Int J Colorectal Dis (2013) 28:537–545 DOI 10.1007/s00384-012-1553-0

ORIGINAL ARTICLE

Long-term results and prognostic factors after resection of hepatic and pulmonary metastases of colorectal cancer Silke Schüle & Yves Dittmar & Thomas Knösel & Peter Krieg & Roland Albrecht & Utz Settmacher & Annelore Altendorf-Hofmann

Accepted: 25 July 2012 / Published online: 11 August 2012 # Springer-Verlag 2012

Abstract Background Resection of colorectal liver or lung metastases is an established therapeutical concept at present. However, an affection of both these organs is frequently still regarded as incurable. Methods All cancer patients are documented in our prospective cancer registry since 1995. Data of patients who underwent liver and lung resection for colorectal metastases were extracted and analysed. Results Sixty-five patients underwent surgery for liver and lung metastases. In 33 cases, the first distant metastasis was diagnosed synchronously to the primary tumour. For the remaining patients, median time interval between primary tumour and first distant metastasis was 18 months (5– 69 months). Complete resection was achieved in 51 patients (79 %) and was less likely in patients with synchronous S. Schüle (*) : Y. Dittmar : U. Settmacher : A. Altendorf-Hofmann Department of General, Visceral and Vascular Surgery, University of Jena, Erlanger Allee 101, 07747 Jena, Germany e-mail: [email protected]

disease (p00.017). Negative margins (p00.002), the absence of pulmonary involvement in synchronous metastases (p00.0003) and single metastases in both organs (p00.036) were associated with a better prognosis. Five- and 10-year survival rates for all patients are 57 and 15 % from diagnosis of the primary tumour, 37 and 14 % from resection of the first metastasis and 20 and 15 % from resection of the second metastasis. After complete resection, 5- and 10year survival rates increased to 61 and 18 %, 43 and 17 % as well as 25 and 19 %, respectively. Long-term survivors (≥10 years) were seen only after complete resection of both metastases. Conclusions Patients with resectable liver and lung metastases of the colorectal primary should be considered for surgery after multidisciplinary evaluation regardless of the number or size of the metastases or the disease-free intervals. Clear resection margins are the strongest prognostic parameter. Keywords Metastases . Hepatic . Pulmonary . Colorectal . Prognosis

Introduction T. Knösel Institute of Pathology, University of Jena, Ziegelmühlenweg 1, 07743 Jena, Germany P. Krieg Department of Cardiothoracic Surgery, University of Jena, Erlanger Allee 101, 07747 Jena, Germany R. Albrecht Department of Visceral and Vascular Surgery, HELIOS Klinikum Aue, Gartenstr. 6, 08280 Aue, Germany

In 2009, colorectal cancer has been the third most common newly diagnosed type of cancer as well as the cause of cancer-related death in the USA [1]. Even after curative resection, between 30 and 40 % of the patients develop distant metastases depending on the stage of the primary tumour [2, 3]. The liver represents the most frequent site of metastases in colorectal cancer. In the presence of sufficient functional remnant parenchyma, liver resection may be regarded as the treatment of choice, resulting in 5-year survival rates ranging from 38 to 58 % [4, 5].

538

Extrahepatic metastases are most often found in the lungs. In this situation, lung resection is generally recommended, if a complete removal of the tumour can be accomplished. Fiveyear survival rates after resection of pulmonary metastases range from 37 to 50 % [6, 7]. Despite these concepts, controversies commonly arise if metastases are discovered in both organs at the same time or if patients have undergone either liver or lung resection in the past and now present with recurrence in the other organ. In view of the results mentioned above, staged liver and lung resection may be considered from the surgical perspective. We present an analysis of all patients who underwent both liver and lung resection in our hospital between 1995 and 2010 for metastases of colorectal carcinoma with a focus on criteria for patient selection and prognostic factors after radical surgery.

Patients and methods

Int J Colorectal Dis (2013) 28:537–545

assess differences in survival. Significant and independent predictors of survival and recurrence were identified by Cox proportional hazard analysis. The stepwise procedure was set to a threshold of 0.05. Statistical significance was defined as a p value <0.05. For observed survival, all deaths were considered as events including patients who died from secondary causes without recurrence.

Results Out of 584 patients who underwent liver surgery for metastases of colorectal carcinoma in the past 15 years, 65 patients (35 males and 30 females) developed additional pulmonary metastases. Median age was 59 years (22– 76 years). In 33 cases, the primary tumour was located in the colon, whereas 32 patients had a rectal primary. Median follow-up from resection of the primary tumour was 54 months, from resection of the first metastasis, 43 months, and from resection of the second metastasis, 30 months.

