Introduction

A uniform international consensus regarding the optimal management of rectal cancer patients has yet to be developed. Recommendations and guidelines have been published [1–11], but guidelines may lag behind current standards in therapy and it is not known if they are used consistently or universally, even within their respective countries. Differences in practice may influence rectal cancer survival and morbidity. Furthermore, adherence to evidence-based clinical practice in colorectal surgery has been shown to be inconsistent in several reports [12, 13].

Scientific comparison of staging and preoperative treatment of rectal cancer may be hindered by the variety of practice patterns within countries and across international borders. In this context, it is important to evaluate inherent international trends and differences in the management and treatment of rectal cancer. We have conducted a survey regarding current practices of preoperative staging and treatment of rectal cancer among an international panel of colorectal surgeons. The aim of the study was to identify regional differences among international experts that may influence our interpretation of current and future studies regarding the preoperative management of rectal cancer. Furthermore, we wanted to detect the impact of multidisciplinary team meetings and department caseload upon the preoperative decision-making.

Materials and methods

The study group

A group of six surgeons from the US, Australia, and Europe constituted the working members of an International Rectal Cancer Study Group (IRCSG). All surgeons had more than 10 years of experience with rectal cancer. IRCSG is an independent group without financial support from any companies or organizations and has representatives from Asia, Europe, and North America. The aim of the study group was to evaluate current international treatment practices of rectal cancer.

Invited colorectal surgeons to IRCSG

Centers and representative surgeons were selected based on prior publications, presentations, or participation at national or international meetings and via a PubMed search for scientific reports on rectal cancer. All respondents were invited to become members of IRCSG. To ensure an experienced international panel, we used two strategies:

  1. 1.

    PubMed search We performed an unsystematic search using the search term rectal cancer combined with staging, treatment, chemotherapy, radiation, surgery. Based upon this search, corresponding authors were identified and included in the survey sample.

  2. 2.

    Oral presentation at an international scientific meeting The working committee of IRCSG participated in several international rectal cancer conferences from 2006 to 2008. Oral presenters in these conferences were identified and included in the survey sample.

The questionnaire

The questionnaire was developed and validated during several meetings of IRCSG. The aim was to cover all aspects of preoperative rectal cancer treatment. Modifications to the survey items were made by literature review and e-mail discussions. The questionnaire consisted of 59 questions. An average of 20 min was needed to complete the questionnaire.

Study logistics

The survey was sent to the identified colorectal surgeons both as a document attached to an e-mail and as a Web-based survey link forwarded to each participant. The Web-based option was offered through SurveyMonkey.com. Two follow-up e-mails were sent to nonresponders, and the survey was open for a total of 16 weeks.

Review of national guidelines

Six central rectal cancer treatment guidelines were reviewed (Table 1) to compare national recommendations for radiological T staging and neoadjuvant treatment. The aim was to gain insight in similarities and differences of guideline recommendations.

Table 1 Guideline recommendations for radiologic T staging and neoadjuvant treatment of rectal cancer [ 2, 6–8, 10, 11]
Full size table

Statistics

Descriptive statistics were performed by percentages, 2 × 2 contingency tables, and Fisher’s exact test. For comparison purposes, respondents were divided into three groups: US-based surgeons, non-US surgeons, and total. Bivariate logistic regression analyses were performed to detect the impact (measured by relative risk [RR]) of department caseload and team meetings upon preoperative decision-making. All tests were two-sided and p < 0.05 was considered statistically significant. All data were analyzed using SPSS v.16.1 (SPSS, Inc., Chicago, IL).

Results

One hundred seventy-three colorectal surgeons were identified and contacted via e-mail. After one initial e-mail and two follow-up e-mails, 123 (71%) surgeons responded. One hundred one responded via the Web-based questionnaire, 20 by returning the questionnaire as an e-mail attachment, and 2 by fax. Of those that returned the survey, 110 (89%) fully completed all parts of the questionnaire.

Demographics (Table 2)

The colorectal centers were located in 28 countries representing five continents: 53 in North America, 43 in Europe, 18 in Asia, 8 in South America, and 1 in Africa. Seventy-eight percent were university hospitals.

Table 2 Hospital affiliation, department caseload, and rectal cancer surgical experience (n = 123)
Full size table

Rectal surgery experience (Table 2)

Ninety-three percent of responding surgeons have more than 5 years’ experience with rectal cancer surgery, and 70% work in departments that manage more than 50 rectal cancers per year. Twenty percent of the surgeons perform more than 50 rectal cancer operations annually.

