Does shared decision making exist in oncologic practice?

We are both clinical oncologists and have between us some 50 years of experience in seeing cancer patients. Countless patients have been evaluated by us during consultations in which decisions have to made about whether (neo-)adjuvant treatment should be given, and, if so, what treatment. During the most recent decade, the majority of the patients have had a rectal cancer referred for preoperative radiotherapy or chemoradiotherapy (RT/CRT) or a radically resected gastrointestinal cancer referred for adjuvant chemotherapy. These consultations have, on the one hand, been very similar and part of a routine where basically the same message has been delivered, but, on the other hand, more and more dissimilar and increasingly challenging. We know the mean survival gains from the trials. We believe we pretty well know the potential gains for the patients considering stage, sub-stage depending on risk factor profile and patient-related factors of importance for treatment-related toxicity, but we do not know beforehand the patient’s own preferences considering potential gains in relation to harms. The ambition has often been to let the patient make the final decision, or at least greatly influence the decision after having been adequately informed about the benefits and harms. This has definitely not always been successful; many patients simply state that “the doctor should decide”. Most patients need more time for contemplation than what is available and many would benefit from a second appointment to mull over their alternatives one more time.

Today the above procedure, i.e. to explain to the patient that there is no clear best choice, that a decision has to be made, and that doing nothing is also an adequate option is called shared decision making (SDM) and has been the subject of research, including a recent systematic overview of the studies [1,2]. After the options have been presented to the patient, the next step is to discuss the benefits and potential harms of each of them, and their respective probabilities. An article about SDM, published in this issue of Acta Oncologica [3] has highlighted this process and its associated challenges. Most oncologists probably agree that SDM is a good thing when it comes to consultations aimed at deciding on (neo-)adjuvant treatments. However, the authors found very few if any of the hundred consultations for preoperative rectal cancer RT or postoperative adjuvant chemotherapy for breast cancer fulfilling minimum criteria that could be set for SDM. According to the authors, the oncologists in the study simply missed the opportunities for SDM by not expressing that a decision about therapy or not had to be made.

The generalizability of the study must be discussed. It was done at one center in one country, it included only two diagnoses and two specific clinical situations, and the numbers of oncologists were limited. Is the situation similar in other countries or at other centers? This is of course not known, although in our view quite possible. The study is well performed including both qualitative and quantitative research methods and the results appear relevant for many other environments and clinical situations. The two clinical situations studied were also well selected and it is our belief that they are relevant for not only other treatments in these diagnoses but also for many other diagnoses.

Balancing gains and losses in rectal cancer radiotherapy

In rectal cancer, the gains and harms from preoperative RT/CRT are well known [4]. Briefly, it reduces the risk of a local failure by 50–70% with the higher value in earlier tumors/with better surgery. There is thus a gain even in early tumors operated by the best surgeons, but the absolute gain is very small (1–3%) if the absolute risk of local failure is small (3–5%), as it is in early stages (cT1-2, 3ab mrf- unless very low, N0; often designated early or good tumors). If the risk is slightly higher (5–10%, cTN+, cT3cd, mrf−; intermediate or bad), the absolute gain is higher (4–8%). In these early and intermediate rectal cancers the absolute gain in local failures is, however, too small to result in a detectable survival gain, whereas this is seen in the more advanced tumors (cT3 mrf+, cT4, designated locally advanced or ugly) (average gain about 10% in the trials). In these tumors, the risk of local failure, either a local recurrence or non-radical surgery can be anything from 15% to 100% if the tumor cannot be removed. The adverse effects, i.e. the harms, both in the short- and long-run, in surviving patients are basically the same irrespective of tumor stage, although they can be influenced by patient-related factors like age and comorbidities. The harms are also well known, particularly after short-course RT [5], albeit with the radiation techniques used in the past. These can be expected to be lower in patients treated today due to the rapid developments in the entire radiation therapy process during the past decades [6]. The profession has balanced the gains and harms, and expressed them in different guidelines [7–9]. An absolute gain of about 5% is by many oncologists considered sufficient to recommend preoperative RT/CRT, although there is no consensus for this figure. The discussion at an MDT should in the early-immediate stages focus on the question “do we recommend this patient preoperative treatment to further diminish the risk of a local failure?” This recommendation should then not only be passed on to the patient by the physician, but the patient should be given the possibility to make an independent decision. The opportunity for such SDM is according to Kunneman et al. [3] almost absent. In several cases, the decision may appear apparent, either to irradiate or not to, but in many cases, particularly in the intermediate group (generally referred to as locally advanced), it is not.

