Deceitful red-flag: angina secondary to iron deficiency anaemia as a presenting complaint for underlying malignancy

Abstract

A 73-year-old man with an 8-week history of angina underwent an exercise tolerance test at the rapid access clinic, which indicated inducible ischaemia and he was subsequently referred for angiogram. His angiogram demonstrated no coronary pathology. It was later discovered that bloods taken on the day of the procedure showed a haemoglobin of 54 g/L (130–180 g/L). His haemoglobin used to book the angiogram 3 months before was 143 g/L. Following angiogram, a mass was identified in the right iliac fossa and CT scan confirmed a caecal tumour. The patient ultimately underwent a curative right hemicolectomy as an outpatient. The case is a reminder of the importance of basic preangiogram investigations, in particularly a full blood count, to rule-out angina secondary to anaemia through a low haemoglobin. Most importantly, it also questions when the appropriate time is for these investigations to be carried out, prior to coronary angiography.

Background

Angina exclusively secondary to anaemia has been widely reported. Frequently ischaemic heart disease manifests as a result of reduced coronary flow due to obstructive atherosclerotic disease. To a lesser extent, an increase in demand, decreased blood volume, decreased oxygenation or decreased oxygen-carrying capacity can result in cardiac ischaemia.^{1 2} Severe anaemia of any aetiology is known to cause a compensatory response of increased cardiac output in an effort to ensure the oxygen supply to the tissue is not compromised.³ Furthermore, the low haematocrit and therefore reduced blood viscosity associated with anaemia help to complement the increased cardiac output and also cause changes in peripheral vascular resistance to selectively divert blood to coronary vascular beds.³ In chronic cases of anaemia, the red blood cells also produce more 2,3-diphosphoglycerate to encourage dissociation of oxygen from haemoglobin.³ Where anaemia persists, these compensatory changes can fail and angina pectoris presents even where the coronary vasculature is not diseased.

Dr Carey Coombs produced an article for the British Medical Journal back in 1926 commenting on 36 of his patients with pernicious anaemia, of which eight developed cardiac pain and two of these had pain on exertion.⁴ While Coombs had a high suspicion that anaemia had caused the chest pain, he could not rule-out the possibility that some of his patients also had coronary vessel disease.⁴ Since Coombs’s work, research has found that angina pectoris due to anaemia is an uncommon but not unusual presentation. However, the true incidence is still unclear as not all patients will undergo coronary angiography to consider the contributing weight of atherosclerotic disease. Cases such as the one presented here often still come as a surprise to clinicians and continue to be a valuable learning point on the diagnostic complexities of chest pain.

Case presentation

A 73-year-old man initially presented to hospital via the Rapid Access Chest Pain Clinic with an 8-week history of chest tightness and associated shortness of breath on exertion, particularly on inclines that resolved with rest after approximately 5 min. His symptoms had begun to worsen in the weeks leading up to his referral. The rapid access chest pain service is set up to provide timely specialist cardiology input for new exertional chest pain, which presents as angina. Primary care services can refer to the clinic and usually an appointment is made within 2 weeks. Patients therefore have a fast route to specialist investigations and interventions if they are suspected of having ischaemic heart disease.

Approximately 2 months prior to his clinic appointment, he had seen his General Practioner with chest tightness and palpitations. He had routine blood tests and a resting ECG completed that did not identify any obvious abnormality. Interestingly, at this point, his haemoglobin level was 143 g/L with a normal mean corpuscular volume (MCV). Although he was referred to cardiology, he arranged to see a cardiologist in a private setting. He was diagnosed with paroxysmal atrial fibrillation and as his CHA₂DS₂-VASc score was recorded as 2, it was deemed appropriate to start anticoagulation as secondary prevention for cardioembolic stroke with edoxaban. Bisoprolol was also started as a rate control agent for his atrial fibrillation. However, prior to his Rapid Access Chest Pain Clinic appointment, he had stopped taking his bisoprolol due to experiencing significant fatigue.

