Category Archives: Haematology

Hodgkin’s Lymphoma – Pitfalls in Diagnosis


Hodgkin’s lymphoma has been known for over 175 years. For about 140 of these the malignant nature of the disease was not certain. The hallmark of Hodgkin’s lymphoma is the Reed-Sternberg cell, which was described independently by Dorothy Reed (1902) and Carl Sternberg (1898). Neither considered Hodgkin’s lymphoma malignant. Diagnosis of Hodgkin’s lymphoma is a challenge even today as the case discussed illustrates.

A 21 year old male presented with a swelling of the right side of the neck and fever. Examination revealed cervical lymphadenopathy. No other nodes were enlarged. An FNAC was reported as chronic lymphadenitis. Anti-tuberculous therapy was started. The lymph nodes increased over the next three months and the fever failed to abate. This was attributed to a paradoxical reaction to treatment and the anti-tuberculous therapy was continued. The patient sought a second opinion after four months of therapy. A lymph node biopsy was performed and a diagnosis of Hodgkin’s lymphoma was made. When staged this patient has stage IIIB Hodgkin’s lymphoma.

The common cause of lymphadenopathy are infections – bacterial, viral, mycobacterial or fungal; malignancy – lymphomas or metastatic; and autoimmune diseases. Lymphadenopathy of acute infection resolves spontaneously or following treatment with antibiotics. These nodes are usually inflamed and tender and are rarely confused with malignancy. Persistent lymphadenopathy is commonly due to tuberculosis, malignancy or autoimmune diseases and needs investigation.

Fine needle aspiration cytology (FNAC) is performed by aspirating material, usually using a 21 or 22 gauge needle, and smearing it onto a slide. It has the advantage of being less invasive than a biopsy. The processing time for FNAC is short and reports are often available on the same day. On the flip side only a small part of the lymph node is sampled and the process of smearing disrupts architecture. FNAC is best for diagnosis of malignancies where a diagnosis can be made by examination of single cells e.g. squamous cell carcinoma or an adenocarcinoma.  FNAC has limitations in the diagnosis of lymphoma. A negative report does not exclude the diagnosis of lymphoma and all lymphomas can not be classified by FNAC.

Though described by Thomas Hodgkin in 1928, it was only in 1967, when the Reed-Sternberg cell was shown be clonal by cytogenetics, that Hodgkin’s lymphoma was proven to be a malignancy. The two terms prevalent for the disease, Hodgkin’s disease and Hodgkin’s lymphoma reflect the uncertainty about the pathogenesis of Hodgkin’s lymphoma. Hodgkin’s Lymphoma is a lymphoma that almost always arising from the B lymphocytes. What makes Hodgkin’s lymphoma different from other common B cell lymphomas like follicular lymphoma or diffuse large B cell lymphoma?

Most of the cells in non-Hodgkin’s lymphomas like diffuse large B cell lymphoma are malignant. Most of the cells in Hodgkin’s lymphoma are reactive cells – lymphocytes, eosinophils, neutrophils, histiocytes and plasma cells (figure 1). The lymphadenopathy of Hodgkin’s lymphoma is because of normal cells whereas the lymphadenopathy of diffuse B lymphocytic lymphoma or a follicular lymphoma is because of malignant cells. The lymphadenopathy of tuberculosis or any other chronic inflammatory process is due to a reactive infiltrate, much like Hodgkin’s lymphoma. FNAC of a lymph node of a diffuse large B cell lymphoma shows malignant cells making diagnosis possible. FNAC of a node involved by Hodgkin’s lymphoma is most like to give a normal inflammatory infiltrate making it impossible to differentiate between Hodgkin’s lymphoma and a cause of chronic lymphadenopathy like tuberculosis. If a Reed-Sternberg cell is seen on FNAC a diagnosis is possible, but given the paucity of the Reed-Sternberg cell in Hodgkin’s lymphoma this rarely happens. FNAC is appropriate for the initial evaluation of a lymphadenopathy, but a biopsy should to be performed when the results of FNAC are not diagnostic.