Balancing gains and losses in breast cancer adjuvant chemotherapy (and radiotherapy)

Adjuvant breast cancer treatment is similar to rectal cancer RT in that for some patients, the (potential) gains are marked and therapy should be recommended, in other patients it is so small that it should not be given and in the remaining, gains are seen but are they sufficient considering the losses? Similar to the situation in rectal cancer, there is no consensus about what gains should be achieved in relation to toxicity for oncologists to recommend and patients to accept adjuvant therapy although several studies have been performed. According to Dutch guidelines, a minimum gain of approximately 4% (absolute) in breast cancer deaths is required in order to advise adjuvant therapy [10]. Other guidelines, e.g. the American National Comprehensive Cancer Network (NCCN) breast cancer guidelines, recommend that adjuvant therapy should be considered for tumors also when the gain is about 1% [11]. In an investigation by a Dutch group [12], most oncologists today consider a gain of 6–10% needed to recommend treatment, a figure not changed since a similar investigation during the 1990s [13]. Since a marked increase in the use of adjuvant chemotherapy has occurred during the time period, the authors suggest that oncologists do not adhere to their own minimally desired treatment effects to recommend therapy. Balancing the gains and losses by both doctors and patients are stressed. However, again, according to Kunneman et al. [3], SDM is virtually not part of routine when adjuvant chemotherapy is discussed, and, although not studied, likely also the same when adjuvant RT for breast cancer is discussed. It would have been valuable to know the stages and tumor characteristics in the article to better evaluate whether the potential gains were limited or more marked for most included patients referred for adjuvant chemotherapy.

Balancing gains and losses for adjuvant chemotherapy in colorectal cancer

Although not studied by Kunneman et al. [3], adjuvant chemotherapy for colon and rectal cancer has common features with adjuvant chemotherapy for breast cancer and for rectal cancer RT/CRT. The gains are, at least in colon cancer well known and so are the losses. In rectal cancer the gains are more uncertain [14,15] whereas the losses are the same since the treatments used are identical.

Adjuvant chemotherapy for rectal cancer has been the subject of much controversy during many years, and the controversy remains even after several recent trials being published and meta-analyses of the trials [14–16]. It was also recently discussed in an editorial in Acta Oncologica [17]. In one of the systematic overviews, we wrote that due to the uncertainties in the magnitude of the gains a “shared decision making should take place, i.e. before an individual decision is made, patients must be informed about the uncertainties of the presence of a small treatment effect, the toxicity and about the lack of consensus between experts. Judgement about whether the benefit outweighs the harm is subjective and should be left to the patient’s discretion”. There was no certainty that those easily written and uncontroversial words would have an impact on how doctors act in front of the patients. After having read the article by Kunneman et al. [3], one interpretation is that it is likely that, in practice, most patients with rectal cancer stage II + III will get adjuvant chemotherapy unless local or national guidelines explicitly state that adjuvant chemotherapy should not be given for that particular stage.

Stage migration has changed the absolute gains from (neo-)adjuvant therapy

The gains in survival from (neo)adjuvant therapy were obvious already in trials run decades ago. Due to improvements in most aspects of medical care since then, e.g. in imaging, surgery and pathology, the stage-specific risk of recurrence is likely much less than it was when the trials were run. It is also likely that the relative gains from therapy seen in the trials are still valid and may potentially even be slightly higher. However, due to this stage migration, the absolute gains have become much smaller and the balance between benefits and harms different. This has been little considered when guidelines have been written/discussions held with individual patients. The risk of local recurrence in rectal cancer in a particular stage may have been 10% some years ago and may today be less than 2–3%. Similarly, the risk of recurrence in colon cancer stage II with no or only one risk factor may have been 20% but can now be less than 10% – a change that is poorly described in literature and thus uncertain [18], but that will have a profound influence on the value of giving adjuvant therapy, increasing the importance of SDM.

However, it should also be mentioned that the losses may be less than was the case in the trials due to developments in supportive care. Further, the technical development in the delivery of RT has substantially decreased the risk of radiation-induced morbidity, both in rectal [4] and breast cancer [19–21].