At the rapid access clinic, an exercise tolerance test was performed and he completed 6 min 8 s of Full Bruce protocol, achieving 72% of his maximum predicted heart rate before the test was stopped due to chest pain and fatigue. There was anterolateral ST depression that was down sloping in recovery. By way of explanation, a normal exercise tolerance test should reach 85% of the maximum predicted heart rate without any associated ST changes. The significance of this result is that the substantial ST depression alongside symptoms of chest pain and shortness of breath at a low work rate, down sloping ST segments and ST depression across five leads all suggest a high probability of coronary artery disease. The cardiology team concluded that there was sufficient evidence to suspect ischaemic heart disease and an outpatient angiogram was booked for a month later. The patient returned home that day with glyceryl trinitrate spray, atorvastatin, aspirin and isosorbide mononitrate with instructions to continue the edoxaban. These medications were started as treatment for stable angina until further investigations were carried out. He underwent his angiogram as a day case, and it showed no obstructive disease with only plaque disease in the proximal right coronary artery.

Unfortunately, it was only noted at this point that his preangiogram haemoglobin was 54 g/L, which was a substantial drop compared with the level taken 3 months previously. On further assessment after angiogram, he reported a vague history of some weight loss but no change of bowel habit or melaena. However, a small mass was palpable in the right iliac fossa. He was transfused, edoxaban discontinued and he was admitted to the cardiology ward.

Investigations

Blood tests to investigate the anaemia showed an MCV of 74.1 fl (85–105 fl), a ferritin of 2 μg/L (30–400 μg/L), a per cent saturated transferrin of 2 (22–550), a normal folate of 8.2 μg/L (3.1–17.5 μg/L) and a B₁₂ of 186 ng/L (200–940 ng/L). This defined the microcytic anaemia as arising due to iron deficiency. The blood film report expanded on this and commented that evidence of very severe iron deficiency anaemia and acute blood loss was seen including acanthocytes, generally marked crenation of many of the red cell membranes, bite cells, schistocytes and rouleaux formations.

A CT abdomen and pelvis showed a large enhancing caecal mass (see figure 1) with small lymph nodes and small sclerotic bony foci. There was also bilateral hydroureteronephrosis with an enlarged prostate. This was further investigated with a CT chest for staging, which revealed no pulmonary metastases, and a nuclear medicine bone scan showed no increased tracer uptake by the sclerotic foci. The colorectal multidisciplinary team arranged upper and lower gastrointestinal tract endoscopy with eight biopsies taken from a caecal mass proving to be a moderately differentiated adenocarcinoma. Duodenal biopsies were not suggestive of coeliac disease contributing to the iron deficiency anaemia. Urology reviewed the hydroureteronephrosis (despite an indwelling catheter at the time) and this was considered to be chronic high-pressure urinary retention secondary to an enlarged prostate, for which they started tamsulosin and planned for intermittent self-catheterisation and a transurethral resection of prostate following surgery on the caecal mass. A prostate-specific antigen level was found to be normal.

Figure 1.

CT abdomen and pelvis with arrow labelling a large caecal tumour.

Differential diagnosis

Primary diagnosis: symptomatic iron deficiency anaemia secondary to a moderately differentiated caecal adenocarcinoma, with blood loss from this likely to have been exacerbated by recently started anticoagulation.

Treatment

After the cardiology team had started oral iron, given several transfusions and completed the aforementioned investigations, he was discharged and brought in under an urgent elective list with the colorectal surgeons for a right hemicolectomy. No neoadjuvant treatment was required.

Outcome and follow-up

He underwent a successful laparoscopic right hemicolectomy to remove the tumour from the caecal pole. The pathology report defined this as T 3 N0 M0 R0 0/21. That is to say the tumour had invaded through the muscle to the outer lining of the bowel wall, there were no involved lymph nodes or distant metastases, there was a complete resection at all surgical margins and in the sample there were 21 non-affected lymph nodes present. No adjuvant therapy was required.

Postoperatively, he developed severe delirium and was taken back to theatre with suspicion about the integrity of the anastomosis. However, the surgical findings were pristine and a CT head and MRI head ruled out any acute intracranial event. His delirium subsequently improved and he was discharged for urology follow-up regarding the prostate. At his most recent colorectal outpatient clinic, he was doing well and has been discharged from their care. He is currently fitted with a long-term catheter with a flip-flow head to manage his chronic retention, his edoxaban has been restarted, but his bisoprolol continues to be held due to his low heart rate. His most recent haemoglobin was 136 g/L.

