Prognostic Factors Associated with Survival After ...

Acrrr Oncdogictr Vol. 30 N o . 5 1991

FROM T H E SECOND DEPARTMENT O F SURGERY, HELSINKI UNIVERSITY CENTRAL HOSPITAL, HELSINKI, FINLAND.

PROGNOSTIC FACTORS ASSOCIATED WITH SURVIVAL AFTER BREAST

Acta Oncol Downloaded from informahealthcare.com by 139.0.25.10 on 05/20/14 For personal use only.

CANCER RECURRENCE M. K. LEIVONENand T. V. K A L I M A

Abstract Factors associated with disease-free interval after the primary treatment and survival after a recurrence of breast cancer were studied in 331 female breast cancer patients treated in 1976 1980. Within five years after the primary treatment. recurrence occurred in 131 patients. The observation time of these patients after recurrence was from few weeks to twelve years. Twenty-nine patients were alive at the end of the follow-up. The average disease-free time was 2 years. The clinical stage of the disease in this material was not significantly associated with the disease-free interval. The median survival time after recurrence was 2.7 years when only breast cancer related deaths were included. Survival was significantly better for patients with primarily stage I disease than for patients with primarily stage 11-IV disease. The size of the primary tumour was not significantly associated with survival after recurrence. The patients with loco-regional recurrence survived almost significantly better than those with distant recurrence. The disease-free time correlated positively with survival after a recurrence. The present study confirms the view that breast cancer includes several subgroups with a diferent type of clinical course.

years after the primary treatment and especially to study the relation of such factors to survival after recurrence.

~

KcJj,iwrc1.s:

Breast cancer. recurrence. disease-free time, survival.

It is universally agreed that recurrence of breast cancer is a poor prognostic sign. Several factors are known to influence local recurrence, such as the extent of surgery ( I , 2), skin oedema (3), and postoperative radiotherapy (4-6). Several factors associated with the occurrence of distant metastases and the prognosis have also been found, such as the primary stage of the disease, histologic grade ( 7 , 8), and ploidy of the tumour (9, 10). The disease-free time is dependent on the growth rate of the tumour ( 1 I ) . Studies of factors associated with survival after recurrence are relatively few. The aim of the present investigation was to study some factors associated with the disease-free time within five

Material and Methods The original material consisted of 331 female breast cancer patients in all stages consecutively treated in the Second Department of Surgery, Helsinki University Central Hospital, in 1976-1980, of whom 131 (39.6%) had their first recurrence during a routine five-year follow-up up to 1985. The patient data are given in Table 1. N o patient was lost to follow-up during this five-year period. The regular follow-up interval was three months for the first two years and then six months up to five years after the primary therapy. The recurrence was ascertained by biopsy or other available method (x-ray, scintigraphy, or fine-needle aspiration biopsy, for example). Of the first recurrences 32 (24.4%) were local (the operation field or the ipsilateral axilla), 28 (21.4%) bone, 16 ( 12.2'%1)lung, 22 ( 16.8%) multiple organs, and 33 (25.2%) miscellaneous distant metastases (Table 2 ) . There were seven contralatera1 recurrences which may have been second primaries but are included in the material. The study population, the 13 I patients with recurrence, was followed with 3-6-month intervals after the secondary therapy or more often when needed due to the clinical situation. The status of these patients was checked in December 1988 and the observation time after first detection of recurrence was from a few weeks to twelve years. The cause of death was determined clinically (when obvious) or by autopsy. At the end of the follow-up 20 out of Submitted 5 March 1990. Accepted for publication 15 October 1990.

583

584

M. K. LEIVONEN A N D T. V. KALIMA

Table 1 Patient data of the total patient group (n = 331) and of the patients with recurrence (n = 131)

Patients Total (n = 331)

With recurrence (n = 131)

Age W s )

56.6 & SD 12.3

55.5 f SD 11.4

Type of operation Mastectomy and axillary clearance Mastectomy and perop. axillary exploration*)

224 (67.7%) 100 (30.2%)

86 (65.6%) 39 (29.8%)

7 (2.1%)

6 (4.6%)


Other Postoperative radiotherapy

181 (54.7%)

70 (53.4?’0)

Clinical stages (pTNM)**’ I I1 Ill IV Not known***)

47 (14.2%) 119 (35.9%) 42 (12.7%) 23 (6.9%) 100 (30.3%)

13 (9.9%) 53 (40.5%) 28 (2 1.4%) 23 (17.6%) 14 (10.7%)

*) with or without a biopsy of axillary nodes. **) x 2 test p < 0.01 between the total patient group and the patients with recurrence. ***) Axillary node status or distant metastases not evaluated.

