Biological Diagnosis of Ocular Toxoplasmosis: a Nine-Year Retrospective Observational Study

DISCUSSION

We performed an observational study over a 9-year period, allowing us to analyze data regarding a great number of samples. However, this kind of study, without prior design, did not allow us to access reliable clinical data about the patients whose samples were analyzed, and we could not ascertain whether the same set of techniques was always used for all analyses of the samples.

Over this 9-year period, our laboratory biologically diagnosed 80 OT cases based on 249 samples analyzed. This relatively low ratio highlights the fact that OT diagnosis primarily relies on ophthalmic examination in most cases. Indeed, in cases where the lesions found are typical of OT, clinicians often avoid performing an ACP (5, 16). Considering that most samples analyzed in our laboratory originated from patients presenting lesions that were compatible with though not typical of a diagnosis of OT, this ratio would indicate that biological confirmation is an essential diagnostic step in atypical OT presentations. In addition, when suspecting other ocular diseases, it is highly probable that clinicians take advantage of the ACP that they perform to assess the possibility of an atypical OT presentation, even when the suspicion is weak (17). This is illustrated by the fact that more than half of the samples were also analyzed for other kinds of microorganisms, either viral or bacterial. That situation artificially decreases the ratio of confirmed biological OT diagnoses, since tests are performed in settings where OT is, in fact, not clinically suspected. Finally, we cannot exclude the possibility that this low ratio might reflect low sensitivity of our diagnosis algorithm.

Of these 80 cases, 52 exhibited a positive PCR. This ratio highlights the relevance of this particular test in our diagnosis algorithm, since two-thirds of our cases could have been diagnosed using a single test on ocular samples. In the present study, the PCR was positive in three patients with an immune profile compatible with a primary infection, as well as in two patients presenting a negative serology, corresponding to either immunocompromised subjects or to very early primary infections—two situations in which PCR is believed to be highly sensitive. However, the PCR results remained negative in two subjects presenting immune profiles compatible with primary infection. The biological diagnosis of these two patients was made because of the CC, since the immunoblotting profiles of the blood and eye samples were similar. This is consistent with the course of a primary infection acquired through per os contamination, that is, systemically. However, since the CC analyzes the differences in the intensities of immunoglobulin synthesis in the eye and blood, rather than differences in antibody repertoires, we were able to detect strong intraocular antibody synthesis corresponding to retinochoroiditis. In this assay, the “mumps” ratio is necessary to assess the impermeability of the BOB. We have chosen to use the parameter of mumps virus immunity since, as previously explained, a large proportion of the French population has become immune to this virus following several vaccination campaigns (18). The same objective could, however, be achieved using virtually any other common microorganism generating long-lasting immunity and not involved in ocular pathology.

The PCR was negative in 28/80 OT cases. OT was then diagnosed using techniques based on detection of different anti-T. gondii antibody expression results, either with respect to the repertoire or the amplitude, between the blood and intraocular fluids. This could reflect the different kinetics of detection of parasites and antibodies by the tests used, during OT. Indeed, antibodies are expected to be expressed in the eye after the onset of infection, even in cases of recurrence (19, 28). In immunocompetent patients at least, under conditions of immune response pressure, given that the parasite presence in situ is probably at a low level and transient, we would expect patients with a positive PCR result to exhibit low ratios of positive immunoblotting or CC and patients with a negative PCR result to exhibit high ratios. Indeed, less than half of the positive PCR samples tested using antibody detection-based tests were positive for the latter, and, similarly, less than a third of the positive CC or immunoblot samples were positive using PCR. Nevertheless, as our analysis relied on real-life data, a small proportion of the samples underwent all three assays, thereby decreasing our statistics’ power and allowing for study biases. It may also be possible that these figures highlight a lack of sensitivity in the tests that we performed. It is, however, likely that the combination of these three assays enables the clinical parasitologists to perform the best test for each step of toxoplasmic retinochoroiditis, resulting in higher overall diagnostic performances, as previously described (13, 14, 20).

In our study, 175/249 of all the ocular samples were not tested using all three techniques; among those 175 samples, 42 displayed a positive PCR. Of the remaining 133 samples with a negative PCR, 86 were not tested with either immunoblotting or CC, and 47 were tested using only one of these tests. Most often, because of insufficient sample volume, these tests were not performed. In regard to the very high specificity of PCR, choosing not to perform antibody detection-based assays on samples already positive by PCR appears understandable, but it remains true that 133/249 (53.4%; 95%CI, 47.2 to 59.6) were not tested with the best of current diagnosis techniques, the use of which might have facilitated the OT diagnosis. This further highlights the importance of ensuring the quality of ocular liquid sampling by ophthalmologists in order to obtain reliable biological test results.

Finally, the performances of the tests that we applied could be compared with the results of a previous study carried out in the same laboratory and published in 2003. For instance, analyzing only the ocular samples tested with all three techniques, we showed a current PCR sensitivity of 40% (10/25) versus 28% PCR sensitivity in the data from 2003 (20). Moreover, the sensitivities of immunoblotting and CC were 33% and 53% in our previous study versus 60% (15/25) and 56% (14/25) in the current study, respectively. At that time, the PCR technique differed in that it used primers targeting a sequence in the B1 gene, with a lower sensitivity (21–23). Thus, these current results may likewise highlight the improvements made in the PCR technique, resulting in an increase in this assay’s sensitivity. However, if the hypothesis that PCR exhibits better performance than antibody detection-based testing at early infection stages proves to be true, the progression of the test’s performances from the earlier study to the later study may also reflect earlier AH sampling in the present study than in the previous one. Increased levels of systematic AH sampling in cases of OT suspicion may explain this sampling being performed at an earlier stage of infection than had previously been the case, reflecting changes in the practice of the ophthalmologists.

In conclusion, our study data remind us that current biological diagnostic tools for OT must be used in combination in order to make the best of the precious ocular fluids sampled by ophthalmologists but also that the accuracy of the results also relies on the volume of the samples. This need for combining tests probably reflects the fact that clinical OT suspicion might occur at different disease stages, rendering parasites or antibodies more or less easily detected depending on the progression of the infection. To gain a better understanding of which parameters are relevant for the performances of the different diagnostic tests, prospective clinicobiological studies focused on correlating the performances of the different tests with clinical characteristics, such as the delay between symptom onset and ACP or the presence of a retinal scar, are critically needed. This would also be an opportunity to assess the performance of the clinical examination within the diagnostic procedure. Moreover, such studies might shed light on fundamental aspects of OT pathophysiology and pave the way to improvements in this infection’s diagnosis and treatment.