Association of Plasma Viral RNA Load with Prognosis in Cats Naturally Infected with Feline Immunodeficiency Virus

Abstract

We measured the quantity of plasma feline immunodeficiency virus (FIV) RNA using a real-time sequence detecting system. Plasma viral RNA load was shown to correlate with the clinical stage, survival time, and disease progression in naturally FIV-infected cats. The present study indicates that the plasma viral RNA load can be used as a clinical marker representing the impairment of the immune system and predicting the clinical outcome in FIV-infected cats.

Feline immunodeficiency virus (FIV) infection is associated with the development of AIDS in domestic cats similar to that caused by human immunodeficiency virus (HIV) infection in humans (1, 13, 24). FIV-infected cats show various degrees of immunological impairment resulting in clinical symptoms. Five clinical stages modified from those in HIV infection have been applied to cats naturally infected with FIV, i.e., acute phase (AP), asymptomatic carrier (AC), persistent generalized lymphadenopathy, AIDS-related complex (ARC), and AIDS stages (6, 14). In general, the criteria for classifying these clinical stages are based on the clinical symptoms of the cats, and an objective clinical marker has been required.

In a recent study, we examined the plasma viral RNA load in naturally FIV-infected cats classified into the AC, ARC, and AIDS stages based on the clinical signs and suggested that the load was increased in the ARC and AIDS stages in comparison to the AC stage (4). Several groups have also measured the plasma viral RNA load for cats experimentally infected with FIV and found its increase in the AP and decrease in the subsequent AC stage of infection (2). But, in most instances, experimental infection with FIV does not induce terminal AIDS in cats. The quantitative competitive PCR (QC-PCR) has been employed in these studies to quantify plasma viral RNA load, a procedure which is complicated and does not readily provide repeatable results. In contrast to QC-PCR, a real-time sequence detecting system was shown to be simple and accurate for quantifying the plasma viral RNA of lentiviruses (21), and its employment enables the viral load to be used as a marker of the clinical stage of FIV-infected cats. In this study, we examined the plasma viral RNA load in cats naturally infected with FIV and investigated its correlation with the clinical outcome.

Thirty-three naturally FIV-infected cats referred to the Veterinary Medical Center of the University of Tokyo for diagnosis and treatment were examined in this study. All of the cases were seropositive for FIV antibody and seronegative for FeLV antigen (IDEXX Laboratories, Portland, Maine). The clinical stage of each cat was determined from the clinical symptoms by the criteria reported previously (6, 14). Thirteen FIV-infected cats (0.5 to 11 years old) did not show any clinical signs associated with FIV infection and were categorized as in the AC stage. Counts of CD4⁺ cells calculated by flow-cytometric analysis were 768 ± 457/μl (mean ± standard deviation [SD]) in these AC cats. Eight cats (3 to 11 years old) infected with FIV had recurrent episodes of infection and inflammation and were categorized as in the ARC stage. The most common clinical signs in these cats were chronic stomatitis and gingivitis, observed in seven of eight animals. Three of the eight cats in the ARC stage had cholangiohepatitis, chronic bronchitis, and pyoderma, respectively. Mean ± SD values of the CD4⁺ cell count in these eight animals were 450 ± 336/μl. All of the 12 cats (3 to 16 years old) categorized as having AIDS showed various signs of marked immunodeficiency. In addition to chronic stomatitis observed in 4 of the 12 animals, 7 cats showed hematopoietic dysfunction, such as neutropenia, anemia, and/or thrombocytopenia. Opportunistic infections with bacteria or fungi were observed in 3 of the 12 cats, and tumors were also detected in 2 animals. CD4⁺ cell counts were 164 ± 132/μl (mean ± SD) in cats in the AIDS stage.

