Initial Surgical Cytoreduction is an Essential Part of the Management of Patients with an Advanced Ovarian Cancer

Nick M. Spirtos, M.D., Scott M. Eisenkop, M.D. Samuel C. Ballon, and John B. Schlaerth, M.D.
Women's Cancer Center

Back to WCC Published Papers

Improvement in median and long-term survival for patients with advanced epithelial ovarian cancer more likely results from the evolution of surgical procedures and strategy that accomplish more complete cytoreductive surgery(CRS) than improved cytotoxic agents. This premise is supported by the observations that: 1) Median survivals in patients with similar volumes of residual disease (microscopic, optimal,<1-3 cm or suboptimal ) have not improved markedly with the development of new cytotoxic agents ; 2) Tumor biology does not influence resectability and influences survival unpredictably; 3) Complete cytoreductive surgery with no visible intra-abdominal and retroperitoneal residual disease is achievable in the vast majority of patients with acceptable morbidity and mortality; 4) interval cytoreductive efforts are associated with poor survival.

Over the last two decades the Gynecologic Oncology Group (GOG) has conducted several phase III trials for patients with optimal and suboptimal residual disease. Consequently, the standard of care for patients with advanced stage ovarian cancer has evolved from a single alkylating agent to combination therapy utilizing a platinum containing agent with taxol. Much attention has been paid to these "advances" yet some very basic scientific principles were ignored. When designing a study it is important to apply treatment arms to the same underlying populations. Individual GOG studies apply different treatment arms to groups having optimal and suboptimal residual disease homogeneously, through randomization. Because the definition of optimal residual disease has become more rigorous, reflecting improved surgical efforts, comparison of survival of patients in recent trials with those in earlier trials is precluded and conclusions regarding "advances" in treatment are invalid.

Among series with identical definitions of "optimal" and "suboptimal" CRS, "advances" in cytotoxic drug regimens have not resulted in improved median and long-term survival. In GOG # 25 (Melphalan + c.parvum) optimal cytoreduction was defined as residual measuring less than 3 cm . However 75% had less than 1 cm residual disease. Median survival for each treatment arm was approximately 32 months. In GOG #52, with a definition of optimal CRS of <1cm, median survival for PAC vs. PC patients was 38 and 31 months respectively. In GOG#104 (optimal CRS <2cm), median survival for intraperitoneal (IP) PC treated patients was 49 vs. 41 months for those treated with IV PC. In GOG #114, (optimal CRS <1cm), survival was compared in patients receiving IV platinum/taxol vs. IP platinum/IV taxol. Among IV platinum/taxol treated patients, those with macroscopic residual had a median survival of 43 months while those visibly disease-free after surgery had a 52 month median survival. In each of these three series, the percentage of patients rendered visibly disease free by CRS has increased. The increase was most notable between 1985 (GOG#52) and 1995 (GOG#114), whereby the percentage of patients beginning treatment without macroscopic residual disease increased from 28% to 37%. Among patients with <1 cm of residual disease, the increase in percentage of patients who were visibly disease-free may have accounted for improvement in median survival as opposed to any change in drug therapy. (Table 1)

Although the median survival for patients with "optimal" cytoreductive operations treated with platinum based regimens increased compared to those treated with single agent alkylators, it is unclear if this improvement is due to more effective cytotoxic drugs or improved cytoreductive surgery. No prospective randomized trial was conducted testing single agent alkylators vs platinum based regimens for patients with residual <1cm. Such a study might have helped determine if the change in surgical endpoint or the change in cytotoxic drugs was responsible for an improved median survival.

Changes in chemotherapy have failed to improve median survival for patients with suboptimal surgery. In GOG#47 (>1cm residual) accrual ended in 1982 and a median survival was 21 months for those treated with PAC. In GOG #111 median survival was 24 months and 38 months for patients receiving CPPD and CPPD/Taxol respectively. In GOG #132 (>1cm) accrual ended in 1994. Median survivals for CPPD vs Taxol vs Taxol /CPPD were 30, 26 and 27 months respectively. Furthermore, it is possible that differences in survival between GOG#47 and Taxol containing protocols (GOG#111 and #132) may be accounted for by inclusion of a larger percentage of patients with approximately 1 cm. of residual disease and fewer patients with massive residual in the more recent series. In GOG #111 (>1cm) patients treated with IV CPPD/Taxol had a better median survival (38months) than patients treated with IV PC (24 months). However, the 48 month Kaplan-Meier estimated survival was nearly identical for both arms. Additionally, other factors such investigator bias may have led to a more "optimal" group of suboptimal patients to be included in this study as the median survival reported in GOG #111 have not been duplicated in GOG #132. The impetus for this occurring, was that GOG #111 was the first group-wide phase III trial using Taxol and at the time the only way to obtain the drug for patients with advanced ovarian cancer. (Table 2)