Figure 1. Histology of Hodgkin’s lymphoma. The malignant cell of Hodgkin’s lymphoma is the Reed-Sternberg (RS) cell. The RS cell is surrounded by a reactive infiltrate consisting of T-lymphocytes along with a varying number of eosinophils, neutrophils, histiocytes and plasma cells. A node involved by Hodgkin’s lymphoma has very few RS cells. The bulk of the node is made up by normal cell. (Modified from http://www.flickr.com/photos/euthman/3884125493/)

Eleven to fifteen percent of the patients with tuberculosis have a paradoxical reaction after starting anti-tuberculous therapy characterized by increasing fever and worsening of clinical and X ray findings (Eur J Clin Microbiol Infect Dis 2003;22:597-602). Paradoxical reaction was first described by Chloremis in 1955 (Am Rev Tuberc 1955;72:527-36) and may be seen 14-270 days after starting anti-tuberculous therapy (Eur J Clin Microbiol Infect Dis 2002;21:803-9). Tuberculoproteins released as a result of rapid killing of bacteria are responsible for paradoxical reaction. Paradoxical reaction in a patient who has been initiated on anti-tuberculous therapy on the basis of diagnosis of tuberculosis made by FNAC, unlike those diagnosed on biopsy, raises the possibility of lymphoma.

Diagnosis of Hodgkin’s lymphoma is not complete without immunophenotyping. This is best done on a biopsy specimen. Given the limitations of FNAC in the diagnosis of lymphoma, it is ideal to perform an excision biopsy of the lymph node if a definitive diagnosis is not possible by FNAC. In case an excision is not possible a large wedge biopsy should be performed. Deep seated lymphomas can be diagnosed by a needle (Tru-Cut) biopsy. The specimen must be sent to a pathologist with experienced and equipment to diagnose lymphoma.

Every oncologist practicing in regions of the world where tuberculosis is prevalent sees patients who have been diagnosed with advanced Hodgkin’s lymphoma after months of anti-tuberculosis therapy. FNAC is an attarctive test in a resource constrained practice but the limitations need to be appreciated. The clinician must understand what the pathologist is trying to say. A report of “chronic lymphadenitis”  must not be treated as tuberculosis. The pathologist must not say more than what the FNAC shows. Every “chronic lymphadenitis” is not tuberculosis!

Mediastinal Adenopathy with Brain lesions


A 26 year old male patient presented with fever of five days duration. The complete haemogram showed a haemoglobin of 13.2g/dL, WBC count of 13500/mm3 with a differential count of 75% polymorphs, 21% lymphocytes, 2% eosinophils and 2% monocytes and a platelet count of 232,000/mm3. The smear did not show plasmodium and the serology for dengue and HIV was negative. The liver and kidney functions and the X-ray Chest was normal.

A CT scan of the chest was performed which showed mediastinal adenopathy which showed patchy enhancement on injection of contrast. The site was deemed as inaccessible for a biopsy

The patient was initiated on four drug anti-tuberculous therapy. The fever subsided in a week.

About two weeks after discharge the patient has a left sided seizure with unconsciousness. He was admitted to the hospital anti-convulsants given and an MRI of the brain preformed.

The MRI showed multiple ring enhancing lesions in the bilateral fronto-parietal temporal lobes with significant peri-lesional oedema and a significant midline structures to the left due to the right frontoparafalcine lesion.

Lymphoma is a highly curable tumour that mimics tuberculous lymphadenitis. Delay in treatment of high grade lymphomas compromises the chance of cure. Lymph node biopsy is an invasive procedure sometimes needing laparotomies and thoracotomies. The issues that arise when a lymph node is relatively inaccessible are:

Is biopsy needed for all patients with lymphadenopathies?

Patients with lymphadenopathy and an infective/inflammatory pathology in the drained areas need not be biopsied. Patients with inflamed nodes need not be biopsied. Patients with deep seated nodes in whom a certain diagnosis can be made by non-invasive tests need not be biopsied. All other patients should be biopsied. FNAC, unless it yields a specific diagnosis, is an inadequate investigation for lymphadenopathy. An excision biopsy is ideal. A large wedge biopsy is acceptable for large nodes and a trucut biopsy is appropriate for deep seated nodes. Many patients have succumbed to an advanced lymphoma because of delayed diagnosis from an inappropriate or inadequate biopsy. From the pathologists point of view the diagnosis of lymphoma may be one of the most challenging. Not everyone is experienced enough to make the diagnosis. Even the experienced pathologists can only report on the material provided. Providing appropriate material is the responsibility of the clinician. When one is treating on a clinical diagnosis the patients needs a close monitoring for signs of treatment failure and an alternative diagnosis.