Patient selection Tumour marker CEA Treatment options for all colorectal cancer patients with distant metastases have been discussed in our multidisciplinary tumour board meetings since 2008 prior to the start of any kind of therapy. If pulmonary and hepatic metastases in colorectal cancer patients are deemed resectable according to the surgeons’ opinion based on pulmonary function as well as the calculated remnant parenchyma, surgical therapy is offered to the patient regardless of the number or size of the metastases or the disease-free interval. Before the establishment of multidisciplinary conferences, most treatment plans were based on individual decisions. Generally, it has been our policy to proceed with liver or lung resection in primarily resectable situations without prior chemotherapy in order to avoid chemotherapy-associated parenchymal damage, particularly in the liver. Methods and statistics Data of all patients who underwent surgery for both liver and lung metastases of colorectal origin at the university hospital of Jena, Germany, between 1995 and 2010 were prospectively documented in our cancer registry. Clinical and tumour-specific information was analysed regarding diagnostic procedures, perioperative course and prognosis. Survival was calculated from the date of diagnosis of the primary tumour as well as from the date of the first and the second metastatic resection. Last update of data was performed on July 2011. Statistical analyses were performed by SPSS 18.0 (SPSS, Chicago, IL) software. Survival curves were calculated by the Kaplan–Meier method. The logrank test was used to

Carcinoembryonic antigen (CEA) was tested preoperatively in 92 % of the patients. An elevated CEA of ≥5 ng/ml was seen in 41 patients (68 %). Site and timing of metastases In 33 patients, distant metastases were present at the time of diagnosis of the primary tumour and were located only in the liver in 18 and both in the liver and the lung in 12 cases. Three patients had lung metastases only at this point in time. The remaining 32 patients developed their first distant metastases after a median time of 18 months (5–69 months) following diagnosis of the primary tumour. For the 48 patients who first developed metastases in one organ, the time interval between diagnosis of the first and the second metastasis ranged from 1 to 72 months with a median of 14.5 months. More details regarding metastases and treatment are given in Table 1. The time interval between diagnosis of the primary tumour and detection of first metastasis will be referred to as “disease-free interval 1” (DFI1) and the time interval between diagnosis of the first and the second metastasis as “disease-free interval 2” (DFI2). Chemotherapy During the course of this study, both chemotherapy regimen and indication for chemotherapy have changed substantially. Only 17 patients never received chemotherapy with 13 of them having been treated in the 1990s. Twenty patients had either neoadjuvant or adjuvant (radio)chemotherapy in the

37 (20–200) 1/4 1/3/1

1

51 (4–170) 1/2 2/1/0

1

42 (13–190) 5/19 7/15/2

1

26 (8–78) 3/9 3/0/0

1

33 (25–40) 0/3 1/2/0

0

58 (22–95) 2 0/18 4/11/3

0 2 3/2 3/2 2/3 23 (12–63)

3/2

0 0 1/2 0/3 2/1 20 (6–27)

1/2

3 3 14/10 13/11 11/13 16 (5–50)

16/8

2 1 10/2 6/6 7/5 0

6/6

0 0 2/1 2/1 0/3 0

3/0

5 10 15/3 10/8 8/10 0

Liver synchronous to primary tumour, n018 Lung synchronous to primary tumour, n03 Liver and lung both synchronous to primary tumour, n012 Liver first metachronous to primary tumour, n024 Lung first metachronous to primary tumour, n03 Liver and lung simultaneously, but metachronous to primary tumour, n05

Solitary/ multiple

8/10

Chemotherapy for first metastatic site Size of lung metastases

context of their primary tumour. Prior to liver resection, 16 of 65 patients (25 %) received chemotherapy with 5fluorouracil and folinic acid, in recent years combined with oxaliplatin or irinotecan. Monoclonal antibodies were added to irinotecan-based therapy in two cases. The indications for neoadjuvant chemotherapy were twofold: It was either administered for synchronous metastases by medical oncologists in initially palliative intent without prior surgical consultation or in order to downsize large or multiple lesions to achieve resectability. Adjuvant chemotherapy was given in ten patients (15 %) as an individual decision after discussion in the multidisciplinary tumour board meeting in case of involved resection margins, short disease-free intervals or depending on the patient’s preference. The number of chemotherapy cycles ranged from 3 to 36 with a median of six cycles. None of the few patients who developed pulmonary metastases first received chemotherapy perioperatively (Table 1). Surgical procedures