The essential nature of initial cytoreductive surgery is best illustrated by examining its impact on median survival when all macroscopic disease is resected. Patients rendered visibly disease free in GOG# 25 had a median survival of 53 months. In GOG#52 those patients with complete cytoreduction had a median survival of 56(CP) and 62(PAC) months. In GOG # 104 patients treated with IP-CDDP and IV-cytoxan who were virtually(< 0.5 cm) disease-free the median survival was 51 months, compared to those absolutely disease free, whose median survival was 76 months. GOG# 114 a similar group of patients treated with IV CPPD/taxol had a median survival of 52 months. No median has been reached in the IP arm of the study. Regardless of the chemotherapeutic regimen utilized, the median survival for patients with complete cytoreduction was essentially the same in these series. It is thought provoking to note that many studies designed to demonstrate the efficacy of dose intensification of chemotherapy have failed. However surgical intensification, resulting in a larger percentage of patients with no visible residual disease, correlates with improved survival. (Table 3)

Some investigators suggest that complete cytoreduction is a result of smaller tumor burdens, and presumably less biologically aggressive disease, rather than increased operative effort. It is also suggested that perceived survival benefits from cytoreduction are explainable by selection of patients with less biologically aggressive neoplasms. Hoskins et al retrospectively stratified data from GOG#52 and observed that patients presenting with minimal metastatic tumor burdens rather than extensive upper abdominal disease experience better survival when undergoing cytoreduction to < 1 centimeter of residual disease. He concluded that "factors other than cytoreductive surgery are more important in predicting survival." In addition to requiring less extensive surgical procedures to achieve optimal (< 1 cm) cytoreduction, the patients who presented with lesser tumor burdens, may have had longer survival because a larger percentage had complete cytoreduction. Unfortunately, the status of the retroperitoneal lymph nodes was not addressed and basic issues as to why disease was left unresected went unanswered. Finally, Hoskins et al did not compare survival of patients rendered visibly disease free by surgery to stage IIIa patients. This comparison is essential as it addresses the issue of truly maximal cytoreductive effort. It is premature to suggest that biological aggressiveness cannot be offset by maximal surgical effort on the basis of existing data. In one of the few papers attempting to correlate biological aggressiveness with disease status, Kimball et al, when using flow cytometry, could not demonstrate a relationship between the DNA ploidy or S-phase fraction and lymph node status in patients with stage III disease. Thus they were unable to conclude that lymph node metastases, a finding commonly said to be indicative of biological aggressiveness, were not associated with corresponding changes in ploidy or s-phase fraction relative to the primary tumor.

In contrast to Hoskins, Le et al compared the survivals of 81 patients with extensive stage III disease who underwent complete cytoreduction to 191 patients with stage IIIA and IIIB who also underwent complete cytoreduction. The 5-year survivals of 56% and 62% respectively were not statistically different. Use of a visibly disease-free outcome as an endpoint for both patients presenting with minimal and extensive metastatic tumor burdens realistically addresses the issue of maximal surgical effort.

Some investigators suggest that the prognosis for patients who undergo complete cytoreduction is adversely affected by the necessity of specific procedures such as bowel resection and lymph node debulking. They suggest that patients with gross lymphadenopathy have biologically more aggressive disease based on second-look data demonstrating persistence of retroperitoneal disease despite complete response of intraperitoneal disease. In contrast, more recent data indicate that the prognosis for patients who undergo complete cytoreduction is not adversely affected by the necessity any specific procedures. Spirtos et al demonstrated that resection of macroscopically positive retroperitoneal nodes for stage IIIc patients resulted in similar survivals to patients with stage III disease and negative lymph nodes. Eisenkop et al have reported that stage IIIC and IV patients whose largest metastases were lymphatic had more favorable survival than other patients. Additionally Eisenknop et al have demonstrated that the requirement of modified posterior exenteration, bowel resection, and diaphragm stripping do not adversely affect the prognosis for patients who had complete cytoreductive surgery(CRS).

