Br. J. Surg. 1990, Vol. 77, November, 1241-1 246
J. Scheele, R. Stangl and A. Altendorf-Hofmann Department of Surgery, University Hospital, Maximiliansplatz I , D 8520 Erlangen, FRG Correspondence to: Dr J. Scheele
Hepatic metastases from colorectal carcinoma: impact of surgical resection on the natural history From 1960 to 1987, 1209 patients with colorectal liver metastases were recorded, and followed until 1 January 1990. In 242 cases the diagnosis was based on external imaging, whereas 967 patients had operative confirmation and staging of their liver disease. Three groups of patients were analysed: group 1 involved 921 cases, of whom 902 were deemed non-resectable whereas 19 could not be unequivocally classgied. Only 21 patients lived f o r longer than 3 years, seven survived for 4 years, but there were no 5-year survivors. Group 2 comprised 62 highly selected patients who at laparotomy demonstrated resectable metastasic spread confined to the liver, but this was not treated mainly because of a formerly different therapeutic approach. These patients had a significantly longer median survival time (14.2 versus 6.9 months), but also failed to achieve 5-year survival. The 226 patients forming group 3 underwent hepatic resection with intent to cure. Nine of them had minimal macroscopic disease left, and 34 with all gross tumour removed had positive margins. Survival of patients with these 43 eventually non-radical resections followed an identical course as in group 2 (median survival 13-3 months, maximum 42 months). Of the 183 patients with potentially curative resection ten died after surgery (5.5 per cent). Actuarial 5 and 10-year survival rates in the remaining 173 patients were 40 and 27 per cent with 25 and seven patients alive at respective periods of time. Until 1 January 1990, 64 patients remained free from recurrent disease for up to 24 years. In three patients the tumour status at death was unclear. The other 106 patients developed definite cancer relapse. Nevertheless they demonstrated a prolongation of survival time by a median of 1 year when compared with the 43 non-radically resected patients or the 62 untreated patients with resectable liver-only metastases, and accomplished a maximum survival time of 8 years. Radical excision of colorectal secondaries to the liver therefore offers effective palliation, and in a small number the chance of a cure. Keywords: Colorectal cancer, liver metastases, curative resection, natural history, prognosis,
treatment failure
Hepatic resection for colorectal metastases has been reported in several recent series with low mortality rates and 5-year survival$igures ranging from 16 to 52 per cent'-9. The concept of aggressive surgical treatment of metastatic disease is, nevertheless, not generally accepted. Because favourable results may be due to patient selection rather than a different therapeutic strategy".' I , prospective randomized trials have been s u g g e ~ t e d ' ~ - 'However, ~. it would be difficult to obtain informed consent from patients randomized to no treatment despite resectable disease and, in view of reported results of surgical resection, it is unlikely that ethical committees would agree to such a trial''.'3. To resolve the uncertainty regarding the value of hepatic resection, we have reviewed a consecutive series of 1221 patients with hepatic metastases from colorectal cancer, who have been prospectively recorded over a 28-year period. The data were analysed to try to answer the following questions. First, can hepatic resection lead to long-term survival and, if so, are there clear selection criteria? Second, can long-term survival also be expected in patients with potentially operable disease who receive no treatment? Finally, has the increased resection rate over the past decade altered the survival pattern?
