereas all cases with tumor volume more than 15,000 cumm showed parametrial involvement. In between tumor volume ranges 5000– 10,000 and 10,000–15,000 cumm, cases did not show any significant difference in parametrial involvement Table 5.

Table 5 Tumor volume correlation with various parameters

Tumor volume (cumm)	Number of cases (%)	Number of cases with bladder/rectum involvement (%)	Number of cases with vaginal involvement (%)	Number of cases with uterine involvement (%)	Number of cases with parametrial involvement (%)
<5000	12 (30)	0 (0)	2 (17)	4 (33)	2 (17)
5000-10,000	6 (15)	2 (33)	4 (67)	4 (67)	2 (33)
10,000-15,000	10 (25)	0 (0)	2 (20)	2 (20)	4 (40)
>15,000	12 (30)	0 (0)	12 (100)	10 (83)	12 (100)

The involvement of various organs in cases of carcinoma cervix was assessed, and the cases involving multiple organs together in the same cases were assessed. No case showed the involvement of all four organs (uterus, vagina, parametrium, bladder/rectum) Table 6.

Table 6 Tumor volume correlation with involvement of other organs

Tumor volume (cumm)	Number of cases (%)	0 out of 4 organs involvement	1 out of 4 organs involvement	2 out of 4 organs involvement	3 out of 4 organs involvement
<5000	12 (30)	8	2	0	2
5000-10,000	6 (15)	2	0	2	2
10,000-15,000	10 (25)	4	6	0	0
>15,000	12 (30)	0	0	2	10

There were eight cases which showed well differentiated SCC (WDSCC). All of them had tumor volume <5000 cumm. Twenty-eight cases had large cell nonkeratinizing SCC, out of which 4 had tumor volume <5000 cumm, 6 had 5000–10,000 cumm, 8 had 10,000–15,000 cumm, and 10 had more than 15,000 cumm of tumor volume. Four cases had adenocarcinoma, out of which two had tumor volume 10,000–15,000 cumm and the other two more than 15,000 cumm Graph 4.

Close

Graph 4: Correlation of tumor volume with tumor differentiation

Export to PPT

Correlation of fibronectin positivity with tumor volume and lymph node metastasis

Fibronectin positivity within tumors was seen in 55% of cases Figure 9. Of these, only 18% showed metastasis to lymph nodes and 82% did not. Among 12 of the tumors with tumor volume <5000 cumm, 10 (83%) cases showed fibronectin positivity whereas only 2 (17%) out of 12 cases with tumor volume more than 15,000 cumm showed fibronectin positivity Graph 5.

Close

Figure 9

Diffuse fibronectin positivity within the tumor cells

Export to PPT

Close

Graph 5: Correlation of tumor volume with fibronectin positivity

Export to PPT

LN5

Immunohistochemical staining using LN5 was done to demonstrate macrophages in tumor infiltrate. One interesting feature in this study is that the macrophages were seen surrounding the groups of tumor cells Figure 10, but their entering into the tumor cell groups is very less, whereas macrophages were seen freely entering into the layers of dysplastic squamous epithelium.

Close

Figure 10

High power view showing macrophages in close proximity to tumor cells (LN5 immunohistochemistry)

Export to PPT

Discussion

Carcinoma of the uterine cervix has been one of the commonest causes of cancer deaths. Unlike other tumors, it can be detected early. The treatment modality is determined by the stage of the disease at the time of diagnosis. Hence, it is highly essential to stage these patients accurately to advise the best treatment to increase their survival. FIGO staging system is effective in predicting the outcome of the disease. However, the disparate prognosis has been observed in patients with tumors in the same stage. Many investigators tried studying various parameters to clear this discrepancy such as pelvic lymph node metastasis,¹ tumor size,² cervical stroma invasion,³ the presence of lymph-vascular space invasion,⁴ histologic subtype,⁵ parametrial invasion,⁶ and corpus invasion⁷ and each was reported to have prognostic value. Fuller et al.⁸ found out that after satisfying for nodal metastasis, the size of the primary tumor, the depth of invasion into the cervix and the histologic grade were associated with an increased risk of recurrence. However, it is more appropriate to select factors with independent prognostic significance. Burghardt and Holzer⁹ claim that bulk or volume of tumor is important in the assessment of prognosis. With this background knowledge, it was planned to study the significance of tumor volume as an independent prognostic factor.