Is there a blood test for the diagnosis of lymphomas?

Lymphomas are of two types Hodgkin’s and non-Hodgkin. Non-Hodgkin’s lymphoma is a collection of about 30 distinct malignancies of the lymphoid cells. Non-Hodgkin’s lymphomas have a leukaemic phase i.e. the phase when the malignant cells spill into the blood. Leukaemic phase occurs early in the course of the disease in patients with low grade lymphomas e.g. small lymphocytic lymphoma, follicular lymphomas and splenic marginal zone lymphomas and very high grade lymphomas e.g. Burkitt’s lymphoma and lymphoblastic lymphoma. The leukaemic phase of small lymphocytic lymphoma is chronic lymphocytic leukaemia and that of very high grade lymphomas is acute lymphoblastic leukaemia. Blood tests can only diagnose lymphomas which are in the leukaemic phase. There are no other blood tests for the diagnosis of lymphomas

Does an MRI scan differentiate a lymphoma from tuberculosis?

A tuberculous granuloma caseates resulting in ring enhancement and oedema around the lesion. The other ring enhancing lesions include primary brain tumour (glioblastoma), metastasis (especially post chemotherapy), abscess, Toxoplasma and Cryptococcus in HIV, resolving hematoma (10-21 days), radiation necrosis, and aneurysm. The MRI features of a primary CNS lymphoma in an immune competent host are quite distinct. The tumour is isointense on T1 and show pronounced uniform enhancement and very little peri-lesional oedema. Ring enhancement is seen rarely. In immunocompromised patients ring enhancement, multiple lesions, spontaneous haemorrhage and non-enhancing lesions are common making it difficult to distinguish CNS lymphoma from infections in these patients. A patient with high index of suspicion of lymphoma must undergo a biopsy.

Could these CNS lesions be from a lymphoma?

The MRI picture is inconsistent with a lymphoma in an immune competent patient. Synchronous CNS and systemic lymphomas are very rare and have been reported only in immunocompromised patients. One report of AIDS related lymphomas described synchronous CNS involvement in 5/19 patients (Annals of Oncology 18 (Supplement 9): ix178–ix182, 2007, abstract 71). Synchronous CNS involvement has also been described in a child with immunodeficiency J Pediatr Hematol Oncol. 2008 Apr;30(4):317-9. Isolated CNS relapse from a lymphoma is seen in less than 1% of the patients. This patient was treated as tuberculosis.

Iron Chelation for Myelodysplastic Syndrome


A 75 year old male was brought to the hospital with breathlessness. On examination he had severe pallor, elevated jugular venous pressure, mild hepatomegaly and bilateral crepitations. The X-ray showed a consolidation.

He was anaemic for three years before presentation and was diagnosed to have a myelodysplastic syndrome (refractory cytopenia with multi-lineage dysplasia) with normal cytogenetics. He needed blood transfusions every four to five weeks. His transfusion requirement had increased over the last few months. He had taken about 20 units of blood.

He had a haemoglobin of 4.5g/dL a WBC count of 3700/mm3. The differential count showed 42% neutrophils, 40% lymphocytes, 3% eosinophils and 5% monocytes. The platelet count was 162,000/mm3. He had been treated with erythropoietin with filgrastim with no benefit and was now on transfusion therapy. He was transfused packed erythrocytes.

The serum ferritin was 1522ng/ml. Should the patient be prescribed iron chelation therapy?

Iron overload is a complication of multiple blood transfusions. Patients with thalassaemia major accumulate iron because of blood transfusion and increased absorption from the gut. Iron accumulation results in cirrhosis, heart failure, arthropathy, skin pigmentation and delayed puberty. Almost all children with thalassaemia major succumb to the consequences of iron accumulation by the end of second decade. Iron chelation by deferoxamine has allowed these patients to lead a near normal lifestyle. It needs to be given as a subcutaneous infusion lasting several hours a day. Two oral iron chelators, deferiprone and deferasirox are available. There are an increasing number of adults with transfusion dependent acquired anaemia who are at the risk of developing iron overload.

