Keywords colorectal cancer - peritoneal carcinomatosis - cytoreductive surgery - HIPEC
Introduction
Peritoneal carcinomatosis (PC) of colorectal origin is considered an advanced terminal
disease. The reported incidence of synchronous and metachronous PC varies widely,
from 3 to 28% and from 4 to 19%, respectively. [1 ] Treatment is based mainly on palliative chemotherapy, but, unfortunately, colorectal
PC does not respond well to systemic chemotherapy like other sites of distant metastases.
[2 ]
[3 ] Hyperthermic intraperitoneal chemotherapy (HIPEC) comprises the direct pumping of
heated chemotherapy into the peritoneal cavity after surgery. Its rationale includes
killing the micrometastatic disease and the minimal residual of gross disease by exposure
of the diseased peritoneum to the higher concentration of chemotherapeutic agents
while keeping its systemic plasma levels low. Another advantage is that the venous
drainage of the peritoneum is via the portal vein to the liver, which provides a detoxifying
effect to the administered drug and helps killing the potential micrometastatic hepatic
deposits.[4 ] Hyperthermia can selectively destroy malignant cells at between 41 and 43°C by different
mechanisms and enhance the cytotoxic effect of the chemotherapeutic agents.[5 ]
[6 ]
Many studies reported that treatment of colorectal PC with cytoreductive surgery (CRS)
and HIPEC was associated with better local control and survival compared with systemic
palliative chemotherapy.[7 ]
[8 ]
[9 ] Identifying patients at high risk for recurrence following CRS and HIPEC at an early
stage could further improve the oncologic outcome. Many studies reported that repeated
CRS and HIPEC were feasible, associated with low morbidity, and with superior oncologic
outcome when compared with palliative chemotherapy alone.[10 ]
[11 ]
[12 ]
The present study aimed to assess the efficacy of CRS and HIPEC compared with CRS
only on the prognosis of colorectal cancer patients diagnosed with PC and to identify
possible clinicopathological factors associated with the recurrence of PC.
Patients and Methods
The present retrospective study included all cases of colorectal carcinoma with PC
from January 2009 to June 2018 operated at the National Cancer Institute (NCI), Cairo
University, Cairo, Egypt.
Inoperable cases and peritoneal disease of noncolorectal origins were excluded. The
medical records of the patients were retrieved from the Epidemiology department, NCI,
Cairo University. The extracted data were demographics, clinicopathological characteristics
of the patients and of the primary tumor, investigation results, PC treatment and
outcome.
All patients were operated on with complete CRS only or with CRS plus HIPEC. A midline
skin incision was made from the xiphoid process to the pubic tubercle, resectioning
the affected parts. Electrosurgery was used for implants on visceral or intestinal
surfaces where resection or excisions of the nodules were done for infiltrative lesions.
For those who underwent HIPEC, the drains and thermal probes were connected to the
extracorporeal circuit of the HIPEC machine (Therma solution 2000).
Intraoperatively, the extent of peritoneal involvement was assessed by the peritoneal
carcinomatosis index (PCI).[13 ] The PCI is calculated as the summation of the size of implants in the abdominopelvic
regions in a score that ranges from 0 to 3 (0: no malignant deposits, 1: nodules < 0.5 cm
in their greatest dimension, 2: nodules of 0.5 to 5.0 cm, 3: nodules > 5.0 cm).
The completeness of surgical resection was assessed by the completeness of cytoreduction
(CC) score.[14 ] A CC-0 is apparent when there is no peritoneal seeding visualized within the operative
field. CC-1 indicates nodules persisting after cytoreduction < 2.5 cm. CC-2 indicates
nodules measuring between 2.5 and 5 cm, whereas CC-3 indicates nodules > 5 cm or a
confluence of unresectable tumor nodule at any site within the abdomen or the pelvis.
After treatment, the patients were followed-up by clinical examination, radiological
imaging, and serum tumor markers (CEA and CA 19.9).
Peritoneal recurrence was defined as any new lesion detected by noninvasive radiological
imaging (computed tomography [CT] or positron emission tomography [PET-CT] scan) with
or without biopsy compared with the first imaging performed 3 months after treatment.
In lesions detected by endoscopy or reoperations, recurrence was defined by pathological
tissue examination. Extraperitoneal recurrence was defined as metastasis to the liver,
the lungs, bone, and as the recurrence in the retroperitoneum and local colonic recurrence.
Follow-up and Survival
The patients were followed-up at 3 monthly intervals by clinical examination, radiological
imaging, and serum tumor markers (CEA and CA 19.9) for 2 years, and then every 6 months
for an additional 3 years.