0007-1323/90/11124146
0 1990 Butterworth-Heinemann
Ltd
Patients and methods From 1960 through 1987, a total of 5249 patiens with colorectal carcinoma were recorded at the Department of Surgery, Erlangen University Hospital. Twelve patients had liver involvement discovered at autopsy, and they are excluded from further analysis. In life, hepatic metastases were detected in 1209 patients: 736 men and 473 women with an age range of 22-91 (median 60)years. They were documented at initial presentation in 713 (59 per cent) patients and during follow-up in the other 496 (41 per cent). In 967 cases (80 per cent) the hepatic involvement was discovered at, confirmed, or treated by means of laparotomy. Of the patients presenting more recently, 242 showed evidence of metastatic spread based on ultrasound examination or computed tomography, often accompanied by carcinoembryonic antigen elevation, but were not considered candidates for surgical exploration. In 179 patients (15 per cent) the primary procedure was performed in other hospitals. In turn, 22 patients (2 per cent), who had initial treatment in our unit, underwent surgery for hepatic metastases elsewhere, involving liver resection in three cases. These 22 patients are included so as not to eliminate eventual long-term survival in untreated patients. Adequate external imaging and detailed operative records allowed a reliable judgement of resectability with respect to both hepatic involvement and overall tumour extent in all but 19 cases. Three subsets
1241
Resection of hepatic metastases from colorectal carcinoma: J. Scheele et al.
of patients were identified: group 1 involved 921 patients (76 per cent) of whom 902 were deemed not resectable. This was due to extensive liver involvement in 509 cases and to extensive extrahepatic spread, or both, in 393 cases. In addition a small subgroup of 19 untreated patients was included in whom disease confined to the liver could not be unequivocally classified as to resectability. Twenty-three patients in group 1 underwent palliative debulking procedures, and 113 received an arterial catheter for subsequent regional chemotherapy, but they all had non-resectable gross tumour left in place. Group 2 comprised 62 patients (5 per cent) in whom hepatic resection seems in retrospect to have been technically feasible and oncologically useful according to currently accepted criteria (i.e. radical excision of the primary growth, no extrahepatic disease, between one and three liver secondaries).Eight of these patients received intra-arterial chemotherapy, but the others remained untreated, reflecting the formerly different approach to metastatic tumour. The remaining 226 patients (19 per cent) formed group 3, and underwent putative curative resection of the liver secondaries. Almost half of them (109 patients) had simultaneous extrahepatic surgery. This involved the primary tumour in 84 patients, a local recurrence in eight, a solitary nodule in the omentum in five, an isolated peritoneal deposit in three, and limited haematogenous spread to other organs in 14 cases. One patient had a second colon cancer and one a renal carcinoma simultaneously removed. Out of the total series of patients seen with colorectal spread to the liver, the percentage of patients resected with the aim of cure was 4 per cent in the 1960s 13 per cent in the 1970s and 24 per cent during the past decade. Foflon-up and sra/isrical evaluation With n o patient lost to follow-up, the tumour status on 1 January 1990, or at the time of death, could be defined in all but three cases (Table I ). Survival figures were estimated using the Kaplan-Meier product limited method, and checked for statistical significance by means of the log rank test.
