The recently published FACS study had planned to investigate OS of intensive surveillance with CEA or CT, alone or in combination, compared to minimal follow-up. Unfortunately, due to recruitment issues, the primary end point was amended to surgical treatment of recurrence with curative intent.
FACS showed that between 12 and 20 patients needed to undergo intensive surveillance to detect 1 potentially resectable recurrence with curative intent, but there was no difference in OS20. The ongoing COLOFOL and GILDA trials are investigating different surveillance schedules without a control no surveillance arm27,28.
Given the uncertain OS benefit of intensive surveillance and outdated clinical trials, the aim of this study was to evaluate outcomes of intensive surveillance in a modern population-based cohort study. Furthermore, we focused the analysis on stage III CC patients, who have the highest risk of recurrence.
A retrospective chart review was conducted on all patients with resected stage III CC patients who initiated AC with either 5-fluorouracil (5-FU) or capecitabine plus oxaliplatin between 2006 and 2011 at the British Columbia Cancer Agency (BCCA). Institutional REB approval was obtained before beginning the review.
Surveillance recommendations at the British Columbia Cancer Agency (BCCA)
Over the study period, the BCCA had a standardized surveillance strategy for all resected stage III CC patients based on guidelines from the NCCN, ESMO and ASCO with the following recommendations: history and physical examination with CEA measured every 3 months for the first 2 to 3 years, and every 6 months thereafter until year 5; abdominal imaging in the form of a computed tomography (CT) of the abdomen and pelvis or ultrasound every 6 months for the first 2 to 3 years, and annually thereafter until year 5.
A yearly chest X-ray was recommended for patients with rectal cancer. Colonoscopy was to be performed within the first year of diagnosis and further endoscopic recommendations were made by the endoscopist.
Patients were referred back to primary care physicians for surveillance follow-up at the discretion of the treating oncologist at the BCCA.
Data was abstracted from the electronic medical record by two authors (Martin Smoragiewicz and Renata D’Alpino Peixoto).
Data collection included baseline patient characteristics, tumors characteristics from pathology and operative reports, whether patients underwent potentially curative metastasectomy, and post-recurrence chemotherapy treatment.
The method of recurrence detection was determined by which modality first prompted investigations leading to a diagnosis of a recurrence: surveillance CT, surveillance rising CEA, or symptoms prompting additional investigations.
Recurrences were classified as locoregional only (recurrence in the mesentery, retroperitoneum, peritoneum, or at the surgical anastomosis), liver and/or lung only, and other metastatic.
OS1 and OS2 were measured from the date of recurrence or date of initial surgery, respectively, to the date of death or last follow-up. Relapse-free survival (RFS) was determined from the date of initial CC surgery to the date of recurrence.
Baseline characteristics of patients with surveillance and symptom-detected recurrences were compared using Pearson chi-square test.
The OS and RFS survival curves were estimated using the Kaplan-Meier method and compared using the log-rank Mantel-Cox test.
All reported P values are two sided and 95% confidence intervals are included. Statistical significance was defined as a P value <0.05. Statistical analyses were performed using SPSS.
The BCCA treated 635 stage III CC patients with oxaliplatin based AC between 2006 and 2011. A total of 175 patients (27.5%) recurred after a median follow-up of 67.7 months. Most recurrences (149 patients, 85%) were identified by surveillance CT (44%) or a rising CEA (41%).
There were no differences in baseline characteristics between patients with recurrences detected by surveillance or symptoms, except for the latter group being slightly older (median age 61 vs. 66, P=0.03) and containing more high-grade tumors (19% vs. 7%, P=0.01).
However, the surveillance-detected recurrences were more likely to be liver and/or lung limited. Furthermore, patients with surveillance-detected recurrences were more likely to undergo potentially curative metastasectomy (39% vs. 7%, P=0.002) and receive any chemotherapy (70% vs. 50%, P=0.03) (Table 1).
Table 1. Patient Characteristics
|Characteristics||Surveillance detected (N=149, 85%)||Symptom detected (N=26, 15%||P|
|Patients, No. (%)|
|Age (median yr)||61||66||0.030|
|Male||72 ||11 ||0.570|
|Follow-up (median mos)||67.7||—||—|
|Right sided||75 ||15 ||0.480|
|T4||49 ||7 ||0.540|
|N2||82 ||14 ||0.910|
|Perineural invasion||30 ||4 ||0.570|
|77 ||15 ||0.570|
|Obstruction||25 ||6 ||0.430|
|Perforation||15 ||1 ||0.310|
|Rising CEA||72 ||—||—|
|Locoregional only||40 ||10 ||0.010|
|Liver and/or lung only||62 ||3 ||0.010|
|Other metastatic||47 ||13 ||0.010|
|Metastasectomy||59 ||2 ||0.002|
|Any chemotherapy||105 ||13 ||0.030|
|2 lines||42 ||5 ||—|
|3 or more lines||11 ||3 ||—|
|Note: CT, computed tomography.|
Compared with symptom-detected recurrences, patients with surveillance-detected recurrences had a shorter median RFS (18.5 vs. 25.3 months, P<0.001, HR 1.82; 95% CI: 1.16-2.85, Figure 1A), and longer median OS post recurrence detection (OS1, 28.5 vs. 6.5 months, P<0.001, HR 0.37; 95% CI: 0.23-0.60, Figure 1B).
However, the median OS from the time of initial surgery was not significantly different between these groups (OS2, 50.9 vs. 39.1, P=0.091, HR 0.66; 95% CI: 0.41-1.07, Figure 1C).
(To view a larger version of Figure 1, click here.)