Objectives
To detect recurrence-free survival (RFS) and its predictors among patients with colorectal
cancer following CRS only versus CRS plus HIPEC.
Statistical Methods
Statistical analysis was done using IBM SPSS Statistics for Windows, version 22 (IBM
Corp., Armonk, NY, USA). Data were expressed as frequency and percentage. The Pearson
chi-squared test or the Fisher exact test were used to test the relationship between
qualitative variables.
Recurrence-free survival was calculated from the date of surgery until the date of
recurrence. Survival analysis was done using the Kaplan-Meier method, and a comparison
between two survival curves was made using the log-rank test.
Multivariate analysis was done using the Cox-proportional hazard regression model
with the forward likelihood ratio method for the factors affecting survival on univariate
analysis. Hazard ratios (HR) with their 95% confidence intervals (CIs) were used for
risk estimation. All tests were two-tailed. A p -value < 0.05 was considered significant.
Results
The present study included 61 patients. Two patients were lost to follow-up very early
after the procedure, so they were excluded from the RFS analysis. About 59% of the
patients were ≥ 45 years old. Females represented 54.1% of the sample. Adenocarcinoma
corresponded to about half of the cases of primary tumors. About three-fourths of
the participants had a low-grade primary tumor. Out of 46 patients with positive nodes,
33 (75.0%) had an extracapsular invasion. About one-third of the patients presented
with a T4 primary tumor. Almost all patients (93.4%) were asymptomatic regarding the
presentation of the primary peritoneal disease. The peritoneal disease was synchronous
in 43.3% of the participants and metachronous in 56.7%. More than 12 lymph nodes (LNs)
were harvested during resection of the primary tumor in 36 cases (61.0%), compared
with 23 cases (39.0%) with < 12 harvested LNs. A PCI ≤ 10 was evident in 37 patients
(61.7%), and > 10 in 23 patients (38.3%). CC0 and CC1 were achieved in 54 patients
(88.5%) and in 7 patients (11.5%), respectively. Most of the patients (73.8%) were
treated by CRS + HIPEC ([Table 1 ]).
Table 1
Demographic, clinicopathological, and treatment characteristics of the studied group
Number
Percentage
Age (years old)
< 45
25
41.0
≥ 45
36
59.0
Gender
Male
28
45.9
Female
33
54.1
Primary tumor characteristics
T stage
T2 + T3
43
70.5
T4a + T4b
18
29.5
N stage (
n
= 59)[* ]
N0
13
22.0
N1
22
37.3
N2
24
40.7
M stage
-ve
33
54.1
+ve
28
45.9
Histopathological type
Signet ring
6
9.8
Mucinous
23
37.7
Adenocarcinoma
32
52.5
Grade
Low (G1,2)
47
77.0
High (G3,4)
14
23.0
LNs harvested during resection (
n
= 59)[* ]
≤ 12
23
39.0
> 12
36
61.0
Lymphovascular invasion (
n
= 51)[* ]
Yes
32
62.7
No
19
37.3
Extracapsular node invasion (
n
= 44)[** ]
Yes
33
75.0
No
11
25.0
Peritoneal disease characteristics
Presentation
Symptomatic
4
6.6
Asymptomatic
57
93.4
Onset
Synchronous
27
44.2
Metachronous
34
55.7
PCI score (
n
= 60)[* ]
≤ 10
37
61.7
> 10
23
38.3
CC score
CC0
54
88.5
CC1
7
11.5
Surgical procedure
CRS only
16
26.2
CRS + HIPEC
45
73.8
Abbreviations: CC, completeness of cytoreduction; CRS, cytoreductive surgery; HIPEC,
hyperthermic intraperitoneal chemotherapy; LN, lymph nodes; LVI, lymphovascular invasion;
PCI, peritoneal carcinomatosis index.
* Some data missing.
** for N1 and N2 only.
The median follow-up period was of 14 months (range: 6 to 72 months). The cumulative
RFS of the whole group was 55.7% at 1 year and 19.9% at 3 years, with a median of
12 months. Most of the recurrences (76.9%) occurred within the 1st year after treatment. Most cases developed peritoneal relapse either alone or in
association with distant metastasis or local relapse; only one patient developed local
relapse only ([Tables 2 ] and [3 ]).
Table 2
Recurrence after treatment of primary peritoneal disease
Recurrence after CRS ± HIPEC
Yes
26
42.6%
No
35
57.4%
Time to recurrence (months) (
n
= 26)
Early < 12 m
20
76.9%
Late ≥ 12 m
6
23.1%
Site of recurrence (
n
= 26)
HIPEC + CRS
CRS
Total
Peritoneum only
6
8
14
Peritoneum plus distant metastasis
4
1
5
Peritoneum plus local recurrence
2
1
3
Local recurrence
0
1
1
Distant metastasis only
2
1
3
Abbreviations: CRS, cytoreductive surgery; HIPEC, hyperthermic intraperitoneal chemotherapy.