Samples of virion-associated RNA were extracted from 140 μl of plasma from these 33 cats using the QIAamp viral RNA kit (Qiagen, Studio City, Calif.) and stored at −80°C until use. For use as a standard control, in vitro-transcribed RNA was prepared from a plasmid containing a 518-bp fragment of the FIV gag gene from clone pFIV14 (11, 12). After linearization of the plasmid with BamHI digestion, in vitro transcription was performed with an RNA transcription kit (Stratagene, La Jolla, Calif.) using the protocol recommended by the manufacturer. Aliquots of the control RNA were stored at −80°C, and the amount of RNA was quantified from the absorbance at 260 nm with a spectrophotometer (UV-160A; Shimadzu, Kyoto, Japan) before use. Sequences of a pair of primers and a probe were chosen using Primer Express software version 1.0 (Applied Biosystems, Foster City, Calif.) from the highly conserved sequence of the FIV gag gene (9): a forward primer, 413F (5′-AAACAGTAAATGGAGCACCACAGTAT-3′, nucleotides 1040 to 1065 in pFIV14 [11, 12]), a reverse primer, 495R (5′-TAGCCCCTCTCTTGCCTTCTC-3′, nucleotides 1122 to 1102), and an internal probe, 440T (5′-TAGCACTTGACCCAAAAATGGTGTCCAATT-3′, nucleotides 1067 to 1095). The 5′ and 3′ ends of the 440T probe were labeled with a reporter dye, 6-carboxyfluorescein, and a quencher dye, 6-carboxytetramethylrhodamine, respectively. Reverse transcription (RT)-PCR was performed with a TaqMan RT-PCR kit (Applied Biosystems) according to the user's manual. As a simple method for screening a large number of clinical samples, we employed a one-step RT-PCR. The reaction mixture (50 μl) contained 22.5 μl of sample RNA (equivalent to the amount of RNA extracted from 52.5 μl of plasma sample) or control FIV RNA (equivalent to 5 × 10¹ to 10⁷ copies of FIV RNA), a 200 nM concentration of the forward primer 413F, a 400 nM concentration of the reverse primer 495R, a 200 nM concentration of the labeled internal probe 440T, 12.5 U of murine leukemia virus reverse transcriptase, 20 U of RNase inhibitor, 1.25 U of AmpliTaq Gold DNA polymerase, 10 mM Tris-HCl (pH 8.3), 5 mM MgCl₂, 50 mM KCl, and 1 mM deoxynucleoside triphosphates. Reverse transcription at 48°C for 30 min and a step for inactivation of murine leukemia virus reverse transcriptase and activation of AmpliTaq Gold DNA polymerase at 95°C for 2 min were followed by 50 cycles of PCR amplification consisting of denaturation at 95°C for 1 min and primer annealing and polymerization at 60°C for 1 min, all carried out as consecutive reactions in a single tube. The RT-PCR and detection of fluorescence intensity were carried out in the ABI PRISM 7700 sequence detection system (Applied Biosystems). All specimens were subjected to the analysis in duplicate. Results were shown by the cycle numbers at which the fluorescence intensity reached the threshold (threshold cycle, C_T). For each reaction, a standard curve was generated from the serially diluted control FIV RNA, and the FIV copy number was calculated by interpolation of the C_T value of the plasma samples to the standard curve. By using the control FIV RNA template, the detectable limit of the real-time sequence detecting system employed in this study was as low as 50 copies per reaction in 52.5 μl of plasma sample, corresponding to 9.5 × 10² copies per ml of plasma. This was much lower than that in our previous study using QC-PCR (4). The amount of plasma viral RNA could be measured in 27 (82%) of the 33 cats but was below the detectable level in 6 cats (18%), which indicates that the system can be used in a laboratory test to examine samples from the clinic.

