The cytology is quite hard to interpret, the cells are overstained, however, it is better than the previous samples.  
 
I don't confirm lymphoblasts infiltration.  
 
There are an increased of erythroblasts  
The myeloid cells show some abnormlalities : increased of precursors, low number of mature granulocytes, abnormalities of ganulations, some atypical blastic cells  
 
I am wondering about some myelodysplasia?  
I don't confirm the diagnosis of leukemia  
 
Could you send pictures of blood cells?  

I agree with Martine's comment, but would like to add that in view of the marked erythroid hyperplasia, probably with megaloblasic features and some abnormalities of myeloid cells including hypersegmentation of neutrophils I would also worry about megaloblastic anemia. In this case a blood film would be helpful.

Unfortunately we could not get a blood smear, the patient was transfered in our city 3 days after blood tranfusion. Should we do it ?

Julien  
 
Yes you can do it, even after the blood transfusion  
 
and that will be a good test to see the improvment of the staining of peripheral blood cells  

I just want to complete some more blood images we did not examine before. This was a Martine Idea. See 5 last images (100 to 90).  
What do you think about the technic?

Dear Julien  
The last pictures show improvments for the technique, we can apprciate the chromatin features of the cells and the basophilia of the cytoplasm  
The cells show some abnomalities for nucleus segmentation of polymophonuclear cells  
We can see an abnormal polychromtophilic erythroblast with heterogeneity of the cytoplasm  
Moreover, the shapes of red blood cells are abnormal with aniscytosis and poikilicytosis with dacryocytes(hématies en larmes) suggesting some myelofibrosis.  
These features from the peripheral blood cell associazted with the features of the bone marrow favoured myelodysplasia.  
It would interested to have Perls staining on bone marrow smear.  
 

I would like to offer an opinion rather late in this discussion. What is the duration of transfusion dependence? What are the CBC (complete blood count) results? Large cells appear increased in the bone marrow, but I can't classify them as the stain is too dark. The neutrophils look unremarkable to me. The peripheral blood (PB) photos at the end are much better than the aspirate photos; is it possible to re-process some aspirate slides? The PB photos show lymphocytes, except for #95, which is a nucleated red blood cell showing slight nuclear irregularity. Dysplasia-like changes in normoblasts occur in many situations of increased erythrocytosis. I don't see the giant normoblasts of parvovirus B19 infection. Given the morphology problems, possibly we should approach this case from a different point of view, namely, what disorders cause transfusion-dependent anemia in children? Possibilities include untreated malaria and congenital anemias such as red blood cell enzymopathies (e.g., G6PD), bone marrow failure syndromes (but neutrophil count is OK; what is platelet count?), hemoglobinopathies, congenital dyserythropoiesis, and other rare disorders. I agree that an iron stain may be helpful for evaluation of congenital sideroblastic anemia. I don't know what the diagnosis is, and would hesitate to call this case myelodysplastic syndrome.

This is to say that I have been thinking about this case over the weekend, and I think I may have been on the wrong track. I am more inclined to agree with a possible diagnosis of MDS, especially if transfusion-dependence is of relatively recent onset. It would be good if more clinical information and laboratory studies would become available.

Last transfusion was done ten days before blood test (latest images. We are waiting for the patient news. She is living far from my city (about 180 km).