BREAST CANCER IN BABYLON: PROGNOSTIC INDEX AND EVALUATION OF TREATMENT RESULTS

Breast cancer is the most common malignancy in females all over the word. It represents about 34% of total malignancies among IRAQI females with high mortality rate representing 1-2% of female mortality &16% of cancer deaths in females in IRAQ. Many prognostic factors which can affect the treatment outcome have been studied to identify patients at high risk of disease relapse who might benefit from post-operative adjuvant therapy. The Aim of this study was evaluation of different prognostic factors to drive a reliable (Prognostic Index) that best fits our breast cancer patients, hoping to give the adjuvant treatment accordingly. This is a study analysis of 566 female patient treated for primary breast cancer between 1992 and 2001 at Oncology Unit, Marjan teaching Hospital. Modified prognostic index (PI.). Was used to identify different prognostic group. We could divide patients into 4 groups : Group 1 with PI <2.5, group 2 with PI > with PI >2.5-3,group 3 with PI>3-3.5 and group 4 with PI>3.5. The 5 year overall survival (OS) and relapse free survival (RFS) were calculated for the whole group and for the different 4 prognostic groups as well as for influence of systemic adjuvant treatment. The 5 year O.S and RES were 75% and 55% respectively for patient with PI< 2.5 and decreased with the increase of the value of PI to reach 50% and 10% respectively in patients with PI> 3.5. The difference in both 5 year O.S, RFS for different prognostic groups was found statistically significant only between patients with PI<3. (Groups 1&2) and those with PI>3. (Groups 3 &4) with P < 0.001. It was shown those patients with PI<3 could benefit from the addition of adjuvant system treatment with better 5 year RFS of 60% in comparison to 40% for patients who did not receive adjuvant systemic treatment (P=0.01). Minimal benefit was obtained in – patients with PI > 3. It was concluded that more intensive adjuvant treatment my be warranted for group 3 and 4 of patients.


Introduction
reast cancer is the most common malignancy in femlaes all over the world [1][2][3][4][5][6][7][8][9][10] . It represents about 34% of total malignancies among Iraqi females with high mortality rate representing 1-2% of females mortality and 16% of cancer deaths in females in Iraq [11][12][13] . Many prognostic factors that can affect the treatment outcome have been studied to identify patients at high risk of disease relapse who might benefit from postoperative adjuvant therapy. The most important factors are tumour size, lymph node status, pathological grade, age of menstrual status and estrogen receptor status [14][15][16][17] . A modified prognostic index has been calculated by Todd et al. (18) for group of patients with longer followup period (minimum 6 year). Also it has been applied prospectively to a further group of 320 patients of 320 patients and shown to be similarly effective in predicting the survival pattern in a group of patients treated by mastectomy.
This prognostic index helped in the selection of patients with excellent prognosis in whom adjuvant therapy was inappropriate and patients with high index score who may benefit from local or systemic adjuvant therapies.
Three prognostic factors were identified by Brown et al. 19  The aim of the present study was to evaluate different prognostic factors and to derive a reliable PI that best fits our breast cancer patients, hoping to give the adjuvant treatment accordingly. We also aimed to compare our P.I. with the NPI that was modified by Brown et al. 19 .

Patients and Methods
This is a clinical pathological analysis of 566 female patients with breast cancer treated at the Oncology Unit Merjan Teaching Hospital Babylon, in the period between 1992-2001. The patient's files were reviewed for detailed information's about age, menopausal status, stage, histopathological type and grade, lymph node status, clinical and pathological size of the tumours and treatment results.
All patients were treated by radical or modified radical mastectomy followed by comprehensive post-operative radiotherapy. The dose given was 45 Gy/20 fractions over 4 weeks (225 cGy/ fraction) to chest wall and peripheral lymphatics using telecobalt-60 machine, similar to the technique used by Fletcher 9 . Systemic adjuvant chemotherapy was given to 250 patients; 203 patients received CMF combination (Cyclophosphamide, 600 mg/m 2 day 1, Methotrexate 50 mg/m 2 day 1 and 5 Fluorouracil 600 mg/m 2 day) and 47 patients received FAC combination (5 Flourouracil 600 mg/m 2 day 1, A draimycin 40 mg/m 2 day 1 and Cyclophosphamide 600 mg/m 2 day 1). The course was repeated every 21 days for 6 courses.
All patients were followed-up regularly, 3 monthly for 2 years, 4 monthly for 5 more years and annually thereafter. The minimum follow-up period was 2 years and the maximum was 8 years.
Relapse free survival (RFS) and overall survival (OS) were calculated from the date of mastectomy to the date of relapse or last followup. A trial of application of the prognostic index used by Brown et al. 1993 in the YBCG study was done with some modifications to suit our group of patients. P1 = 0.1 x pathological tumours size (cm.) + 0.5 x grade + 0.6 * L.N. status. As the majority of our patients had positive axillary lumph nodes and high grade tumours, the coding for the various prognostic factors were modified as follow: The patients in our study were then divided into groups according to the P1: good (P1 < 1. 25

Statistical methods
All data were tabulated and statistically studied by descriptive analysis as well as survival analysis using the life test procedure (product limit survival estimates) 21 .
Comparisons between groups were performed using the long rank test. AP value 0.05 was considered significant. The analysis was performed for the whole studied groups as well as different prognostic groups after application of PI. The OS and RFS were fperformed in correlation to different tumours size, L.N. status, grade as well as the use of adjuvant systemic treatment.