Results Pathological class~fication following hepatic resection Of the 249 patients undergoing hepatic resection, 23 (9 per cent) had deliberate palliative debulking, leaving residual tumour in the liver ( n = 10) or at various extrahepatic sites ( n = 13) (primary tumour, five; local recurrence, one; limited peritoneal seedings, four; distant metastases, eight). Of the 226 group 3 patients, in one case each a non-resectable tumour nodule in the liver remnant, an extensive tumour thrombus in the portal vein, and massive hepatic vein invasion became evident at the very end of the procedure, whereas in six cases some extrahepatic disease could not be completely removed. Thus 217 patients were clinically made tumour-free, but histological examination revealed positive margins in 34 cases (liver, 31 ; extrahepatic sites, three). Therefore 183 patients (8 1 per cent ofgroup 3) fulfilled the prime criterion of potentially curative intervention, i.e. removal of all detectable gross tumour with negative margins. Operarice mortality The 30-day mortality rate for 679 patients who proved to have non-resectable disease at laparotomy was 9 per cent. This includes two deaths in patients undergoing debulking of large masses. The corresponding figure in 242 cases having external
imaging only was 8 per cent. However, 31 of the 60 patients who died within 1 month of surgical exploration, and all 20 cases undergoing external imaging, succumbed due to advanced tumour stage, not as a result of treatment or complications. In group 2, three patients (5 per cent) died within 1 month of surgery, two from anastomotic leakage and one from age-related progressive cardiopulmonary deterioration. The mortality rate following putative curative hepatic resection was 5 per cent (ten of 183 patients undergoing radical procedures and two of 43 patients having non-radical intervention). Death was related to the procedure performed on the liver in six patients (surgical complications, three; hepatic failure, three). Anastomotic leakage was the major cause in two cases, acute renal failure in one, and isolated cardiopulmonary complications in three patients. The mortality rate fell from 13 per cent (five out of 39 patients) before 1978 to 4 per cent (seven out of 187 patients) during the past 10 years. Prognosis As expected, the 921 group 1 patients had a poor outlook. None survived for 5 years from operative staging or first external imaging, and only 21 (2 per cent) and seven (1 per cent) lived for longer than 3 and 4 years, respectively (Figure 1 ) . Nine patients are presently alive at 25-57 months from diagnosis, three of them in the fifth year. There was no significant difference between patients who could and could not be classified with regard to resectability (P=0.78), and only a trend in favour of the 679 patients undergoing laparotomy compared with the 242 with external imaging only (P=0.06). Patients with metastatic disease confined to the liver did significantly better than those with distant extrahepatic spread (P=0.0004), but the differences in median survival times (7.3 uersus 6.4 months) or 2-year survival figures (12 uersus 5 per cent) were marginal. The 62 group 2 patients with untreated resectable disease confined to the liver did considerably better than group 1 patients in terms of median survival time (14.2 cersus 6.9 months) but also failed to achieve 5-year survival. Three patients are alive at 27, 29 and 46 months.
t 0
I
1
2
.
,
,
.
,
3 4 5 6 7 6 9 Time from laporotomy/ imaging (years)
1
0
Figure 1 Surciiul of patienrs classified inro group I (-----. n = 921). group 2 (---; n = 62) and group 3 (-; n = 226). including operarire deaths. The number a/ rhe end of [lie curoe indicuies rhr nuniher of' parienis suri>ii:ingheyond I0 years. P < 0.0001
Table 1 Ouri,onie and turnour sfarus of 1209 parients witli colorecru1 liver nietas~asesa / I January 1990
Group 3 Group 1
Patients
(n=921)
Alive with no evidence of disease Alive with disease
-
Dead with no evidence of disease Dead with disease Dead, tumour status unknown
-
30-day deaths
1242
9
832
Group 2 (n=62) ~
3 ~
56 ~
80
3
(curative) (n=183)
54 14
Group 3 (non-radical) (n=43) 0
10
92 3
41
10
Br. J. Surg., Vol. 77, No. 11, November 1990
Resection of hepatic metastases from colorectal carcinoma: J . Scheele et al.
