Eosinophilia with bladder cancer

The words of the 14th-century English philosopher, William of Ockham, have classically implored physicians to search for a unifying diagnosis when encountering a patient. Therefore, one should not posit multiple disease processes when a simpler, singular explanation is evident. However, in the face of compelling evidence, multiple possible concomitant diagnoses can sometimes be at play in a patient.

A medical oncology consultation was requested for a 61-year-old man, who presented to the hospital with a non-ST elevation myocardial infarction. His past medical history was significant for hypertension and hypercholesterolemia. Also of note was that he did not speak English fluently; Romanian was his first language.

Interestingly, a subsequent cardiac catheterization did not reveal any lesions that required an interventional procedure. As such, the patient was placed on medical management for heart disease, including antiplatelet therapy. While he was being monitored after the catheterization, he developed gross hematuria. Initially, this was thought to be secondary to the combination of high-dose clopidogrel, aspirin, and glycoprotein IIb/IIIa receptor blockers. However, the hematuria persisted, and on further questioning, the patient related that the onset of hematuria dated back to two months prior to admission.

Initial laboratory data included a complete blood count, which showed a white blood cell count of 9,900 /mcL, hemoglobin 13.1 gm/dL, mean corpuscular volume 94.9, and platelet count 361,000/mcL. Of note, the differential of the white blood cells revealed an eosinophilia, with 16% eosinophils, 44% neutrophils, 28% lymphocytes, 10% monocytes, and 2% basophils.

Amit Mehta

Inspection of the peripheral blood film did not demonstrate any immature or otherwise abnormal cells. The remainder of the laboratory studies, including electrolytes, kidney function, liver function tests, and coagulation parameters, were largely unremarkable. The only exceptions were hyperglycemia and an elevated HbA1c percentage of 7.8%, confirming the new diagnosis of diabetes.

Based on this development, he was evaluated by an urologist. A cystoscopy detected the presence of multiple diverticula within the bladder (see figure). Within one of these diverticula was a mass; it was biopsied. On account of the unusual anatomy, taking a biopsy with muscle in the specimen was not possible. The pathologic interpretation was high-grade urothelial carcinoma; strong histologic evidence of chronic cystitis was also seen.

A staging evaluation with a CT scan of the chest, abdomen, and pelvis did not show distant disease. Therefore, at this point, the patient had a high-grade urothelial carcinoma, although the crucial criterion of muscle-invasiveness could not be judged. However, given the cystoscopic exam, and after an extensive discussion with the urologists, it did not appear that there was any muscle in the particular bladder diverticulum. Therefore, it was unlikely that this tumor had a muscle-invasive component. As such, the decision was made to plan for a cystectomy, with possible bacillus Calmette–Guérin therapy adjuvantly, depending on the final pathology.

An official phone translation service was utilized to explain all of the findings to the patient, along with the implications. His questions were answered in his native language, and he expressed understanding. An appointment was set up on his behalf with the urologist, affording the patient time to recover from the myocardial infarction, and to then prepare for a cystectomy. Unfortunately, the patient did not follow-up at the outpatient clinic. To compound matters, the physicians who cared for him in the hospital were unable to reach the patient by way of the contact information he had provided during his admission.

Second visit

Six months later, the patient was readmitted to the hospital with symptoms of unstable angina. History uncovered that he had been unable to take the medications previously prescribed to him, as he could not afford them. He was again taken for a cardiac catheterization, and three bare metal stents were placed at sites of atherosclerotic disease. The status of the patient’s bladder carcinoma was reassessed. A CT scan revealed a new 2.3 cm right adrenal lesion, but no lymphadenopathy. Also, quite interestingly, his peripheral blood eosinophilia had increased markedly, with a total white blood cell count of 21,000 /mcL, with 55% eosinophils.

Given the substantial absolute eosinophilia, a careful evaluation was performed. The patient was not taking any medications, prescribed or over-the-counter, which could explain this finding. He had no history of asthma or allergy symptoms. Other than his emigration from Romania several years ago, he had no recent travel history that may predispose him to a parasitic infection. The physical examination was negative for any lymphadenopathy, organomegaly, or dermatologic manifestation; therefore, along with a repeated CT scan, this argues against lymphoma (eg, eosinophilia can be seen in a number of Hodgkin’s lymphoma cases).

A few possible explanations remained. The leading candidate diagnoses are the hypereosinophilic syndromes, systemic mastocytosis, and chronic myelogenous leukemia. Importantly, eosinophilia has been well described as a paraneoplastic syndrome with bladder cancer as well.

To sort through this differential diagnosis, a few laboratory studies were performed. Results were positive for a deletion in chromosome 4q12 (resulting in a fusion of the FIP1L1 and PDGFRα genes), and the absolute eosinophilic count: tryptase ratio was 1020. Testing for the BCR-ABL mutation and flow cytometry was negative for abnormalities. In concert, this points to a diagnosis of a hypereosinophilic syndrome (myeloproliferative variant).

Although the patient ideally should receive treatment, the more pressing issue was addressing the bladder cancer. The additional concern now, six months later, was the new adrenal lesion. We recommended biopsy of this lesion, to see if the lesion was truly metastatic. Given the unstable angina and his status post-catheterization, arrangements were made for the patient to see both urology and medical oncology as an outpatient, and compliance to pursue treatment was stressed.

