Impact of Surgical Therapy on Overall Survival

The use of surgical therapy for the primary tumor has been associated with prolongation of survival in essentially all of the published multi-institutional studies. These studies have large numbers and high death rates,

As shown in Table 74-1. The HR hovers around 0.6 in all four analyses, with an approximately 18% difference in 3-year survival across the studies. In the NCDB report, the 3-year survival was 35 months in the surgically resected patients with free margins, compared with 26 months in women undergoing resection with involved margins and 17 months in the nonsurgical group. In both analyses of SEER data, surgically treated patients lived 11 to 15 months longer than nonsurgically treated patients (P < 0.001). Tumor-free margins were achieved more frequently in women undergoing total compared with partial mastectomy, and this may explain the somewhat better outcomes associated with total mastectomy in some analyses.22,24 These data are shown in Figure 74-1. Two additional groups of authors recorded the percentage of free margins (achieved in just under 50% of patients) but did not include this in the analysis of survival.

A

Years following diagnosis


P < 0.0001

Years following diagnosis

B


No. patients at risk at beginning of each period

Years


Local surgery

0

1

2

3

4

5

None

173

77

51

37

26

17

Yes, negative margins

61

49

39

29

21

14

Yes, positive margins

33

21

13

9

6

5

Yes, unknown margins

33

19

13

9

5

4

Figure 74-1 Survival by resection margin status for patients with de novo stage IV breast cancer reported to the National Cancer Database, 1990-1993. A, Five-year observed survival rates of cases treated with partial mastectomy, by margin status. B, Five-year observed survival rates of cases treated with total mastectomy, by margin status. C, Five-year adjusted disease-specific survival by surgical margin status in women reported to the Geneva Tumor Registry, 1977-1996. (A and B, From Khan SA, Stewart AK, Morrow M: Does aggressive local therapy improve survival in metastatic breast cancer? Surgery 132:620-626, 2002; C, From Rapiti E, Verkooijen HM, Vlas-tos G, et al: Complete excision of primary breast tumor improves survival of patients with metastatic breast cancer at diagnosis. J Clin Oncol 24:2743-2749, 2006.)


Overall survival (mos)

Figure 74-2 Adjusted and unadjusted overall survival by definitive surgery status among patients treated at Baylor College of Medicine, 1973-1991. Kaplan-Meier curves were estimated by definitive surgery status and were compared using the log-rank test. Survival was adjusted for estrogen receptor, progesterone receptor, and number of metastases, but using the average of each variable in the Cox regression model. (From Blanchard DK, Shetty PB, Hilsenbeck SG, Elledge RM: Association of surgery with improved survival in stage IV breast cancer patients. Ann Surg 247:732-738, 2008.)


Indolent course of osseous metastases. However, if therapy of the primary tumor is beneficial in the setting of metastatic disease, it is not possible to rule out a benefit for women with visceral metastases based on present evidence. On the other hand, women with bone-only disease may have the same relative (but longer absolute) benefit than women with more quickly lethal disease (e. g., visceral metastases).

The presence of hormone receptor expression or use of endocrine therapy was also associated independently with prolonged survival in some analyses,27,31 As was the use of systemic therapy.22 Younger women and those with regional lymph node involvement fared worse in one study.31 In the two analyses that have included resection margin data, free margins largely explained the benefit of surgery.22,31 Variables that did not have an independent impact on outcome were tumor size, extent of lymph node dissection, and number of positive axillary nodes. It is notable that despite the differences in the variables included in multivariate models (mainly disease-related in the NCDB and Geneva analyses and mainly demographic in the SEER analyses), the HR of death was strikingly similar across studies (about 0.6).

In single-institution reviews, the survival differences are not consistent. The trend favors the surgical group, either significantly27,28 Or nonsignificantly,25 But the failure to demonstrate a difference may be attributable to smaller numbers or higher overall survival rates. The MDACC study population is somewhat unique in its excellent survival of 83%. In this study, an improvement in metastatic progression-free survival for surgically treated patients was observed (relative risk 0.54, 95% CI 0.38-0.77). This, along with the shorter follow-up (32.1 months) means that there were too few deaths during the study period to demonstrate an overall survival difference between groups. The Baylor College of Medicine analysis provided a comparison of overall survival between surgical and nonsurgical groups before and after statistical adjustments for confounding variables (estrogen receptor, progesterone receptor, number of metastatic sites; Fig. 74-2). It is of interest that there was very little difference in the curves.

  • Contact
  • Category: Women's diseases