The 226 group 3 patients had an overall actuarial survival rate of 31 per cent at 5 years, operative mortality included (Figure 1 ). In the subgroup of 43 patients with involved margins (n=34) or with small amounts of gross disease remaining ( n = 9), prognosis was poor with median and maximum survival of 13.3 and 42 months, respectively, figures that are similar to those of the resectable but untreated group 2 patients. This was found despite a 63 per cent incidence of solitary metastases among patients undergoing these non-radical procedures. In contrast, patients who met the criteria for curative resection (i.e. no residual tumour and negative histology margins) had an actuarial 5-year survival rate of 38 per cent, rising to 40 per cent if operative mortality is excluded (Figure 2 ) , with 29 per cent or 31 per cent, respectively, considered tumour-free after that time. Twenty-five of the 80 patients undergoing hepatic resection before 1985 lived for more than 5 years, five of the 31 patients operated on before 1980 survived for 10 years, and the two who had been resected in the 1960s both lived for longer than 23 years with one still alive. Of the three patients who underwent hepatic resection in other hospitals, two died from recurrent disease after 14 and 66 months respectively, and one is alive without cancer relapse at 61 months. For the potentially curative group the impact of various
tumour-related and therapeutic features on prognosis is summarized in Table 2. With respect to crude actuarial survival, mesenteric lymph node involvement ( P = 0.0002), synchronous detection of hepatic involvement ( P =0.031), presence of
7 LL-
,
$-7
0
I
2
3
4
5
6
7
8
9
1
0
Time from liver resection (yeors)
Figure 2 Signijkance of extent of the procedure in patients resected with curative intent (group 3), excluding operative mortality. Curative procedure (n = 173); ---, non-radical procedure (n = 4 / ) . The number at the end of the curve indicates the number of patients surviving beyond 10 years. P 5 years
NED survival
Crude survival
NED survival
18 7
13 7
50 30
36 32
22 20
41 36
32 30
16 9
12 8
46 34
39 33
20 23
41 38
29 32
16 9
12 8
42 38
32 33
24 18
45 33
32 29
22 2
18
64
1
15
39 30
23 20
43 25*
37 13t
14 9 2
14 5 1
26 30 23
73 39 27
44 25 I1
53 4713 14
50 2913 7
17 8
12 8
53 27
31 47
20 26
32 45t
22 39t
18 2
14 4 2
53 17 10
33 37 33
21 22 17
40 37 39
30 32 35
20 5
18 2
68 12
42 27
24 14
44 19:
35 13f
17 8
13
58 22
46 29
25 13
42 34
33 24t
14
11
11
9
45 35
39 26
26 I2
42 30
34 23*
11
4 10
8 4 8
39 15 26
30 40 58
14 22 29
28 34t 50
23 29* 39
12 13
9 11
31 49
29 36
20 22
40 39
29 31
5
NED survival, survival with no evidence of disease;
Br. J. Surg., Vol. 77,
5-year actuarial survival (YO)
Median length of survival (months)
No. of 5-year survivors
7
*
No. 11, November 1990
trend at 0.05
J. Scheele, R. Stangl and A. Altendorf-Hofmann Department of Surgery, University Hospital, Maximiliansplatz I , D 8520 Erlangen, FRG Correspondence to: Dr J. Scheele
Hepatic metastases from colorectal carcinoma: impact of surgical resection on the natural history From 1960 to 1987, 1209 patients with colorectal liver metastases were recorded, and followed until 1 January 1990. In 242 cases the diagnosis was based on external imaging, whereas 967 patients had operative confirmation and staging of their liver disease. Three groups of patients were analysed: group 1 involved 921 cases, of whom 902 were deemed non-resectable whereas 19 could not be unequivocally classgied. Only 21 patients lived f o r longer than 3 years, seven survived for 4 years, but there were no 5-year survivors. Group 2 comprised 62 highly selected patients who at laparotomy demonstrated resectable metastasic spread confined to the liver, but this was not treated mainly because of a formerly different therapeutic approach. These patients had a significantly longer median survival time (14.2 versus 6.9 months), but also failed to achieve 5-year survival. The 226 patients forming group 3 underwent hepatic resection with intent to cure. Nine of them had minimal macroscopic disease left, and 34 with all gross tumour removed had positive margins. Survival of patients with these 43 eventually non-radical resections followed an identical course as in group 2 (median survival 13-3 months, maximum 42 months). Of the 183 patients with potentially curative resection ten died after surgery (5.5 per cent). Actuarial 5 and 10-year survival rates in the remaining 173 patients were 40 and 27 per cent with 25 and seven patients alive at respective periods of time. Until 1 January 1990, 64 patients remained free from recurrent disease for up to 24 years. In three patients the tumour status at death was unclear. The other 106 patients developed definite cancer relapse. Nevertheless they demonstrated a prolongation of survival time by a median of 1 year when compared with the 43 non-radically resected patients or the 62 untreated patients with resectable liver-only metastases, and accomplished a maximum survival time of 8 years. Radical excision of colorectal secondaries to the liver therefore offers effective palliation, and in a small number the chance of a cure. Keywords: Colorectal cancer, liver metastases, curative resection, natural history, prognosis,