Eisenkop et al in a landmark paper have recently demonstrated that complete cytoreduction is achievable in 86% of 163 consecutive stage IIIc and IV patients. Unlike GOG trials where the true operative denominator is unknown, this patient population was unselected. On the basis of age, grade, stage, performance status and extent of intra-abdominal disease this series had patients with disease at least as advanced as other widely cited series. The overall group median survival of was 54 months and the median survival for patients who had complete cytoreduction was 62 months. Median survival in this series compares favorably to all other reports of patients who were selectively rendered visibly disease-free. This result is remarkable as survival benefits from complete CRS previously reported for small percentages of overall series were reproduced for a large percentage of unselected patients which is more representative of the ovarian cancer population.

These results also suggest that as in most aspects of life , outcomes are describable in terms of a bell curve. This is further illustrated by a review of GOG #8812. Forty-two patients with <1cm residual disease were treated with sequential chemotherapy (PC) and hyper- fractionated WAR. The Women's Cancer Center(WCC) entered 9/42 patients for whom the surgical plan was to attain no visible residual disease rather than terminating the surgery once 1 cm of residual was attained. The median survival of patients accrued by the WCC was 54 months compared to 31 months for the rest of the institutions. Questions regarding the locations of residual disease and reasons for leaving it are important to consider in future investigations.

Recent studies have provided insight about the role of interval as opposed to initial cytoreductive surgery. Surwit et. al. reported a median survival of 27 months for patients treated with neoadjuvant chemotherapy and 28 months for those undergoing primary operation. Of the 29 patients in this study 13 were determined to be unresectable based on "Nelson's criteria" which are CT findings. Nelson et al never describe just what makes findings such as diaphragmatic or gallbladder fossa disease e.g. unresectable. Of the remaining sixteen patients eight have pleural effusions and eight have omental disease and cul de sac nodularity. It is not immediately obvious why disease in the omentum or cul de sac cannot be resected and it is incumbent on those making such statements to explain this phenomenon. which are never Van der Burg et al similarly reported median survivals of 26 and 20 months in 319 stage IIb-IV patients with and without interval cytoreductive surgery. These studies prove that less than complete cytoreduction in general and suboptimal cytoreduction in particular, done either primarily or following three courses of chemotherapy, remains "suboptimal" and reinforces the implications of the surgical bell curve.

The results of most interval cytoreductive papers are no different than those associated with initial suboptimal surgery (Table 4) and no rationale has been put forth explaining why the order of the surgery and chemotherapy favors giving the chemotherapy first. In fact, given that diagnostic errors would arise more frequently when giving chemotherapy first there appears to be reason to continue to perform laparotomy first.

In conclusion it is clear that no other treatment modality can impact median survival as does optimal CRS and complete CRS in particular. If complete CRS were a new drug it would be considered a major breakthrough in the care of patients with advanced ovarian cancer.

>References

Gall S; Bundy B; Beecham J; Whitney C; Homesley; Lifshitz S; Adcock LL:

Therapy of stage III (optimal) epithelial carcinoma of the ovary with melphalan or melphalan plus Cornynebacterium parvum (a GOG study): Gynecol Oncol 1986 Sep: 25 (1) 26-36

Omura GA; Morrow CP; Blessing JA; Miller A; Buchsbaum HJ; Homesley HD; Leone L: A randomized comparison of melphalan versus melphalan plus hexamethylmelamine versus adriamycin plus cyclophosphamide in ovarian carcinoma: Cancer 1983 Mar 1,51(5):783-9

Omura GA; Blessing JA; Ehrlich CE; Miller A; Yordan E; Creasman WT; Homesley HD: A randomized trial of cyclohosphamide and doxorubicin with or without cisplatin in advanced ovarian cancer: Cancer 1986 May1; 57(9):1725-30

OmuraGA; Bundy BN; Berek JS; Curry S; Degado G; Mortel R : Randomized trial of cyclophosphamide plus cisplatin with or without doxorubicin in ovarian carcinoma: a GOG study: J Clin Oncol 1989 Apr;7(4):457-65

