Introduction

Strategies for solving the inequalities in cancer care and disappointing survival rates have triggered the formation of multidisciplinary team (MDT) of specialists that have improved processes and outcomes [1]. The MDTs have been recommended as a new standard of care for malignant diseases to improve diagnosis, treatment planning and outcomes [2, 3]. Few pieces of evidence, however, exist to demonstrate the direct benefit of MDTs in cancer survivorship [4]. The most important advantage of the MDT approach is that it allows a wide range of diagnosis to be discussed while providing physicians information to facilitate treatment planning in challenging scenarios.

Many hospitals have also recently adopted MDTs in the management of all patients with colorectal cancer (CRC). Physicians who treat CRC must consider all potential treatment plans such as neoadjuvant and adjuvant chemotherapies, surgery, and palliative therapies under the exact diagnosis using multiple modalities. It is very difficult for an individual specialist to make a decision considering the wide variety of these treatment strategies. Some papers have described the effects of MDT discussions on survival [5] and improved decision-making processes in patients with CRC [6].

The MDTs demand greater time and resources. Medical team members would need to render uncompensated hours, whereas treatment duration may be prolonged [7, 8]. There was also a negative opinion about the implementation of MDTs for all patients with CRC. Predominant benefits of the MDT process were achieved with advanced diseases [5], whereas much of the time spent for early or localized diseases may be futile [9, 10]. Many studies to date have shown improvements in the standardization of care and an increased proportion of patients receiving this standard [11,12,13]. It is reasonable to suggest that a better patient outcome would result from more patients receiving the standard of care.

Despite the belief that MDT meetings play an important role in the management of specific CRC categories considering the expenditure of time and resources, there have been no data to elucidate this. Our hypothesis was that this CRC category, which more often requires multimodal therapy, would be frequently influenced by MDT discussions. The purpose of this study was to assess the impact of MDT meetings on the treatment plan of patients with complex CRC. In addition, we evaluated the number of MDT recommendations implemented and the factors that influenced MDT discordance.

Materials and methods

We analysed records of patients referred to MDTs from 1 January 2011 to 31 December 2014. All consecutive cases referred for discussion presented locally advanced, metastatic and recurrent CRCs and colonic metastasis of other cancers for which clinical specialists had difficulty making decisions by themselves. For consistency, only patients with histologically confirmed CRC were included in the analysis, whereas those presenting early CRC and miscellaneous colonic lesions from other organs were excluded. We did not assess individual encounters in patients who had previously been admitted to MDTs and were being re-evaluated as part of a continuing decision process. Considering the high volume of patients in our tertiary referral centre, only complicated cases and diagnosis, for which individual clinicians had difficulty making a therapeutic strategy, were included in the MDT clinics.

MDT

MDT meetings for colorectal malignancies occur four times a week and continue until decision-making is completed for each patient. The meeting is led by a colorectal surgeon and starts when all participants are present. These participants consist of specialists including colorectal surgeons, medical oncologists, radiation oncologists, radiologists and oncological nurse coordinators, as well as gastroenterologists, hepatobiliary surgeons and chest surgeons as required. After the discussion of each individual patient, a consensus is reached on the treatment plan or further diagnostic work-up.

Data collection

Specific information was prospectively collected on patients and disease characteristics, including demographic data, histological type, disease extent and radiological stage, metastatic or recurrence site and previous treatments. Clinical specialists’ initial diagnosis and proposed treatment plan and the MDT’s consensus recommendations for assessment and treatment were also collected. Consensus recommendations were subsequently compared with the referring clinician’s initial treatment plan, which was recorded before the MDT meeting.

All processes done by the MDT were digitally recorded in the prospectively designed medical data base form

After case discussions, assessment and treatment plans were recorded on the remnant part of the case form by the moderator. Differences recorded between the referring clinician’s plan and MDT’s consensus recommendations included changes in assessment, such as additional lesions, disease extent and misdiagnosis, and treatment, such as changes from nonoperative to operative and operative to nonoperative treatment and changes in nonoperative treatment approaches.

For this analysis, the referring doctor’s initially proposed treatment plans were compared with MDT recommendations; ‘complete application’ meant that definitive treatment plans were the same as the proposed treatment plans, whereas ‘optional application’ meant that one of the proposed plans was accepted into the definitive treatment plans. ‘Alteration’ meant that an alternative treatment plan was added to the proposed plans or that none of the proposed plans were accepted into the definitive treatment plans. We followed the cohort of patients assessed and treated by the MDT to assess the implementation of MDT recommendations. Data were collected after institutional review board approval from the Asan Medical Center.

Statistical analysis

Data have been reported as percentages and means (standard deviations). To evaluate factors influencing MDT decision and treatment implementation, univariate logistic regression was used. Results were analysed using the SPSS software (IBM Corp. Released 2011. IBM SPSS Statistics for Windows, Version 20.0. Armonk, NY: IBM Corp). P < 0.05 was considered statistically significant.