Table 3
Patients and disease characteristics in relation to 1-year recurrence-free survival
(RFS) after treatment of colorectal peritoneal carcinomatosis
n
Failures (n )
RFS (%)
Median survival (months)
p -value
Whole Group
59
26
55.7
12
Surgery for primary peritoneal disease
CRS
15
12
23.8
9
0.002
CRS + HIPEC
44
14
66.2
35
Age (years old)
< 45
23
14
48.8
11
0.415
≥ 45
36
12
58.2
35
Gender
Female
32
16
36.6
11
0.105
Male
27
10
69.0
56
Primary tumor characteristics
T stage
T2 & T3
31
14
62.4
35
0.035
T4 a & b
28
12
37.2
11
N stage
N0
13
2
76.9
NR
0.087
N1
22
9
51.0
35
N2
24
14
44.0
12
M stage
Negative
31
7
75.5
35
0.009
Positive
28
19
33.0
9
Final stage
Stages 1 & 2
10
1
80.0
NR
0.024
Stage 3
21
6
73.4
35
Stage 4
28
19
33.0
9
Histopathological type
Adenocarcinoma
31
9
73.6
35
0.061
Signet ring
6
4
40.0
9
Mucinous
22
13
38.5
20
Grade
Low grade
46
17
64.5
16
0.007
High grade
13
9
13.2
9.0
LNs harvested during resection (n = 59)[* ]
≤ 12
23
11
24.8
10
0.06
> 12
36
14
67.4
35
LVI (n = 51)[* ]
Yes
32
20
34.2
10
0.001
No
19
4
89.7
35
Extracapsular node invasion (n = 44)[* ]
Yes
32
21
39.0
10
0.005
No
11
3
83.3
35
Peritoneal disease characteristics
Onset
synchronous
26
16
38.9
11
0.073
metachronous
33
9
68.3
35
PCI score
≤ 10
37
17
48.3
12
0.546
> 10
22
9
61.2
35
CC score
CC0
53
23
55.4
13
0.270
CC1
6
3
66.7
NR
Abbreviations: CC, completeness of cytoreduction; CRS, cytoreductive surgery; HIPEC,
hyperthermic intraperitoneal chemotherapy; LN, lymph node; LVI, lymphovascular invasion;
PCI, peritoneal carcinomatosis index.
NR: No median RFS because more than half of the patients of this group did not develop
recurrence until the end of the study.
* Some data missing.
In the univariate analysis, the following variables were associated with worse RFS:
T4 tumor (1 year RFS: 62.4% for T2,3 versus 37.2% for T4, p = 0.035), presence of lymphovascular invasion (1 year RFS: 34.2% for LVI+ versus
89.7% for LVI-, p = 0.001), presence of extracapsular nodal invasion (1 year RFS: 39 versus 83.3%,
p = 0.005), high grade tumors (1 year RFS: 64.5 for low grade versus 13.2% for high
grade, p = 0.007), and patients treated for their PC with CRS only (1 year RFS: 23.8 for CRS
only versus 66.2% for CRS + HIPEC, p = 0.002). Other variables that tend toward statistical significance of worse RFS
include patients with N2 nodal stage at the initial surgery (1 year RFS: 76.9, 51,
and 44% for N0, N1, and N2, respectively, p = 0.087), harvesting < 12 LNs during resection of the primary tumor (1 year RFS:
24.8 versus 67.4%, p = 0.060), patients with mucinous histology (1 year RFS: 73.6, 40, and 38.5% for adenocarcinoma,
signet ring carcinoma, and mucinous carcinoma, respectively, p = 0.061), and patients with synchronous peritoneal disease (1 year RFS: 38.9 versus
68.3% for synchronous and metachronous disease, respectively, p = 0.073). ([Table 3 ]). Using the Cox-regression model, RFS was independently affected by tumor grade
and type of surgical management of the primary peritoneal disease ([Table 4 ] and [Figs. 1 ] and [2 ]).
Table 4
Independent factors affecting recurrence-free survival of the whole studied group
p -value
HR
95%CI for HR
Lower
Upper
Management of primary peritoneal disease (CRS versus CRS + HIPEC)
0.011
2.924
1.282
6.667
Grade of primary colonic disease (high grade versus low grade)
0.037
2.488
1.057
5.857
Abbreviations: CI, confidence interval; CRS, cytoreductive surgery; HIPEC, hyperthermic
intraperitoneal chemotherapy; HR, hazard ratio.