Alberts DS; Liu PY; HanniganEV; O'Tool R; Williams SD; Young JA; Franklin EW; Clark-Pearson DL; Malviya VK; DuBesshter B: Intraperitoneal cisplatin plus intravenous cyclophosphamide vs intravenous cisplatin plus intravenous cyclophosphamide for stage III ovarian cancer: N EnglJMed 1996 Dec 26;335(26):1950-5

McGuire WP; Hoskins WJ; Brady MF; Kucera PR; Partridge EE; Look KY; Clark-Pearson DL; Davidson M: Cyclophosphamide and cisplatin compared with paclitaxel and cisplatin in patients with stage III and stage IV ovarian cancer: N Engl J Med 1996 Jan 4; 334 (1) 1-6

Hoskins WJ; Bundy BN; Thigpen JT; Omura GA: The influence of cytoreductive surgery on recurrence -free interval and survival in small-volume stage III epithelial ovarian cancer: a GOG study: Gynecol Oncol 1992 Nov; 47 (2):159-66

Kimball RE; Schlaerth JB; Kute TE; Schlaerth AC; Santoso J; Ballon SC; Spirtos NM: Flow cytometric analysis of lymph node metastases in advanced ovarian cancer: Clinical and biological significance: Am J Ob Gynecol 1997: 176: 1319-27

Le T, Krepart GV, Lotocki RJ, Heywood MS: Does debulking surgery improve survival in biologically aggressive ovarian carcinoma?: Gynecol Oncol 1997 Nov: 67(2): 208-214

Spirtos, NM; Gross GM; Freddo JL; Ballon SC: Cytoreductive surgery in advanced epithelial cancer of the ovary: the impact of aortic and pelvic lymphadenectomy: Gynecol Oncol 1995 Mar; 56 (3): 345-352

Eisenkop SM;Friedman RL;Wang HJ: Complete cytoreductive surgery is feasible and maximizes survival in patients with advanced ovarian cancer: A prospective study: Gynecol Oncol 1998 June; 69, 103-108

Nelson, BE, Rosenfield AT, and Schwartz,PE: Preoperative abdominopelvic computed tomographic prediction of optimal cytoreduction in epithelial ovarian carcinoma: J Clin Oncol 11 Jan 1993; 11: 166-172

van der Berg MEL, van Lent M, Kobierska A, et al: The effect of debulking surgery after induction chemotherapy on the prognosis in advanced epithelial ovarian cancer (EORTC): NEJM 1995 332,629-34


Table 1 - Relationship between Median Survival and Cytoreductive Status (% NED) in "Optimal" GOG Studies

GOG #

% NED

Chemotherapy

Median Survival

25

(1977)

32%

Melphalan +

C. Parvum

36 months

52

(1981)

28%

PAC vs

PC

38 months

31 months

104 (1988)

35%

IP-PC vs

IV-PC

49 months

41 months

114 (1992)

37%

IP-PT vs

IV PT

NA

52 months

Table 2 - Relationship Between Survival and Criteria of Suboptimal CRS in GOG Trials

GOG # Definition
Suboptimal
Chemotherapy Median Survival

22

(1977)

> 3 cm

Melphalan

Melphalan/Hex

AC

 

12-14 months

47

(1981)

> 1 cm

PAC

AC

 

19 months

111 (1988)

> 1 cm

CDDP/Taxol

CDDP/CTX

38 months

24 months

132 (1992)

> 1 cm

CDDP

Taxol

CDDP/Taxol

30 months

26 months

26 months

Table 3 - Consistency of Survival Outcomes Among Subgroups with Complete CRS Despite Changes in Chemotherapy

GOG Protocol Chemotherapy

%NED Median Survival
25 (1977) Melphalan +

C. Parvum

 

32

53 months
52 (1981) PC vs

PAC

28 55 months

62 months

104 (1988) IP-PC

IV-PC

35 76 months
114 (1992)

 

 

WCC (1998)
PC

IV-PT

IP-PT

IV Platin Combo
37

 

 

85

 
 

52 months

Not reached

62 months

Table 4 - Neoadjuvant Chemotherapy Studies

  Survival
Author # Pts Age Stage Median 4 Yr

Wils, 1986 (The Netherlands)

18

56

III/IV

NA (16 mo)

