Catastrophic Complications After a Man’s Acute Myocardial Infarction

Ronald Rubin, MD—Series Editor

A 60-year-old man sustained an acute myocardial infarction (AMI). A brief summary of his history revealed that he was a classic case of metabolic syndrome, with obesity, type 2 diabetes mellitus, and hypercholesterolemia. His adherence regarding weight loss and lifestyle changes had been poor. He took a statin and an oral hypoglycemia agent. He had been a heavy smoker for 30 years and still smokes approximately ½ to 1 pack of cigarettes daily.

Five days ago, he had experienced myocardial ischemic pain following dinner, his family convinced him to go to the emergency department, where the results of an electrocardiogram and troponin tests were consistent with his having had an AMI.

Emergency catheterization revealed multivessel occlusions, and 3 stents were placed. All went well initially, but on day 4, pain in both feet ensued, and discoloration in his toes was noted. By the morning of day 5, the great toe and second toe of his left foot and the fourth and fifth toe on his right foot had become purple. This occurred during therapy with aspirin, clopidogrel, and heparin.

Results of enzyme-linked immunosorbent assay for heparin-induced thrombocytopenia were negative. Doppler ultrasonography revealed strong pulsations in the dorsalis pedis and posterior tibial arteries bilaterally. Results of a completed blood cell count showed normal and unchanged platelets, but the hemoglobin was 8.3 g/dL.

Echocardiography revealed an ejection fraction of 45%. There was a suggestion of “vegetations” around the aortic valve; however, repeated blood cultures remained negative.

On day 10, new shortness of breath was noted. A chest radiograph revealed the presence of a new, large, left-sided pleural effusion. Anticoagulation was interrupted, and thoracentesis was performed, which revealed a hemorrhagic, exudative effusion. Cultures were negative, but the next day the pathologist called to report that the cytology results were class V definitive for malignant cells in the effusion. A computed tomography (CT) scan with contrast did not reveal mass lesions in the lungs. However, the liver was diffusely mottled with filling defects consistent with metastases. Results of a virtual colonoscopy were negative for mucosal lesions, and repeated stool samples were negative for occult blood.

Which one of the following statements is incorrect regarding the evaluation and management of this patient?

  1. A wide variety of neoplasms are capable of causing the clinicopathologic syndrome being manifested.
  2. Even with the use of modern immunohistochemical staining and molecular profiling techniques, identification of the primary site of a neoplasm frequently is not possible.
  3. Despite the presumed and potential heterogeneity in tissue origins, empiric therapy schemes continue to approach the syndome as a single entity.
  4. Positron emission tomography (PET) scanning is the radiologic procedure of choice and will demonstrate the primary lesion in most cases, especially in the presence of widely metastatic disease.

Answer on next page

Correct Answer: PET scanning is the radiologic procedure of choice and will demonstrate the primary lesion in most cases, especially in the presence of widely metastatic disease.

The case presented here is an example of the clinically difficult but scientifically interesting entity of cancer of unknown primary site. Over the years, this entity has generated discussion about whether any primary tumor can, on occasion, present with widespread and ongoing diffuse metastatic disease while the initial primary lesion remains small (even tiny), asymptomatic, and not obvious or clinically diagnosable with usual techniques. Or, can any neoplasm, regardless of initial source, possess biologic characteristics of ultra-early shedding of tumor cells with the capability to establish widespread, viable, metastatic lesions that, in essence, result in a characteristic virulent syndrome and natural history independent of the original tissue source? I favor the latter, certainly as it regards the clinical behavior and prognosis of these patients’ condition.

Going backwards from our usual view of approaching disease from a presentation-epidemiology-diagnostics-therapy methodology, one can start by saying that once a diagnosis of “cancer of unknown primary site” is made using the current, quite refined diagnostic techniques, therapeutics is quite problematic, and prognosis is very poor. Thus, even in 2016, this essentially still assumes that these neoplasms share biology as a single entity, and empiric similar combination chemotherapies (mainly platinum-based) are used with mediocre results. Most studies demonstrate response rates of 25% to 35% and survival ranges of 6 to 16 months, with longer survival for nodal/pleural/peritoneal disease presentations and shorter when visceral disease is the dominant presentation.1,2

Returning to the clinical presentation: As noted above, the patient most commonly presents with adenopathy and/or pleural/pericardial effusions, discovered either clinically or with imaging scans, that quickly reveal a tissue diagnosis of malignancy in the absence of obvious clinical or radiographic primary. Or the presentation can be a widely metastatic visceral disease picture, with explosive pulmonary, hepatic, and cerebral metastatic tumors, again without obvious primary lesion. A particularly virulent presentation, as was seen in the case above, is the presence of nonbacterial thrombotic endocarditis, with diffuse tumor emboli in the microcirculation, with microangiopathic hemolytic anemia, disseminated intravascular coagulation due to tumor-related tissue factor, and thrombotic microangiopathy with subsequent tissue ischemia.