treatment failure
Hepatic resection for colorectal metastases has been reported in several recent series with low mortality rates and 5-year survival$igures ranging from 16 to 52 per cent'-9. The concept of aggressive surgical treatment of metastatic disease is, nevertheless, not generally accepted. Because favourable results may be due to patient selection rather than a different therapeutic strategy".' I , prospective randomized trials have been s u g g e ~ t e d ' ~ - 'However, ~. it would be difficult to obtain informed consent from patients randomized to no treatment despite resectable disease and, in view of reported results of surgical resection, it is unlikely that ethical committees would agree to such a trial''.'3. To resolve the uncertainty regarding the value of hepatic resection, we have reviewed a consecutive series of 1221 patients with hepatic metastases from colorectal cancer, who have been prospectively recorded over a 28-year period. The data were analysed to try to answer the following questions. First, can hepatic resection lead to long-term survival and, if so, are there clear selection criteria? Second, can long-term survival also be expected in patients with potentially operable disease who receive no treatment? Finally, has the increased resection rate over the past decade altered the survival pattern?
0007-1323/90/11124146
0 1990 Butterworth-Heinemann
Ltd
Patients and methods From 1960 through 1987, a total of 5249 patiens with colorectal carcinoma were recorded at the Department of Surgery, Erlangen University Hospital. Twelve patients had liver involvement discovered at autopsy, and they are excluded from further analysis. In life, hepatic metastases were detected in 1209 patients: 736 men and 473 women with an age range of 22-91 (median 60)years. They were documented at initial presentation in 713 (59 per cent) patients and during follow-up in the other 496 (41 per cent). In 967 cases (80 per cent) the hepatic involvement was discovered at, confirmed, or treated by means of laparotomy. Of the patients presenting more recently, 242 showed evidence of metastatic spread based on ultrasound examination or computed tomography, often accompanied by carcinoembryonic antigen elevation, but were not considered candidates for surgical exploration. In 179 patients (15 per cent) the primary procedure was performed in other hospitals. In turn, 22 patients (2 per cent), who had initial treatment in our unit, underwent surgery for hepatic metastases elsewhere, involving liver resection in three cases. These 22 patients are included so as not to eliminate eventual long-term survival in untreated patients. Adequate external imaging and detailed operative records allowed a reliable judgement of resectability with respect to both hepatic involvement and overall tumour extent in all but 19 cases. Three subsets
1241
Resection of hepatic metastases from colorectal carcinoma: J. Scheele et al.
of patients were identified: group 1 involved 921 patients (76 per cent) of whom 902 were deemed not resectable. This was due to extensive liver involvement in 509 cases and to extensive extrahepatic spread, or both, in 393 cases. In addition a small subgroup of 19 untreated patients was included in whom disease confined to the liver could not be unequivocally classified as to resectability. Twenty-three patients in group 1 underwent palliative debulking procedures, and 113 received an arterial catheter for subsequent regional chemotherapy, but they all had non-resectable gross tumour left in place. Group 2 comprised 62 patients (5 per cent) in whom hepatic resection seems in retrospect to have been technically feasible and oncologically useful according to currently accepted criteria (i.e. radical excision of the primary growth, no extrahepatic disease, between one and three liver secondaries).Eight of these patients received intra-arterial chemotherapy, but the others remained untreated, reflecting the formerly different approach to metastatic tumour. The remaining 226 patients (19 per cent) formed group 3, and underwent putative curative resection of the liver secondaries. Almost half of them (109 patients) had simultaneous extrahepatic surgery. This involved the primary tumour in 84 patients, a local recurrence in eight, a solitary nodule in the omentum in five, an isolated peritoneal deposit in three, and limited haematogenous spread to other organs in 14 cases. One patient had a second colon cancer and one a renal carcinoma simultaneously removed. Out of the total series of patients seen with colorectal spread to the liver, the percentage of patients resected with the aim of cure was 4 per cent in the 1960s 13 per cent in the 1970s and 24 per cent during the past decade. Foflon-up and sra/isrical evaluation With n o patient lost to follow-up, the tumour status on 1 January 1990, or at the time of death, could be defined in all but three cases (Table I ). Survival figures were estimated using the Kaplan-Meier product limited method, and checked for statistical significance by means of the log rank test.