Disease-free interval Liver after primary tumour metastases (DFI1) in months; median (range) Solitary/ multiple

Lung Size of liver metastases metastases

539

Site and timing of metastases

Table 1 Characteristics of metastases, treatment and survival in relation to site and timing of metastases (all patients, n065)

Surgery for liver metastases

Surgery for lung Complications Survival after metastases first operation (range) <5 cm/≥5 cm <2 cm/≥2 cm Neoadjuvant Adjuvant Segmentectomy/ Lobectomy/ hemihepatectomy/ wedge resection radiofrequency ablation

Int J Colorectal Dis (2013) 28:537–545

A major liver resection (three segments or more) was performed in 41 patients (63 %); 18 patients underwent a minor resection. In six patients, the lesions were deemed irresectable due to pre-existing damage of the liver parenchyma. These patients received open radiofrequency ablation, which was classified as a non-curative procedure in this analysis. With respect to the lung metastases, the most common procedure was wedge resection in 44 patients (68 %). Laser enucleation of metastases was performed as a parenchymasparing procedure in 11 cases (17 %) in the presence of multiple lesions. Ten patients underwent lobectomy. In those patients who were diagnosed with liver and lung metastases at the same time, liver resection was the first procedure in 11 and lung resection in three cases. The remaining three patients underwent simultaneous liver and lung surgery. During liver resection, regional lymphadenectomy in the hepatoduodenal ligament was not routinely performed. Nodal metastases were detected in 1 out of 19 patients in whom suspicious lymph nodes were removed. With respect to pulmonary resection, six patients underwent regional lymphadenectomy. Positive lymph nodes were found in two cases. Complications Minor surgical complications were noted in two patients with superficial wound infections and two patients who required drainage of symptomatic pleural effusion. One patient underwent re-laparotomy for leakage of the biliodigestive anastomosis. With respect to non-surgical major

540

complications, one patient experienced pulmonary embolism. Postoperative in-hospital mortality was 0 %.

Int J Colorectal Dis (2013) 28:537–545 Observed rate of recurrence rectum colon

Resection status Complete tumour resection was possible in 51 patients (78 %); resection margins were involved in four patients (after liver and lung resection in two cases each). In six patients, macroscopic residual tumour was left behind because of the intraoperative detection of more extensive metastatic disease than anticipated or rapid tumour progression that led us to withhold further surgery in the setting of a planned staged resection. Complete resection of both hepatic and pulmonary metastases was less frequently achieved in synchronous disease (DFI100; 66 vs. 91 %, p00.017). Number and size of metastases A single liver metastasis was diagnosed in 31 cases (48 %). Fourteen patients had two, 7 patients had three and the remaining 13 patients had four or more liver lesions. The size of the metastasis (or of the largest metastasis in case of multiple lesions) was 5 cm or more in 31 (48 %) patients. Both liver lobes were involved in 12 cases (18 %). With respect to pulmonary metastases, solitary lesions were seen in 34 cases (52 %), whereas 31 patients had two or more pulmonary metastases. The size of the lesion (or of the largest in case of multiple metastases) was 2 cm or more in 29 patients (45 %). Bilobar disease was present in 25 patients (38 %). Re-resection and recurrence Following resection of the second metastasis, 42 patients (65 %) developed tumour recurrence after a median of 7 months (0–36 months). The most common location of recurrence was the lungs in 34 %, followed by the liver in 27 %, both liver and lungs in 15 % and the peritoneum as well as the brain each in 5 %. Less common locations were the mediastinum, the bones and locoregional recurrence. Repeat liver resection was possible in 18 (35 %, two liver resections in 11 and three or more in seven cases) and repeat pulmonary resection in 14 patients (27 %, two lung resections and three or more each in seven cases). We saw a significantly earlier recurrence in patients with a rectal primary (patients after complete tumour resection, n051, p00.040, Fig. 1). Survival Calculated from the date of the primary tumour resection, we currently have 31 actual 5-year and six 10-year survivors (eighteen 5-year and six 10-year survivors counting from the resection of the first metastasis as well as nine 5-year and

p=0.026

Months after R0 resection of second metastasis

Fig. 1 Time to recurrence for 42 patients with recurrence after R0 resection depending on the site of primary tumour