Discussion

This is a case of significant symptomatic iron deficiency anaemia, which was investigated as primary coronary artery disease when in fact the underlying pathology was a caecal malignancy.

Interestingly, there are a handful of cases published with a similar clinical scenario. A case report of a 65-year-old man who had exertional chest pain and underwent extensive cardiovascular workup was published in 1994. His angiogram was normal, and it was later revealed that his full blood count showed a severe macrocytic anaemia in-keeping with pernicious anaemia. The report concluded that there was a need to use a more sequential approach to investigations with basic blood tests to be completed first, prior to invasive and costly interventions.⁵

Another case report published in 2017 discussed a 75-year old with watershed cerebral infarct alongside exertional angina. His admission haemoglobin was 50 g/L and he had ischaemic ECG changes and positive troponins. Correction of his anaemia, which was ultimately found to be secondary to a gastrointestinal stromal tumour, resulted in reversal of his exertional angina suggesting the absence of significant coronary disease and thus avoiding the need for angiography and possibly subsequent antiplatelet therapy.⁶

The requirement for testing baseline haemoglobin prior to cardiac catherisation is important. Anaemia is known to be associated with an increased in-hospital and short-term mortality post percutaneous coronary intervention.⁷ In fact, a 2012 review of 6528 patients who underwent percutaneous coronary intervention over 7 years identified a clear correlation between severe anaemia and increased risk of mortality.⁸ Baseline haemoglobin may also be used as a marker for other comorbidities such as advanced age, undiagnosed malignancies, diabetes, chronic kidney disease and autoimmune disease.⁹ A paper published in 2018 looked at 2055 patients who underwent coronary angiography and demonstrated that the severity of anaemia and periprocedural drop in haemoglobin were strong predictors of subsequently developing acute kidney injury.¹⁰ The National Health Service Inform website quotes the risk of a serious complication, which includes damage to coronary vessels, as 2 in 1000. Furthermore, periprocedural administration of pharmacological agents such as dual-antiplatelet therapy may exacerbate any underlying bleeding tendencies. Both factors could potentially lead to a significant drop in haemoglobin.^{11 12} Overall in accumulation, the incidence of anaemia post percutaneous coronary intervention can be as high as 9% as observed in 2009.⁷

At present, there are no fixed guidelines as to when haemoglobin should be tested prior to elective coronary angiogram. The Society for Cardiovascular Angiography and Intervention (SCAI) produced an expert consensus on best practice expected in the cardiac catherisation laboratory in 2016. This suggests that a complete blood count and renal profile should be tested within 30 days of the procedure. It further specifically states that significant anaemia should be assessed prior to percutaneous coronary intervention.¹²

In this particular case, the cause of his anaemia was likely secondary to occult bleeding from his caecal tumour, which has been well established,^13–15 and this bleeding was likely exacerbated by anticoagulation with edoxaban. The time course of events is challenging to track completely, given the 3-month gap between the haemoglobin levels. The clinical history provided on the electronic request for bloods, which included the first haemoglobin (143 g/L), states that he had been suffering from chest tightness and palpitations for some time but this is distinctly different from the angina pain he subsequently developed. Later this month he was diagnosed with atrial fibrillation and an echo scheduled after the angiogram showed no valvular disease or systolic dysfunction with only a mildly hypertrophied left ventricle. This tightness and palpitations he was experiencing seem to be due to paroxysmal atrial fibrillation through rate-related discomfort. The commencement of edoxaban for secondary prevention does appear to coincide with the worsening of his chest pain and it progressing to exertional angina, indicating that it clearly contributed to further bleeding from an as yet-to-be discovered caecal tumour. This caused decreased blood volume and oxygen-carrying capacity causing angina and resulting in the initial referral to the chest pain clinic with more developed exertional fatigue, central chest pain and jaw discomfort. He attributed these symptoms to the recently started beta-blocker and despite stopping this, still felt symptomatic, to a lesser extent. A study showed that anaemia causes compensatory haemodynamic adaptability of the left ventricle to increase cardiac output and causes remodelling of the left ventricle itself and the coronary arteries resulting in intima-media thickening.^{1 11} In hindsight, the beta-blocker would have been dampening his compensatory haemodynamic adaptability.

