Difficult case.  
Let us start with the (cortical CD1a+!?) T-ymphoblastic lymphoma/ALL..., considering that some details would be of interest:  
lymphocytosis?  
blasts in the periphery?  
thymic tumor? lymphadenopathy?  
Here I would perform some additional stainings (CD2, CD4, CD7 and CD8) in order to completely phenotypically characterize the lymphoblasts.  
Considering the second part/tumor and the information here, I think that it is still not possible to call it dendritic cell sarcoma since S100 is negative and other markers like CD35, CD11c, CD68 etc. are still not done. On the other hand there are not that much malignant spindle cell tumors expressing cyclin D1 (leimyosarcoma, synovial sarcoma, epitheloid sarcoma...). CD99, CK7, CD57, CD56, D2-40? In the face of the clinical history I would also perform some viral (EBV, HHV8) studies.  
Finally, I would address the prvious CLL smears and try to get as much as possible clincal data on CLL diagnosis, treatment and history from 2001.

Thank you for your opinion.  
Unfortunately, we have no Cd35. CD68 was negative. Other immuno are still in progress.

At first glance the histology of the femoral tumor remembers to a Kaposi-sarcoma (tumor stage). Despite a negative immunostain for CD34 it should be considered in the diff. diagnosis. HHV8 and D2-40 are of particular interest. Additional informations about tumor size, its exact localisation and the nature of the inracytoplasmic pigment (Fe-stain) might be helpful for the diagnosis as well.

I agree with Dr.Tzankovs comment, we need more information  
a. on the CLL, how solid is this diagnosis, on what is it based apart f cytology.  
Morphology and immunophenotype of the lymphoid infiltrate is well consistent with T-LBL/Leukemia, we should know if there are blasts in the peripheral blood, thymic tumor, a bone marrow biopsy is certainly indicated.  
Regarding the sarcomatous part, an X-ray should be appended, and let's wait for the pending immunostains.

It is absolutely necessary to perform BM biopsy - in theory the patient can have a T-precursor ALL in the bone marrow, which infiltrates a non-related bone tumor. Since S-100 , CD68 are negative (as well as CD34 and CD117) the sarcoma diagnosis (and classification) still cannot be established. I agree that this is a difficult case that needs further investigation.

Please find aspirate photos: atypical neoplastic large cells grouped in clusters and rosettes (probably metastasis). The GRANULARITY is prominent (reminescent of brown pigment in HE?). Background" small lymphos with multiple Humprecht's shadows (SLL/CLL). BM is comming. After pathological fracture of femur the resection will be performed.

Please find out additional IH:  
Lymphoid cells positive CD99 (+) (!!!!), Cd4 (+), CD2 (+), CD8/Cd7 (-) (positive reactive T cells), CD56/CD57 (-).  
spindle cells negative for CD99 (-), Ck7 (-).  

do not forget to stain for HHV8

CD99 positivity is common in T-ALL.

Dear Dr Meskauskas, could you please be so kind and let us know what additional investigations in this case are still pending ? Do you plan to ask for a bone marrow biopsy ? will you be able to provide an x-ray of the femoral lesion? I would like to have this information before I ask others to comment on this case.  
Regarding the T-lymphoblastic part I found one case report of a T-ALL occuring in a CLL {PMID: 8535207}

My dear friends. On behalf of dr.Meskauskas I would like to point out, that:  
1. In BM biopsy: the similar spindling cell proliferation is seen with some perivascular organisation and maybe "luminina" (like in Kaposi sarcoma) with TdT+ cell nests in capillaries. CD34 was negative, so we will procede with CD31, WF.  
2. Pathological fracture of femur was performed: histology is comming. Macro: pathological tissues are bloody and brown colloured.  
3. HHV8 unfortunatelly is out of my possibilities now and there.  
 
So we conclude, that "spindle cell" CK(-) tumor/sarcoma is widelly metastatic (primary fracture and threpine biopsy shows the same picture). Mostly interesting is the presence of irregular CD3+ TdT+ cell's nests in the sarcoma. Maybe it represent residual progenitors in occupied bone marrow. I cannot find any explanation for that. Any signs of CLL are found in BM space. I just asked about HIV status of the patient. Maybe you will comment this situation?  
 
I will append BM pictures in minutes. Loading...

HE is placed in. Appologies for suboptimal quality: we will reverse on NASDE decalc process soon. I hope the quality will improve. IH is loading in minutes...

