The FOCCUS study: a prospective evaluation of the frequency, severity and treatable causes of gastrointestinal symptoms during and after chemotherapy

In 2015, 359,960 people in the UK were newly diagnosed with cancer. Mortality rates have fallen over the last 10 years for most cancers, for example, oesophageal cancer by 7%, bowel cancer 15% and 32% for gastric cancer [1]. Research continues to develop better therapies.

However, now that effective treatments for many forms of cancer exist, it has become increasingly important to maintain and improve quality of life after diagnosis rather than focussing solely on reducing cancer mortality. To improve or maintain quality of life often means treating the symptoms caused by the cancer or the anti-cancer therapies. To do this effectively may require the development of new services providing expert supportive care which would not only improve patient well-being but also enhance delivery of the optimal cancer treatment, uncompromised by toxicity [2].

GI symptoms are common in patients with cancer. They may predate the cancer or arise because of the tumour or de novo as a side effect of treatment. It is not unusual for GI symptoms to impinge on patients’ quality of life and to interfere with cancer treatment prompting dose reduction or cessation of therapy potentially resulting in less effective tumour control. When abnormal GI symptoms persist long term, they can have a devastating impact [3].

Some GI symptoms are well managed. For example, severe vomiting once prevented delivery of effective but extremely emetogenic chemotherapeutic regimens. Intensive research produced potent anti-emetic drugs. Consequently, vomiting is rarely a significant problem for patients with cancer. Much less attention has been paid to the incidence, severity or optimal management of other GI symptoms commonly reported by patients such as anorexia, altered taste, borborygmi, bloating, constipation, diarrhoea, dysphagia, early satiety, frequency of defaecation, incontinence, mucus discharge, nausea, nocturnal defaecation, pain/cramps, reflux, regurgitation, steatorrhoea, tenesmus, urgency of defaecation, weight loss and wind.

It is widely believed that GI symptoms arise as a result of pathological changes caused by chemotherapy. However, the evidence to support this hypothesis is weak. In the upper GI tract, chemotherapy causes mucositis characterised by increased apoptosis and pathological changes such as decreased villous height. The intensity of chemotherapy correlates with the severity of the histological change [4]. However, the severity of the symptoms patients develop with similar degrees of mucositis can vary widely. Evidence from non-malignant disorders also suggests that histological change correlates poorly with symptoms [5].

Instead, we have proposed an alternative model for the development of GI symptoms (Fig. 1a, b) namely, that abnormal GI symptoms developing as a result of cancer therapies are caused primarily by disrupted GI physiological function(s). Physiological upset may be triggered by different factors including pathological changes. The differing physiological reserves between individuals explain why patients with the same degree of pathological change do not always develop the same symptoms. We tested this hypothesis in a large randomised trial in patients with new-onset GI symptoms developing after pelvic radiotherapy [6]. In that study, we demonstrated that if the clinician focused on identifying which abnormal physiological processes had developed, treatable abnormalities—e.g. bile acid malabsorption, carbohydrate malabsorption, exocrine pancreatic insufficiency, small intestinal bacterial overgrowth (SIBO) and vitamin deficiency—were frequently detected. In addition, we showed that in the majority of patients, several different physiological abnormalities developed at the same time. However, clinical judgement predicted poorly which physiological abnormalities were present [7]. Instead, if clinicians arranged specific investigations for each symptom, using checklists for accuracy, correct diagnoses could be made rapidly, appropriate treatments prescribed and significant clinical benefit achieved at a relatively low cost [6, 8]. Other studies support these findings [9‐12]. Indeed, a nurse can be trained to assess and manage the large majority of such patients by following such checklists no less effectively than a gastroenterologist [6]. This approach seems to be valid in patients treated with a wide variety of therapies for cancer not just radiotherapy [8].

In patients with GI cancers, which comprise more than 20% of all new cancer diagnoses, chemotherapy is an integral part of treatment for many, often in combination with surgery and/or radiotherapy and/or biological therapies. A number of treatable physiological abnormalities have been described following surgery, radiotherapy and treatment by anti-cancer biological agents [8, 10, 11], but there has been virtually no research to identify the physiological mechanisms which lead to similar symptoms in patients receiving chemotherapy.

The FOCCUS (Focussing on Cancer Chemotherapy’s Undiagnosed Symptoms) study was carried out to quantify prospectively the incidence, severity, frequency and impact on quality of life of chemotherapy-induced GI symptoms. It aimed to document how frequently treatable causes for those symptoms could be identified.

This was a prospective, observational, single-centre cohort study in patients aged 18 years or above undergoing chemotherapy for GI malignancies in the GI unit of one specialist cancer hospital. The study was reviewed and approved by the institutional clinical research committee and by the appropriate ethics committee. The study was conducted in accordance with the Helsinki Declaration, 1975 (revised 1983). Patients gave written informed consent before study enrolment.

This is the first study assessing the frequency of life-changing GI symptoms in patients receiving cancer chemotherapy. We show that standard oncological toxicity scoring commonly used in oncology underestimated GI morbidity. Chemotherapy improved symptoms which could be attributed to a cancer causing partial obstruction to the GI lumen but many other symptoms worsened. Significant numbers reported unremitting symptoms.

