During 2006–2009, the 267 samples were collected from 3 groups of patients (Table). Group 1 comprised patients who had traveled from Asia to Germany; location of their permanent residency was unknown. Groups 2 and 3 and the control group comprised only persons from northern Germany. From group 1, a total of 75 sputum and nasal swab specimens were collected from patients who had unconnected cases of RTI and who had recently traveled by air (11). From group 2, a total of 31 bronchoalveolar lavage (BAL) samples were collected from patients with chronic obstructive pulmonary disease (defined by a forced expiratory volume in 1 second/forced vital capacity <70% and forced expiratory volume in 1 second <80% of the predicted value) who had signs of RTI. From group 3, a total of 161 BAL and tracheal secretion samples were collected from patients with severe RTI and immunosuppression as a result of solid organ or bone marrow transplantation. From the control group, throat swabs were collected from 52 healthy persons and BAL samples were collected from 10 healthy volunteers who had no signs of RTI and no known underlying disease.

All samples were analyzed by culture for pathogenic bacteria and fungi and by PCR for rhinoviruses, adenoviruses, enteroviruses, influenza viruses A and B, parainfluenza viruses 1–3, respiratory syncytial virus, cytomegalovirus, Epstein-Barr virus, and human metapneumovirus. All samples were tested in duplicates obtained by individual RNA extractions. XMRV RNA was reverse transcribed from total RNA, after which nested PCR or real-time PCR were conducted as recently described (1,12). No serum samples were available from these patients to confirm the results by serologic testing.

For group 1, XMRV-specific sequences were detected with relatively low frequency (2.3%). For group 2, XMRV-specific sequences were amplified in 1 BAL sample, which was also positive for Staphylococcus aureus by routine culture methods. For group 3, XMRV-specific sequences were detected at a frequency of 9.9%, which was significantly higher than that for the healthy control group (3.2%) at the 90% confidence level but not at the 95% level (p = 0.078, 1 sample t-test). Of 16 group 3 samples, 10 showed no signs of co-infection. The remaining 6 samples showed co-infection with rhinovirus or adenovirus (1 sample each); S. aureus (3 samples); or mixed infection with pathogenic fungi, Candida albicans and Asperigillus fumigatus (1 sample).

All samples that were positive for XMRV by gag-nested PCR, together with a set of those that were negative for XMRV, were retested by real-time PCR. Results showed low XMRV RNA concentrations, 10³ –10⁴/mL of specimen.

To confirm the validity of XMRV detection, a subset of 6 specimens (3 XMRV positive and 3 XMRV negative) were tested by using an alternative PCR assay for viral RNA (3) and a C-Type RT Activity Kit (Cavidi, Uppsala, Sweden) for type C reverse-transcription activity. XMRV sequences from alternative targets in the gag and env regions were confirmed in 2 of the 3 XMRV-positive samples but in none of the controls. One XMRV-positive BAL specimen showed an 8-fold increase above background of specific type C retroviral reverse-transcriptase activity, suggesting presence of active type C retrovirus within this sample. This assay is substantially less sensitive than reverse transcription–PCR.

All XMRV gag sequences (390-bp fragment) were 98%–99% identical to previously published XMRV sequences from persons with prostate cancer (1,2). Phylogenetic analysis showed close clustering (Figure).

XMRV, originally identified in RNase L–deficient patients with familial prostate cancer, has gained interest since recent work showed its protein expression in as many as 23% of prostate cancer cases (10) and XMRV-specific sequences were detected in PBMCs of 67% patients with chronic fatigue syndrome (5). These results, however, could not be confirmed by others (6–8). Both studies also detected XMRV protein or sequences in their control cohorts with frequencies of 6% and 4%, respectively.

Among the most pressing information gaps with regard to XMRV is its preferred route of transmission. Detection of XMRV in PBMCs and plasma of patients with chronic fatigue syndrome raises the possibility of blood-borne transmission; sexual transmission has also been hypothesized on the basis of indirect evidence (5,9). We detected XMRV in respiratory secretions of immunocompetent patients with and without RTI at a frequency of ≈3.2%, which is in good concordance with the recently reported prevalence in the general population of up to 4% (5). Frequency of XMRV detection in group 1 patients (2.25%) was comparable to that of human metapneumovirus and rhinovirus within this group and considerably less frequent than that of parainfluenzavirus (15.5%) or influenza A virus (7.6%) detection (11).

Our findings indicate that XMRV or virus-infected cells might be carried in and transmitted by the respiratory tract. Attempts to isolate infectious virus from XMRV sequence–positive respiratory samples failed, possibly because of inadequate storage of samples before virus culturing attempts or relatively low copy numbers of the virus within the samples. Thus, whether the respiratory tract serves as a putative transmission route for XMRV cannot be determined at this time. The observed increase in prevalence among immunosuppressed patients with RTI suggests that XMRV might be reactivated in absence of an efficient antiviral defense. Together with earlier observations on increased XMRV replication in RNase L–deficient cells (1,12), this finding implies that the immune system plays a role in controlling XMRV replication. It remains unknown whether immunosuppression predisposes a patient to secrete infectious XMRV from the respiratory tract or whether presence of virus might be meaningless for epidemiology in a way similar to HIV-1 (15). Future studies should address whether the respiratory tract might serve as a source of XMRV infection or whether immunosuppression might cause an increased risk for primary infection.