Preoperative staging (Table 3)

Fifty-five percent of surgeons prefer CT scan, 35% MRI, 29% rectal ultrasound, 12% digital rectal examination, and 1% PET scan in all rectal cancer patients for preoperative staging. All use more than one imaging modality in all rectal cancer patients for preoperative staging. Significantly more non-US surgeons use MRI on all rectal cancer patients compared to US surgeons (42% vs. 20%, p = 0.03). US surgeons prefer ERUS significantly more than non-US surgeons do (43% vs. 21%, p = 0.01). Similarly, we detected a difference in PET scan use: 53% of non-US surgeons and 25% of US surgeons never use PET scan (p = 0.005).

Table 3 Preference of image modality for staging of rectal cancer
Full size table

Indications for preoperative chemoradiotherapy (CRT) (Table 4)

Eighty-one surgeons (73%) consider threatened circumferential margin (CRM) an indication for neoadjuvant treatment. Of those, 40% consider 1 mm or less, 39% consider 2 mm or less, and 21% consider 3 mm or less as threatened CRM. Sixty-seven (61%) always give neoadjuvant treatment to both stage II and stage III rectal cancer patients. However, two US surgeons and two non-US surgeons answered “others” for “give preoperative CRT treatment to all T3 cancers and greater,” which means stage II or higher.

Table 4 Indications for preoperative chemoradiotherapy
Full size table

Ninety-two percent of US surgeons compared to 43% non-US surgeons give CRT for stage II and stage III rectal cancers (p = 0.0001). A significantly higher proportion of US surgeons consider rectal cancer with poor histological differentiation an indication for CRT (10/39 vs. 5/71, p = 0.008). Cancer in the distal third of the rectum is an indication for preoperative CRT for 15 (14%) surgeons, while cancer in the distal two-thirds is used by five (5%) surgeons, and one (1%) gives neoadjuvant treatment to all rectal cancer patients.

Neoadjuvant treatment and other preoperative considerations (Table 5)

Ninety-two percent prefer 5-fluorouracil (5-FU)-based long-course neoadjuvant CRT. Ten percent (4) of the US surgeons and 16% (12) of non-US surgeons prefer short-course radiation therapy. A significantly higher proportion of the US surgeons have radiation therapy available at their hospital (89% vs. 74%, p = 0.04), and significantly more of non-US surgeons have regular rectal cancer audits (74% vs. 51%, p = 0.01).

Table 5 Neoadjuvant treatment and preoperative considerations
Full size table

Institutional radiation rate (Fig. 1)

Twenty-nine of 39 US surgeons have an institutional irradiation rate (IRR) >50% and 31/71 non-US surgeons have an IRR >50% (p = 0.001). However, nine (8%) surgeons answered that their institution offers radiation to more than 90% of all rectal cancer patients, and a similar proportion (8%) have an IRR of 10–29%.

Fig. 1
figure 1

Institutional irradiation rate (IRR) for rectal cancer at 123 international centers. Significantly more US centers (29/39) have IRR >50% compared to non-US centers (31/71) (p = 0.001)

Full size image

Impact of multidisciplinary teams and caseload (Table 6)

Departments with regular multidisciplinary team meetings are more likely to prefer MRI for local staging (RR = 3.62), and there is a trend toward significance (p = 0.06). Similarly, patients with threatened circumferential margin are more likely to receive neoadjuvant treatment in departments with team meetings (RR = 5.67, p = 0.03). Other significant impacts of team meetings were found upon pathology report quality (RR = 4.85, p = 0.01), new chemotherapy regimen if there are liver metastases (RR = 6.41, p = 0.02), and one-stage surgery when there are liver metastases (RR = 0.25, p = 0.02). Similar influences of caseload upon preoperative decision-making were not observed.

Table 6 Impact of caseload and multidisciplinary teams upon preoperative descision-making
Full size table

Discussion

This is the first survey of current practice among international colorectal centers regarding preoperative management of rectal cancer. The results demonstrate a wide variation in preoperative staging procedures, inconsistencies in indications for preoperative CRT, and differences in treatment procedures for identical rectal cancers. These variations highlight the need for more and better scientific evidence to help guide rectal cancer treatment as well as the need for international focus upon development of guidelines.

There exist several guidelines for rectal cancer treatment, at both national and international levels. New guidelines have recently been published [9]. However, these guidelines vary in their recommendations (Table 1). Similarly, practice may vary between countries and continents because surgeons prefer treatment according to their own guidelines. Implementation of national guidelines of clinical practice seems slow, as pointed out in other surveys [12, 13] and has been demonstrated among colorectal surgeons [14–18]. The differences in preoperative staging and treatment options cannot be explained by the heterogeneity of the institutional affiliations of the participants. As the vast majority of the responding colorectal surgeons work in academic medical centers, the differences in current practices reflect the lack of clinical evidence or the slow implementation of such evidence. Evidence from the US about adherence to preoperative staging and treatment options suggests marked variation. For the eight centers that participate in the National Comprehensive Cancer Network (NCCN), adherence to guidelines and quality measures is variable [18]. Concordance with guidelines is recently discussed in an editorial by Browman [19]. Certainly, it is important to indicate how well a recommendation is aligned with the evidence from which it is derived, but it is uncertain how practitioners interpret, respond to, or act upon a recommendation. We think our study reflects this argument, showing inconsistent practice also within continents and national borders.