131 patients (15.3%) were alive and free of disease, nine patients (6.9%) were alive with a recurrence, 89 patients (67.9%) had died of breast cancer, and 13 patients (9.9%) for another reason; these 13patients were disease-freeor had recurrence which was probably not the cause of death. One hundred and thirty reoperations were performed. The pTNM classification according to UICC 1987 was used. Disease-free time and cumulative survival rates were estimated by the life-table method and Mantel-Cox’s statistics. The frequency distributions in the patient material were tested using x 2 statistics. BMDP statistical programs using a VAX 8600 computer were used for all computations. Table 2 Site of recurrence

Local Local skin area Ipsilateral axilla Distant Bone Viscera Pleura Lung Liver Abdomen Brain Soft tissue Supraclavicular region Other lymph node Contralateral breast Skin Other Multiple sites*) Not known

32 16 16 75 28 25 1

The crude median survival of the 131 patients with a recurrence after the primary therapy was 3.9 years when all the causes of death were included. The life-table estimated survival rates after the primary operation were 68% for three years, 47% for five years, and 30% for ten years when only breast cancer related deaths were included. The disease-free time averaged 2.1 f SD 1.4 years (range 0-5.9 years). There was no significant difference in diseasefree time between the clinical stages I, I1 and 111. The disease-free time was not significantly associated with the primary tumour size ( p = 0.80) and was not significantly different as between loco-regional and distant recurrence ( p = 0.46). The median survival after recurrence was 2.5 years (range 0- 1 1.7) when all the causes of death were included, and differed significantly between stages I-IV ( p = 0.01). The five-year life-table survival rates for stage I-IV were 62%, 2870, 13% and 17% respectively. The corrected survival rates for the whole study population were 46% three years after recurrence, 31% five years, and 27% ten

16 5

Table 3

1

2 22 6 2

Disease-jiree time (mean f S D yrs) after primary treatment and crude survival after recurrence (mean f SD years) in stages I-IV

7

2 5 22 2

multiple organs, not including synchronous recurrences within the local skin area and ipsilateral axilla

*)

Results

Disease-free time

Survival (median)

2.0f 1.1 2.3 f 1.7 1.7 1.3

5.7 f 3.6 (5.4) 3.0 f 2.7 (2.1) 2.6 f 1.8 (2.4) 2.8 f 2.9 ( I .7)

~

Stage I I1 Ill IV

+

585

BREAST CANCER RECURRENCE

.'* 1

2


0

4

,

\

\m

6 TIME (years)

, 8

, \\El, 10

12

significant differences were found concerning disease-free time or survival after recurrence. However, considerable differences were found when single various sites of the recurrences were compared, p < 0.001 (Fig. 2). The prognosis was better in patients with recurrence in the contralateral breast, in the operation field, and ipsilateral axilla (the corrected five-year survival rates after the recurrence were 83%, 44% and 49% respectively) than in patients with metastases in the brain, abdomen, or multiple organs (all were dead within four years). The corrected survival after recurrence did not differ significantly between different therapy groups (local vs systemic vs combination vs nihil) in stages I-IV ( p = 0.43). The disease-free time correlated with survival after recurrence with a p-value of 0.008, though with wide spread (correlation coefficient 0.23) (Table 3).

Fig. I . Corrected survival after recurrence, by pTNM-stage at diagnosis. p > 0.05.

Discussion years after recurrence when only breast cancer related deaths were included. Breast cancer related survival after the recurrence was related to primary clinical stage (Fig. 1) and patients with stage I disease survived significantly better than those with stages 11-IV ( p = 0.01). The corrected five-year survival rates after the recurrence in stages I-IV were 62%, 32% 20% and 22% respectively. In patients with histological axillary node metastases ( N 1) the primary tumour size was not significantly associated with survival after the recurrence (the corrected five-year survival rates were TI 29%, T 2 17% and T3 20%. p = 0.81). For loco-regional and distant recurrences, the survival rates after the recurrence differed almost significantly. p = 0.04 (corrected five-year rates were for local recurrence 46% and for distant recurrence 27%). If the distant metastases were divided into three dominant sites (bone, viscera and soft tissue), no

I

0

I

\

..................................................

\

\3 2

4

1

8

8

10

\2

.

,

12

TIME (pars)

Fig. 2. Survival after recurrence in different sites with correction for deaths not related to breast cancer. p