Priorities between diagnoses

The Swedish Board of Health Care regularly presents national guidelines as a means of supporting those who make decisions concerning the allocation of resources within Health and Medical Care (https://www.socialstyrelsen.se/nationalguidelines). Priorities are given for different interventions in specific clinical situations after a systematic overview of the evidence base. Besides gains and losses, also financial aspects play a role. The simultaneous evaluation of breast-, colorectal- and prostate cancers recently allowed a calibration of priorities for similar interventions between diagnoses (https://www.socialstyrelsen.se/nationalguidelines/nationalguidelinesforthetreatmentofbreast-prostateandcolorectalcancers). It became apparent that breast cancer oncologists value much smaller gains higher than gastrointestinal cancer oncologists do, and the priority figures for adjuvant chemotherapy in both colon and rectal cancer became much higher and more similar to breast cancer for similar gains, than was the case in the previous evaluation done by the board separately some years ago. Representatives from patient organizations were part of the process, but whether the differences in traditions reflect opinions of differences among the patients or not are not known. Much more research is then needed about patient preferences. Even so, the SDM process should take place in every consultation.

The role of the MDT conference

The MDT decision ought to be “referral to a (medical/radiation) oncologist” and the recommendation behind it “for a discussion with the patient balancing the gains and losses of the particular treatment based upon their preferences”. The latter appears not frequently fulfilled [3]. Further support of limited practice of SDM may come from data from the Swedish National Colorectal Cancer Quality Register [22]. In colon cancer patients younger than 75, the MDT recommended treatment (referral to an oncologist) in 26% in stage II (likely those with risk factors) and in 83% in stage III. Treatment was actually initiated in the majority of those patients (87% in stage II and 93% in stage III). The reasons for not given adjuvant chemotherapy are not known, but are of course multifactorial. Medical reasons like complications to surgery with delayed recovery and comorbidities not fully recognized at the MDT conference could explain most of the reasons for not giving chemotherapy, but doctors and patients preferences could play a role, i.e. a SDM process could have taken place in a few individuals resulting in a decision for no treatment. In rectal cancer, the proportions receiving therapy were slightly lower (stage II 67% of 24% planned for therapy, stage III 84% of 70% planned). Besides the reasons given above, several Swedish hospitals participated in the SCRIPT study [23] and a reason could thus be that they were randomized to no treatment. Taken together, these figures indirectly may indicate that a SDM process has not frequently taken place. The MDT conferences promote treatment according to guidelines, at least in rectal cancer [24].

We have no personal statistics on how frequently the MDT recommendation has been modified, but changes have occurred, and in both directions. The SDM process takes time, no doubt, but feels very rewarding and is part of modern health care. Kunneman’s article [3] raises the question of why we miss out, where we go wrong. We believe that clinical studies constitute one area where the doctor is probably more inclined to persuade the patient to accept treatment, but few patients are included in trials. Another is very difficult to objectively measure but lies within the area of communication and clinical experience; when patients are perceived to have low health literacy and/or have difficulty making medical decisions. It is sometimes very obvious in consultations that a patient has limited understanding of the concept of cancer and cancer treatment and will have difficulty in making a decision even after having been well informed. Sometimes these kinds of patients will express openly that they do not wish to make such a difficult decision and that it is better left to the physician. Studies in this area have showed that patients of low socioeconomic status or low health literacy not always desire active involvement in clinical decision making. One interesting study on patients undergoing RT treatment has shown that those who did not want to partake in decision making concerning treatment were still satisfied with this decision six weeks into the treatment, while some of those who were actively involved in their own treatment had anxiety and doubted that the right choice had been made after a few weeks of treatment [25].

Conclusions

The article by Kunneman et al. [3] and subsequent/similar articles will hopefully influence oncologists and inspire us to actively involve our patients in the decision making process. This is not only needed in the referral for (neo-)adjuvant treatment but also if there is a choice to be made between different active alternatives or best supportive care only in metastatic disease. While we strive towards tailored oncological treatment for our patients it is also important to tailor information to the individual, and to that person’s resources. SDM need not imply that all patients should receive the same, standardized information but rather that information should be given on that person’s terms. When possible, patients should be offered to partake in a decision process leading to the best possible option for that person. It is not the intention of the article, neither of this editorial, to be patronizing but rather to encourage working towards involving our patients more in medical decision making.