four 10-year survivors after resection of the second metastasis). A difference between overall and disease-specific survival was not seen, as all but one of the deceased patients who succumbed to endocarditis died of tumour recurrence. Median survival for all patients was 62 months from the resection of the primary tumour, 48 months from the resection of the first metastasis and 37 months from the resection of the second metastasis. Five- and 10-year survival rates for all patients are given in Table 2. Prognostic factors 1. All patients, n065 Patient’s age as well as location, stage and grading of the primary tumour; location of synchronous metastases; DFI1; DFI2; location of the first metastasis; number, distribution and diameter of metastases and resection status were evaluated regarding prognostic value. On univariate analysis, patients with clear margins had a statistically significant better survival (p00.038 calculated from the primary tumour, p00.002 from the first metastasis and p00.016 from the second metastasis), turning resection status into the strongest prognostic factor (Fig. 2a). Furthermore, patients with metachronous disease (DFI1 >0) survived longer than patients with synchronous metastases (DFI100). A diameter of 5 cm or more in the largest liver metastasis (Fig. 2b) as well as a short disease-free interval between liver and lung metastases (DFI2 <12 months) was also associated with a reduced survival (Table 3, p00.021 and p00.006, respectively). None of the other parameters had a statistically significant impact on survival. On multivariate analysis, the metastatic pattern of one lesion each in the liver and in the lungs vs. more than one lesion in either organ (hazard ratio 2.5, 95 % confidence

Int J Colorectal Dis (2013) 28:537–545

541

Table 2 Survival of all patients after surgical therapy for hepatic and pulmonary colorectal metastases in comparison to data in literature Author

Number

From primary tumour

From first metastases

From second metastases

Median survival

Median survival

Median survival

Survival rate (%) 5 years

10 years

Survival rate (%) 5 years

10 years

Survival rate (%) 5 years

10 years

Brouquet Headrick (clear margins only)

112 58

– –

– –

– –

– –

– –

– –

58

50 30a

~45 16a

Kawano Marudanayagam Miller Neeff Own data (all patients) Own data (clear margins only) Own data (involved margins)

35 43 131 44 65 51 14

67 – 83 – 62 68 37

65 – 65 64 57 61 39

32 – 32 – 15 18 0

– 42 60 – 48 53 37

– 54 49 42 37 43 13

– – 20 17 14 17 0

40a – 40 – 37 39 26

37a – 31 27 20 25 0

18a – 19 – 15 19 0

a

Survival given from first lung resection

survival depending on the analysed parameters is displayed in Table 4. On multivariate analysis, we saw the diameter of the (largest) metastasis (hazard ratio 2.2, 95 % CI 1.1 to 4.3, p00.029), DFI2 (hazard ratio 3.1, 95 % CI 1.5 to 6.5, p00.003) and the presence of solitary metastases both in liver and lungs (hazard ratio 3.5, 95 % CI 1.5 to 8.2, p00.005) as independent statistically significant parameters regarding survival.

interval (CI) 1.2 to 5.4, p00.021, Fig. 2c) and DFI2 (hazard ratio 2.5, 95 % CI 1.3 to 4.8, p00.006) proved to be independent statistically significant parameters. Liver metastases <5 cm and complete resection both showed a trend to better survival (hazard ratio 2.0, 95 % CI 0.9 to 4.5, p00.074, and hazard ratio 1.8, 95 % CI 1.0 to 3.2, p00.068, respectively). Long-term survivors (≥10 years) were seen only after complete resection of both metastases. No further statistical evaluation was performed for those cases with infiltrated resection margins due to the small sample size of 14 cases. None of these patients survived the second metastasis for more than 52 months.

Discussion Despite encouraging results after resection of hepatic metastases of colorectal cancer, the debate about the reasonability of surgery in the presence of both hepatic and pulmonary metastases continues. Data in literature are

2. Patients after complete metastatic resection, n051 For the 51 patients in whom complete resection had been possible, univariate analysis of median, 5- and 10-year

a

c

b Observed survival

1,0

Observed survival

1,0

0,8

Observed survival

1,0

0,8

0,8

R0

< 5cm

0,6

0,6

0,6

0,4

0,4

0,4

solitary both

41% one site multiple

R1/2 17%

0,2

5cm and more

0,2

9%

p=0.018 0,0

0,2

0,0

0,0 0

1

2

3

4

5

6

7

8

9

10

0

Years after resection of first metastasis

at risk

8%

p=0.036

1

2

3

4

5

6

7

8

9

10

0

1

R0

51

38

17

6

< 5cm

26

19

12

R1/R2

14

4

1

0

5cm

25

19

5

4

2

3

4

5

6

7

8

9

10

Years after resection of first metastasis

Years after resection of first metastasis solitary both

18

15

7

4

2 one site mult.

