The strengths and limitations of routine staging before treatment with abdominal CT in colorectal cancer

Advanced colorectal carcinoma (CRC), defined as locally advanced or metastasized disease or both, is present in a relevant proportion of patients diagnosed with colorectal cancer. Common localizations for distant metastases are the liver, the peritoneal cavity and the lung. Staging with chest CT as a routine procedure before surgery has not shown to be of clinical benefit, mainly due to the low incidence of clinically relevant lung metastases and low specificity of chest CT [1‐3]. Pre-operative staging with abdominal CT might be beneficial when the ability to detect advanced CRC is high enough and the findings offer information that may change the treatment plan. Such findings include liver metastases (LM), peritoneal carcinomatosis (PC) and locally advanced colon cancer (LACC). In the past, these conditions were frequently regarded as incurable and suitable for palliative measures only. Nowadays various multi-modality treatments offer a chance of cure to selected patients [4‐15]. For patients with incurable advanced CRC, alternative options are sometimes available and staging may change the treatment plan towards the 'best palliative care', that may consist of avoidance of surgery in selected patients [16‐18]. At the present, adequate staging with CT before treatment is already considered as the standard in international CRC guidelines. However the evidence for this advice in literature is mostly limited to studies focused on the radiological accuracy and also are usually conducted in selected patient groups. The actual findings of staging after implementation of such guideline has not been described before. An observational study may be helpful in the debate on actual clinical relevance and provide lead points to optimize staging in clinical practice.

The aim of this study is to describe the outcomes of routine staging with abdominal CT in a unselected hospital population (n = 612) after the introduction of a regional guideline recommendation in 2007.

The data were collected in the Medical Spectrum Twente, a large community teaching hospital in the regional capital of a foremost rural area in the eastern part of the Netherlands. It functions as a regional referral center for liver and lung surgery, but has no facilities for the treatment of peritoneal carcinomatosis with hyperthermic intraperitoneal chemotherapy (HIPEC).

Routine pre-operative staging with a CT of chest and abdomen for patients with CRC was introduced as a regional CRC guideline in 2007 and preceded similar national guideline recommendations (2008). CT scanning was performed on a 16 and 64 slice scanner (Toshiba Aquillion 16 and 64) after intravenous contrast injection (visipaque 320, 90 ml, 3 ml/s.) in the portal venous phase, with a slice thickness of 1 mm and a reconstruction of 0.8 mm. When preoperative scanning was omitted, staging with abdominal CT was intended within 3 months after surgery. Patients with rectal cancer, defined as localization below the peritoneal reflection, were additionally staged with a pelvic MRI for determination of the local invasion and possible lymph node metastases (cTN stage) and received neo-adjuvant (chemo)radiation according to the Dutch guidelines on rectal cancer. Follow-up after curative treatment of non-metastatic CRC consisted of serum CEA measurements every 3 months combined with bi-annual ultrasound of the liver.

Pathological staging was based upon the TNM classification 2002 (6th edition) and classified according to the American Joint Committee on Cancer (AJCC) stages. Advanced CRC was defined as either locally advanced disease, presence of distant metastases or both. Locally advanced colon cancer (LACC) was defined by pT4 stage; meaning the tumor showed invasion through the serosal layer or into surrounding organs. Locally advanced rectal cancer (LARC) was defined as all patients that had either a T4 tumor or a T3 tumor with a threatened circumferential margin on pelvic MRI. The final diagnosis of liver metastases was based upon radiological (CT, contrast-enhanced ultrasound and/or PET scanning) and per-operative findings. In case of resection or in case of persistent uncertainty, histological confirmation was obtained. The final diagnosis of peritoneal carcinomatosis was by histological confirmation. The final diagnosis of lung lesions was on radiological grounds (chest CT). In case of resection or indeterminate lung lesions, histological confirmation was obtained when feasible. Several lung lesions remained indeterminate: these were followed by repeat CT scanning and considered positive when growth was observed [3]. Incurable CRC was defined as all macroscopical irradical (R2) resections of the primary tumor, when the patient had no resection of the primary tumor, or when no intended curative treatment of distant metastases was done.

Emergency presentation in the surgical patients was defined as all non-planned admissions to the hospital due to symptoms related to the tumor, with a subdivision into 'urgent' defined as surgery imperative within 5 days and 'acute' procedures within 6 hours.

The ability of the staging abdominal CT to detect advanced disease was analyzed in surgical patients that were staged with CT before treatment; the gold standard for PC and LACC were per-operative findings confirmed with histology. For LM the findings on CT (negative for liver metastases or indeterminate lesions) were related to per-operative findings and follow-up.