IH features: Spindle cell population: CAM5.2/PanCK/BerEp4(-). Blastic islands: CD3+; CD5-; TdT+; maybe some- Pax5+. The nuclei are polymorphic, lobated, "flower like", slightly cleaved.  
There are no CD79a+ cells at all. Interstitial multiple Pax5+ cells, some of them- intravascular and with slightly irregular nuclei. Some intravascular cells are totally negative for all markers (apoptotic, necrotic?).  
So these "hemato" findings are strange despite enough silent "sarcoma" IH phenotype. The larger amount of tissue in bone resection maybe will expose the process in more detailed fashion. Thank you for collaboration.

D2-40 is absent in the moment too. Appologies for both. I will be back with resection material day by day.

Form the additional pictures, I am convinced that we are facing a Kaposi sarkoma (therfore I recommended at the very beginning to look for HHV8). Contrary to hematogones (TdT+/CD79a+/CD10+/CD34+), the co-occurence of TdT and CD3 in groups of cells was in my opinion always associated with T-ALL/LBL. There are still some infos that can be easily get and you still did not reported:  
lymphocytosis?  
blasts in the periphery?  
thymic tumor? lymphadenopathy?  
how was the CLL treated? alkylating agents, steroids, fludara?

Please find your requested info: CT report/conclusions:  
 
Disseminated lymphadenopathy: neck bilateral (<3.8cm), submandibular, submental, supraclavicular (<1.5cm ), axilar (<3cm, mediastinal (<1,3cm), groin (<3,5cm), visceral/ retroperitoneal/paraaortic/iliacal (<6cm). Spleen 12,3cm with ciscumscribed 9,5cm tumor. Intraossal tumor in the intertrochanteric region of right femur (osteosarcoma vs mts?) ~7cm.  
right groin node: lipomatous tumor 4cm (lipoma vs liposarcoma?). Small hepatic hypodense nodules <2,3 cm and cavernous haemangioma (?) in S8 segment (2cm). Hernia of abdominal wall.  

CLL was diagnosed for 10 years. She was on chlorambucil therapy. 2006: 4 courses of fludarabin/cyclophosphan. some moths ago: 2 courses of fludarabinu due to progression. In blood: absolete lymphocytosis (see attached file).  

I have no data about aspirate now. After histology: we have no haematopoiesis at all in BM space.

Thanks Ugnius for uploading of all the additional information, which is very important for evaluation of this unusual case. Could you please find out if there are blasts in the peripheral blood?  
Obviously there are two malignancies the T-LBL and a most probably vascular tumor which is either disseminated in bones or is a primary multicentric vascular bone tumor. Kaposi sarcoma has been mentioned as an option. Multicentric epithelioid angiosarcoma, or hemagioendothelioma might be considered as well. However the lack of cytokeratinexpression ca be taken as an argument against epithelioid AS. Anyway there is little doubt that this is a malignant vascular neoplasm. For further characterization lets' wait for the material from the resection and the IHC.

Blasts were not mentioned/found in the blood test. The patient is migrating between orthopedics and hematology. And we see that full hemato evaluation was not done. Really it's difficult to catch some aspirate fluid from such obliterated BM space... To be cont'd.

Yes ,but a simple peripheral blood smear should be possible !

It means actually that we face not 2 but 3 tumors simultaneously. And some very difficult treatment decisions.  
 
I would start with peripheral blood cytology and flow just to find out if we are still dealing with the old CLL, a new ALL or their combination. Bone marrow seems to be heavily infiltrated on cytology, monotonous cell morphology with narrow cytoplasm and the evident lack of Gumprechts are not typical for CLL, there could be a blood mirror of the process. In addition, secondary tumors arising in CLL patients are often infiltrated by malignant B-cells; that is not the case here. Of course, the big setback is that we cannot be completely sure if the initial process has been of the B-cell lineage.  
 
And there seem to be indications for a lymph node biopsy both for diagnostics and for staging.

In BM blasts was not mentioned/found. Only prominent lymphocytosis and atypical (MTS) cells in aspirate. I've appended aspirate evaluation and flow (in lithuanian, but the meanings are clear I hope).

In short: CLL/SLL flow picture with prominent lymphocytosis in aspirate with Ig lambda/kappa 4/1 and mts cells.

So there is a bona fide B-CLL present.  
 
What I mean is that there could be a small population of CD3+/TdT+/CD99+ cells in periphery, say 2-3%. It would make the whole thing a bit easier. In my personal experience flow is much more sensitive than morphology in finding peripheral blasts, for some of the ALL cells could be (and usually are)morphologically unrecognisable. Possible T-lymphoblastic infiltration in lymph nodes is another option.  
 
It is very probable that there are two clonaly unrelated lymphoproliferations, together with the soft tissue tumor. It should be decided which of them to treat or to consider palliation. So the more confirmation for the diagnoses, the better (CLL diagnosis is without doubt, I think).