33

23

10

2

Fig. 2 Observed survival. a For 65 patients depending on resection status. b For 51 R0 resected patients depending on diameter of largest liver metastasis. c For 51 R0 resected patients depending on number of metastases in the liver and lung

0.002 0.009 (0.006) 0.059

Patient selection In the presence of resectable metastases (i.e. of sufficient remnant functional parenchyma), it is our policy to suggest metastatic resection to all patients fit enough to undergo surgery and to proceed with liver or lung resection without prior chemotherapy in order to avoid chemotherapyassociated parenchymal damage, particularly in the liver. We do not base our decision to offer surgical therapy on response to chemotherapy, although this has been discussed controversially for several years. Chances for cure may be reduced in patients with a poor response to chemotherapy, yet there are no high-quality data to prove the futility of a surgical approach in this situation. Therefore, we find it hard to justify an exclusion of these patients from a potentially curative therapy. In case of borderline or irresectable metastases, “neoadjuvant” chemotherapy is used for downsizing in order to achieve resectability.

0.140 0.083

0.216

n.a. 0.10 (0.012) 0.56

0.085

0.021

<0.0001 (<0.0001) n.a. 0.02 0.005 (<0.0001) 0.007

n.a. n.a. n.a. n.a. 0.711

0.12

n.a.

0.02

n.a. 0.001 (<0.001) n.a. n.a. 0.01 0.85

0.52

n.a.

n.a. n.a. n.a. 0.03 n.a. n.a.

n.a.

n.a.

0.03 (0.007) n.a. 0.783 0.604 0.897

n.a.

n.a.

n.a.

n.a. n.a. 0.01 (0.004) 0.1 (0.04) n.a. n.a.

0.11

scarce, rendering it difficult for surgeons to bring forward solid arguments in favour of metastasectomy. To the best of our knowledge, only three other series including more than 50 [8–10] and another three recent studies reporting on more than 30 patients [11–13] with liver and lung metastases of colorectal origin have been published. The long-term results of our study show similar survival rates after resection of both hepatic and pulmonary metastases compared to those after surgery only for hepatic metastases, adding impact to previous publications from other centres. Therefore, we strongly suggest that metastasectomy should be offered to patients with hepatic and pulmonary metastases of colorectal cancer if a complete resection is deemed feasible.

51.79 %

Survival

n.a. not assessed

65 2011 Own data

≤113, ≤86 % 131 2007 Miller

Unknown 43 Marudanayagam 2009

≤42, ≤95 % 44 2009 Neeff

58 58 2001 Headrick

35 2011 Kawano

<100 %

Resection of last metastastic site 1st pulmonary resection 1st pulmonary resection Resection of 1st metastasis Resection of 1st metastasis Resection of 1st metastasis Resection of 1st metastasis ≤92, ≤82 % 112 2011 Brouquet

(Largest) liver Pulmonary Interval, first to metastasis involvement at second ≥5 cm first metastasis metastasis Number of liver metastases CEA ≥5 ng/ml Location of primary tumour Year of Number Number of patients Starting point for Age publication of with negative survival analysis patients margins Author

Table 3 Prognostic factors on univariate and multivariate analysis in comparison to data in literature (significant p values on multivariate analysis given in brackets)

0.09 (0.08)

Int J Colorectal Dis (2013) 28:537–545 Positive resection margins

542

The outcome of patients solely undergoing liver resection for colorectal liver metastases in our hospital has recently been published [14]. When comparing these data with the 51 patients after complete metastasectomy described in this manuscript, we found similar results (median survival, 41 vs. 53 months; 5-year survival rates, 43 vs. 43 %, and 10-year survival rates, 28 vs. 17 %, calculated from the time of liver resection and resection of the first metastasis, respectively). The majority of other reports match these findings for the presence of either only lung or liver metastases vs. metastases in both organs [8, 10, 12, 15–17]. Nevertheless, the presence of further metastases in other organs must be regarded as an additional risk factor for diminished survival. Two studies including 16 and 17 patients detected reduced survival rates for those with liver and lung metastases compared to pulmonary metastases only [17, 18]. This difference to other series and our own data might be explained by small sample sizes in the respective studies.

