The presence of HHV6 DNA was detected in 16/50 (32%) of our lymphoma patients, 6/16 (37.5%) of these patients were viremic. Our results were close to those reported by Di Luca et al. who detected HHV6 DNA in 29% of their cases . Sumiyoshi et al.  have reported presence of HHV6 in 50-68% in patients with different malignant lymphomas. In the current study, the frequency of HHV6 DNA in HD cases was 21%. This figure lies between the very higher frequencies reported by Valente et al. (73%)  and the low frequency (12%) reported by Torelli et al. . None of our healthy volunteers showed HHV6 infection.
Discrepancies in the results could be attributed to several reasons: i) All these studies have investigated the presence of HHV6 in lymphoid tissues only which indicate that the virus could be activated at sites other than peripheral blood compartment , ii) The wide spectrum of sensitivity of the techniques used in different laboratories , the amount of DNA analyzed, and the presence of variants of HHV6 which did not hybridize well to the primers used could lead to false negative results, iii) most of our patients had absolute neutropenia which may affect the positivity of PCR assay due to the presence of a low number of infected cells.
In contrast with Tailor et al. , we observed that there was a significant difference in frequency of HHV6 infection between NHL and HL cases (p = 0.023). Such difference in frequency of HHV6 in both NHL and HL might be linked to different treatment regimen given according to type of lymphoma .
CMV is considered to be one of the most important opportunistic infections occuring in lymphoma patients as a result of impairment in cell mediated immunity . CMV pneumonia (CMVp) is to be one of the most common clinical presentations of CMV disease and is associated with considerable morbidity and mortality in patients with hematological malignancies [30, 31]; disease-specific mortality rate reaches up to 30% in lymphoma patients . Moreover, CMV retinitis  encephalitis and oesophagitis have been observed among patients with HL and NHL diseases . In contrast to Hingmire et al (34), who reported that CMV infections are less common in HL and NHL patients who did not undergo allogeneic BMT, we could detect CMV infection in 34% of our lymphoma patients. Furthermore, it is noteworthy that the majority of the patients in our NHL series derived from a B-cell lineage 16/17 (94%). More than half of our NHL patients of B cell subtype (9/16, 56%) showed CMV infection. This might be due to exposure to more selective suppressive chemotherapy that leads to diminished T cell function with the disappearance of CD8 cytotoxic population . We also observed that the antibody titer to CMV was consistent with its reactivation from latency in lymphoma (median value in patients with and without CMV infection was 4, 2.6 respectively, p = 0.006).
It has been documented that the risk of developing severe CMV disease depends on the degree of immunosuppression and underlying disease . Previous studies have reported that HHV6 infection is one of the major contributions for induction of an immunosuppression state in patients with BMT and solid organ transplantation. Multiple studies suggested association between HHV6 and CMV infection following organ transplantation with effects on both clinical picture and prognosis [25, 36]. These observations encouraged us to investigate frequency of lymphoma patients positive for both herpes viruses (HHV6/CMV) and secondly, determine if there is a relationship between presence of both viruses and severity of lymphoma disease. Our data showed that 20% of lymphoma patients were positive for both herpes viruses. Moreover, the presence of both viruses was more common among NHL cases (8/17, 47%). Previous studies have addressed explanations for such observations which could be due to: (i) immunosuppression from both NHL disease and its treatment may predispose patients to higher risk of coinfection, (ii) An immunomodulating effect of HHV6 since it can induce production of interleukin-1β and tumor necrosis factor-alpha, suppress T lymphocyte function due to reduced interleukin-2 synthesis , and suppress bone marrow by inducing interferon-alpha production. [37, 38], (iii) HHV-6 can directly infect CD4+ T-cells and induce apoptosis, thus altering key immune activation molecules pathways and subsequently disturbing the cytokine network. (iv) HHV-6 can also infect thymic epithelial cells, hematopoietic stem cells, and natural killer cells, which are critical for immune maturation and protection against cancer and viral infections.
All these factors could contribute to pathologic effects of other viral infections like CMV as a result of HHV6 reactivation , and also create an environment suitable for persistence of HHV6 latency .
The presence of both herpes viruses was associated with uncontrolled lymphoma disease (p = 0.021) and salvage phase of treatment (p = 0.034) but not with advanced stage (p = 1.000).
Griffiths et al.  showed that the combination of both HHV6 and CMV infection after organ transplantation was more likely to be associated with CMV disease than with CMV infection alone. Our observations extended to demonstrate the immunosuppressive effect of HHV6 and its associated clinical manifestations among cases who had CMV infection (reactivation or reinfection), these include: i) patients who were positive for CMV and negative for HHV6 were older than those with negative CMV and positive HHV6, median age were 15 y and 5.5 Y respectively. ii) 70% of patients with positive herpes viruses (HHV6, CMV) showed clinical manifestations of severe chest infection and were significantly associated with more frequent episodes of febrile neutropenia (median 3 episodes), long duration of febrile neutropenia > 10 days, absolute neutrophil count (ANC) of < 0.8, lymphopenia (< 0.5), and low Hb concentration (Hb < 9.1). iii) The presence of HHV6 single infection was significantly associated with longer duration of febrile neutropenia > 10 days when compared to HHV6 negative patients and did not show any significant difference when compared to those positive for both herpes viruses. However, we could not rely completely on such observations because all of our patients were subjected to lymphoma treatment which could aggravate the suppressive effect of HHV6, this favoring development of CMV infection and disease and because of the lower number of studied cases. Therefore, the role of HHV6 as a predictor for CMV syndrome in cases with CMV infection is needed to be carefully evaluated in a larger number of lymphoma cases. It will be also worthy to follow the patients before and after treatment to consider the role of iatrogenic immunosuppression.
In the present exploratory study, a single sample was analyzed. This could not indicate the timing, duration and the course of infection. More frequent sampling with quantitative analysis of peripheral blood and tissues in follow up studies may be necessary for better determining the causal effect of this unique β-herpesvirus relationship.