Join us for another installation of STC (technically international) with Ramey Bajwa.  Play along and see if you can guess the diagnosis any earlier than we did (or did not).

 

Episode 68 Concepts Review
By Sadid Khan

The case is such an excellent example of a few fundamental concepts such as

1. when in doubt, go back and re-take the history
2. the importance of an exposure history
3. the importance of considering pre-test probability when ordering investigations, demonstrated through the serological testing results in this case.
4. the difficulties in serological diagnosis of infectious diseases

The appropriate ordering and accurate interpretation of investigations of any sort is such a fundamental part of diagnostic reasoning and safe medical practice. I have seen many occasions of diagnostic / investigation spirals occurring after chasing “red herring” laboratory results.

I do not have an answer for this case, but I wanted to share some thoughts about the serological testing from my perspective as infectious diseases and microbiology trainee.

Clinically this 33 years old female presenting with about 10-14 days of a febrile systemic illness with mild hepatitis, but notably no lymphocytosis or blood film abnormalities would be strongly suggestive of CMV or related syndromes (atypical lymphocytes). It is not uncommon to see some lymphopenia or leukopenia, but left-shifted granulocytes would be an unusual pattern for CMV related syndromes.

The duration of illness is important because serology needs enough time to become positive.
The age is important, especially with things like EBV and CMV because many people beyond adolescence are seropositive (and for example, you should always be very cautious about diagnosing acute EBV infection in a 70 year old).

Importantly she has cell-mediated immunosuppression through use of 6MP for her underlying Crohns disease. This is important because serological diagnosis relies upon measuring the body’s immune response to infections, so an immunosuppressed individual (in particular affecting CD4 helper T cells or B cells) – depending on their degree of immunosuppression – may not have normal serological responses. Another important consideration in immunosuppression are patients who have received IVIG as part of therapy – where the pooled infusion of IgG from the general population makes serological diagnosis not reliable.

The other difficulty is this patients background of autoimmunity + the current “inflammed” state, so the chances of throwing false positive IgM due to non-specific or cross-reactive antibodies are quite high. This is a patient where you would not trust any form of an IgM to make a serological diagnosis – you want to demonstrate an IgG seroconversion or significant rise in IgG titre over an interval period (or use a non serologically based test like PCR).

Specific to this case, the patient had positive:

• CMV IgM and IgG
• EBV IgM and IgG
• Parvovirus IgM and IgG
• Bartonella IgG 512
• Phase 2 Q Fever IgM 1280

These scenarios are important to speak with the microbiology laboratory – they often have a lot of behind the scenes information that can help guide interpretation e.g. CMV IgMs and IgGs are often reported as Yes/No, but the assays they are tested on give strengths of those values. This can be helpful in ways such as a very high IgG titre, with a low positive IgM titre (while not definitive) may be more suggestive of distant past infection with a cross-reactive IgM.

The low CMV avidity is interesting. Avidity is how strongly the antibody binds to the antigen (how “avid”), and older antibodies have had more time to mature so bind much more strongly (high avidity) compared to newly developed IgG antibodies from recent infection (low avidity). There is an intermediate avidity where the test cannot reliably distinguish. An important consideration in this case is the patient’s immunosuppression – this will affect how the antibody matures over time, and it can take much longer than an immunocompetent individual for the CMV IgG to mature and become highly avid (and it may not ever). I emphasise this because listening to this case I am not entirely convinced of acute CMV infection as the diagnosis (though still could be possible), and relying on avidity alone for this diagnosis may not be reliable in the face of other less suggestive clinical and laboratory features or more likely alternative diagnoses.

The EBV serology result provided in the case is likely incomplete (most laboratories would have additional markers being tested that can differentiate stage of infection reliably –> EBV NA IgG).

Similar concepts with Parvovirus apply, though many adults are seropositive by that age or have an exposure history of a young child. I suspect this is also past infection with a cross-reactive IgM.

Q fever is one of the most complicated serologies to interpret. Very broadly, Phase 2 IgM and IgG represents acute infections and Phase 1 IgM and IgG represent chronic infection. As mentioned above, I would not make a confident serological diagnosis based upon an IgM alone but the case did have some compatible clinical features and relevant epidemiology (noting that Q fever transmission is via aerosols and can be transmitted many kilometres downwind from high risk areas like abattoirs, etc). I cannot explain the positive Q fever PCR in this patient and may indeed be from laboratory error (but is unusual). The failure to demonstrate an IgG seroconversion goes against this, but keeping in mind attenuated serological responses in immunosuppression, and that early treatment can also delay or prevent seroconversion (though it sounds like her illness was long enough to probably do so). The dropping in the IgM titre over time likely represents the resolution of the underlying inflammatory state with a reduction in the cross-reactive antibody rather than treatment response. Even with an effective treatment I probably would have expected them to seroconvert their IgG.

I am surprised that Bartonella was not considered further. There was relevant epidemiological exposure, a reasonably compatible clinical illness (subacute systemic febrile illness) and noting that immunosuppression may alter disease presentations and may not present with the classical cat scratch syndrome. I also make note that an IgG titre of 512 (not an IgM) is much less likely to be cross-reactive (but still may be).  The absence of lymphadenopathy I agree however is unusual, but I certainly would have thought about repeat Bartonella serology to demonstrate a rise in IgG (or a Bartonella PCR is also available, but limited utility in blood)

Acute Toxoplasmosis was the other important consideration with cat exposures that I thought would have been an important diagnosis to consider. This is another common cause of a subacute systemic febrile illness with hepatitis. Often lymphocytosis is also seen, but certaintly not always. The illness is often self limiting apart from quite significant immunosuppression, and I think is quite compatible with her presentation.

Probably by far the most useful thing in interpreting complex serology is checking if they’ve had any previous serological testing, or have any historical sera stored (many laboratories are required by law to store them for at least 3 months, but often 1 year). A patient that has seen a gastroenterologist and undergone immunosuppression will almost certaintly have had some form of serological testing in the past, and things like knowing if the patient was CMV positive years ago, or having the serum from different time points to test in parallel is the most accurate ways to make serological diagnosis.

The email turned out to be much longer than anticipated so I apologise. I write mainly because the case itself is very interesting from both a clinical and laboratory perspective, and it highlights some of the great difficulties in accurate interpretation of laboratory results, and in particular serology. You both are in a great position to educate many doctors and in particular trainees, so discussing these issues and their complexities is an excellent educational endeavour.

Appropriate ordering and interpretation of investigations is often such an underappreciated element of safe diagnosis, and keeping in mind the limitations of the tests we order and always considering pretest probability – especially in serological diagnosis where there is such a tendency often for non-specific results – is important.