No previous study has investigated systematically why patients have these difficult symptoms. When routine gastroenterological investigative tests were used, one in 5 patients had treatable but missed comorbidities before chemotherapy, particularly thyroid dysfunction, vitamin B₁₂ deficiency and exocrine pancreatic insufficiency. During and after chemotherapy, frequently identified causes which are curable with antibiotics include small intestinal bacterial overgrowth in one in every two patients with suggestive symptoms and unsuspected urinary tract infection in 3 in 10 patients with abnormal urine analysis. In other patients’ symptoms, e.g. those with bile acid malabsorption (4 in 10 patients with loose stool), treatments either completely cure or significantly improve symptoms [8].

While this was a heterogeneous population, the mechanisms causing GI symptoms appeared to be similar across tumour types. Our findings are consistent with animal studies, case reports, small series and even data from patients with non-malignant GI disorders which suggest that inflammatory processes predispose to loss of function by the barrier layer of the gut and that specific biochemical functions of receptors and enzyme symptoms in the GI tract are altered temporarily or permanently [21, 22]. This leads to exactly the same types of physiological dysfunction of the GI tract as we report in patients after treatment with radiotherapy [23].

Many clinicians believe that chemotherapy causes predictable, usually mild GI symptoms, and that investigations are unnecessary as patients improve adequately when given symptomatic treatments. This perception may partly result from the inadequacy of the questionnaires measuring toxicity used in clinical practice which are often not comprehensive and fail to reflect accurately the patient experience [24‐26]. There has been little focus by clinicians on the symptoms which patients consider particularly difficult [27]. Possibly, this is because even specialists with an interest in the GI tract frequently do not understand which symptoms matter to the patient and the extent to which these symptoms affect quality of life [28]. This is why a rapid shift to the routine use of patient-reported outcome measures (PROMs) is important for clinicians and patients alike: PROMs increase the frequency of appropriate discussion during consultations and are associated with improved symptom control, better supportive care and patient satisfaction [29].

It is frequently stated that clinical experience is adequate at identifying symptoms which require investigation and those which can be treated empirically. However, we increasingly understand that this approach is naïve in patients with GI benign disorders [30] let alone those symptoms which develop after complex anti-cancer therapies, where often multiple causes contribute. The treatable GI diagnoses found in this study, bile acid malabsorption, carbohydrate malabsorption, exocrine pancreatic insufficiency and SIBO, produce similar symptoms (Table 3) and frequently coexist. Even expert clinicians struggle to identify the correct diagnoses for such symptoms from the history alone [7].

When so many patients had on-going difficult symptoms, it is puzzling why the majority declined the investigations offered. Various reasons may have contributed. It is widely believed that difficult GI symptoms are inevitable during chemotherapy. For decades, the priority has been to develop effective treatments for cancer with almost no research into improved symptom management unless that symptom (e.g. neutropaenia or intractable vomiting) threatened delivery of chemotherapy. Patients are offered symptomatic treatments and are told that things will rapidly improve once chemotherapy is completed. Possibly, patients were preoccupied with getting their cancer treated and feared that our tests might delay/stop their treatment. Or else, patients who do not feel well and are already spending a lot of time attending the hospital may become reluctant to undergo “research” tests which they feel are unlikely to benefit them and which were proposed by researchers, not their treating clinician. It was not unusual for patients to postpone recommended investigation(s) until the symptoms had persisted for several months. This is understandable for endoscopic investigations which are invasive, unpleasant and probably undertaken in this patient group a number of times already. However, this is less understandable for a breath test or a SeHCAT scan which is not onerous, carries minimal risk, is not unpleasant and has such potential to reduce symptom burden by diagnosing treatable conditions [6, 8].

This study was not designed to show the benefit to the patient from diagnosing treatable disorders. Nor were the research team conducting the study trained to deliver the treatments required to address the abnormalities found. After arranging investigations, we informed the patient, oncology team and the GP of what we had found and recommended treatment, but did not record whether those treatments were given or whether they were given in the way we recommended.

If people with cancer present with bowel dysfunction before treatment, they are more likely to suffer severe bowel toxicity from treatment [26]. We know that targeted, early supportive care significantly improves outcomes [31, 32] [32] and when symptom control is integrated with the provision of oncological treatment, this improves quality of life [33]. It is therefore an intuitive assumption that better gastroenterological supportive care will improve difficult GI symptoms and quality of life for patients and diminish the requirement to consider chemotherapy dose reduction. Between one in eight and one in three patients, each month required their initially prescribed regimen to be changed because of toxicity. While we are unable to determine how many of these changes were due to GI toxicity alone, our study suggests that potentially many patients are having their chemotherapy intensity reduced potentially unnecessarily if more effort was devoted to diagnosing and treating the underlying cause of GI toxicity adequately. However, it will require a randomised trial to determine whether survival is enhanced by timely gastroenterological supportive care.

Our approach in this study was not radical; it followed long-established principles of good general medical practice. Our findings suggest that a much more robust and holistic assessment of patients at presentation to the oncology unit should be routine. Urgent consideration needs to be given how best to manage the large numbers of patients with difficult bowel function quickly. Probably, this requires a nurse practitioner trained in gastroenterology with rapid access to the relevant diagnostic facilities to be embedded in the oncology clinic.

The FOCCUS study demonstrates that difficult bowel symptoms before, during and after chemotherapy are common and for the first time has demonstrated that there are frequently treatable causes for these symptoms. A randomised controlled trial is required to demonstrate the size of the survival benefit and the impact on quality of life by offering these patients’ expert gastroenterological supportive care.