Int J Colorectal Dis (2013) 28:537–545

543

Table 4 Prognosis after complete resection of hepatic and pulmonary colorectal metastases (clear resection margins in all locations), n051 Item

Number Observed survival from primary tumour Median Survival Survival p survival rate (%), rate (%), (months) 5 years 10 years

Age at primary tumour

Observed survival from first metastasis

Observed survival from second metastasis

Median Survival Survival p survival rate (%), rate (%), (months) 5 years 10 years

Median Survival Survival p survival rate (%), rate (%), (months) 5 years 10 years

0.032

0.023

0.142

<60 years

28

84

71

27

72

54

28

44

36

31

≥60 years

23

61

50

10

43

30

6

34

12

–

Colon

28

73

63

25

68

51

23

39

27

27

Rectum

23

61

59

11

50

35

12

38

23

12

Site of primary tumour

Site of synchronous metastases Liver

0.199

a

0.361

0.0003

0.466

0.0003

0.019 a

12

73

73

0

73

73

0

41

18

Lung

2

33

0

0

33

0

0

23

0

0

Both

7

50

32

0

50

32

0

50

32

0

Interval primary tumour to first metastases (DFI1) <12 months ≥12 months

0.202

33

68

54

10

18

71

72

31

Interval of first to second metastases (DFI2) <12 months 27 ≥12 months

24

–

0.974

68

51

10

46

31

31

0.061

0.844

43

22

12

30

28

28

0.031

0.874

54

43

19

40

22

17

43

23

15

84

79

21

73

62

21

39

27

21

Site of first metastases

0.422

0.672

0.553

Liver

36

72

70

18

57

49

17

39

22

19

Lung

5

40

40

20

40

20

20

23

20

20

Both

10

50

32

16

50

35

18

50

36

18

53

43

43

31

16

8

Number of metastases

0.020

Solitary both

18

84

74

45

One site, multiple

33

65

55

7

Number of liver metastases Solitary

25

73

73

29

Multiple

26

54

50

10

Distribution of metastases Unilateral, both One site, bilateral Diameter of liver metastases <5 cm ≥5 cm a

0.036 84

60

41

48

35

8

0.141

0.007

0.193 68

54

25

46

33

11

0.323

0.040 48

33

17

31

26

12

0.497

0.178

27

61

59

33

54

47

30

41

32

32

24

71

64

5

51

40

6

39

18

6

0.052

0.018

0.034

26

86

71

30

78

66

28

48

43

30

25

61

52

9

46

22

9

38

9

9

No patient at risk for longer than 7 years in this group

Caution is warranted when comparing survival rates and median survival times in literature because different starting points for calculation were chosen in all larger series. We have therefore calculated our own data for the complete group of patients accordingly, essentially leading to similar results (Table 2). Some authors calculated their survival rates from the date of the first pulmonary resection [9, 11]. For better comparability, we chose to display these data in

Table 3 under the heading “From second metastasis”, since the first metastatic lesion was located in the liver in approximately 90 % of the cases. Five-year survival rates range around 60 % after resection of the primary tumour, between 40 and 50 % after resection of the first and between 20 and 30 % after resection of the second metastasis. In contrast to this, Brouquet et al. published exceptionally good results with a 50 % 5year survival rate and a 45 % 10-year survival rate [8].

544

We would like to point out the fact that the 10-year survival rates counting from the date of the primary tumour and the first and the second metastasectomy are equal (18 vs. 17 vs. 19 %), while 5-year survival rates differ significantly (61 vs. 43 vs. 25 %). Since we are able to prolong survival for years even in a palliative situation with the use of modern chemotherapy protocols, we know that 5-year survival rates are no longer a measure for cure. There is ample evidence for the hypothesis that the actual 10-year survival means cure, whereas up to one third of the 5-year survivors will suffer a later cancer-related death [14, 19]. Prognostic factors The main problem in detecting reliable prognostic factors that all studies faced are the small sample sizes in combination with a broad variety of potential prognostic factors. This is particularly evident with respect to multivariate analyses. Takahashi [20], Brouquet [8] and Neeff [13], for example, found statistically significant differences in survival depending on the site of primary tumour. However, in the first study, patients with rectal cancer reached a 10-year survival rate around 70 %, while none of the patients with rectal cancer in the two other studies survived in 10 years. This even led to a statistically significant result on multivariate analysis in favour of colonic primaries in one of these reports. Multi-centre data collection and analysis might be helpful in order to avoid this statistical problem. Chemotherapy Most publications about the resection of liver and lung metastases of colorectal origin cover a time period of 12 years or more [8–11, 13, 20]. Both the indications for chemotherapy and the protocols used have changed frequently during the respective study periods. Furthermore, they depend on the preferences of the medical oncologists as well as on the patient’s general condition and age. We are aware of the fact that chemotherapy has influenced our survival data, but it is impossible to stratify our analysis according to the protocols used or analyse prognosis in relation to the response to chemotherapy with any statistically meaningful result. To the best of our knowledge, there are no data to prove a statistically significant impact of chemotherapy on long-term survival in this setting. Disease-free intervals The disease-free interval between the primary tumour and the development of metastases is widely accepted as one of the most important aspects regarding prognosis in patients with liver metastases only [14, 21, 22]. In our own patients, however, we saw a better prognosis for those with