The major aim of this study was to investigate the relation between plasma viral RNA load and prognosis in cats naturally infected with FIV. First, we examined the relation between the viral load and clinical stage in the FIV-infected cats. Of the 13 FIV-infected cats in the AC stage, 8 had 5.54 × 10⁴ to 2.46 × 10⁶ copies of FIV RNA per ml of plasma and 5 did not have a detectable level of plasma FIV RNA (Fig. 1). Of the eight FIV-infected cats in the ARC stage, seven had 1.26 × 10⁵ to 2.34 × 10⁶ copies of FIV RNA per ml of plasma and one cat did not have detectable plasma FIV RNA. Twelve FIV-infected cats in the AIDS stage had 2.78 × 10⁵ to 1.98 × 10⁸ copies of FIV RNA per ml of plasma. These results indicate that plasma viral RNA levels increase with the progression of clinical stage. The plasma viral RNA load was significantly higher in the cats in the AIDS stage and in the ARC and AIDS stages than in the cats in the AC stage (Mann-Whitney rank sum test; P = 0.000 and P = 0.002, respectively). The differences between the AC and ARC stages and between the ARC and AIDS stages were not significant. There was no significant correlation between the plasma viral RNA load and total lymphocyte count, CD4⁺ cell count, CD8⁺ cell count, or CD4⁺/CD8⁺ cell ratio (Spearman rank correlation test).

FIG. 1.

Plasma viral RNA loads in 33 cats naturally infected with FIV. Clinical stages were estimated based on the clinical signs. Each point represents the value for one cat. The threshold of sensitivity for the real-time sequence detecting system was 9.50 × 10² copies/ml.

Next, we examined the relation between the plasma viral RNA load and survival time. The survival times for these 33 cats ranged from 1 month to more than 41 months (median, 7 months). The survival times for 16 FIV-infected cats with a high viral load (⩾10⁶ copies/ml) were 1 month to more than 41 months (median, 2.5 months), whereas those for 17 FIV-infected cats with a low viral load (<10⁶ copies/ml) were 1 month to more than 41 months (median, 14 months). The survival times for these two groups were significantly different (Fig. 2) (nonparametric log rank test, P = 0.0006). From these results, it can be concluded that the high viral load is indicative of the highly immunodeficient state at later stages, such as the AIDS stage, and poor prognosis in cats infected with FIV.

FIG. 2.

Kaplan-Meier survival curves of 33 cats naturally infected with FIV, divided into those with a high plasma viral RNA load (⩾10⁶ copies/ml) and those with a low plasma viral RNA load (<10⁶ copies/ml).

Furthermore, we carried out a longitudinal assessment of the change of plasma viral RNA load and clinical symptoms in eight naturally FIV-infected cats. The duration of the observation ranged from 2 to 46 months. During the observation period, the cats with signs of illness were treated with supportive therapies, including administration of antibiotics and/or corticosteroids, fluid therapy, and/or blood transfusion.

In three cats (cases 1, 2, and 3) which showed no change in clinical symptoms during the period, the plasma viral RNA load exhibited no obvious increase.

Clinical symptoms worsened in five cats (cases 4 to 8) during the observation period. Two (cases 4 and 5) of these five cats showed a progression in the clinical stage: in case 4 the disease progressed from the AC to the ARC stage, and in case 5 it progressed from the ARC to the AIDS stage. Three (cases 6, 7, and 8) of the five cats showed a progression of clinical symptoms although they remained in the same clinical stage: cases 6 and 7 in the ARC stage and case 8 in the AIDS stage. Plasma viral RNA load did not change in case 4 with progression from the AC to the ARC stage: 2.46 × 10⁶ copies/ml at month 0 and 8.97 × 10⁵ copies/ml at month 35. However, in case 5, in which the disease progressed from the ARC to the AIDS stage during the period from month 8 to month 11, the viral load showed a more than 100-fold increase: 3.31 × 10⁵ copies/ml at month 8 and 3.76 × 10⁷ at month 11 (Fig. 3a). This cat died 12 months after the first examination from lymphoma. In two (cases 6 and 7) of three cats with disease progression within the ARC stage, the plasma viral RNA load increased by more than 10-fold: 1.82 × 10⁶ copies/ml at month 0 to 2.02 × 10⁷ at month 14 in case 6 (Fig. 3b), 1.26 × 10⁵ copies/ml at month 0 to 1.87 × 10⁷ copies/ml at month 8 in case 7 (Fig. 3c). These two cats both died within 2 months after the last examination. In case 8 in which the clinical symptoms were aggravated within the AIDS stage, the viral load remained very high: 2.58 × 10⁷ copies/ml at month 0 and 2.07 × 10⁷ copies/ml at month 2. This cat died soon after the last examination. Therefore, in three of five cats which showed obvious disease progression, an increase of the viral load of up to 150-fold was observed in parallel with the changes in clinical symptoms. Longitudinal assessment of the viral load in each FIV-infected cat can directly reveal the kinetics of viral replication. Therefore, the present findings indicate the usefulness of measuring plasma viral RNA load in understanding the immunological status and prognosis of FIV-infected cats.