In this study, the tumor volume is compared with lymph node metastasis, parametrial extension, vaginal and uterine corpus involvement. It is evident from the present study that the tumor volume correlates well with lymph node metastasis, parametrial extension, vaginal, and uterine corpus involvement. Tumor volume on fresh specimens was estimated according to Dargent et al.¹⁰ According to several investigators, lymph node involvement was considered to be one of the prognostic factors in cervical carcinoma.^11,12 Hence, a correlation was made between lymph node metastasis and tumor volume, to see if any importance can be given to tumor volume as an independent prognostic factor in cases of carcinoma cervix. The observations of our study indicate that the tumor shows lymph node metastasis as the tumor volume increases. There is a clear-cut difference between the number of cases with lymph node metastasis grouped under <5000 cumm tumor volume and those with more than 15,000 cumm tumor volume. Because cases are few in number, a detailed distinction of various groups of lymph nodes could not be made. However, in general, there is a similarity between the observations of this study and the study done by Dargent et al.¹⁰ Similar correlation was also observed by Baltzer and Koepcke.¹³ Burghardt and Pickel¹⁴ have used the tumor area rather than the tumor volume to correlate with lymph node metastasis and found that it was significant. However, still it is felt that the estimation of the tumor volume will be more appropriate to find the risk cases.

Bleker et al.¹⁵ showed the importance of detection of parametrial involvement to stage and plan the treatment of carcinoma of cervix. He showed the significance of parametrial involvement in stage Ib and IIa cases. Hence, in this study, correlation was made between parametrial involvement by the tumor and the tumor volume. In accordance with Burghardt and Pickel,¹⁴ in the present study parametrial involvement was observed with an increasing trend in tumors as they increase in volume. There is a clear-cut difference between the number of cases with parametrial involvement grouped under <5000 cumm tumor volume and those with more than 15,000 cumm tumor volume. The smallest tumor volume with parametrial involvement was 3670 cumm. In some of the cases, parametrial involvement was observed microscopically as small tumor deposits in the parametrial tissue around the small vasculature, although they were negative for the tumor involvement on gross examination. This indicates that the tumor involves the parametrium not only by direct spread but also by lymphatic route.

Extension of cervical cancer in the uterine body was associated with a higher incidence of peritoneal carcinomatosis and distant metastasis resulting in decrease survival rates by a factor of 10–20%.⁷ This finding was observed by Grimard et al.¹⁶ who studied tumors belonging to stage Ib. Noguchi et al.¹⁷ reported uterine body invasion in 7.8% of stage Ib patients, 25.5% of stage IIa patients and 38.2% of stage IIb patients, indicating increasing incidence of uterine involvement as the stage of disease increases. The observation of the present study was quite in accordance with those of Noguchi. In this study, 43% of stage Ib patients, 50% of cases in stage IIa and 62.5% of stage IIb patients showed uterine involvement. Similar findings were observed with vaginal involvement also. A good correlation was also observed between the tumor volume and histological type of tumor. All the WDSCC had tumor volume <5000 cumm whereas 80% of moderately differentiated and 89% of poorly differentiated carcinomas had tumor volume more than 5000 cumm. All the adenocarcinomas showed tumor volume more than 10,000 cumm. This supports the findings observed by others.¹⁸

Conclusion

Tumor volume can be considered as an independent prognostic factor to assess the spread of the tumor. Cases with tumor volume <5000 cumm show low risk in terms of parametrial involvement and lymph node metastasis and those with tumor volume more than 15,000 cumm are definitely going to behave badly. Second, fibronectin positivity of tumor carries some importance in identifying the low-risk cases. A detailed study on macrophages in neoplasia needs to be carried out to observe whether their presence in the inflammatory infiltrate has positive or negative effect in the tumor spread.

Finally, it can be concluded from the present study that the tumor volume assessment, detection of fibronectin in carcinoma of cervix would give a clear idea regarding the behavior of tumor in vivo. A brief attempt on detection of macrophages in these tumors is just to get an insight into the immunological aspect of these tumors so that further study in aspect can be planned to see whether tumor immunology is related with tumor progression or regression.