This is a peripheral T-cell lymphoma, unspecified. In Kikuchi disease, apoptotoc bodies would be prominent. The aberrant phenotype also favours NHL.  
In our series, we observed lack of CD5 (or CD7) as most frequently lost, coexpression of CD4/CD8 was frequent, altough the CD8 positive lymphocytes here are eventually at least in part reactive. CD4/CD8/GcB positive cases are very rare (we did not see it), still cytotoxic proteins (most frequently TIA) can occasionally be observed in PTCL/U.

Dear dr.Went, thank you suporting this actual diagnosis for young girl. With best wishes. Mery Christmas. I hope the virtual communication with you and collegues will be continued in the New Year. Sincerelly yours.

To the incisive comments of the previous reviewer, I would add that the present of disseminated lymphadenopathy (how big are those adenopathies?) would also favor a PTCL over Kichuki. Generalized lymphadenopathy is rare in KD (1-20%) and the mediastinal involvment is particularly rare. Mild fever can be seen in KFD but true "B" symptoms are also uncommon. CT and NMR are also useful and seem able to identify KFD with some accurancy.  
 
However, in this case, the diagnosis, at leastat distance is complicated by several factors:  
 
I cannot see very well the inside of the necrotic areas in the pics and I am unable to check for presence/absence of karyorrhectic debris.  
Plasmacytoid monocytes can also facilitate the diagnosis of KFD. Is the CD68 result shown in the fig. truly representative? One large pos. cell is present in the upper left field. CD4, which stains histiocytes, seems positive in cells which resemble plasmacytoid monocytes. You should consider adding CD123. Also, since the histiocytes in KFD are peculiarly MPO/CD68 positive as opposed to those of PTCL which are usually MPO negative, you sould perform MPO. The immunophenotype results are not totally supportive of a PTCL, at least of a typical angiocentric cytotoxic type. There is limited information concerning the eventual loss of CD5 outside malignant lymphoma. KFD has not been extensive investigated with CD5 at least by flow. So I cannot comment on this.  
 
In conclusion, the clinical findings in conjunction with the morphology seems to suggest PTCL. However, as shown by the British study by Menasce et al in 1998 in Histopathology, most cases of KFD have a false positive presuntive diagnosis of lymphoma. Since we practice evidence based pathology, clonality studies are certainly indicated in this case. I would be happy to help out in this regard, if you are interested in sharing with us tissue material.

Dear dr.A.Orazzi. It's honour to us to hear from you. And thank you for a chance to make molecs and get consult in your lab. We really will do it on the occassion.  
MPO for sure is requested and will appear in a short time here. I've appended some add pictures with CD68+ cells, debris and fragment from cortex with visualisation of ratio CD3/CD5.  
In this case Ki67 was so impressive- exceeding "reactive" values (90%). And the proliferation of atypical T cells are solid enough spreading to paracortical zones. I agree that Kikuchi demonstrates the solid sheets of cells but maybe more in non necrotic and cellular phase. In this case large area of necrosis is environed by solid sheets of atypical T cells.  
It's true, tha the minority of CD68+ cells are larger monocyoid ones. But "dendritic" macrophages or histyocytes with small reniform nuclei are predominant. I fully accept that it's very difficult to reflect the fullness of slide digitally.  
With best wishes to you in Christmas time. Sincerelly yours.

IH 2006: MPO+ monocytoid cells in paracortical and perinecrotic areas.

The case was sent to dr.A.Orazzi for external consultation.  
Thank you all for participation.

Additional BM biopsy: there are some atypical medium sized CD3+ cells with wrincled nuclei and apoptotic bodies near it in the BM.

Dr.A.Orazi consultation results: TCR and Ig gene rearrangement were not detected. Monocytoid T lympho's proliferation CD123+. FINAL DIAGNOSIS: KIKUCHI FUJIMOTO NECROTIC LYMPHADENOPATHY.  
Thank you all for comments especially dr.A.Orazi for consultation, IH stains (CD123) and molec's.