Fig. 1 Recurrence-free survival in relation to the surgical management.
Fig. 2 Recurrence-free survival in relation to the grade of the primary colonic disease.
Discussion
The present study demonstrated that the RFS was 52.2% at 1 year in patients with peritoneal
carcinomatosis on top of CRC treated with CRS and/or HIPEC. The RFS was independently
worsened by higher tumor grade and primary peritoneal disease management with CRS
only. Other factors that appeared to worsen the prognosis were T4 tumor, lymphovascular
invasion, and extracapsular nodal invasion.
These results agreed with the study by Verwaal et al.,[15 ] who reported a recurrence rate of ∼ 64% after a median follow-up of 47.5 months
with a median time to recurrence of 13.7 months, despite the longer follow-up period
in their study. They found that most recurrences occurred within the 1st year. A systematic review and meta-analysis including 27 studies reported a recurrence
rate of between 22.5 and 82% after CRS and HIPEC for PC of colorectal origin.[16 ]
In the present study, the most common histopathological type was adenocarcinoma (52.5%),
followed by the mucinous (37.7%) and signet ring (9.8%) types. Many studies confirmed
the higher frequency of adenocarcinoma (between 70 and 90%) compared with the mucinous
(between 10 and 20%) and signet ring (between 1 and 7%) types.[17 ]
[18 ]
[19 ]
We observed a higher recurrence rate in patients with T4 primary tumors than in those
with T2 and T3 tumors (RFS: 37.2 versus 62.4%, p = 0.033). Our results are in line with those of Segelman et al.[20 ] reported that the Independent predictors for metachronous PC were colonic cancer
(hazard ratio (HR) 1.77, 95 per cent confidence interval 1.31 to 2.39; P = 0.002 for right-sided colonic cancer), advanced tumour (T) status (HR 9.98, 3.10
to 32.11; P < 0.001 for T4) as compared to 0.60 (0.15, 2.32) for T2 in the multivariate analysis.
Taylor et al.[21 ] also reported a lower recurrence rate with a higher 5 year disease free survival
for T1, T2, and T3a.
The present study demonstrated that the onset of peritoneal metastases negatively
affected recurrence with borderline significance. Synchronous onset tends to have
higher recurrences after treatment than the metachronous counterpart (RFS: 38.9 versus
68.3%, respectively, p = 0.071). Hentzen et al. described earlier recurrence after CRS with HIPEC for metachronous
PC compared with synchronous PC (HR: 1.63; 95%CI: 1.18–2.26); however, Hompes et al.
reported no impact of the onset of peritoneal metastasis on disease-free survival.[22 ]
[23 ]
In the present study, the use of HIPEC with CRS resulted in better RFS compared with
CRS only (RFS: 23.8 versus 66.2%, p = 0.002). These findings are consistent with those of Chua et al., who found a poorer
RFS in a group of 2,298 patients with pseudomyxoma peritonii treated with CRS compared
with those treated with CRS and HIPEC (HR: 0.65; p = 0.030). The results obtained by Quenet et al. reported a median RFS of 11.1 months
(95%CI: 9.0–12.7) in the non-HIPEC arm and of 13.1 months (95%CI: 12.1–15.7) in the
HIPEC arm (HR: 0.90; 95% CI: 0.69–1.90; p = 0.486), while the 1-year RFS rates were 46.1 and 59% in each arm, respectively.[24 ]
[25 ]
In the present study, we observed a 1-year RFS of 64.5% for low grade versus 13.2%
for high grade tumors (p = 0.007) both in the univariate (p = 0.007) and in the multivariate analysis (p = 0.037). Günther et al. reported that primary tumor grading reflected the individual
tumor phenotype and its biological behavior better than the immunohistochemical studies,
and it was the independent predictor of metachronous distant metastasis in the studied
group.[26 ]
The present study demonstrated that the presence of lymphovascular invasion negatively
affected recurrence (1-year RFS: 34.2 for LVI+ versus 89.7% for LVI-; p = 0.001). Another single-center analysis of 1,616 patients also reported a negative
impact of lymphatic invasion on the RFS (HR: 2.0449; 95%CI= 1.4932–2.8365; p = 0.01).[27 ]
Conclusion
The clinicopathological characteristics of the primary tumor appear to be significant
predictors of recurrence following treatment of PC. These include the T4 stage, high
grade, +ve lymphovascular, +ve extracapsular nodal invasion, and treatment with CRS
only. On the multivariate analysis, the RFS was independently affected by the primary
tumor grade and by the type of surgical management of the primary peritoneal disease.