Results
The characteristics of the 566 patients included in the study are shown in Table  I. Their ages ranges between 26 and 75 years with a median age of 43 years, 374 patients (66%) were premenopausal while 192 patients (34%) were postmenopausal.
Patients were staged according to TNM classification (UICC/AJCC). The 5 years OS and RFS for all patients were 75% and 55% respectively. Locoregional relapse occurred in 20 patients (3.5%), distant relapse in 303 patients (53.5%) while 243 patients (43%) had both locoregional and distant relapses. The majority of relapses were observed in the first 2 years postoperatively with gradual decrease in the incidence thereafter. Bone metastasis was the commonest site of distal relapse and was observed in 311 patients (55%) followed by lung metastasis in142 patients (25%). Other sites of distant metastases included liver, brain, pleura, contralateral breast and axilla.
The effect of adjuvant systemic treatment was studied in relation to PI. There has been better 5 year RFS for those patients who received than for those who did not receive adjuvant systemic treatment only in the group of patients with PI3.2 (36% and 46% respectively with p=0.017). The 5 year RFS for patients who received adjuvant treatment with PI>3.2 was 16% which is nearly similar to 15% for those who did not receive treatment.

Discussion
In the present study of 566 patients with breast cancer, the median age was 43 years, which is younger than Western series where the median age was reported to be 54 years 22 . The frequency of T1 tumours in the present study was 4% only. This finding differs from that reported in western series as T1 constitute about 30% of cases 22,23 . The majority of our patients presented with T2 tumours (44%) which is similar to western series 21 . T3 and T4 tumours constitute about 35% and 17% of our patients respectively. The frequency is higher than 20% and 5% reported in western series 22 .
The majority of patients in the present study had positive axillary lymph nodes (90%). About 64% of patients had 4 or more positive nodes. These figures are different from western series where the incidence of positive lymph nodes was reported to be about 45% with low incidence of 4 or more positive lymph nodes about 17% 19,22,24 . These findings show that our patients presented in more advanced stage than western series. The most frequent pathological type in this study was invasive duct carcinoma (88%), which is similar to most western series 6,15 . Grade I tumour was recorded in 2.5% of our patients. This frequency is considered very low compared to about 30% reported in western series 19,25 . The majority of our patients had grade 2 tumour (74%). Grade III was recorded in 23.5% of our patients, which is similar to most reported series.
Locoregional relapse occurred in 18% of patients, the majority within the 1 st 3 years after treatment. Chest wall relapse constituted 85% of locoregional relapse while axillary recurrence was found in 10.4% of patients. The incidence of locoregional recurrence in our study is similar to that reported in Person reported series 26  Reviewing our available materials led us to introduce some modification to suit best our patients. In our modified PI, we used pathological rather than clinical tumour size because we think it is more accurate. Grade was recorded as 1 and 2 versus 3 as the incidence of grade 1 in our patients was very low. As the majority of our patients had positive lymph nodes, nodal status was coded as 1 for negative lymph nodes, 2 for 1-3 positive lymph nodes, 3 for 4-7 positive nodes and 4 for more than 8 positive nodes. Thus new cuts off values (2.5-3-3.5) were testes statistically to separate groups of patients with different survival.
Statistical analysis of patients with PI 3 compared to those >3 showed difference in 5 years OS and RFS which was statistically highly significant (p0.001). Thus, our PI succeeded in dividing the patients into two main groups with different prognosis.
There had been better RFT for patients who received than for those who did not receive adjuvant systemic treatment. This difference was more significant for the patients PI 3(p=0.017). These results suggest that patient with PI 3 probably need more intensive adjuvant therapy than those with PI3. This is in agreement of most reported series who use intensive chemotherapy for those high risk patients 25,30 . Further validation of the PI in a randomised prospective study should be considered for proper evaluation of its value and addition of other prognostic factors.

Conclusion
We may conclude from the present study that our patients with breast cancer present with advanced stage so the results of treatment were inferior to that reported in western series. Using PI we succeeded in dividing patients into two main groups. The prognosis of patients PI> 3 in poorer than those patients with PI3. More intensive adjuvant systemic treatment should be considered for patients with PI>3.