In a 3 year period (2007-2009), 612 patients were diagnosed with colorectal carcinoma in the study center. Staging with abdominal CT before treatment was done in 513 patients (84%). Surgery, either palliative or curative, was performed in 551 patients (90%). The proportion of surgical patients with an emergency presentation was 21% (115/551).

Advanced CRC is common, occurring in 58% of CRC patients in the study. It raises little question at the present that staging before treatment is required. It is presumed the abdominal CT can accurately identify the patients with advanced CRC. For liver metastases and LACC the abdominal CT indeed is an adequate first-line imaging technique, however for PC it grossly lacks sensitivity. Appreciating the relationship between LACC and PC, LACC may serve as a warning sign for the possible presence of PC. The incidence of advanced CRC was especially high in patients with emergency presentations, while staging tended to be omitted more often in this group.

The present study was done in a single hospital and therefore has a limited number of patients as compared to true population-based studies. The additive value of this analyses is the prospective design, that is inherently more accurate and provides more detailed clinical data. The outcomes are expected to represent population based incidences fairly good, due to the fact the study was done in an relatively isolated region in the Netherlands that knows very little referral or selection bias. In comparison to a large population-based study from the Netherlands [19] this study reports a higher incidence of mCRC (colon cancer 34% vs 25% and rectal cancer 22% vs 24%). This variance in incidence of mCRC probably has several causes; such as due to differences in staging routines, mode of data registration, CT scanning protocols and the adverse regional socio-economic status.

Considering potentially curable mCRC, the relevance and outcomes of staging before treatment is most well-known for liver metastases. Staging with CT before treatment provides the opportunity to change the treatment strategy; various studies have reported promising results of alternative approaches in terms of eligibility of resection and the oncological outcome [10, 11, 20]. Previous reports show a high accuracy of CT for liver metastases, with an initial sensitivity of approximately 85% which is similar in this study [21]. One study concluded that also small and indeterminate liver lesions should be reported, to optimize the detection rates of true metastases[22]. Our observations agree with that conclusion; Indeterminate lesions in the liver were not a major concern and discrimination of indeterminate liver lesions with additional imaging caused no major diagnostic uncertainties or resources. Contrast-enhanced abdominal MRI is an equivalent alternative to multi-slice CT scanning for LM, [23] however may cause more logistic problems.

Concerning PC, the benefits of staging before treatment are much less outspoken. In the recent past PC was regarded as a virtually incurable condition with little treatment options. Therefore accurate staging of PC was considered less important. This has changed since the introduction of HIPEC offering a chance for cure in selected patients [6, 12, 13]. Since HIPEC is performed in specialized centers only, accurate pre-operative staging of PC has become vital to improve the outcome of these patients. The present study confirms the finding of previous studies that PC is poorly visualized on a CT scan [24]. Other imaging techniques may be more sensitive to detect and estimate the extent of PC, such as diffusion-weighted imaging (DWI) combined with MRI [25]. This however remains to be proven for its clinical value in colorectal cancer. An alternate approach may be to utilize the observed close relationship between PC and LACC, e.g. by performing a diagnostic laparoscopy in patients with a cT4 tumor on CT; estimation of T stage on abdominal CT in this and another study [26] seems to be fairly reliable.

Standardized radiology reports on T stage in colon cancer and on localization, sizes, spread and preferably 'weighed' suspicion of distant metastases may enhance the ability to find advanced CRC with the staging abdominal CT.

All authors have approved on the final version of this manuscript

IG: Surgeon, fellow in oncological gastrointestinal surgery. From february 2010 till july 2011 she worked as a fellow at the Catharina Hospital and starting from July 2011 she continues her fellowship at the Medical Spectrum Twente (MST). Further she holds a position as researcher at the UMCG. She previously (also) worked at the MST as a resident in general surgery. This study was part of her thesis "Searching for metastases in colorectal cancer" (University of Groningen).

JK: Oncological gastrointestinal surgeon, specialized in gastrointestinal and hepatobiliary surgery. Tutor for scientific research at the Medical Spectrum Twente.

JA: Staff radiologist at the Medical Spectrum Twente

IdH: Oncological gastrointestinal surgeon, specialized in treatment of peritoneal carcinomatosis (PC) and head of the research group on PC at the Catharina Hospital.

WM: Oncological and gastrointestinal surgeon, specialized in colorectal and head/neck surgery. Responsible tutor of the trainees in general surgery at the Medical Spectrum Twente.

TW: Professor in oncological surgery and former head of department at the UMC Groningen.