Results Pathological class~fication following hepatic resection Of the 249 patients undergoing hepatic resection, 23 (9 per cent) had deliberate palliative debulking, leaving residual tumour in the liver ( n = 10) or at various extrahepatic sites ( n = 13) (primary tumour, five; local recurrence, one; limited peritoneal seedings, four; distant metastases, eight). Of the 226 group 3 patients, in one case each a non-resectable tumour nodule in the liver remnant, an extensive tumour thrombus in the portal vein, and massive hepatic vein invasion became evident at the very end of the procedure, whereas in six cases some extrahepatic disease could not be completely removed. Thus 217 patients were clinically made tumour-free, but histological examination revealed positive margins in 34 cases (liver, 31 ; extrahepatic sites, three). Therefore 183 patients (8 1 per cent ofgroup 3) fulfilled the prime criterion of potentially curative intervention, i.e. removal of all detectable gross tumour with negative margins. Operarice mortality The 30-day mortality rate for 679 patients who proved to have non-resectable disease at laparotomy was 9 per cent. This includes two deaths in patients undergoing debulking of large masses. The corresponding figure in 242 cases having external
imaging only was 8 per cent. However, 31 of the 60 patients who died within 1 month of surgical exploration, and all 20 cases undergoing external imaging, succumbed due to advanced tumour stage, not as a result of treatment or complications. In group 2, three patients (5 per cent) died within 1 month of surgery, two from anastomotic leakage and one from age-related progressive cardiopulmonary deterioration. The mortality rate following putative curative hepatic resection was 5 per cent (ten of 183 patients undergoing radical procedures and two of 43 patients having non-radical intervention). Death was related to the procedure performed on the liver in six patients (surgical complications, three; hepatic failure, three). Anastomotic leakage was the major cause in two cases, acute renal failure in one, and isolated cardiopulmonary complications in three patients. The mortality rate fell from 13 per cent (five out of 39 patients) before 1978 to 4 per cent (seven out of 187 patients) during the past 10 years. Prognosis As expected, the 921 group 1 patients had a poor outlook. None survived for 5 years from operative staging or first external imaging, and only 21 (2 per cent) and seven (1 per cent) lived for longer than 3 and 4 years, respectively (Figure 1 ) . Nine patients are presently alive at 25-57 months from diagnosis, three of them in the fifth year. There was no significant difference between patients who could and could not be classified with regard to resectability (P=0.78), and only a trend in favour of the 679 patients undergoing laparotomy compared with the 242 with external imaging only (P=0.06). Patients with metastatic disease confined to the liver did significantly better than those with distant extrahepatic spread (P=0.0004), but the differences in median survival times (7.3 uersus 6.4 months) or 2-year survival figures (12 uersus 5 per cent) were marginal. The 62 group 2 patients with untreated resectable disease confined to the liver did considerably better than group 1 patients in terms of median survival time (14.2 cersus 6.9 months) but also failed to achieve 5-year survival. Three patients are alive at 27, 29 and 46 months.
t 0
I
1
2
.
,
,
.