Int J Colorectal Dis (2013) 28:537–545

metachronous disease (DFI1 >0) if all surgically treated patients were included (n065, p00.017). This difference disappeared for the subgroup of patients in whom complete tumour resection was possible (n051, p00.974). This phenomenon may be explained by the influence of residual tumour on prognosis, taking into account the fact that we found complete resection to be less likely in synchronous disease (DFI100, p00.017). Surprisingly, none of the other studies detected any difference in outcome for a DFI1 shorter or longer than 3 [9] or 12 months [8, 10, 23]. An interval of 12 months or more between the resection of the first and the second metastasis, on the other hand, was an independent prognostic factor both in the complete patient collective (n065, p00.006) and in the subgroup after complete resection (n051, p00.031), which supports the results in the publications mentioned above [10–12]. Although the disease-free interval may be regarded as a parameter of aggressiveness of the tumour, we did see longterm survivors even if both DFI1 and DFI2 were zero. Therefore, we believe that the surgical option should not be restricted to patients with a longer disease-free interval. Metastatic tumour burden and biology Preoperative level of CEA as well as number and size of metastases are widely accepted parameters for metastatic tumour burden. In two studies, patients with normal preoperative CEA had a better prognosis than patients with an elevated marker [8, 9]. We could not confirm this in our own data, but patients with solitary and small liver metastases did have a better survival than patients with multiple and larger metastases (Table 3). Similar observations led other authors to formulate the concept of “oligometastases” [24], describing a state in which tumour recurrence is limited to few regions and lesions. These patients do not develop diffuse cancer dissemination because the metastatic capacity of the tumour has not fully evolved, giving local therapy a chance to remove all gross neoplastic tissue. Additional perioperative systemic chemotherapy might destroy micrometastases, culminating in the cure of the patient [25]. The inconsistency of the prognostic factors described in literature suggests strongly that we will not be able to predict prognosis based on clinical data only. Differences in literature are most likely explained by the large number of variables in this complex patient collective, particularly because tumour biology is not taken into account. In truth, tumour biology is likely to be the most important factor to determine outcome apart from complete surgical tumour resection, but unfortunately also the least deciphered. Further research with a special focus on tumour genetics is ongoing in many cancer centres, including our own, and will hopefully bring forward more accurate prognostic parameters in the near future.

Int J Colorectal Dis (2013) 28:537–545

Conclusions After complete resection of liver and lung metastases in patients with metastatic colorectal cancer, a clear resection margin is the strongest prognostic factor on univariate analysis, while the diameter of the (largest) liver metastasis as well as the DFI2 proved to be independent statistically significant parameters. Differences in literature are most likely explained by the large number of variables in this complex patient collective, particularly because tumour biology is not taken into account. In view of our 10-year survival rates of nearly 20 %, suggesting a chance for cure in every fifth patient, we strongly recommend to offer the option of metastatic surgery to all patients with hepatic and pulmonary colorectal metastases, if complete resection is considered feasible.

Conflict of interest The authors declare that they have no conflicts of interest.