Diagnostic Techniques

Regarding attempts at establishing the primary site, needless to say that if routine radiology and CT scanning had been revealing, the entity would not be termed “cancer of unknown primary site.” Thus when “routine” studies do not demonstrate a primary lesion, more-refined and focused studies can be used, such as magnetic resonance breast imaging in women with nodal or pleural effusion presentation, and directed, local, and more-invasive endoscopic studies in patients with significant clinical clues (eg, smoker, cough, occult blood in stool).

A very exciting, evolving technique is PET scanning, and a review of studies has yielded presumed primary lesions in as many as 88% but in as few as 39% of cases.1 The consensus is that PET scanning has great potential in this arena but that more uniform, prospective studies need be done to confirm when and where to incorporate such testing. Answer D overstates the utility of PET scanning at this time and is the incorrect statement. Tumor markers should always be evaluated and include carcinoembryonic antigen, CA-125, prostate-specific antigen, human chorionic gonadotropin, and α-fetoprotein. However, considerable variability and mixtures are found such that markers are not particularly helpful. When present, they do have some utility in monitoring response to therapy, however.1

Tissue immunohistochemical staining and molecular profiling seem obvious in such evaluations and are done routinely, but again they have a surprisingly low yield (25%), since most “unknown primary” tumors will manifest a diffuse mixture of abnormalities that is nonspecific and thus nondiagnostic.3,4 Conversely, the analysis of patient collections with the diagnosis of “cancer of unknown primary” also show no grouping of staining or molecular profiling or markers that coincide with the unifying observed clinical behavior, regardless of where the cancer originated (unlike, for example, the JAK2 mutation, which results in a shared marrow myeloproliferative behavior that can be a bit clinically differentiable as polycythemia vera, essential thrombocytosis, or myelofibrosis, but all still manifest the core behavior of myeloproliferation). Nonetheless, we are in the era of molecular profiling, and this technique has vast potential in the precise diagnosis of any neoplastic lesion; hopefully, in time, cancer of unknown primary will be no exception.

For now, therefore, as the case in the patient described above, a small subset of patients present with often widely metastatic disease, in whom no definitive primary lesions can be identified, despite an inclusive routine and advanced radiologic testing, immunohistochemical testing and even vanguard molecular profiling of good tissue specimens. There appears to be a unifying biologic behavior to these tumors, and that behavior most often is very aggressive and morbid. Rather mediocre empiric chemotherapies are offered, but the response rate and survivorship is quite poor. The prognosis is quite poor, and most patients will demonstrate a short and lethal course, as was seen in the presented case.

Patient Follow-Up

The patient continued to deteriorate clinically at an even accelerated rate. The malignant pleural effusion continually required drainage to relieve respiratory tract symptoms. Another round of thromboembolic events ensued, including a cerebral infarction that resulted in cortical blindness. During this period, laboratory parameters consistent with profound consumptive coagulopathy (persistent severe hypofibrinogenemia and thrombocytopenia) were present, and CT scanning demonstrated hemorrhagic foci within the emboli. Subsequent major gastrointestinal tract bleeding requiring frequent blood transfusions also occurred. After 72 hours of these events, discussions with the family resulted in the institution of palliative care, and the patient died approximately 2 months after the onset of his illness.

Ronald Rubin, MD, is a professor of medicine at the Temple University School of Medicine and is chief of clinical hematology in the Department of Medicine at Temple University Hospital, both in Philadelphia, Pennsylvania.


  1. Varadhachary GR, Raber MN. Cancer of unknown primary site. N Engl J Med. 2014;371(8):757-765.
  2. Hainsworth JD, Greco FA. Treatment of patients with cancer of an unknown primary site. N Engl J Med. 1993;329(4):257-263.
  3. Greco FA, Spigel DR, Yardley DA, Erlander MG, Ma X-J, Hainsworth JD. Molecular profiling in unknown primary cancer: accuracy of tissue origin prediction. Oncologist. 2010;15(5):500-506.
  4. Hainsworth JD, Rubin MS, Spigel DR, et al. Molecular gene expression profiling to predict the tissue of origin and direct site-specific therapy in patients with carcinoma of unknown primary site: a prospective trial of the Sarah Cannon Research Institute. J Clin Oncol. 2013;31(2):217-223.