FIG. 3.

Changes of plasma viral RNA load at several time points in three FIV-infected cats. Case 5 progressed from the ARC stage to the AIDS stage (a). Cases 6 and 7 showed progressions of clinical symptoms within the ARC stage (b and c, respectively).

Plasma viral RNA load has also been used as an indicator of the efficacy of antiviral therapy in HIV infection (10). Although we did not estimate the change of plasma viral RNA load after antiviral therapy in this study, a previous report showed a decline of viral load after administration of an RT inhibitor in FIV-infected cats (22). Since the real-time sequence detecting system employed here was very sensitive and easy to carry out, it may also be suitable for evaluating the efficacy of antiviral drugs in a large number of clinical samples. Antiviral chemotherapy for FIV infection in cats has yet to be established. But, when an effective antiviral chemotherapy becomes available, a laboratory examination for the plasma viral RNA load will be a powerful tool for both timing the initiation of chemotherapy and monitoring its efficacy.

Several studies have shown that plasma viral RNA load increases shortly after infection to an initial peak in AP (2), decreases in the subsequent AC (23) stage, and increases again in ARC and AIDS stages (4). Furthermore, disease progression after experimental infection with a peculiar FIV strain was reported to be rapid in cats with a high initial peak of viral load (3). Plasma viral load was also used to examine the efficacy of FIV vaccines under trial (16, 17). A high viral load accompanies the development of clinical signs due to immunodeficiency, suggesting that the increase is closely related to the development of immunodeficiency in FIV infection. On the other hand, clinical data indicating immunological dysfunction in symptomatic FIV-infected cats were also reported. Histopathological studies on the lymph node in naturally FIV-infected cats showed a serially depletive change from the ARC stage toward the AIDS stage (7). Severe lymphopenia in the peripheral blood is one of the major findings with FIV-infected cats (19). In addition to plasma viremia, immunological dysfunction coexists with disease progression and appears to be in close association with disease progression in FIV infection.

In HIV infection, the change in plasma viral RNA load from the AP to the AIDS stage is similar to that in FIV infection (18). The viral load has an obvious relation to both clinical symptoms and prognosis in patients infected with HIV (8). Although the load is low in the AC stage in HIV infection, several groups have shown that viral replication is active even at this stage (5). By quantifying the change in plasma viral RNA load after administration of protease inhibitors in HIV-infected patients, the life cycle of HIV in vivo was calculated to be 2.6 days on average (15). On the other hand, though the CD4⁺ cell count gradually decreases during HIV infection, several reports have indicated that the replication of lymphocytes was enhanced in HIV-infected patients (20). Thus, the gradual decline in the lymphocyte count seems to occur because more of these cells are destroyed than are produced. From these reports, the plasma viral RNA load is hypothesized to be a result of the balance between active replication of HIV and active elimination of the virus by the immune system. According to this hypothesis, it is conceivable that the balance between replication and elimination of virus is maintained during the period of low viral RNA loads, but a shift to the replication dominance results in a high viral RNA load and the progression to AIDS.

The mathematical dynamics of FIV replication have not been reported; thus, it is unclear whether replication occurs throughout the period of infection as in HIV infection. It is possible that the plasma viral RNA load is a result of the balance between active replication and continuous elimination of the virus; however, further investigation of the viral dynamics in FIV infection will be required. Even if this hypothesis were true, it still remains unclear whether a high plasma viral RNA load is the cause or the result of the disease progression to AIDS in FIV infection. Considering that in this study the plasma viral RNA loads in FIV-infected cats were elevated in the terminal AIDS stage just before death, it may be reasonable to speculate that the high viral load in FIV-infected cats is a result of an impairment in viral elimination due to the destruction of the immune system.