The high mortality and morbidity associated with anaemia in ischaemic heart disease is well documented in the literature. Therefore, it is fundamental to not only exclude anaemia as the primary cause of ischaemic chest pain before undergoing an angiogram but also important to investigate the haemoglobin in the knowledge of its clinical significance should anaemia be found in the context of an angiogram showing coronary vessel disease. It is clear in retrospect that the cause of angina in this instance was the significant anaemia. However, the normal blood count 3 months prior to the angiogram and a very significant exercise tolerance test result would favour discounting anaemia as the primary cause in an elective haemodynamically stable patient. Ultimately, however, the change of his symptoms from tightness and palpitations to angina pain should have prompted a repeat set of bloods, particularly a haemoglobin level to consider whether it was low. The electronic booking of the angiogram should be fail-safe, in that it demands a haemoglobin level to be added before proceeding. However, the system allowed and still allows haemoglobin levels to be entered that may not necessarily be recent and the angiogram can still be booked. Booking systems for significant investigations or procedures should be changed such that they only allow blood results to be entered that are taken within the last month before the request is processed. Clarifying the appropriate time for measuring the haemoglobin is an important learning point. In instances where elective surgery is delayed, a repeat preassessment appointment is scheduled to confirm continued patient suitability for the procedure. Therefore, a similar approach should be adapted for elective coronary angiography, whereby a repeat full blood count, as a minimum, should be reorganised within approximately 4 weeks of the procedure, if there is a longer interval between booking the angiogram and the procedure itself.

Patient’s perspective.

I was of course concerned with the finding of anaemia following my angiogram which was done as a daycase in the clinical investigation unit. I had expected the cardiology team to find a cause for my chest pain but thought this would be due to a blocked vessel, so when I was told the angiogram was normal and the cause was due to low haemoglobin this came as a bit of a shock. My immediate thought was why did I have anaemia and what was to happen about all the new medications I had recently started including the edoxaban?

The cardiology junior doctors who saw me explained that new anaemia had to be fully investigated as it could be due to a source of bleeding within the body and that the blood thinning medications would be stopped. I was made aware of what the possible causes for bleeding could be. As they took a more in-depth history the puzzle began to come together and I informed them of the slight change to my bowels that had been the case for a few months.

After having several transfusions in the clinical investigation unit I was moved up to the cardiology ward and things moved quickly with scans being ordered the following day, much to my family’s relief. When the news was broken that this was indeed bowel cancer, it was not unexpected but naturally I was eager to get things moving and further scans including a bone scan thankfully showed it had not spread. Plans were soon made to get me home and within a week I had heard of a date to be assessed for theatre after which the operation was scheduled. I was quite unwell and confused after the operation but grateful to the surgeons, upon recovering, to hear that the section of the bowel where the cancer had been found had been removed. I continue to slowly get used to having the stoma.

Looking back it still seems hard to take it all in as things moved so quickly. I am grateful of the outcome. However, I do think that the outcome could have been so different had the anaemia not produced such significant chest pain, leading to investigation. You often hear about what to look out for with regards to bowel cancer, but chest pain in relation to this is not something I have been made aware of until now.

Learning points.

• Angina is an unusual but well-documented presentation of symptomatic anaemia and therefore a haemoglobin level should be checked for all patients with angina prior to more invasive tests such as coronary angiogram.

• A timely haemoglobin check prior to an elective coronary angiogram is essential as anaemia is a strong predictor of postprocedural acute kidney injury, its presence has implications on the use of antiplatelets and anticoagulants and, if found in the presence of coronary artery disease, there is a well-recognised high morbidity and mortality.

• While there is no definitive time as to when it is best to perform blood tests prior to elective coronary angiography, a haemoglobin level taken 4 weeks preceding the procedure would be appropriate. This may have implications for electronic booking of elective angiograms, given a notification process would be required to remind clinicians to repeat a haemoglobin level if the procedure is delayed and demand a relatively recent sample to enable the procedure to be booked in the first instance.