Variations in preoperative staging imaging will select identical rectal cancer patients to undergo different treatment regimens depending on the sensitivity and specificity of the selected imaging modality. MRI and rectal ultrasound have better diagnostic properties than CT for rectal cancer [20, 21], but still CT is used for local staging by 54.5% of the surgeons compared to 34.5% that prefer MRI for all rectal cancer patients. The distance to the CRM has been increasingly recognized as an important factor and a surrogate marker for local recurrence. In this context, MRI has increased in popularity because of its ability to help decide the distance to the CRM and it should be used widely for preoperative treatment planning. CT has little or no place in describing the distance to the CRM because of its low spatial resolution. A recently published review recommends the use of MRI for all rectal cancer patients [22]. In our study, 11% of respondents never use MRI and approximately 50% use it in selected cases. In our opinion these numbers are surprisingly low and might reflect the slow implementation of evidence-based medicine among colorectal surgeons. All centers in this study use more than one staging procedure. CT has a role in evaluating the infiltration of other organs, usually for large T4 tumors. In addition, CT is used to determine metastatic disease in the liver and/or lungs (M stage in the TNM classification). Endoscopic rectal ultrasound (ERUS) has a role in evaluating small tumors for which local excision might be feasible. Retrospectively, it might be argued that some surgeons participating in the survey answered the question with respect to M staging and that this might bias the results regarding CT use (54% use it on all patients). However, the questionnaire specifically says “CT scan of the pelvis,” i.e., local staging, which reflects the actual use of CT for local staging (either alone or in combination with other modalities). In contrast, we do believe that CT is the most common modality for M staging.

Generalization of the results from this study must be made with care. The invited surgeons were selected based upon publications listed in PubMed and presentation or participation at national or international meetings, thus respondents could be prone to selection bias. However, the respondents were from large university hospitals throughout the world (78%) as well as from other large hospitals with a relatively high volume of rectal cancer patients, and the participants are all published authors and teachers at national meetings. Similarly, the surgical experience among the respondents is high: 93% of the responding surgeons have experience with rectal cancer treatment for more than 5 years, and 35% have experience for more than 20 years (Table 2). Thus, in our opinion the respondents are thoroughly experienced and good representatives for their national rectal cancer practice.

This survey has revealed prominent variations in practice standards in the preoperative staging and treatment of rectal cancer. The survey emphasizes the need for establishing an international consensus for the management of patients with rectal cancer and for identifying areas for future research. An international consensus is needed in order to develop staging and preoperative treatment standards if outcomes are to be compared between institutions and countries. Focusing on standardization and documentation as a process may also improve results [23]. The survey also points out the need for agreement on the staging and neoadjuvant treatment modalities used if we are to compare outcomes in rectal cancer treatment.

The different indications for neoadjuvant treatment will select noncomparable groups of patients in outcome studies. Different staging procedures and treatment algorithms could certainly reflect the published differences in local recurrence rates and 5-year cancer-specific survival rates [24, 25]. Heald [26] standardized the surgical technique for rectal cancer surgery with a significant impact on the risk of local recurrence. Probably an evidence-based standardization of preoperative practices will also benefit rectal cancer patients [27].

The wide variation in indications for neoadjuvant treatment and radiation rates should alert both national and international rectal cancer expert organizations as well as health-care administrators. The individualized treatment plans for rectal cancer patients should reflect the complexity of the disease and not geography or the surgeons’ or institutions’ preferences and traditions. Establishment of multidisciplinary teams (MDTs) comprising surgeons, oncologists, radiologists, and pathologists can hopefully elevate the decisions regarding the individual rectal cancer patient to a level where documentation and up-to-date knowledge are focused. In our study we have shown that MDTs significantly influence preoperative decision-making (Table 6). Interestingly, regular MDT meetings significantly influence decisions on choice of staging modality, neoadjuvant treatment, and several other critical factors in the preoperative planning of rectal cancer treatment. We believe that regular MDT meetings will improve guideline adherence and quality of rectal cancer care, as recently addressed by Taflampas [22]. According to our analysis, department caseload does not have as much influence on preoperative decision-making.

Over- and understaging as well as over- or undertreatment will have an impact on health-care costs, frequency of side effects of CRT, and surgery, including a patient’s quality of life and local recurrence and cancer-specific survival. Continuous education of rectal cancer surgeons seems necessary to increase implementation of evidence-based clinical practice. Heath-care authorities should develop quality control systems in order to ensure adherence to national and international recommendations and guidelines. The national and international societies associated with rectal cancer treatment have the challenge of developing such updated evidence-based recommendations.