Results

Throughout the study period, data from 1383 patients after 467 consecutive MDT meetings were included. Clinical details and MDT decisions of therapeutic plans are summarized in Table 1. Most of the patients were presented to the panel by the medical oncologist (47.4%), followed by the colorectal surgeon (39.3%) and gastroenterologist (9%). A total of 549 (39.8%) database forms were completed for patients with newly diagnosed CRC. Two hundred and ninety-seven of newly diagnosed CRC cases (54.1%) were referred for treatment of metastatic lesions. In a total of 833 recurrent diseases, the most common site was the liver (42.7%), followed by the lungs, distant lymph nodes and local recurrence. In addition, 129 (15.5%) cases had recurrence in more than two sites.

Table 1 Patient characteristics
Full size table

Initially proposed treatment plans by referring physicians were changed in 179 (12.9%) patients. The rate of change in therapeutic plan increased depending on the reason for presentation, with newly diagnosed cancer patients having an overall change in 7.6% (42 of 549 presentations) and recurrent cancer patients in 16.4% (137 of 833 presentations) of their MDT presentations (P < 0.001). All 179 patients were presented to MDTs to determine resectablility and oncological benefit of resection in advanced and recurrent lesions.

Changes in therapeutic plans according to MDT recommendations included a change to nonsurgical treatment in 119 patients (66.5%), modifications to the surgical approach and a change to no treatment in 54 patients (Table 2). In the 119 cases re-evaluated for disease extent, treatment plans were altered to palliative chemotherapy in 45 cases due to unresectable lesions and to neoadjuvant chemotherapy in 45 cases for marginally resectable lesions. A total of 10 patients with marginally resectable lesions (10/45) underwent primary lesion resection and neoadjuvant chemotherapy for metastasis. In 29 of the 119 cases re-evaluated for disease extent, another treatment modality was decided upon without altering the intent of the treatment. Either radiofrequency ablation (RFA) or stereotactic body radiation therapy (SBRT) was finally decided for anatomically difficult lesions in eight patients referred for surgery. Moreover, surgery combined with RFA or SBRT for curative purposes was decided in six other cases that were also confirmed to have unresectable lesions. In 54/179 (30.2%) cases that were altered to have no treatment, the diagnosis of metastasis or recurrence was altered to benign lesions in 43 patients. Hence, diagnostic follow-up was decided upon. Re-evaluation of the general condition in 11 patients meant that the MDT decided to use a conservative treatment based on the assumption that the patient’s systemic condition could not withstand the suggested treatment.

Table 2 Reasons for changes in the treatment plan (n = 179)
Full size table

The impact of radiological review is shown in Table 3. Among the 179 cases with therapeutic plan alteration, the review of the radiologist under full clinical information from MDT members affected the diagnosis and decision in 85 (47.1%) cases. The most common change in the diagnosis after radiologist review was the alteration from metastasis or recurrence to benign lesions in 40/85 (47.1%) cases. After reviewing radiological findings, resectability considering disease extent and general condition of patients was altered in 45 (52.9%) cases.

Table 3 Change in radiology reports (n = 85/179)
Full size table

Although follow-up of all patients receiving the therapeutic plan showed 89% adherence to MDT recommendations, 152 MDT decisions were not implemented. Logistic regression analysis showed that treatment intent, nonsurgical approach and alteration of decision were important reasons for not following MDT recommendations (Table 4).

Table 4 Factors for the discordance between MDT decision and implementation of the treatment plan
Full size table

Discussion

The benefits of MDTs haves made it a new standard of care for cancer patients despite limited supporting data [2]. However, conducting MDTs cost a significant amount of time and money for medical personnel and hospitals. In addition, the MDT approach is often fragmented, resulting in the lack of communication between providers, which causes a significant delay in the treatment [14]. In our high-volume tertiary cancer center, MDTs could not accommodate all of the cases which included over 7000 patients treated for newly diagnosed CRC throughout the study period, because of restrictions on time, despite being performed four times a week.

For efficiency, we also determined whether all patients with CRC should be discussed at MDT meetings. The lack of evidence regarding the benefit of a MDT in early-stage colon cancers suggests that they might be managed by protocols and guidelines, with referring clinicians having the option of MDT inclusion. One study analysing the large number of patients with low-stage colon cancer demonstrated no alteration of management after the MDT process and advised that patients can be managed using protocols [15]. On the other hand, a MDT for early cancers may have a role in certain aspects. A study on a specialist MDT for early colon cancers demonstrated an improvement in preoperative staging, a reduction in margin positivity and an increase in the use of local excision following the implementation of recommendations [16]. Therefore, MDT approach was appropriate for diagnostic and therapeutic uncertainty, not for certain disease types or cancer stages. The potential advantage of a MDT consisting of expert clinicians is believed to be because of the better results derived by discussions than by a single expert. Comparing decisions from an MDT to that of a single physician can be an important index in determining the utility of a MDT.

Our MDTs were held for four sessions a week and at least five medical providers participated. One session was allocated to 2 h, and the maximum of were patients were discussed at 30-min intervals. MDTs conducted discussion for an average of three patients (1387 cases out of 467 MDT meetings) for each session. The mean full-time equivalent (FTE) unit that indicates the workload of provider on our MDTs, assuming a full-time employee works 40 h per week, is 0.75. We did not take into account the fact that participating medical staff gave information about patients to be discussed at least 2 working days before the meetings and participated in the discussion after each review, but 0.75 of FTE could be considered to be very effective when compared with the results obtained through our MDT meetings.