Please find BONE RESECTION slides: CLL is obvious in BM near tu the tumor. "Kaposiform angiomatous tumor/Kaposi sarcoma" picture the same. The blastic islands the same. I will procede with additional IH on BM: the tumor CD34+ (patchy) CD31+, so angiosarcoma diagnosis is clear: Kaposi? Other? "Blastic" quiz is not resolved for today... RESECTION IH pending.

The sarcoma looks at least Kaposifom out... D2-40 and HHV8 can resolve some open questions, though I must admitt that the papillary endothelial proliferations on picture BONE V1 look like hobnail hemangioendothelioma and finally should lead to the designation of the tumor as hemangioendothelioma with Kaposiform features (I must admitt that this should be reserved to tumors of infncy) at the morphological basis alone, but HHV8 has to be excluded, particulalry since the patient is immunocompromized... CLL is obviuos, the TdT/CD3 positive blast are aparently now detectable in the peripheral blood (3%?) and should in my opinion be regarded as T-ALL.

Please find BM IH round No.2. I'm sorry for HHV8 and D2-40 once more. I will try to organize it in the next year, if Wallstreet financial problems do not explode the whorld.

The novel images, incl. immunostains for CD34 and CD31 support the diagnosis of a Kaposi-sarcoma (iatrogenic type in an immunocompromized patient). If at all, the diff. diagnosis is an angiosarcoma.

Dear Ugnius,  
 
Actually you can ask to check HHV-8 integration in peripheral WBC's. This is a more or less common PCR technique and could be informative, for the viral load is usually not limited to the tumor itself.

I agree with almost everything that has been said on this case. However, I will add my comments to further complicate the matters :-). In the first place, the presence of pigment in many spindle cells is a feature that needs to be clarified. Is it iron or is it melanin? CD31 and CD34 expression have been described in melanoma (and even Fli-1) and there may be biological transdifferentiation towards endothelium in such case. The staining for the two endothelial markers is usually focal (CD31>CD34) in such cases. I understand that S-100 was interpreted as negative. However, I would want to confirm that the pigment is iron. If it is not all iron, please do Melan-A and HMB-45 (or other rather specific melanoma marker) to make sure that they are negative.  
Regarding lymphoblastic population; it is not clear from the images how widespread this population is in the bone marrow or blood. There is great difference in how to interpret these findings regarding the extent of the disease. If the findings indicate that this is a substantial process, you need to suggest lymphoblastic lymphoma. However, if these are small and discrete, I would not jump to make additional diagnosis without further evidence of T-lymphoblasts. You also see that Pax-5 strongly+ cells appear increased suggesting an increase in B-lymphoblasts or hematogones. I have seen clusters of TdT+/CD10+/CD3+ cells in some conditions which after all said and done did not represent T-lymphoblastic process. I do not think that this was published. I just want to recommend to be careful before you make an additional diagnosis.

Thanks Emina for your down-to earth comment. After all the question of the ALL is of practical importance, since this is a treatable condition. I would like to see high power images of the 3 areas I marked on an image of the BMB. On low power I get the impression that there is a marked nodular and interstitial infiltrate by small lymphoid cells, consistent with CLL. In the centers of the nodules accumulations of large lymphoid cells (lymphoblasts?)can be suspected, if this is correct the lymphoblastic part is probably quite substantial. What is the immunophenotype of these cells? There is also a disseminated lymphadenopathy, is it due to dissemination of the lymphoblastic process? Depending on the patients condition and the relevance of this information an FNA of an enlarged lymph node can be considered.  
From the therapeutic point of view, the exact nature of the vascular tumor seems to me less relavant.

My Dears, HMB45 was negative. I will append virtual slide and additional photos when IH will be completed on surg material with blasts, CLL and tumor soon. I will ask our hematologists about add clin data. Fe stain will be done additionally.

Please find HE and TdT virtual slides ith exact topography of blasts/CLL/angiotumor. PLease comment if any technical problems persist with opening of these slides. Thank you.

PLease find Fe stain: positive in tumor cells and background.

Thanks Ugnius for the nice iron stain. Sorry I could not open the virtual slides.  
Still I would appreciate a high power view and immunostain of the fields I marked on the BMB.

Appologies for temporary technological troubles: we are just learning IT:) PLease check once more the links to virtual slides and reply (printscreen) if any troubles persist. Thank you for patience.

I tried your virtual slides again the TdT one I could open, but only low power was clear, when I tried to enlarge the image dissolved in pixels.  
May I remind you of my previous simple request to post high power images of the fields I marked on the BMB?

Please find add CLL documentation. Appologies for delay (case was not mine).