The issues to be considered before initiating chelation therapy in patients with myelodysplastic syndromes are:

  1. Are the principles of chelation in thalassaemia applicable to acquired anaemia?
  2. Is there evidence of deleterious effect of iron on the survival of MDS?
  3. Does evidence from clinical trials support iron chelation?
  4. What are the guidelines for use of iron chelators in MDS?
  5. Which of the iron chelators is suitable for use in MDS?
  6. Is there a role for chelation in other transfusion dependent anaemias?

Are the principles of chelation in thalassaemia applicable to acquired anaemia?

Myelodysplastic syndrome (MDS) is the commonest cause of acquired transfusion dependent anaemia. MDS is a group of haematological disorders characterised by cytopenias, ineffective erythropoiesis and a risk of transformation to leukaemia. It is classified into five risk categories depending on the type of MDS, presence or absence of bone marrow fibrosis, chromosomal changes present and transfusion requirements viz. very low, low, intermediate, high and very high risk. The survival in these groups is >10 years, >5 years, 4 years, 2 years and 1 year respectively. The higher the risk category higher is the risk of leukaemia. The incidence of MDS increases with age. Patients with thalassaemia major develop consequences of iron overload develop over a period of years. More than 80% of the patients of thalassaemia live for 10 years. The paradox in MDS patients is that those who need the most blood transfusions are not likely to live for long enough to develop consequences of iron overload.

Is there evidence of deleterious effect of iron on the survival of MDS?

Iron overload and transfusion dependency is associated with a shorter overall survival increased risk of leukaemic transformation in patients with MDS. Iron overload has been implicated in cardiac dysfunction, increased risk of leukaemic transformation, increased infections and increased mortality in patients who undergo stem cell transplantation. The most obvious cause of iron overload in patients with MDS would appear to be repeated blood transfusion. MDS is associated with ineffective erythropoiesis. Patients with anaemia associated with ineffective erythropoiesis like thalasssaemia and congenital dyserythropoietic anaemia develop iron absorption because of increased absorption from the gut. Iron overload may also in part represent a more aggressive disease characterised by more ineffective erythropoiesis. Decreased survival and iron overload may not have a cause and effect relationship.

Does evidence from clinical trials support iron chelation?

There are no randomized controlled trials to support the concept of chelation in MDS. Retrospective studies have shown benefits of chelation. In the absence of solid evidence one has to fall back on collective wisdom of experts in finding an answer to the question of chelation in MDS.

What are the guidelines for use of iron chelators in MDS?

Patients with high risk MDS do not survive long enough to develop complications of iron overload, Patients with low risk MDS are at a greater risk to develop complications of iron overload. The consensus among experts is that patients with low risk MDS should be give irons chelation is the serum ferritin is more than 1000ng/ml or the patient has received 20 units of blood.

Which of the iron chelators is suitable for use in MDS?

The three chelators appear to be equally effective in patients with thalassaemia. MDS patients are prone to infections and have cytopenias. Deferoxamine needs parenteral administration and is associated with increased risk of fungal infections. As these patients may be already thrombocytopenic and are prone to infections deferoxamine is not appropriate for MDS. Deferiprone is contraindicated in patients with neutropenia that may be present in many patients of MDS. Iron chelation may improve the haematopoiesis in some patients. For deferoxamine and deferiprone the effect correlates with degree or iron chelation. Deferasirox may have effects independent of iron chelation activity though this is no solid data to support this contention.

Is there a role for chelation in other transfusion dependent anaemias?

The concept of iron chelation evolved in thalassaemia major, which today is an established indication for chelation. Transfusion has been used in patients with sickle cell syndrome pre-operatively, to prevent acute chest syndrome and to prevent strokes. Patients with sickle cell disease may be relatively protected against damage by iron by ill understood mechanisms. SCD has a strong inflammatory component. Ferritin being an acute phase reactant may not reflect body iron accurately in patients with SCD. Chelation should be considered for adults who have transfused 20-30 units of blood, children who have transfusion induced load of 100mg/kg and/or hepatic iron concentration of 7-9mg/g (this corresponds better with a ferritin >3000ng/ml than the 1000ng/ml cut off used for thalassaemia and MDS). Chelation can reduce serum ferritin in overtransfused patients of other anaemias but the survival benefit of chelation is not clear.

This patients has a low risk myelodysplastic syndrome is a candidate for iron chelation preferable with deferasirox.