,
3 4 5 6 7 6 9 Time from laporotomy/ imaging (years)
1
0
Figure 1 Surciiul of patienrs classified inro group I (-----. n = 921). group 2 (---; n = 62) and group 3 (-; n = 226). including operarire deaths. The number a/ rhe end of [lie curoe indicuies rhr nuniher of' parienis suri>ii:ingheyond I0 years. P < 0.0001
Table 1 Ouri,onie and turnour sfarus of 1209 parients witli colorecru1 liver nietas~asesa / I January 1990
Group 3 Group 1
Patients
(n=921)
Alive with no evidence of disease Alive with disease
-
Dead with no evidence of disease Dead with disease Dead, tumour status unknown
-
30-day deaths
1242
9
832
Group 2 (n=62) ~
3 ~
56 ~
80
3
(curative) (n=183)
54 14
Group 3 (non-radical) (n=43) 0
10
92 3
41
10
Br. J. Surg., Vol. 77, No. 11, November 1990
Resection of hepatic metastases from colorectal carcinoma: J . Scheele et al.
The 226 group 3 patients had an overall actuarial survival rate of 31 per cent at 5 years, operative mortality included (Figure 1 ). In the subgroup of 43 patients with involved margins (n=34) or with small amounts of gross disease remaining ( n = 9), prognosis was poor with median and maximum survival of 13.3 and 42 months, respectively, figures that are similar to those of the resectable but untreated group 2 patients. This was found despite a 63 per cent incidence of solitary metastases among patients undergoing these non-radical procedures. In contrast, patients who met the criteria for curative resection (i.e. no residual tumour and negative histology margins) had an actuarial 5-year survival rate of 38 per cent, rising to 40 per cent if operative mortality is excluded (Figure 2 ) , with 29 per cent or 31 per cent, respectively, considered tumour-free after that time. Twenty-five of the 80 patients undergoing hepatic resection before 1985 lived for more than 5 years, five of the 31 patients operated on before 1980 survived for 10 years, and the two who had been resected in the 1960s both lived for longer than 23 years with one still alive. Of the three patients who underwent hepatic resection in other hospitals, two died from recurrent disease after 14 and 66 months respectively, and one is alive without cancer relapse at 61 months. For the potentially curative group the impact of various
tumour-related and therapeutic features on prognosis is summarized in Table 2. With respect to crude actuarial survival, mesenteric lymph node involvement ( P = 0.0002), synchronous detection of hepatic involvement ( P =0.031), presence of
7 LL-
,
$-7
0
I
2
3
4
5
6
7
8
9
1
0
Time from liver resection (yeors)
Figure 2 Signijkance of extent of the procedure in patients resected with curative intent (group 3), excluding operative mortality. Curative procedure (n = 173); ---, non-radical procedure (n = 4 / ) . The number at the end of the curve indicates the number of patients surviving beyond 10 years. P 5 years
NED survival
Crude survival
NED survival
18 7
13 7
50 30
36 32
22 20
41 36
32 30
16 9
12 8
46 34
39 33
20 23
41 38
29 32
16 9
12 8
42 38
32 33
24 18
45 33
32 29
22 2
18
64
1
15
39 30
23 20
43 25*
37 13t
14 9 2
14 5 1
26 30 23
73 39 27
44 25 I1
53 4713 14
50 2913 7
17 8
12 8
53 27
31 47
20 26
32 45t
22 39t
18 2
14 4 2
53 17 10
33 37 33
21 22 17
40 37 39
30 32 35
20 5
18 2
68 12
42 27
24 14
44 19:
35 13f
17 8
13
58 22
46 29
25 13
42 34
33 24t
14
11
11
9
45 35
39 26
26 I2
42 30
34 23*
11
4 10
8 4 8
39 15 26
30 40 58
14 22 29
28 34t 50
23 29* 39
12 13
9 11
31 49
29 36
20 22
40 39
29 31
5
NED survival, survival with no evidence of disease;
Br. J. Surg., Vol. 77,
5-year actuarial survival (YO)
Median length of survival (months)
No. of 5-year survivors
7
*
No. 11, November 1990
trend at 0.05