51% 1pt

Chambers, 1989 (Yale)

17

63

X

15 months

NA

Jacob, 1991 (MD Anderson)

22

58

III/I

16 months

NA

Schwartz, 1993 (Yale)

11

73

X

15 months

0%

Lim, 1993 (Clatterbridge,UK)

30

56

III/IV

10 months

7% 2pt

Goodman, 1992 (Harvard)

7

53

IV

30 months

8% 1pt

van der Berg, 1996 (EORTC)

160

55

II-IV

18 months

12%*

Variation of Survival by Institution: GOG 8812 RT in Optimally Debulked Ovarian Cancer

Institute

Number

Median Survival

95% CT

WCC

(N. Calif.)

9

 

53.6 mo

25.5-64.9

All

Others

33

31.4 mo

23.7-43.5

Primary Cytoreductive Surgical Outcomes

Institution &
Def of "Optimal"

%Optimal

Optimal
Median

Suboptimal

Median

Overall

Median

5-Yr

UCLA:

< 1cm

60%

41 mo

20 mo

31 mo

20%

HPO (Madrid): <2cm

27%

47 mo

22 mo

31 mo

20%

Indiana U Med : < 2 cm

30 %

45 mo

22 mo

26mo

23%

Roswell Park: <1 cm

40%

50 mo

25 mo

38 mo

31%

Vanderbilt U:

< 2 cm

36 %

72 mo

13 mo

45 mo

~35%

MDAnderson: < 2 cm

65%

43 mo

20 mo

33 mo

na

Multivariate Analysis of Survival

Covariate

Risk ratio

P value1

Age: < 61 years

> 61 years

1.00

2.52

0.003

Stage: IIIC

IV

1.00

2.10

0.04

Ascitic volume: < 1000 ml

1.00

0.01

> 1000 ml

2.41

Cytoreductive outcome: Not NED

2.14

0.02

NED

1.00

Histology: All Other

Mucinous/Clear cell

1.00

2.24

0.03

Woman’s Cancer Center (Southern California)
Stage IIIC/IV Ovarian Cancer 1990-1996 Patient Characteristics

Parameter

Number

Percentage

Total

163

---

Stage IIIC

IV

136

27

83.4%

16.6%

GOG Performance 2/3

113

69.3%

Grade 3

115

70.6%

> 1 Liter Ascites

97

59.5%

> 10 cm Metaseases

108

66.3%

Extensive Pelvic Disease

85

52.1%

Bulky Retroperitoneal Nodes

84

51.5%

Woman’s Cancer Center (Southern California)
Stage IIIC/IV Ovarian Cancer 1990-1997: Procedures

Procedure

Number

Percentage

Excision Reproductive Organs

162

99.3%

Omentectomy

163

100%

Modified Posterior Exenteration

85

52.1%

Extra-pelvic Bowel Resection

32

19.6%

Diaphragm Stripping/Resection

66

40.5%

Retroperitoneal LND

152

93.3%

Peritoneal Implant Ablation

145

89.0%

Other (splenectomy, distal pancreactomy, urologic, etc)

31

19.0%

Woman’s Cancer Center (Southern California)
Stage IIIC/IV Ovarian Cancer 1990-1997 Outcomes

Parameter

Outcome

Cytoreductive Outcome

Complete Cytoreduction

< 1 cm Residual Disease

> 1 cm Residual Disease

139 (85.3%)

22 (13.5%)

2 (1.2%)

Mean Operating Time

254 minutes (75-435)

Median Hospital Stay

12 days (2-58)

Median Estimated Blood Loss

1190 ml (100-6000)

Death Within 30 Days

3 (1.8%)

Death Within 1 Year

19 (11.7%)

Neoadjuvant Chemotherapy Studies

Relationship Between Survival and Criteria of Suboptimal CRS in GOG Trials

GOG #

Definition

Suboptimal

Chemotherapy

Median Survival

22

(1977)

> 3 cm

Melphalan

Melphalan/Hex

AC

 

12-14 months

47

(1981)

> 3 cm

PAC

AC

 

19 months

111 (1988)

> 1 cm

CDDP/Taxol

CDDP/CTX

38 months

24 months

132 (1992)

> 1 cm

CDDP

Taxol

CDDP/Taxol

30 months

26 months

26 months

Back to WCC Published Papers