References 1. Jemal A, Siegel R, Ward E et al (2009) Cancer statistics, 2009. CA Cancer J Clin 59:225–249 2. Ballantyne GH, Quin J (1993) Surgical treatment of liver metastases in patients with colorectal cancer. Cancer 71:4252–4266 3. Valentini V, van Stiphout RG, Lammering G et al (2011) Nomograms for predicting local recurrence, distant metastases, and overall survival for patients with locally advanced rectal cancer on the basis of European randomized clinical trials. J Clin Oncol Off J Am Soc Clin Oncol 29:3163–3172 4. Abdalla EK, Vauthey JN, Ellis LM et al (2004) Recurrence and outcomes following hepatic resection, radiofrequency ablation, and combined resection/ablation for colorectal liver metastases. Ann Surg 239:818–825, Discussion, 825–817 5. Dhir M, Lyden ER, Wang A et al (2011) Influence of margins on overall survival after hepatic resection for colorectal metastasis: a meta-analysis. Ann Surg 254:234–242 6. Lee WS, Yun SH, Chun HK et al (2007) Pulmonary resection for metastases from colorectal cancer: prognostic factors and survival. Int J Color Dis 22:699–704 7. Yedibela S, Klein P, Feuchter K et al (2006) Surgical management of pulmonary metastases from colorectal cancer in 153 patients. Ann Surg Oncol 13:1538–1544 8. Brouquet A, Vauthey JN, Contreras CM et al (2011) Improved survival after resection of liver and lung colorectal metastases compared with liver-only metastases: a study of 112 patients with

545 limited lung metastatic disease. J Am Coll Surg 213:62–69, Discussion, 69–71 9. Headrick JR, Miller DL, Nagorney DM et al (2001) Surgical treatment of hepatic and pulmonary metastases from colon cancer. Ann Thorac Surg 71:975–979, Discussion, 979–980 10. Miller G, Biernacki P, Kemeny NE et al (2007) Outcomes after resection of synchronous or metachronous hepatic and pulmonary colorectal metastases. J Am Coll Surg 205:231–238 11. Kawano D, Takeo S, Tsukamoto S et al (2012) Prediction of the prognosis and surgical indications for pulmonary metastectomy from colorectal carcinoma in patients with combined hepatic metastases. Lung Cancer 75:209–212 12. Marudanayagam R, Ramkumar K, Shanmugam V et al (2009) Long-term outcome after sequential resections of liver and lung metastases from colorectal carcinoma. HPB (Oxford) 11:671–676 13. Neeff H, Horth W, Makowiec F et al (2009) Outcome after resection of hepatic and pulmonary metastases of colorectal cancer. J Gastrointest Surg 13:1813–1820 14. Settmacher U, Dittmar Y, Knosel T et al (2011) Predictors of longterm survival in patients with colorectal liver metastases: a single center study and review of the literature. Int J Colorectal Dis 26:967–981 15. Borasio P, Gisabella M, Bille A et al (2011) Role of surgical resection in colorectal lung metastases: analysis of 137 patients. Int J Colorectal Dis 26:183–190 16. Joosten J, Bertholet J, Keemers-Gels M et al (2008) Pulmonary resection of colorectal metastases in patients with or without a history of hepatic metastases. Eur J Surg Oncol 34:895–899 17. Limmer S, Oevermann E, Killaitis C et al (2010) Sequential surgical resection of hepatic and pulmonary metastases from colorectal cancer. Langenbecks Arch Surg 395:1129–1138 18. Landes U, Robert J, Perneger T et al (2010) Predicting survival after pulmonary metastasectomy for colorectal cancer: previous liver metastases matter. BMC Surg 10:17 19. Tomlinson JS, Jarnagin WR, DeMatteo RP et al (2007) Actual 10year survival after resection of colorectal liver metastases defines cure. J Clin Oncol 25:4575–4580 20. Takahashi S, Nagai K, Saito N et al (2007) Multiple resections for hepatic and pulmonary metastases of colorectal carcinoma. Jpn J Clin Oncol 37:186–192 21. Fong Y, Fortner J, Sun RL et al (1999) Clinical score for predicting recurrence after hepatic resection for metastatic colorectal cancer: analysis of 1001 consecutive cases. Ann Surg 230:309–318, Discussion, 318–321 22. Nordlinger B, Guiguet M, Vaillant JC et al (1996) Surgical resection of colorectal carcinoma metastases to the liver. A prognostic scoring system to improve case selection, based on 1568 patients. Association Francaise de Chirurgie. Cancer 77:1254–1262 23. Shah SA, Haddad R, Al-Sukhni W et al (2006) Surgical resection of hepatic and pulmonary metastases from colorectal carcinoma. J Am Coll Surg 202:468–475 24. Hellman S, Weichselbaum RR (1995) Oligometastases. J Clin Oncol 13:8–10 25. Niibe Y, Hayakawa K (2010) Oligometastases and oligorecurrence: the new era of cancer therapy. Jpn J Clin Oncol 40:107–111