Figure. CT scan of the patient’s bladder on presentation. Note the bladder mass situated within one of the bladder diverticula. Courtesy of A Mehta, MD

Third visit

After being discharged, the patient again presented to the emergency department, within a matter of days. His chief complaint was of abdominal pain, and labs showed that he was significantly hypercalcemic at 15 mg/dL. A combination of intravenous hydration, pain control, and bisphosphonate therapy normalized the serum calcium level and his symptoms. Staging scans were repeated. Along with a larger mass within the bladder diverticulum, a metastatic deposit was also identified just posterior to the bladder, and another in the cecum. Lastly, the right adrenal lesion had enlarged in size to 5 cm.

Given these new clinical developments, the patient now had metastatic bladder cancer. Even without the metastatic lesions, it did not appear that the bladder mass was any longer clearly resectable. We had an extensive discussion with the patient about treatment options, which included chemotherapy with gemcitabine and cisplatin, and palliative care.

The patient expressed interest in pursuing chemotherapy, but the physicians had significant concerns about the patient’s compliance. However, the patient consistently stated that he wished to receive therapy, and we also enlisted the hospital social support services to assist with expediting insurance paperwork. Hopefully, this will facilitate the patient seeking further medical care, if that is what he wishes to do.

Case Discussion

This was a challenging case for many reasons. Even at presentation, the patient had an unusual bladder anatomy, making a clear diagnosis (ie, muscle-invasive status) and surgical plan difficult. Second, and perhaps paramount, was the patient’s approach and ability to deal with his diagnosis. In spite of speaking with the patient via a translation service, and with the help of his friends, it was not clear what was holding him back from seeking further medical attention. Unfortunately, he did not have any immediate family in the United States, and perhaps that may have made a difference.

A psychiatric evaluation was also requested to try and explore this further; finances seemed to be playing a role in his hesitancy in seeking treatment. Still, given that the hospital was facilitating his insurance application, and also arranging for transportation, this did not seem to fully explain the situation. Although it is still not totally apparent what the true underlying issues were, hopefully, continuing dialogue with the patient with ease some of his concerns or reservations.

Also of considerable interest to us was the concomitant peripheral blood eosinophilia. There are a number of published reports describing eosinophilia as a paraneoplastic syndrome with bladder cancer. In some of these articles, it is described that the eosinophilia is abrogated with treatment. In our patient’s case, the eosinophilia consistently and steadily increased with time, resulting in a persistent absolute eosinophilia. Working through the differential diagnosis, a deletion of chromosome 4q12 was discovered. This abnormality causes a fusion of the FIP1L1 and PDGFRα genes, which is seen in the myeloproliferative variants of hypereosinophilic syndrome, such as with chronic eosinophilic leukemia.

Also supporting the diagnosis was the ratio of the absolute eosinophilic count to tryptase level. This can help distinguish between hypereosinophilic syndrome and systemic mastocytosis; the ratio is classically >100 in hypereosinophilic syndrome (in our patient the ratio was 1020).

This second diagnosis was helpful in our understanding of the patient’s disease process. First of all, the hypereosinophilic syndromes are associated with other systemic manifestations, including cardiac ones. Although one would prefer not to give a patient multiple diagnoses, in this case, matters were different. Additionally, chronic eosinophilic leukemia can transform into an acute leukemia. Therefore, if the eosinophilia had been reflective of an acute eosinophilic leukemia, this would have changed our management, as it would have been more urgent to treat the acute leukemia.

Thus, if any of the complications along the spectrum of a hypereosinophilic syndrome occurred during the course of chemotherapy, it would surely be advantageous to have knowledge a priori about a second distinct disease process. Moreover, the potential course of treatment would be clearer.

Generally, the FIP1L1-PDGFRα fusion gene rearrangement is very sensitive to therapy with imatinib. As there is a possibility of cardiac involvement with hypereosinophilic syndrome (vs. being secondary to traditional cardiovascular risk factors), concomitant glucocorticoid therapy may also be considered.

Amit Mehta, MD, is a third-year Fellow in the Division of Hematology and Oncology at University of Medicine and Dentistry, New Jersey, and is a member of the HemOnc Today Editorial Board.

I generally agree with the dictum of the 14th century philosopher William of Ockham that multiple diagnosis are less attractive than a unitary etiology for unique disorders. The above case would seem to demand more than one diagnosis, but I am not convinced that our contributing fellow, Dr. Mehta, has it neatly wrapped up. In fact, ‘Harry of Minneapolis’ suggests that another scenario should be considered. To wit: Paraneoplastic eosinophilia as well as adrenal metastases are very rare with bladder cancers, but not at all unusual with lung carcinomas. Actually, the eosinophilopoietin, IL-5, has been isolated from lung neoplasms. Moreover, hypercalcemia in this patient suggests a tumor is producing parathormone-like material — not uncommon with lung neoplasia, but vanishingly rare with urothelial tumors. The Eastern European background of this patients makes it likely that cigarettes have been a significant part of his past life. Finally, the chromosomal abnormality affecting 4q12 is not, per se, a pathognomonic sign of eosinophilic myeloproliferative disease unless actual F1P1L1/PDGFR fusion is actually verified by molecular biologic assay. Therefore, this patient may well have multiple diagnoses, to be sure, but an occult (PTH/IL-5 secreting) pulmonary tumor plus a more indolent urothelial tumor may explain his unique syndrome.

– Harry S. Jacob, MD

HemOnc Today Chief Medical Editor

For more information:

• Bain BJ. Immunol Allergy Clin North Am. 2007;27:377-388.
• Jovanovic JV. Blood. 2007;109:4635-4640.
• Maric I. J Allergy Clin Immunol. 2007;120:680-687.