Overall, this analysis revealed that MDT meetings have a significant impact in approximately one out of 10 patients with complex CRC. This is a unique study that looks into the efficacy of MDT meetings in changing the management of advanced and recurrent CRCs. Review of imaging by specialized radiologists under efficient and comprehensive communication with clinicians greatly influences changes in the management plan. Of the 43 patients having discrepancies between the diagnoses of the referring physician and the MDT meeting in terms of discriminating ambiguous relapse and metastatic lesions, 40 underwent no treatment after a MDT review of imaging because of the lower possibility of relapse and metastasis.

The rate of diagnostic discrepancies in this study was not as high as that in previous reports, which had studied a variety of malignancies to evaluate the discrepancies in diagnosis. One of the most remarkable discrepancies was reported from the breast cancer literature, wherein review of imaging and review of pathology led to interpretation changes in 45 and 29% cases, respectively. An MDT approach in head and neck cancers demonstrated that staging refinement was affected in over two-thirds of the patients [17]. The reason for the lower discrepancies in our study was thought to be because of the ease in colonic lesion biopsy using colonoscopy and the rarity of ambiguous pathological findings in CRCs. In addition, clinicians who referred patients to MDTs were also CRC specialists at our institution.

The high volume of patients with recurrent CRC being referred to our MDT for discussion may suggest that individual physicians, even those specializing in CRC, might not feel confident in diagnosing and/or managing such patients, especially when medical decisions regarding the possibility and effects of surgery are difficult to make. The results of this study showed that changes in the therapeutic plan were more frequently observed in those having recurrent CRC than those who were newly diagnosed.

Factor that had the greatest influence on the change of primary treatment plan was due to the radiographic review as in Table 3. The second cause of change in the treatment plan (64/179 of the changes) was a reassessment of the possibility of resection by surgeons. Changes to palliative chemotherapy for an unresectable lesion were 23 cases. Neoadjuvant chemotherapy for a marginally resectable lesion was planned in 29 cases. RFA or SBRT was performed in seven patients with lesions difficult to undergo surgical approach. And RFS or SBRT was combined with surgery in five cases. Third, 19 cases with surgically resectable lesions were changed to minimally invasive RFA or SBRT with similar oncological outcome. Another reason was that in the case where the clinician integrated the patient’s condition and the course up to the present and judged that both surgery and chemotherapy were difficult, it was eight cases. Although the metastasis was suspected on image findings, three cases were clinically diagnosed as inflammatory lesions rather than metastasis.

Implementation rates can be an index of the effects of MDT discussions. The rate of implementation in our study was 1231/1383 (89.0%). This high rate of implementation suggests that the colorectal MDT in this setting is an effective forum for making appropriate decisions. A few papers have studied the implementation rates for MDT recommendations. In the study from Plymouth [10], the implementation rate was 44/47 (93.6%), whereas in the study from Bristol [18] it was 137/157 (87.3%). In the current paper, the most common reason for not following MDT recommendations was patient choice, which considered the recommended treatment unacceptable or not beneficial for survival. This was because our MDT mainly included newly diagnosed patients who had expected a bad prognosis and recurrent cancer patients who had multiple lesions. It is essential to know the factors affecting implementation in special types of diseases and MDT settings to improve implementation rates. The implementation of our colorectal MDT’s recommendations was attributed to palliate intent, nonsurgical approach and change in treatment plan. Through this, low implementation can be inferred when patients received unsure recognition and expected unclear therapeutic effects.

The weaknesses of this study included the retrospective study design, despite prospective data collection, and inherent limitations in the heterogeneous cohort, which was selected based on the discretion of individual physicians. This resulted in a potential selection bias, wherein more difficult and controversial cases were included. Despite patient selection, changes in the treatment plan were not higher than those in studies of other organs. This study of selected complex group may not be generalizable outside of high volume, highly specialized centres. However, the results suggest that our MDTs may be an effective option in hospitals that are difficult to implement for all CRC patients. The heterogeneous cohort in this study makes it difficult to analyse the improvement in oncological outcomes of this type of MDT.

The strengths of this study comprised the inclusion of more than 1300 patients with complex CRCs and the limitation to single-institution analysis of our own data wherein our MDT had consistent referral indications despite the heterogeneous group. This unique study has objectively determined whether the introduction of MDT meetings altered the diagnosis and management of patients with complex CRC. Future studies should include outcome measurement to evaluate whether MDT contributes to the improvement of oncological prognosis for patients with metastatic and recurrent colorectal diseases.

Conclusion

Discussion of patients with complex colorectal cancer at MDT meetings results in the alteration of clinical decisions. Half of these alterations were based on radiologist reviews under efficient and clear communication with clinicians. The present study demonstrates that imperfection decisions by individual physicians can be decreased by the process of multidisciplinary decision-making. Moreover, these types of MDTs, which include only selected patients with complex diseases, could be an efficient alternative to the regular MDT.